diff --git a/.gitignore b/.gitignore
index cde006900de80075ac87a6047ad3d26dfd695f1c..7bf42561d7a277cd5e24d9a6a9839fc01e917321 100644
--- a/.gitignore
+++ b/.gitignore
@@ -1,5 +1,5 @@
.DS_Store
.RData
.Rhistory
-.ipynb_checkpoints
+analysis_and_scripts/.ipynb_checkpoints
sources_ignored
\ No newline at end of file
diff --git a/Kandi.pdf b/Kandi.pdf
index 9eda4ee7669d08ec9b9b8b9f64c52a787a3e7d58..1facc56b54cca32436a80631a974e6dc84ad3841 100644
Binary files a/Kandi.pdf and b/Kandi.pdf differ
diff --git a/Kandi.synctex.gz b/Kandi.synctex.gz
index f312ea141e71cb34ffc0370ee740a942515eff00..dc722e5833bb2d2381319c53779cb24561b2a1e9 100644
Binary files a/Kandi.synctex.gz and b/Kandi.synctex.gz differ
diff --git a/Kandi.tex b/Kandi.tex
index 4b597c1c3a5eb59745775b37747ac76119441434..c95be60260ea7dabab50d5be22a06b74e5b304e1 100644
--- a/Kandi.tex
+++ b/Kandi.tex
@@ -44,6 +44,7 @@
\newcommand{\pr}{\mathbb{P}} % tn merkki
\newcommand{\D}{\mathcal{D}} % aineisto
\newcommand{\s}{\mathcal{S}} % tn merkki
+\newcommand{\M}{\mathcal{M}} % tn merkki
\newcommand{\R}{\mathbb{R}}
\newcommand{\C}{\mathbb{C}}
@@ -74,8 +75,8 @@
\addtolength{\voffset}{0.45cm}
\addtolength{\textheight}{-0.9cm}
-\title{Kandidaatin tutkielma\\ {\Large Rikoksenuusinnan ennustaminen kausaalipäättelyllä}} % Parempi otsikko
-\author{Riku Laine\\ Valtiotieteellinen tiedekunta, Helsingin yliopisto}
+\title{Kandidaatintutkielma\\ {\Large Kausaalipäättely ja valikoitumisharha}} % Parempi otsikko
+\author{Riku Laine\\ Valtiotieteellinen tiedekunta \\ Helsingin yliopisto}
\date{\today}
%%%%%%%%%%%%%%
@@ -103,7 +104,7 @@ Tämän tutkielman on tarkastanut XYZ. Haluan kiittää kaikkia edellä mainittu
\bigskip
-\rightline{Helsingissä \today}
+\rightline{Helsingissä \today,}
\rightline{Riku Laine}
\bigskip
@@ -114,56 +115,35 @@ Tämän tutkielman on tarkastanut XYZ. Haluan kiittää kaikkia edellä mainittu
\noindent I would like to wholeheartedly thank assistant professor Michael Mathioudakis from University of Helsinki's Department of Computer Science for numerous things. He provided me this extremely interesting thesis topic and provided insightful and encouraging comments throughout the process. Antti Hyttinen from the same department also gave important insight in the causal modelling and commented on the content.
-
-%%%%%%%%%
-%%%%%%%%%
-%%%%%%%%%
-
-\chapter{Tiivistelmä -- Kypsyysnäyte?}\label{tiiv}
-
-% refillä pelkät numerot
-
-\emph{\nameref{johd}}-luvussa esittelen ongelman asettelun ja tilanteen yleisen viitekehyksen. Keskustelemme rikoksenuusinnan ennustamisesta yhdysvaltalaisessa oikeusjärjestelmässä. Esitän kappaleessa yleisen kuvauksen takuukäsittelyn etenemisestä oikeusprosessina, jonka jälkeen pohdin hieman takuukäsittelyn yhteiskunnallista merkitystä ja motivaatiota hyvään ennusteeseen. Kappaleen lopussa kirjoitan hieman kausaalipäättelystä uutena tilastotieteellisenä paradigmana \cite{pearl10}.
-
-Kappaleessa \emph{\ref{aineisto}} esittelen käyttämäni aineistolähteet ja niiden ominaispiirteet. Esitän COMPAS-tietojen ominaispiirteet ja \emph{jotain muuta}. Esitän myös kuinka olen luonut analyyseissä myöhemmin käytettävän aineistosetin mukaillen Lakkarajun vuoden 2017 konferenssijulkaisua \cite{lakkaraju17}.
-
-\emph{\nameref{metodit}}-kappaleessa esitän käyttämäni mallit ja menetelmät. Teen lyhyen katsauksen aikaisempaan kirjallisuuteen ja tutkimuksiin tällä sovellusalalla. Käyn lisäksi läpi tässä tutkielmassa myöhemmin käytettäviä matemaattisia sekä verkkoteoreettisia merkintöjä ja määritelmiä. Teen joitakin osoituksia ja osoitan kuinka mallimme ei riipu havaitsemattomista (unobservables) muuttujista. % Mallin robustius?
-
-Luvussa \emph{\ref{tulokset}} esitän algoritmillani saavuttamani tulokset ja vertailen niitä Lakkarajun \cite{lakkaraju17} saavuttamiin. Olen eritellyt erillisiin alalukuihin synteettisellä ja COMPAS-aineistoseteillä saavutetut tulokset.
-
-Viimeisessä kappaleessa \emph{\nameref{diskussio}} esitän mallien ja tutkielmani virhelähteet ja muut ongelmat sekä keskustelen tulosten mahdollisesta vaikutuksesta, sikäli niitä sovellettaisiin sikäläisen oikeuslaitoksen toimintaan.
-
-%%%%%%%%%
-%%%%%%%%%
-%%%%%%%%%
-
\chapter{Johdanto}\label{johd}
-Tässä kappaleessa esittelen tutkielman taustaa ja yhdysvaltalaisen oikeuslaitoksen takuukäsittelyprosessin yleisellä tasolla. Sen jälkeen paneudun hieman vangitsemispäätöksen yhteiskunnalliseen merkitykseen: minkä takia ihmisiä vangitaan ja mitä perusteita on vangitsemattajättämispäätökselle. Pyrin luvun aikana myös hieman selvittämään takuujärjestelmän käyttöä Suomessa ja kappaleen lopussa pohdin hieman kausaalipäättelyä paradigman muutoksena tilastotieteen kentällä. Jätän kuitenkin tarvittavien merkintöjen esittämisen kappaleeseen \emph{\nameref{kausaalimerk}} ja mallin esittelyn \emph{\nameref{kausaalimalli}}-lukuun.
+Tämän tutkielman tavoitteena on luoda kausaalipäättelyn avulla algoritmi, jolla voimme arvioida ennustavien mallien tarkkuutta, kun käytettävissä on ainoastaan valikoitumisharhasta kärsivää aineistoa. Samankaltaista asetelmaa ovat julkaisuissaan käsitelleet muun muassa Lakkaraju ja Madras \cite{lakkaraju17, madras18}. Pyrin tutkielmassani luomaan joustavamman ja tarkemman vaihtoehdon Lakkarajun luomalle supistusalgoritmille, mutta esitän ensin yleistä taustaa kausaalipäättelystä ja valikoitumisharhasta.
-% https://julkaisut.valtioneuvosto.fi/bitstream/handle/10024/76171/omkm_2009_2.pdf
-
-%%%%%%%%%
-
-\section{Takuukäsittely prosessina}\label{pros}
+%Tässä kappaleessa esittelen tutkielman taustaa ja yhdysvaltalaisen oikeuslaitoksen takuukäsittelyprosessin yleisellä tasolla. Sen jälkeen paneudun hieman vangitsemispäätöksen yhteiskunnalliseen merkitykseen: minkä takia ihmisiä vangitaan ja mitä perusteita on vangitsemattajättämispäätökselle. Pyrin luvun aikana myös hieman selvittämään takuujärjestelmän käyttöä Suomessa ja kappaleen lopussa pohdin hieman kausaalipäättelyä paradigman muutoksena tilastotieteen kentällä. Jätän kuitenkin tarvittavien merkintöjen esittämisen kappaleeseen \emph{\nameref{kausaalimerk}} ja mallin esittelyn \emph{\nameref{kausaalimalli}}-lukuun.
-% Johdanto, yhdysvallat, Suomi, kritiikki
-
-Yhdysvalloissa, kuten monissa muissa anglosaksisissa maissa, on käytössä järjestelmä, jota nimitetään takuu- tai vakuusjärjestelmäksi. Takuujärjestelmä on epäillyn vaihtoehto tutkintavankeudelle hänen odottaessaan oikeudenkäyntiä ja Yhdysvalloissa oikeus takuuseen periytyy maan perustamisen ajalta \cite{okm, zaniewski14}. Suomen oikeus- ja sisäasiainministeriön alaisen esitutkinta- ja pakkokeinotoimikunnan mukaan takuujärjestelmiä on kolmenlaisia: kahdessa niistä epäilty maksaa itse käteisellä vakuuden tai asettaa omaisuuttaan vakuudeksi ja kolmannessa jokin ulkopuolinen taho ''menee takuuseen epäillyn velvollisuuksien täyttämisestä'' \cite{okm}.
-
-Yhdysvalloissa epäillyn pidätyksen jälkeen hänet viedään paikallisen oikeusviranomaisen järjestämään takuukuulemiseen (bail hearing) \cite{zaniewski14}. Kuulemisessa päätetään takuun myöntämisestä, eli voidaanko epäilty vapauttaa, vai halutaanko hänet asettaa vankeuteen ennen oikeudenkäyntiä. Kuulemisessa päätetään myös mahdollisen takuun määrästä sekä vapauttamisen ehdoista \cite{zaniewski14}. Takuu voidaan suorittaa taattuna tai takaamattomana maksusitoumuksena tai maksaa suoraan -- erityistapauksissa epäilty voidaan vapauttaa myös pelkällä kirjallisella sitoumuksella (release on personal recognizance (ROR)) \cite{zaniewski14}.
-
-% Tilastoja?
+% https://julkaisut.valtioneuvosto.fi/bitstream/handle/10024/76171/omkm_2009_2.pdf
%%%%%%%%%
-\section{Yhteiskunnallinen merkitys ja kritiikki}\label{ykmerk}
-
-Zaniewski toteaa lyhyessä kirjallisuuskatsauksessaan, että takuujärjestelmän vuoden 1982 uudistus ei onnistunut laskemaan tarpeettomia vangitsemisia -- päinvastoin niiden suhteellinen määrä kaksinkertaistui 22\%:sta 49\%:iin vuodesta 1984 vuoteen 2007. Nykyisellään sikäläinen oikeusjärjestelmä suosii suoraan rahalla maksettavia tai taatuilla maksusitoumuksilla hoidettuja takuita, mikä asettaa huonossa taloustilanteessa olevat epäillyt eri tilanteeseen. \cite{zaniewski14}
-
-Suomessa vakuusjärjestelmää ei ole käytetty, vaikka aiemmin mainittu toimikunta toteaakin sen sisältyvän tullilain 44 §:ään. Kyseisessä pykälässä ''- - säädetään mahdollisuudesta asettaa pidätetyn tai vangitun vapaaksi päästämi[s]en ehdoksi, että hän asettaa vakuuden, jonka harkitaan takaavan hänen saapumisensa oikeudenkäyntiin ja ehkä tuomittavien seuraamusten suorittamisen''. Kuten he tarkentavat, lisäksi usein edellytetään, että epäilty ei asu Suomessa, ja epäillään hänen pakenevan maasta ennen oikeudenkäyntiä tai rangaistusta \cite{okm}. Sekä yhdysvaltalaiselle että suomalaiselle järjestelmälle on yhteistä, että takuu tuomitaan menetettäväksi valtiolle, jos vapauden ehtoja rikotaan.
-
-Kritiikkiä on esitetty molemmissa maissa osaltaan samoihin asioihin. Suomessa pykälää ei ole sovellettu, koska luultavasti sen tulkintaohjeet ovat niin niukat, kuten myös sääntely \cite{okm}. Yhdistävänä kritiikkinä sekä Zaniewski että esitutkinta- ja pakkokeinotoimikunta mainitsevat muun muassa sen, kuinka takuumaksujen toimeenpano vaikuttaa tai Suomen tapauksessa vaikuttaisi pienituloisten taloustilanteeseen \cite{zaniewski14, okm}. Suomalainen toimikunta esittää lisäksi monia muitakin ongelmakohtia, sikäli takuujärjestelmä haluttaisiin ottaa Suomessa käyttöön, esimerkkinä he toteavat, että vakuusmaksujen maksamiseen tulisi todennäköisesti liittymään ''epätoivottavia lieveilmiöitä'' \cite{okm}. Tähän ongelmaan on Yhdysvalloissa jo osittain reagoitukin, sillä esimerkiksi Californian osavaltio päätti viime vuonna poistaa takuumaksut käytöstä \cite{cnn}.
+%\section{Takuukäsittely prosessina}\label{pros}
+%
+%% Johdanto, yhdysvallat, Suomi, kritiikki
+%
+%Yhdysvalloissa, kuten monissa muissa anglosaksisissa maissa, on käytössä järjestelmä, jota nimitetään takuu- tai vakuusjärjestelmäksi. Takuujärjestelmä on epäillyn vaihtoehto tutkintavankeudelle hänen odottaessaan oikeudenkäyntiä ja Yhdysvalloissa oikeus takuuseen periytyy maan perustamisen ajalta \cite{okm, zaniewski14}. Suomen oikeus- ja sisäasiainministeriön alaisen esitutkinta- ja pakkokeinotoimikunnan mukaan takuujärjestelmiä on kolmenlaisia: kahdessa niistä epäilty maksaa itse käteisellä vakuuden tai asettaa omaisuuttaan vakuudeksi ja kolmannessa jokin ulkopuolinen taho ''menee takuuseen epäillyn velvollisuuksien täyttämisestä'' \cite{okm}.
+%
+%Yhdysvalloissa epäillyn pidätyksen jälkeen hänet viedään paikallisen oikeusviranomaisen järjestämään takuukuulemiseen (bail hearing) \cite{zaniewski14}. Kuulemisessa päätetään takuun myöntämisestä, eli voidaanko epäilty vapauttaa, vai halutaanko hänet asettaa vankeuteen ennen oikeudenkäyntiä. Kuulemisessa päätetään myös mahdollisen takuun määrästä sekä vapauttamisen ehdoista \cite{zaniewski14}. Takuu voidaan suorittaa taattuna tai takaamattomana maksusitoumuksena tai maksaa suoraan -- erityistapauksissa epäilty voidaan vapauttaa myös pelkällä kirjallisella sitoumuksella (release on personal recognizance (ROR)) \cite{zaniewski14}.
+%
+%% Tilastoja?
+%
+%%%%%%%%%%
+%
+%\section{Yhteiskunnallinen merkitys ja kritiikki}\label{ykmerk}
+%
+%Zaniewski toteaa lyhyessä kirjallisuuskatsauksessaan, että takuujärjestelmän vuoden 1982 uudistus ei onnistunut laskemaan tarpeettomia vangitsemisia -- päinvastoin niiden suhteellinen määrä kaksinkertaistui 22\%:sta 49\%:iin vuodesta 1984 vuoteen 2007. Nykyisellään sikäläinen oikeusjärjestelmä suosii suoraan rahalla maksettavia tai taatuilla maksusitoumuksilla hoidettuja takuita, mikä asettaa huonossa taloustilanteessa olevat epäillyt eri tilanteeseen. \cite{zaniewski14}
+%
+%Suomessa vakuusjärjestelmää ei ole käytetty, vaikka aiemmin mainittu toimikunta toteaakin sen sisältyvän tullilain 44 §:ään. Kyseisessä pykälässä ''- - säädetään mahdollisuudesta asettaa pidätetyn tai vangitun vapaaksi päästämi[s]en ehdoksi, että hän asettaa vakuuden, jonka harkitaan takaavan hänen saapumisensa oikeudenkäyntiin ja ehkä tuomittavien seuraamusten suorittamisen''. Kuten he tarkentavat, lisäksi usein edellytetään, että epäilty ei asu Suomessa, ja epäillään hänen pakenevan maasta ennen oikeudenkäyntiä tai rangaistusta \cite{okm}. Sekä yhdysvaltalaiselle että suomalaiselle järjestelmälle on yhteistä, että takuu tuomitaan menetettäväksi valtiolle, jos vapauden ehtoja rikotaan.
+%
+%Kritiikkiä on esitetty molemmissa maissa osaltaan samoihin asioihin. Suomessa pykälää ei ole sovellettu, koska luultavasti sen tulkintaohjeet ovat niin niukat, kuten myös sääntely \cite{okm}. Yhdistävänä kritiikkinä sekä Zaniewski että esitutkinta- ja pakkokeinotoimikunta mainitsevat muun muassa sen, kuinka takuumaksujen toimeenpano vaikuttaa tai Suomen tapauksessa vaikuttaisi pienituloisten taloustilanteeseen \cite{zaniewski14, okm}. Suomalainen toimikunta esittää lisäksi monia muitakin ongelmakohtia, sikäli takuujärjestelmä haluttaisiin ottaa Suomessa käyttöön, esimerkkinä he toteavat, että vakuusmaksujen maksamiseen tulisi todennäköisesti liittymään ''epätoivottavia lieveilmiöitä'' \cite{okm}. Tähän ongelmaan on Yhdysvalloissa jo osittain reagoitukin, sillä esimerkiksi Californian osavaltio päätti viime vuonna poistaa takuumaksut käytöstä \cite{cnn}.
%Kritiikkiä on esitetty niin itse takuun rahallisesta määrästä (lähde?) kuin perusteista (propublica).
@@ -174,69 +154,87 @@ Kritiikkiä on esitetty molemmissa maissa osaltaan samoihin asioihin. Suomessa p
\section{''Kausaalipäättely uutena paradigmana''}\label{para}
% miksi halutaan siirtyä (frekventistisen/bayes-ppäättelyn ongelmat), edut, esiintyminen, erot, käyttö
-Kuten Pearl ja Mackenzie esittävät kirjassaan Miksi, ihmisillä on luontainen kausaalisen päättelyn taito \cite{miksi}. Tavalliset tilastollisen päättelyn menetelmät eivät tarjoa tapaa määritellä kausaalista yhteyttä: aineistosta voidaan päätellä erilaisia \emph{korrelaatioita}, mutta päättely \emph{A johtuu B:stä} vaatii uudenlaista lähestymistapaa. Käytännön tutkimuksessa kausaaliset yhteydet kiinnostavat erityisesti lääketieteen alalla. Kuten Kalisch toteaa, aiemmin päättely on perustunut jonkin biomarkkerin ja taudin samanaikaiseen ilmaantumiseen. Jos markkeri ja tauti ilmaantuvat samanaikaisesti, voidaanko markkerin arvoa muuttamalla hoitaa tautia? \cite{kalisch14}
-
-Syy-seuraussuhteen vahvuuden matemaattinen määrittely vaatii uutta lähestymistä myös todennäköisyyslaskennan merkintöihin. Pearl käyttää alkuperäisessä, englanninkielisessä kirjallisuudessa merkintää 'do' ilmaisemaan interventiota. Merkinnällä halutaan erottaa tavanomainen ehdollinen todennäköisyys $\pr(Y|X=x)$ interventiosta, jossa pakotamme muuttujan $X$ arvoon $x$: $\pr(Y|\text{do}(X=x))$. Kimmo Pietiläinen käyttää kirjan suomennoksessa do-operaattorista käännöstä \emph{tee}, mutta seuraan tässä tutkielmassa Pearlin merkintöjä, ellen erikseen muuta mainitse \cite{miksi}. Esittelen käyttämäni merkinnät tarkemmin kappaleessa \ref{kausaalimerk}.
-
-
-* Esimerkkejä Miksi-kirjasta väärin määritellyistä malleista? Esimerkkejä aloista, joila jo käytetty, oleellisimmat pointit historiasta
-%%%%%%%%%
-
-\section{Valikoitumisharha}\label{sl}
+Kuten Pearl ja Mackenzie esittävät kirjassaan Miksi, ihmisillä on luontainen kausaalisen päättelyn taito \cite{miksi}. Tavalliset tilastollisen päättelyn menetelmät eivät tarjoa tapaa määritellä kausaalista yhteyttä: aineistosta voidaan päätellä erilaisia \emph{korrelaatioita}, mutta päättely \emph{A johtuu B:stä} vaatii uudenlaista lähestymistapaa. Käytännön tutkimuksessa kausaaliset yhteydet kiinnostavat erityisesti lääketieteen alalla. Kuten Kalisch toteaa, aiemmin kausaalisuuden päättely on perustunut korrelaatioiden havaitsemiseen. On hypotetisoitu, että biomarkkerin ja taudin samanaikainen ilmaantuminen viittaisi siihen, että markkeri aiheuttaa taudin. Voimmeko siis markkeria käsittelemällä vaikuttaa tautiin tai jopa parantaa se? \cite{kalisch14}
-% aiempaa tutkimusta, miten voidaan muissa tutkimuksissa tassoittaa -> Tässä tutkimkssa
+Syy-seuraussuhteen matemaattinen määrittely vaatii uutta lähestymistä myös todennäköisyyslaskennan merkintöihin. Pearl käyttää alkuperäisessä, englanninkielisessä kirjallisuudessa merkintää 'do' ilmaisemaan interventiota. Merkinnällä halutaan erottaa tavanomainen ehdollinen todennäköisyys $\pr(Y|X=x)$ interventiosta, jossa asetamme muuttujan $X$ arvoon $x$: $\pr(Y|\text{do}(X=x))$. Kimmo Pietiläinen käyttää kirjan suomennoksessa do-operaattorista käännöstä \emph{tee}, mutta seuraan tässä tutkielmassa Pearlin merkintöjä, ellen erikseen muuta mainitse \cite{miksi}. Alalla käytetään myös muita, alaindekseillä rikastettuja merkintätapoja \cite{pearl10}. Esittelen käyttämäni merkinnät tarkemmin kappaleessa \ref{kausaalimerk}.
-Tässä tutkielmassa yritän määrittää rakenteen, jonka avulla voidaan tehdä ennusteita aineston harhaisuudesta huolimatta. Meidän tapauksessamme harha syntyy tuomarien päätöksistä -- jos tuomari päättää evätä epäillyltä takuut, emme voi tehdä havaintoja epäillyn rikoksen uusinnastaan. Tällöin voidaan puhua ei-satunnaisesta puuttuneisuudesta, koska on selvää että tulosten puute ei ole minkäänlaisen satunnaisprosessin tulos: vaarallisimmat rikolliset halutaan ottaa talteen ja vaarattomimmat päästää pois \cite{laaksonen13}.
+Kausaalipäättelyssä mallit voidaan esittää graafeina, eli verkkoina. Verkoista voidaan suoraan lukea eri muuttujien relaatiot kausaalisuuden suuntien ja riippuvuuksien suhteen.
-Lakkaraju käyttää termiä harhasta \emph{''selective labels''} \cite{lakkaraju17}.
-
-%%%%%%%%%
%%%%%%%%%
-%%%%%%%%%
-
-\chapter{Aineistot}\label{aineisto}
-Tässä luvussa kuvaillaan käytetyt aineistot ja niiden ominaispiirteet.
+\section{Valikoitumisharha -- seulotun aineiston ongelma}\label{sl}
-%%%%%%%%%
+Aineiston luova mekanismi on esitetty kuvassa \ref{valikoitumisharha} ja toimii siten, että aluksi jokin henkilö tai muu entiteetti saapuu päätöksentekijän eteen seulottavaksi. Päätöksentekijän tavoitteena on estää haitallinen tulos ($y=0$) pitäen samalla myönteisten päätösten ($t=1$) määrä mahdollisimman pienenä. Seuloja pyrkii siis antamaan kielteisen päätöksen kaikille niille, joilla epätoivottava tulos on todennäköisin. Päätöksen jälkeen henkilö siirtyy vaiheeseen, jossa Kohtalo määrittää hänelle tuloksen $y\in\{0, 1\}$. Kielteisen päätöksen saaneille tulos voidaan merkitä puuttuvaksi tai onnistuneeksi, koska haitallista tapahtumaa ei havaita.
-\section{COMPAS}\label{compas}
+Aineiston generoivaa mekanismia voidaan havainnollistaa lääke- ja oikeustieteen alan esimerkeillä. Henkilöllä viitataan ensin mainitussa potilaaseen ja jälkimmäisessä epäiltyyn. Seuloja voi olla esimerkiksi lääkäri, joka päättää annetaanko potilaalle vahvempaa ja samalla kalliimpaa lääkettä, jolloin relapsia ei havaita. Oikeudellisessa asetelmassa seulojalla voidaan tarkoittaa tuomaria, joka päättää epäillyn vapauttamisesta takuita vastaan ilman pelkoa rikoksen uusimisesta. Molemmilla päättäjillä on selkeä kannustin estää haitalliset tulokset -- sairauskohtaukset tai rikokset -- pitäen samalla päätöksistä aiheutuvat rasitteet yhteiskunnalle ja yksilöiden elämille mahdollisimman pienenä.
-COMPAS-aineisto (Correctional Offender Management Profiling for Alternative Sanctions) on alun perin ProPublica-julkaisun koostama aineisto yhteensä 18 610 amerikkalaisesta. Aineistossa on muun muassa heidän demografiset tiedot, kuten ikä, sukupuoli ja rotu, ja rikoshistoriaan liittyvät tiedot. Oikeammin COMPAS viittaa Northpointe-yhtiön työkaluun, joka antaa arvion epäillyn rikoksenuusintariskistä. Arvio perustuu epäillyn vastauksiin kyselyyn, jossa tiedustellaan hänen taustoistaan, kuten lähipiirin huumeidenkäytöstä ja epäillyn taipumuksesta väkivaltaisuuteen. ProPublica kokosi aineiston alun perin paljastaakseen arvion tuottavan algoritmin mustia syrjivän luonteen. ProPublican analyysi osoitti, että mustat saivat järjestelmällisesti korkeamman riskiarvion kuin valkoihoiset. \cite{propublica16}
+Havaintoja voi puuttua erilaisissa tutkimuksissa useista eri syistä. Kyselytutkimuksissa vastauskatoa voi syntyä esimerkiksi vastaajan haluttomuudesta vastata kysymykseen tai yksinkertaisesti siitä syystä, että vastaajaa ei tavoiteta. Jos aineiston puuttuneisuusmekanismi on luonteeltaan täysin satunnainen, eli vastauksen puuttuneisuus ei liity millään tavalla mitattuihin muuttujiin, voidaan sanoa aineistoa puuttuvan \emph{täysin satunnaisesti}. Käänteisessä tapauksessa voidaan puhua \emph{ei-satunnaisesta puuttuvuuudesta}. \cite{laaksonen13}
-ProPublica esittää artikkelinsa metodologiaosiossa, kuinka he ovat päätyneet lopulliseen aineistoon, joka käsittää tiedot 6172 henkilöstä. Pääpiirteissään he ovat siistineet aineistoa siten, että se yhdistää oikeat henkilöt oikeisiin pisteytyksiin ja oikeisiin uusintatuomioihin. Joitakin johdettuja mutujia luotiin, kuten tekstuaalinen kuvaus desiilipisteytyksestä scoretext joka ryhmittää etc etc.
+Tässä tutkielmassa tarkasteltavasssa asetelmassa havaintojen puuttuminen liittyy sekä havaittuihin että havaitsemattomiin muuttujiin. Puuttuneisuuden voidaan sanoa olevan \emph{satunnaista ehdollisesti}, koska aineistoa puuttuu vain yksilöiltä, joilla on korkea todennäköisyys haitalliseen tulokseen. (Erilaisia aineiston puuttuneisuusmekanismeja esitelllään laajemmin esimerkiksi Laaksosen kirjassa \emph{Surveymetodiikka}.) Puuttuneisuutta voidaan korvata imputoinnilla, jolla yritetään tehdä mahdollisimman hyvä arvaus puuttuvasta arvosta. Todistan tutkielmassani myöhemmin, että kausaalipäättelyä hyödyntämällä voimme estimoida havaitusta, valikoitumisharhaisesta aineistosta haluttuja tunnuslukuja ilman imputointia harhattomasti. \cite{laaksonen13} Englanninkielisessä kirjallisuudessa seulotun aineiston ongelmasta on alettu käyttää Lakkarajun esittämää termiä \emph{selective labels} \cite{lakkaraju17}. % se lähde, missä näin väitettiin
-\begin{table}[h!]
+\begin{figure}%[H]
\centering
-\begin{tabular}{lrrrrrrrrrr}
-\hline \hline
- Muuttujan nimi & $\bar{x}$ & Keskihajonta & Min & 25\% & 50\% & 75\% & Max \\
-\hline \hline
- age & 34,5 & 11,7 & 18 & 25 & 31 & 42 & 96 \\
- priors\_count & 3,25 & 4,74 & 0 & 0 & 1 & 4 & 38 \\ \hline
- days\_b\_screening\_arrest & -1,74 & 5,08 & -30 & -1 & -1 & -1 & 30 \\
- decile\_score & 4,42 & 2,84 & 1 & 2 & 4 & 7 & 10 \\
- is\_recid & 0,484 & 0,500 & 0 & 0 & 0 & 1 & 1 \\ \hline
- two\_year\_recid & 0,455 & 0,498 & 0 & 0 & 0 & 1 & 1 \\
- length\_of\_stay & 14,6 & 46,7 & -1 & 0 & 1 & 5 & 799 \\
-\hline \hline
-\end{tabular}
-\caption{COMPAS-aineiston numeeristen muuttujien hajontalukuja}
-\label{table:1}
-\end{table}
+\includegraphics[scale = 0.4]{valikoitumis_iso}
+\caption{Valikoitumisharha aineiston generoivana mekanismina \cite{lakkaraju17}}
+\label{valikoitumisharha}
+\end{figure}
+%%%%%%%%%
+%%%%%%%%%
%%%%%%%%%
-\section{Synteettinen}\label{synteettinen}
-
-Synteettinen aineisto luotiin Lakkarajun artikkelissaan selostamalla tavalla \cite{lakkaraju17}. aineistoan simuloitiin kolme muuttujaa $X$, $Z$, ja $W$. Näistä muuttujista $X$ vastaa informaatiota, joka on sekä mallin että tuomarin havaittavissa, eli informaatiota, joka on kirjattu oikeuden pöytäkirjoihin tai on kerättävissä muista rekistereistä, kuten vastaajan sukupuoli. Muuttujalla $Z$ kuvataan tietoa, jonka vain tuomari voi havaita: kuten Lakkaraju havainnollistaa, tällaista voi olla esimerkiksi tieto siitä, onko vastaajalla perhettä mukana oikeussalissa \cite{lakkaraju17}. $W$ on mallissa havainnollistamassa reaalimaailmaa. Muuttujalla esitämme aineistossa informaatiota, joka ei ole saatavilla päätöksentekijöille eikä mallille mutta vaikuttaa silti rikoksenuusimisriskiin. aineistossa nämä ovat kaikki riippumattomia standardinormaalijakautuneita satunnaismuuttujia, eli $X, W, Z \sim N(0, 1) \independent$.
-
-Yhdistämme henkilöt satunnaisesti kuhunkin $M = 500$ tuomariin, joista jokaiselle määritellään hyväksymisprosentti $r \in [0,1]$. Tuomarin hyväksymisprosentti määritetään ottamalla arvoja tasajakaumasta suljetulta väliltä [0,1; 0,9] ja sitten pyöristämällä ne 10 desimaalin tarkkuuteen. Tulosmuuttuja Y simuloidaan määrittämällä sen ehdollinen todennäköisyys seuraavasti: $\pr(Y=0|X, Z, W)=\frac{1}{1+\text{exp}\{-(\beta_XX+\beta_ZZ+\beta_WW)\}}$, missä kertoimet $\beta_X$, $\beta_Z$ ja $\beta_W$ on asetettu arvoihin 1, 1 ja 0,2 vastaavassa järjestyksessä. \cite{lakkaraju17}
+\chapter{Aineiston generointi}\label{aineisto}
-Päätösmuuttujan $T$ ehdolinen todennäköisyys $\pr(T=0|X, Z)=\frac{1}{1+\text{exp}\{-(\beta_XX+\beta_ZZ)\}} + \epsilon$ missä $\epsilon \sim N(0, 0,1)$ vastaa pientä määrää kohinaa. Henkilöltä $i$ kielletään takuut, eli $T_i=0$ jos muuttujan $T$ ehdollinen todennäköisyys on tuomarin $j$ suurimman $(1-r)\cdot 100\%$ joukossa. Lopuksi koulutusaineisto suodatettiin siten, että saatavissa oli vain yksilöt, jotka päästettiin vapaaksi $(T=1)$. \cite{lakkaraju17}
-\begin{table}[h!]
+%%%%%%%%%%
+%
+%\section{COMPAS}\label{compas}
+%
+%COMPAS-aineisto (Correctional Offender Management Profiling for Alternative Sanctions) on alun perin ProPublica-julkaisun koostama aineisto yhteensä 18 610 amerikkalaisesta. Aineistossa on muun muassa heidän demografiset tiedot, kuten ikä, sukupuoli ja rotu, ja rikoshistoriaan liittyvät tiedot. Oikeammin COMPAS viittaa Northpointe-yhtiön työkaluun, joka antaa arvion epäillyn rikoksenuusintariskistä. Arvio perustuu epäillyn vastauksiin kyselyyn, jossa tiedustellaan hänen taustoistaan, kuten lähipiirin huumeidenkäytöstä ja epäillyn taipumuksesta väkivaltaisuuteen. ProPublica kokosi aineiston alun perin paljastaakseen arvion tuottavan algoritmin mustia syrjivän luonteen. ProPublican analyysi osoitti, että mustat saivat järjestelmällisesti korkeamman riskiarvion kuin valkoihoiset. \cite{propublica16}
+%
+%ProPublica esittää artikkelinsa metodologiaosiossa, kuinka he ovat päätyneet lopulliseen aineistoon, joka käsittää tiedot 6172 henkilöstä. Pääpiirteissään he ovat siistineet aineistoa siten, että se yhdistää oikeat henkilöt oikeisiin pisteytyksiin ja oikeisiin uusintatuomioihin. Joitakin johdettuja mutujia luotiin, kuten tekstuaalinen kuvaus desiilipisteytyksestä scoretext joka ryhmittää etc etc.
+%
+%\begin{table}[H]
+%\centering
+%\begin{tabular}{lrrrrrrrrrr}
+%\hline \hline
+% Muuttujan nimi & $\bar{x}$ & Keskihajonta & Min & 25\% & 50\% & 75\% & Max \\
+%\hline \hline
+% age & 34,5 & 11,7 & 18 & 25 & 31 & 42 & 96 \\
+% priors\_count & 3,25 & 4,74 & 0 & 0 & 1 & 4 & 38 \\ \hline
+% days\_b\_screening\_arrest & -1,74 & 5,08 & -30 & -1 & -1 & -1 & 30 \\
+% decile\_score & 4,42 & 2,84 & 1 & 2 & 4 & 7 & 10 \\
+% is\_recid & 0,484 & 0,500 & 0 & 0 & 0 & 1 & 1 \\ \hline
+% two\_year\_recid & 0,455 & 0,498 & 0 & 0 & 0 & 1 & 1 \\
+% length\_of\_stay & 14,6 & 46,7 & -1 & 0 & 1 & 5 & 799 \\
+%\hline \hline
+%\end{tabular}
+%\caption{COMPAS-aineiston numeeristen muuttujien hajontalukuja}
+%\label{table:1}
+%\end{table}
+
+%%%%%%%%%
+
+%\section{Synteettinen}\label{synteettinen}
+
+Synteettinen aineisto luotiin Lakkarajun selostamalla tavalla. Aineistoon simuloitiin kolme muuttujaa $X$, $Z$, ja $W$. Näistä muuttujista $X$ vastaa informaatiota, joka on sekä mallin että päätöksentekijän havaittavissa. Käytännössä muuttuja $X$ vastaa kirjallista informaatiota, joka on kirjattu erilaisiin pöytäkirjoihin tai rekistereihin. Muuttujalla $Z$ kuvataan tietoa, jonka vain päätöksentekijä voi havaita: kuten Lakkaraju havainnollistaa, tällaista voi olla oikeudessa esimerkiksi tieto siitä, onko vastaajalla perhettä mukana oikeussalissa. $W$ tuo malliin kohinaa. Muuttujalla esitämme aineistossa informaatiota, joka ei ole saatavilla päätöksentekijöille eikä mallille, mutta vaikuttaa silti epätoivottavan tuloksen riskiin. Aineistossa nämä ovat kaikki riippumattomia standardinormaalijakautuneita satunnaismuuttujia, eli $X, W, Z \sim N(0, 1) \independent$. \cite{lakkaraju17}
+
+Aineistossa jyvitämme jokaiselle $M=100$ päätöksentekijälle 500 arvioitavaa. Kaikille päättäjille arvotaan hyväksymisprosentti ottamalla arvoja tasajakaumasta suljetulta väliltä [0,1; 0,9] ja sitten pyöristämällä saadut arvot 10 desimaalin tarkkuuteen. Tulosmuuttuja Y määritetään ehdollisen todennäköisyyden
+\begin{equation} \label{y_ehd}
+\pr(Y=0|X, Z, W)=\dfrac{1}{1+\text{exp}\{-(\beta_XX+\beta_ZZ+\beta_WW)\}}
+\end{equation}
+mukaisesti. Jos $\pr(Y=0|X, Z, W) \geq 0,5$, tulosmuuttujan arvoksi asetetaan 0 ja vastaavasti jos $\pr(Y=0|X, Z, W) < 0,5$ muuttujan arvoksi asetetaan 1. Lausekkeissa \ref{y_ehd} ja \ref{t_ehd} olevat kertoimet $\beta_X$, $\beta_Z$ ja $\beta_W$ on asetettu arvoihin 1, 1 ja 0,2 vastaavassa järjestyksessä. \cite{lakkaraju17}
+
+Päätösmuuttuja $T$ määritetään kaksivaiheisesti: ensin määritetään todennäköisyys kielteiselle päätökselle ja sitten muuttujan arvo asetetaan näiden todennäköisyyksien keskinäisen suuruuden mukaisesti. Muuttujan $T$ ehdollinen todennäköisyys
+\begin{equation} \label{t_ehd}
+\pr(T=0|X, Z)=\frac{1}{1+\text{exp}\{-(\beta_XX+\beta_ZZ)\}} + \epsilon,
+\end{equation}
+missä $\epsilon \sim N(0, 0,1)$ vastaa pientä määrää kohinaa. Henkilölle $i$ annetaan kielteinen päätös, eli $T_i=0$, jos muuttujan $T$ ehdollinen todennäköisyys on päättäjän $j$ suurimman $(1-r)\cdot 100\%$ joukossa. Toisin sanoen tuomari $j$ antaa myönteisen päätöksen $r$ prosentille hänen arvioitavakseen annetuista henkilöistä, joilla on alin todennäköisyys kielteiseen päätökseen. \cite{lakkaraju17}
+
+Kun aineisto saatiin simuloitua, se jaettiin koneoppimisen käytäntöjen mukaisesti kahteen yhtä suureen osaan, niin sanottuihin koulutus- ja testiaineistoihin. Lopuksi koulutusaineistoa muokattiin siten, että tulosmuuttujan arvo oli saatavissa vain yksilöille, joille oli annettu positiivinen päätös $(T=1)$. Kielteisen päätöksen saaneille tulosmuuttujan arvo asetettiin arvoon NA, kuten kuvassa \ref{valikoitumisharha}. Syntetisoidun aineiston keskeisimmät hajontaluvut on esitetty taulukossa \ref{synt_hl}. \cite{lakkaraju17}
+
+\begin{table}[H]
\centering
\begin{tabular}{lrrrrrrrrrr}
\hline \hline
@@ -247,12 +245,12 @@ Muuttuja & Keskiarvo & Keskihajonta & Minimi & 25\% & 50\% &
Z & 0.01 & 1.00 & -4.85 & -0.67 & 0.00 & 0.68 & 4.24 \\
W & 0.01 & 1.00 & -4.03 & -0.67 & 0.01 & 0.68 & 4.29 \\
result\_Y & 0.50 & 0.50 & 0.00 & 0.00 & 0.00 & 1.00 & 1.00 \\
- probabilities\_T & 0.50 & 0.28 & -0.34 & 0.28 & 0.50 & 0.72 & 1.30 \\
+% probabilities\_T & 0.50 & 0.28 & -0.34 & 0.28 & 0.50 & 0.72 & 1.30 \\
decision\_T & 0.48 & 0.50 & 0.00 & 0.00 & 0.00 & 1.00 & 1.00 \\
\hline
\end{tabular}
\caption{Synteettisen aineiston muuttujien hajontalukuja}
-\label{table:2}
+\label{synt_hl}
\end{table}
%%%%%%%%%
@@ -261,63 +259,60 @@ Muuttuja & Keskiarvo & Keskihajonta & Minimi & 25\% & 50\% &
\chapter{Menetelmät}\label{metodit}
-Tässä kappaleessa selostan analyyseissa, mallinnuksessa ja validoinnissa käyttämäni menetelmät.
-
-%%%%%%%%%
-
-\section{Aiemmat tutkimukset?}\label{aiemmat}
+Tässä kappaleessa selostan mallin laatimisessa ja arvioinnissa käyttämäni teoreettisen taustan. Koska kausaalinen malli esitetään verkkona, käyn aluksi läpi vaadittavat verkkoteoreettiset määritelmät. Esitän sen jälkeen mallini graafina ja osoitan kausaalisen vaikutuksen olevan identifioituva.
-Aiemmat tutkimukset ovat lähestyneet monesta näkökulmasta, mutta ilman kausaatiota.
-
-%%%%%%%%%
-
-\section{Validointimetodit}\label{validointi}
-
-Tulosten arvioinnissa käytetään visuaalista tarkastelua ja XZY. Laskemme arvioista vapaaksi päässeiden uusijoiden suhteen kaikkiin tuomittuihin, eli niin sanotun virhesuhteen (failure rate).
-
-%%%%%%%%%
+%%%%%%%%%%
+%
+%\section{Aiemmat tutkimukset?}\label{aiemmat}
+%
+%Aiemmat tutkimukset ovat lähestyneet monesta näkökulmasta, mutta ilman kausaatiota.
+%
+%%%%%%%%%%
+%
+%\section{Validointimetodit}\label{validointi}
+%
+%Tulosten arvioinnissa käytetään
+%
+%Tulosten arvioinnissa käytetään visuaalista tarkastelua ja XZY. Laskemme arvioista vapaaksi päässeiden uusijoiden suhteen kaikkiin tuomittuihin, eli niin sanotun virhesuhteen (failure rate).
+%
+%%%%%%%%%%
\section{Verkkoteoria}\label{verkot}
-Kausaalipäättelyn mallit määritellään verkkoina. Esitän tässä kappaleessa lyhyesti kaikki tarvittavat verkkoteoreettiset määritelmät, joita tulen hyödyntämään. Noudatan määritelmissä Oinosta \cite{oinonen16}.
-
+Verkot koostuvat \emph{solmuista} ja \emph{kaarista}, joita voidaan havainnollistaa pisteinä ja viivoina tai nuolina näiden pisteiden väliilä. Kaaret ovat järjestettyjä pareja, kuten verkot itsekin, mutta oletan tavallisimmat joukko-opin merkinnät ja käsitteet tunnetuiksi. Noudatan määritelmissä Oinosta \cite{oinonen16} ja erikseen merkityissä kohdissa Kivistä \cite{tira}.
% TiRan materiaalit??
% Ota esimerkki verkko ja kirjoita siitä lyhyet havainnollistavat kommentit
-\begin{figure}[H]\label{esverkko}
+\begin{figure}[H]
\centering
\includegraphics[scale = 0.5]{full_model}
-\caption{Esimerkkiverkko $H = (V, E)$, missä $V = \{R, X, Z, T, Y\}$.}
+\caption{Eräs verkko $H = (V, E)$, missä $V = \{R, X, Z, T, Y\}$.}
+\label{esverkko}
\end{figure}
\begin{maar}[Suunnattu verkko] \label{suun_verkko}
-\emph{Suunnattu verkko G} on pari $(V, E)$, missä $V \neq \emptyset$ on solmujen joukko ja $$E = \{(a, b) \in V \times V | \text{ solmusta } a \text{ on nuoli solmuun } b \} $$ on \emph{kaarien} joukko.
+\emph{Suunnattu verkko G} on pari $(V, E)$, missä $V \neq \emptyset$ on \emph{solmujen} joukko ja $$E = \{(a, b) \in V \times V | \text{ solmusta } a \text{ on nuoli solmuun } b \} $$ on \emph{kaarien} joukko.
-\end{maar}
+\end{maar}
-\noindent Kuvassa \ref{esverkko} näkyvässä verkossa esimerkiksi $(X, R) \in E$, mutta $(T, Z) \notin E$, koska solmusta $T$ ei ole nuolta solmuun $Z$. Lisäksi voidaan todeta, että kaarien joukkoon kuuluu yhdeksän järjestettyä paria ja solmujen joukko $V$ käsittää viisi alkiota, jotka on lueteltu kuvatekstissä.
+\smallskip
+
+\noindent Kuvassa \ref{esverkko} näkyvässä verkossa esimerkiksi $(X, R) \in E$, mutta $(T, Z) \notin E$, koska solmusta $T$ ei ole nuolta solmuun $Z$. Lisäksi voidaan todeta, että kaarien joukkoon kuuluu yhdeksän järjestettyä paria ja solmujen joukko $V$ käsittää viisi alkiota.
+
+\smallskip
\begin{maar} % Lähtösolmu, maalisolmu, vierussolmu
Oletetaan, että $G=(V, E)$ on suunnattu verkko ja $a, b \in V$. \\
-\noindent Merkintä $a \rightarrow b$ tarkoittaa, että $(a, b) \in E$. Tällöin sanotaan, että $a$ on kaaren $(a, b)$ \emph{lähtösolmu} ja $b$ on kaaren $(a, b)$ \emph{maalisolmu}. Sanotaan myös, että solmu $b$ on solmun $a$ \emph{vierussolmu}. \\
+\noindent Merkintä $a \rightarrow b$ tarkoittaa, että $(a, b) \in E$. Tällöin sanotaan, että $a$ on kaaren $(a, b)$ \emph{lähtösolmu} ja $b$ on kaaren $(a, b)$ \emph{maalisolmu}. Sanotaan myös, että solmu $b$ on solmun $a$ \emph{vierussolmu} tai että solmut $a$ ja $b$ ovat \emph{vierekkäisiä}. \\
\noindent Jos $(a, a) \in E$, sanotaan suunnatussa verkossa olevan \emph{silmukka} solmussa $a$.
\end{maar}
-\noindent Esimerkkiverkossa $H$ kaaren $(Z, T)$ lähtösolmu on solmu $Z$ ja maalisolmu solmu $T$. Lisäksi huomataan, että verkossa $H$ ei ole yhtään silmukkaa.
-
-\begin{maar}[Vierekkäisyys] \label{vierekkaisyys}
-
-Oletetaan, että $G=(V, E)$ on suunnattu verkko ja $a, b \in V$. \\
-
-\noindent Jos solmujen $a$ ja $b$ välillä on nuoli, niin solmujen $a$ ja $b$ sanotaan olevan \emph{vierekkäisiä}.
-\end{maar}
-
-\noindent Kuvan \ref{esverkko} verkosta havaitaan, että melkein kaikki solmut ovat toistensa vierussolmuja. Ainoa poikkeus on solmut $R$ ja $Y$, joiden välillä ei ole nuolta ja jotka eivät siten ole vierekkäisiä.
+\noindent Esimerkkiverkossa $H$ kaaren $(Z, T)$ lähtösolmu on solmu $Z$ ja maalisolmu solmu $T$. Lisäksi huomataan, että verkossa $H$ ei ole yhtään silmukkaa. Kuvan \ref{esverkko} verkosta havaitaan, että melkein kaikki solmut ovat toistensa vierussolmuja. Ainoa poikkeus on solmut $R$ ja $Y$, joiden välillä ei ole nuolta ja jotka eivät siten ole vierekkäisiä.
\begin{maar}[Yksinkertainen suunnattu verkko] \label{yk_suun_verkko}
@@ -330,17 +325,24 @@ Oletetaan, että $G = (V,E)$ on suunnattu verkko, jossa ei ole yhtään silmukka
\begin{maar}[Polku ja suunnattu polku] \label{polku}
-Oletetaan, että $G$ on yksinkertainen verkko ja $n \in \N, n \geq 1$. \\
+Oletetaan, että $G$ on yksinkertainen verkko ja $n \in \N, n \geq 1$.
-\noindent Verkon $G$ solmujen jono $v_1, \ldots, v_n$ on \emph{polku} solmusta $v_1$ solmuun $v_n$, jos jonon jokaisesta solmusta on kaari jonon seuraavaan solmuun. Polkua voidaan merkitä $v_1 \leadsto v_n$. \\
-
-\noindent Jos verkko $G$ on suunnattu verkko, $a, b \in V$ ja kaikki polun $a \leadsto b$ kaaret kulkevat kaarien suuntien mukaisesti, voidaan täsmentää, että polku $a \leadsto b$ on \emph{suunnattu polku}.
+\begin{enumerate}[(a)]
+\item Verkon $G$ solmujen jono $v_1, \ldots, v_n$ on \emph{polku} solmusta $v_1$ solmuun $v_n$, jos jonon jokaisesta solmusta on kaari jonon seuraavaan solmuun. Polkua voidaan merkitä $v_1 \leadsto v_n$.
+\item Jos verkko $G$ on suunnattu verkko, $a, b \in V$ ja kaikki polun $a \leadsto b$ kaaret kulkevat kaarien suuntien mukaisesti, voidaan täsmentää, että polku $a \leadsto b$ on \emph{suunnattu polku}.
+\item Polku on \emph{yksinkertainen}, jos kukin solmu esiintyy polussa vain kerran, paitsi että viimeinen ja ensimmäinen saavat olla sama solmu. \cite{tira}
+\item Yksinkertainen polku on \emph{sykli} eli \emph{kehä}, jos viimeinen ja ensimmäinen solmu ovat samat. \cite{tira} %Suuntaamattomassa verkossa lisäksi vaaditaan, että syklissä pitää olla vähintään kolme solmua.
+\end{enumerate}
\end{maar}
-\noindent Huomataan, että esimerkkinä käytetyssä verkossa $H$ on useita polkuja solmusta $R$ solmuun $Y$. Polku $R \rightarrow T \rightarrow Y$ on suunnattu polku ja $R \leftarrow X \rightarrow Y$ on tavallinen polku, sillä solmujen $R$ ja $X$ välillä kuljetaan nuolen suunnan vastaisesti.
+\smallskip
+
+\noindent Huomataan, että verkossa $H$ on useita polkuja solmusta $R$ solmuun $Y$. Polku $R \rightarrow T \rightarrow Y$ on ainut suunnattu polku ja $R \leftarrow X \rightarrow Y$ on tavallinen polku, sillä solmujen $R$ ja $X$ välillä kuljetaan nuolen suunnan vastaisesti. Verkossa ei ole yhtään sykliä eli se on \emph{syklitön}. Suunnatuista ja syklittömistä verkoista voidaan käyttää englannin kielestä juontuvaa lyhennettä DAG \emph{(directed acyclic graph)} \cite{tira}.
-\begin{maar} \label{sukulaisuus}
+\smallskip
+
+\begin{maar}[Jälkeläisyys] \label{sukulaisuus}
Oletetaan, että $G=(V, E)$ on suunnattu verkko ja $a, b \in V$. \\
@@ -349,114 +351,168 @@ Oletetaan, että $G=(V, E)$ on suunnattu verkko ja $a, b \in V$. \\
\noindent Esimerkiksi kuvan \ref{esverkko} verkossa solmulla $Y$ ei ole jälkeläisiä ja solmun $Z$ jälkeläiset ovat kaikki muut verkon solmut poislukien se itse, eli solmun $Z$ jälkeläiset on joukko $V \setminus \{Z\}$.
+Kausaalipäättelyssä kausaalisten vaikutusten identifiomiseksi tarvitaan usein selvittää niin sanotut \emph{haarukka-} ja \emph{käänteiset haarukkasolmut}. Määritellään ne seuraavaksi.
+
+\begin{maar}[Haarukkasolmu] \label{haarukka}
+
+Oletetaan, että suunnatussa verkossa on polku $A \leftarrow B \rightarrow C \leftarrow D$. Tällöin solmua B sanotaan \emph{haarukkasolmuksi} ja solmua C \emph{käänteiseksi haarukkasolmuksi}.
+
+\end{maar}
+
%%%%%%%%%
\section{Kausaalipäättely}\label{kausaali}
-Erityisesti \cite{pearl10}. Esittele merkunnät, määritelmät ja mallli. Käännökset Miksi-kirjaa mukaillen?
+Judea Pearl esittää artikkelissaan \cite{pearl10}, että kaikessa tutkimuksessa, joka hyödyntää kausaalipäättelyä, tulisi edetä järjestelmällisesti neljässä vaiheessa:
+
+\begin{enumerate}
-\subsection{Johdanto?}\label{kausaalijohd}
+\item Määrittely: Määritetään tavoitesuuruus Q funktiona Q($\M$), joka voidaan laskea kaikille malleille $\M$.
+\item Oletuksien esitys: Esitä kausaaliset oletukset luonnollisella kielellä ja ilmaise niiden rakenteellinen osa verkkona.
+\item Identifioituvuus: Osoita, onko tavoitesuuruus määritettävissä (ilmaistavissa estimoitavina parametreina).
+\item Estimointi: Estimoi tavoitesuuruutta, jos se on identifioituva tai approksimoi sitä jos se ei ole. Tarkista mallin mahdolliset (tilastolliset) oletukset ja implikaatiot ja muuta mallia, jos oletukset osoittautuvat paikkaansa pitämättömiksi.
-Kausaalipäättelyssä mallit määritellään usein yksinkertaisina suunnattuina verkkoina. Mallin määrittämästä verkosta voidaan suoraan lukea kausaaliset riippuvuussuhteet ja malliin kuuluvat muuttujat. Jos mallissa on solmut $A$ ja $B$ ja jos solmu $B$ on solmun $A$ jälkeläinen, niin muuttujalla $A$ on mallin mukaan jonkinlainen kausaalinen vaikutus muuttujaan $B$. Jos verkossa muuttujien välillä ei ole jälkeläisyyssuhdetta, niin ne ovat toisistaan riipumattomat. Kausalisen vaikutuksen funktionaalista muotoa ei usein määritellä.
+\end{enumerate}
+
+\noindent Tutkielmani tavoitteena on esittää algoritmi, jolla voimme paremmin ennustaa riskiä populaatiotasolla, kun muutamme myönteisten päätösten osuutta jakun käytössä on valintaharhasta kärsivää aineistoa. Todennäköisyyslausekkein ilmaistuna haluamme siis selvittää vapautusprosentin muutoksen vaikutusta epätoivottavan tapahtuman $Y=0$ todennäköisyyteen, mikä voidaan kirjoittaa muotoon
+\begin{equation} \label{q_m}
+\pr(Y=0 | \text{do}(R=r)).
+\end{equation}
-* Usein funktionaalista muotoa ei määritellä,, lisää tähän ne nuoliversiot yhtälöistä havainnollistamaan, että siirrytään yhtäsuuruudesta määräytymiseen \cite{kalisch14}
+\noindent Huomataan, että lauseke \ref{q_m} ei riipu mistään mallista $\M$, joten se täyttää Pearlin tavoitesuuruuden Q määritelmän mukaiset ehdot.
-\subsection{Merkinnät}\label{kausaalimerk}
+Kausaalipäättelyssä mallit määritellään usein yksinkertaisina suunnattuina verkkoina. Mallin määrittämästä verkosta voidaan suoraan lukea kausaaliset riippuvuussuhteet ja malliin kuuluvat muuttujat. Jos mallissa on solmut $A$ ja $B$ ja jos solmu $B$ on solmun $A$ jälkeläinen, niin muuttujalla $A$ on mallin mukaan jonkinlainen kausaalinen vaikutus muuttujaan $B$. Jos verkossa muuttujien välillä ei ole jälkeläisyyssuhdetta, niin ne ovat toisistaan riipumattomat. Kausaalisen vaikutuksen funktionaalista muotoa ei usein määritellä.
-Kausaalipäättelyssä käytettävät merkinnät noudattelevat pitkälle tavallisia todennäköisyyslaskennan merkintöjä. Kun selvitetään muuttujan $X$ vaikutusta muuttujaan $Y$ ja tehdään interventio asettamalla muuttuja $X$ arvoon $x_0$, sitä merkitään $\pr(Y| \text{do} (X=x_0))$.
+\subsection{Merkinnät ja keskeiset lauseet}\label{kausaalimerk_laus}
-\subsection{Määritelmät}\label{kausaalimäär}
+Kausaalipäättelyssä käytettävät merkinnät noudattelevat pitkälle tavallisia todennäköisyyslaskennan merkintöjä. Kun selvitetään muuttujan $X$ vaikutusta muuttujaan $Y$ ja tehdään interventio asettamalla muuttuja $X$ arvoon $x_0$, sitä merkitään $\pr(Y| \text{do} (X=x_0))$.
-\begin{maar}[Takaovikriteeri, \emph{back-door criterion}]\label{d_sep}
+Käydään seuraavaksi läpi kausaalilaskennan kannalta keskeisimmät lauseet. Lauseiden todistukset sivuutetaan, mutta ne on löydettävissä Pearlin artikkelin lähteistä \cite{pearl10}. Määritelmät \ref{d_sep} ja \ref{takaovi} \textbf{JNE}.
-Joukko $\s$ sulkee / katkaisee (blocks) polun $p$, jos vähintään toinen seuraavista ehdoista on voimassa:
+\begin{maar}[d-separoituvuus \cite{pearl10}]\label{d_sep}
+
+Joukko $\s$ katkaisee (blocks) polun $p$, jos vähintään toinen seuraavista ehdoista on voimassa:
\begin{enumerate}[(a)]
-\item Polku $p$ sisältää vähintään yhden solmun, joka on jonkin kaaren lähtösolmu ja kuuluu joukkoon $\s$. (arrow-emitting)
+\item Polku $p$ sisältää vähintään yhden solmun, joka on jonkin polun kulkusuuntaisen kaaren lähtösolmu ja kuuluu joukkoon $\s$. (arrow-emitting)
\item Polku $p$ sisältää vähintään yhden käänteisen haarukkasolmun (collision node), joka ei kuulu joukkoon $\s$ ja jolla ei ole jälkeläisiä joukossa $\s$.
\end{enumerate}
+\noindent Jos joukko $\s$ katkaisee kaikki polut muuttujasta $X$ muuttujaan $Y$, sanotaan joukon $\s$ d-separoivan muuttujat $X$ ja $Y$. Tällöin $X$ ja $Y$ ovat riippumattomia ehdolla $\s$, eli $X \independent Y | \s$.
+
\end{maar}
-\begin{maar}\label{adjustment}
-Oletetaan, että halutaan selvittää (satunnais)muuttujan X kausaalista vaikutusta muuttujaan Y. Joukko $\s$ on \emph{riittävä} tasoitukseen (adjustment), kun seuraavat ehdot ovat voimassa: \textbf{sufficifient to adjusment = identifioituva?}
+
+\begin{maar}[Takaovikriteeri (\emph{back-door criterion}) \cite{pearl10}] \label{takaovi}
+
+Oletetaan, että halutaan selvittää muuttujan X kausaalista vaikutusta muuttujaan Y. Joukko $\s$ on \emph{riittävä} vaikutuksen selvittämiseen (sufficient for adjustment), kun seuraavat ehdot ovat voimassa:
\begin{enumerate}[(1)]
\item Yksikään joukon $\s$ alkioista ei ole solmun X jälkeläinen.
-\item Joukon $\s$ alkiot katkaisevat kaikki märitelmän \ref{d_sep} mukaiset polut / ''takaovireitit'' solmusta X solmuun Y.
+\item Joukon $\s$ alkiot katkaisevat kaikki määritelmän \ref{d_sep} mukaiset polut solmusta X solmuun Y.
\end{enumerate}
\end{maar}
+
+
\subsection{Malli}\label{kausaalimalli}
-Mallimme esitellään alla. Mallissamm
+Malli sisältää viisi muuttujaa, jotka on esitelty lyhyesti taulukossa \ref{syntmjat}. Muuttujalla $R$ kuvataan päätöksentekijän hyväksymisprosenttia, eli sitä prosentuaalista osuutta henkilöistä, joilla on pienin vaara epätoivottavaan tulokseen ja joille siten voidaan antaa myönteinen päätös. $X$ ilmentää henkilön henkilökohtaisia ominaisuuksia, jotka ovat sekä päätöksentekijän että mallin havaittavissa. Muuttuja $X$ voi olla esimerkiksi jonkinlainen rekisteritieto, kuten ikä tai sukupuoli. Muuttuja $Z$ on muuttuja, jonka tuomari tai muu asiantuntija voi havaita, mutta joka on mallilta piilotettu. Muuttujan $Z$ voidaan ajatella esimerkiksi oikeuskäsittelyjen tapauksessa kuvaavan epäillyn kääytöstä oikeussalissa. Tulosmuuttuja $Y$ ja päätösmuuttuja $T$ ovat kaksiarvoisia ja niiden määrittelyt on esitelty kuvassa \ref{valikoitumisharha}: myönteistä päätöstä merkitään $t=1$, kielteistä $t=0$. Vastaavasti myönteinen tulos määritellään muuttujan $y$ arvoksi 1, kielteinen arvoksi 0.
+Mallin määrittelevä graafi on estetty kuviossa \ref{final_model} ilman virhemuuttujia. Graafista voidaan suoraan lukea oletukset: oletetaan, että $Z \independent X, R$ mutta laajennetaan Lakkarajun oletuksia sallimalla muuttujan X vaikutus muuttujaan R \cite{lakkaraju17}. Mallin oletetuilla kausaalisilla vaikutuksilla on lisäksi selkeästi ilmaistavat realisaatiot: kuinka osuuden $R$ muuttaminen vaikuttaa päätökseen ja edelleen päätös tulokseen ja niin edelleen.
-\begin{table}[h!]
+\begin{table} %[H]
\centering
\begin{tabular}{rl}
\hline \hline
Muuttuja & Kuvaus \\
\hline
- R & Vapautusprosentti, vapautumiskynnys \\
- X & Henkilökohtaiset muuttujat, kirjalliset \\
- Z & Henkilökohtaiset muuttujat, tuomarin havaiitsemat\\
- W & Henkilökohtaiset muuttujat, havaitsemattomat, \emph{kohtalo}\\
- Y & Uusinta, $Y=0$ uusi, 1 niin ei uusinut\\
- T & 0 on jail, 1 on bail\\
+ R & Myönteisten päätösten osuus prosentteina $r \in [0, 1]$ \\
+ X & Kirjatut muuttujat, havaittavissa kaikille \\
+ Z & Kirjaamattomat muuttujat, vain päättäjän havaitsemat\\
+ Y & Tulosmuuttuja, $y \in \{0, 1\}$\\
+ T & Päätösmuuttuja, $t \in \{0, 1\}$\\
\hline \hline
\end{tabular}
-\caption{Mallin muuttjienn selitteet}
+\caption{Mallin muuttujien selitteet}
\label{syntmjat}
\end{table}
-\begin{figure}[H]
+\begin{figure}% [H]
\centering
\begin{subfigure}[b]{0.4\textwidth}
\includegraphics[width=\textwidth]{final_model}
- \caption{lopullinen malli}
+ \caption{Malli ilman interventiota.}
\label{final_model}
- \end{subfigure}
+ \end{subfigure}
~ %add desired spacing between images, e. g. ~, \quad, \qquad, \hfill etc.
%(or a blank line to force the subfigure onto a new line)
\begin{subfigure}[b]{0.5\textwidth}
\includegraphics[width=\textwidth]{intervention_model}
- \caption{interventio}
+ \caption{Malli, johon interventio on merkitty.}
\label{intervention_model}
\end{subfigure}
~ %add desired spacing between images, e. g. ~, \quad, \qquad, \hfill etc.
%(or a blank line to force the subfigure onto a new line)
- \caption{Kasuaalimallit graafina}\label{mallikuvat}
+ \caption{Kausaalimallit graafeina.}\label{mallikuvat}
\end{figure}
+Johdetaan muuttujan $R$ kausaalivaikutus muuttujaan $Y$ yli kaikkien ositteiden X. Huomataan, että osuuden $R$ kausaalinen vaikutus voidaan ilmaista suoraan lausekkeella \ref{q_m}, sillä $\pr(Y=0|\text{do}(R=0))=0$ ja siten edelleen
+\begin{equation*}
+ \pr(Y=0|\text{do}(R=r))-\pr(Y=0|\text{do}(R=0)) \\
+% =\: \pr(Y=0|\text{do}(R=r))-0 \\
+ =\: \pr(Y=0|\text{do}(R=r)).
+\end{equation*}
+
+Osoitetaan seuraavaksi, että X on riittävä vaikutusten korjaamiseen määritelmän \ref{takaovi} mukaisesti, kun selvitetään muuttujan R kausaalista vaikutusta muuttujaan Y. Mallista voidaan suoraan lukea, että takaovikriteerin ensimmäinen ehto on voimassa: X ei ole muuttujan R jälkeläinen. Polut, jotka muuttujan X pitää katkaista ollakseen riittävä vaikutusten korjaamiseen ovat $R \leftarrow X \rightarrow Y$, $R \leftarrow X \rightarrow T \rightarrow Y$ ja $R \leftarrow X \rightarrow T \leftarrow Z \rightarrow Y$. Muuttuja X täyttää kuitenkin määritelmän \ref{d_sep} (a)-kohdan ehdon ja siten d-separoi muuttujat R ja Y. Tällöin X on riittävä vaikutusten korjaamiseen ja voidaan hyödyntää Pearlin kaavaa 25 \cite{pearl10}:
+\begin{subequations} \label{derivation}
+\begin{align}
+ \pr&(Y=0|\text{do}(R=r)) = \sum_x \pr(Y=0| R=r, X=x) \pr(X=x) \label{derivation1} \\
+ &= \sum_x \left( \sum_t \pr(Y=0, T=t| R=r, X=x) \right) \pr(X=x) \label{derivation2} \\
+ &= \sum_x \left( \sum_t \pr(Y=0| T=t, R=r, X=x)\pr(T=t| R=r, X=x) \right) \pr(X=x) \label{derivation3} \\
+ &= \sum_x \pr(Y=0| T=1, R=r, X=x) \pr(T=1| R=r, X=x) \pr(X=x) \label{derivation4} \\
+ &= \sum_x \pr(Y=0| T=1, X=x) \pr(T=1| R=r, X=x) \pr(X=x) \label{derivation5}
+\end{align}
+\end{subequations}
-\begin{algorithm} % enter the algorithm environment
-\caption{Kausaalialgoritmi} % give the algorithm a caption
-\label{causal_alg} % and a label for \ref{} commands later in the document
-\begin{algorithmic}[1] % enter the algorithmic environment
-\REQUIRE aineisto $(\mathbf{x}, t, y) \in \D_t, \D_v$ ja hyväksymisaste $r \in [0, 1]$, missä $\D_t$ on testiaineisto ja $\D_v$ validointiaineisto.
-\ENSURE $\pr(Y=0|\text{do}(R=r))$
-
-\STATE Määritä $f(x) = \pr(X=x)$ testiaineistosta.
-\STATE Ennusta vastetta $Y$ selittävillä muuttujilla $X$ käyttäen harjoitusaineiston havaintoja, joilla $T=1$.
-\STATE Määritä harjoitusaineiston jokaiselle havainnolle $P(Y=0|X=x)$ käyttäen yllä olevaa mallia.
-\STATE Järjestä havainnot nousevaan järjestykeen edellisen kohdan todennäköisyyksien mukaan.
-\STATE Alusta muuttuja \texttt{summa} = 0.
-\FORALL{Jokaiselle parametriavaruuden pisteelle}
- \STATE $p_x \leftarrow P(X=x)$
- \STATE $\mathcal{D_x} \leftarrow \{\mathcal{D} | X = x\}$
- \STATE Assign first $r\cdot 100\%$ observations from $\mathcal{D_x}$ to $\mathcal{D}_{rx}$
- \STATE $p_t \leftarrow \dfrac{|\{\mathcal{D}_{rx}|T=1\}|}{|\mathcal{D}_{rx}|}$
- \STATE $\mathcal{D}_{tx} \leftarrow \{\mathcal{D}_x | T = 1\}$
- \STATE $p_y \leftarrow \dfrac{|\{\mathcal{D}_{tx}|Y=0\}|}{|\mathcal{D}_{tx}|}$
- \STATE Lisää muuttujaan \texttt{summa} tulo $p_y \cdot p_t \cdot p_x$
-\ENDFOR
-\RETURN \texttt{summa}
-\end{algorithmic}
-\end{algorithm}
+Yllä oleva lauseke on yhtäpitävä myös jatkuville muuttujan $x$ arvoille, kun korvaamme summaukset integraalilla parametriavaruuden yli: $$\pr(Y=0|\text{do}(R=r)) = \int_x \pr(Y=0| T=1, X=x) \pr(T=1| R=r, X=x) \pr(X=x).$$
+
+\subsection{algo}
+
+
+Pearlin mukaan:
+
+$$P(Y=0|do(R=r), X=x)=P(Y=0|R=r, X=x)=P(Y=0|R=r, X=x, T=1)P(T=1|R=r, X=x)$$
+
+Mallit vaikutukset laskettiin Pythonilla versio 3.6. Syötteett sklinear mallliin , joka fitattiin testi dataan ja sitten integroitiin eri leniencyn tasoilla muuttujan X parametriavaruuden eli reaaliakselin ylitse.
+
+%\begin{algorithm} % enter the algorithm environment
+%\caption{Kausaalialgoritmi} % give the algorithm a caption
+%\label{causal_alg} % and a label for \ref{} commands later in the document
+%\begin{algorithmic}[1] % enter the algorithmic environment
+%\REQUIRE aineisto $(\mathbf{x}, t, y) \in \D_t, \D_v$ ja hyväksymisaste $r \in [0, 1]$, missä $\D_t$ on testiaineisto ja $\D_v$ validointiaineisto.
+%\ENSURE $\pr(Y=0|\text{do}(R=r))$
+%
+%\STATE Määritä $f(x) = \pr(X=x)$ testiaineistosta.
+%\STATE Ennusta vastetta $Y$ selittävillä muuttujilla $X$ käyttäen harjoitusaineiston havaintoja, joilla $T=1$.
+%\STATE Määritä harjoitusaineiston jokaiselle havainnolle $P(Y=0|X=x)$ käyttäen yllä olevaa mallia.
+%\STATE Järjestä havainnot nousevaan järjestykeen edellisen kohdan todennäköisyyksien mukaan.
+%\STATE Alusta muuttuja \texttt{summa} = 0.
+%\FORALL{Jokaiselle parametriavaruuden pisteelle}
+% \STATE $p_x \leftarrow P(X=x)$
+% \STATE $\mathcal{D_x} \leftarrow \{\mathcal{D} | X = x\}$
+% \STATE Assign first $r\cdot 100\%$ observations from $\mathcal{D_x}$ to $\mathcal{D}_{rx}$
+% \STATE $p_t \leftarrow \dfrac{|\{\mathcal{D}_{rx}|T=1\}|}{|\mathcal{D}_{rx}|}$
+% \STATE $\mathcal{D}_{tx} \leftarrow \{\mathcal{D}_x | T = 1\}$
+% \STATE $p_y \leftarrow \dfrac{|\{\mathcal{D}_{tx}|Y=0\}|}{|\mathcal{D}_{tx}|}$
+% \STATE Lisää muuttujaan \texttt{summa} tulo $p_y \cdot p_t \cdot p_x$
+%\ENDFOR
+%\RETURN \texttt{summa}
+%\end{algorithmic}
+%\end{algorithm}
%%%%%%%%%
%%%%%%%%%
@@ -464,15 +520,20 @@ Muuttuja & Kuvaus \\
\chapter{Tulokset}\label{tulokset}
-%%%%%%%%%
+- se pääkuvaaja vertailuineen
-\section{Synteettinen}\label{synttulokset}
+- beta ztan vaikutus?
-%%%%%%%%%
-
-\section{Compas}\label{compastulokset}
+- erilaiset mallit ja koko käyrä aina 1 asti -> kuinka meillä parempi
+- voidaanko antaa estimaateille mitään luottusvälejä tjsp?
+\begin{figure}[H]
+\centering
+\includegraphics[width = \textwidth]{tulos_kuva_placeholder_en}
+\caption{Tulokset kuvana}
+\label{tuloskuva}
+\end{figure}
%%%%%%%%%
%%%%%%%%%
@@ -480,20 +541,26 @@ Muuttuja & Kuvaus \\
\chapter{Diskussio}\label{diskussio}
+- Jatkosuunnitelmat: tutkitaan beta zetan vaikutusta tuloksiin, kuinka hyvin estimoituu. Sovelletaan oikeaan data settiiin. Mielenkiintoiseksi on osoittautunut propublica julkaisun artikkelissa machine bias käyttämä COMPAS-aineisto.
+
+- Ongelmat / muut huomiot: Tällä aikataululla en ole tehnyt mallin validointeja: onko kausaaliset pathwayt reasonable. Malli itsessään on suhteellisen yksinkertainen joten (KÄSIENHEILUTTELU) on jokseenkin luultavaa, että sinällään mallin spesifionnissa tuskin on mitään virheitä. Voitaisiin ehkä tietenkin koostaa jokseenkin hienosyisempi malli (erilaiset rikoshistoria yms erikseeen) ja jotain. Jvat muuttujat? P-uloitteinen parametriavaruus???
+- Mallin validointi epäeettistä, koska vaatisi huonoja päätöksiä > meillä kyllä synteettinen?
-\begin{verbatim}
-# R-koodi, tulos sama
-library(igraph)
-library(causaleffect)
-# simplify = FALSE to allow multiple edges
-g <- graph.formula(X -+ R, X -+ D, X -+ Y, R -+ D , D -+ Y, D -+ Y, Y -+ D, simplify = FALSE)
-# Here the bidirected edge between X and Z is set to be unobserved in graph g
-# This is denoted by giving them a description attribute with the value "U"# The edges in question are the fourth and the fifth edge
-g <- set.edge.attribute(graph = g, name = "description", index = c(6,7), value = "U")
+- Implikaatiot: parempia malleja???
-res <- causal.effect("Y", "R", G = g)
-\end{verbatim}
+%\begin{verbatim}
+%# R-koodi, tulos sama
+%library(igraph)
+%library(causaleffect)
+%# simplify = FALSE to allow multiple edges
+%g <- graph.formula(X -+ R, X -+ D, X -+ Y, R -+ D , D -+ Y, D -+ Y, Y -+ D, simplify = FALSE)
+%# Here the bidirected edge between X and Z is set to be unobserved in graph g
+%# This is denoted by giving them a description attribute with the value "U"# The edges in question are the fourth and the fifth edge
+%g <- set.edge.attribute(graph = g, name = "description", index = c(6,7), value = "U")
+%
+%res <- causal.effect("Y", "R", G = g)
+%\end{verbatim}
%%%%%%%%%
diff --git a/analysis_and_scripts/Analysis_07MAY2019_new.ipynb b/analysis_and_scripts/Analysis_07MAY2019_new.ipynb
new file mode 100644
index 0000000000000000000000000000000000000000..aabbd42e4084f328c8bf084ed8a4a4918d52d1b0
--- /dev/null
+++ b/analysis_and_scripts/Analysis_07MAY2019_new.ipynb
@@ -0,0 +1,645 @@
+{
+ "cells": [
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "toc": true
+ },
+ "source": [
+ "<h1>Table of Contents<span class=\"tocSkip\"></span></h1>\n",
+ "<div class=\"toc\"><ul class=\"toc-item\"><li><span><a href=\"#Causal-model\" data-toc-modified-id=\"Causal-model-1\"><span class=\"toc-item-num\">1 </span>Causal model</a></span><ul class=\"toc-item\"><li><span><a href=\"#Notes\" data-toc-modified-id=\"Notes-1.1\"><span class=\"toc-item-num\">1.1 </span>Notes</a></span></li></ul></li><li><span><a href=\"#Synthetic-data\" data-toc-modified-id=\"Synthetic-data-2\"><span class=\"toc-item-num\">2 </span>Synthetic data</a></span></li><li><span><a href=\"#Algorithms\" data-toc-modified-id=\"Algorithms-3\"><span class=\"toc-item-num\">3 </span>Algorithms</a></span><ul class=\"toc-item\"><li><span><a href=\"#Contraction-algorithm\" data-toc-modified-id=\"Contraction-algorithm-3.1\"><span class=\"toc-item-num\">3.1 </span>Contraction algorithm</a></span></li><li><span><a href=\"#Causal-algorithm\" data-toc-modified-id=\"Causal-algorithm-3.2\"><span class=\"toc-item-num\">3.2 </span>Causal algorithm</a></span></li></ul></li><li><span><a href=\"#Performance-comparison\" data-toc-modified-id=\"Performance-comparison-4\"><span class=\"toc-item-num\">4 </span>Performance comparison</a></span><ul class=\"toc-item\"><li><span><a href=\"#Predictive-models\" data-toc-modified-id=\"Predictive-models-4.1\"><span class=\"toc-item-num\">4.1 </span>Predictive models</a></span></li><li><span><a href=\"#Visual-comparison\" data-toc-modified-id=\"Visual-comparison-4.2\"><span class=\"toc-item-num\">4.2 </span>Visual comparison</a></span></li></ul></li></ul></div>"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "## Causal model\n",
+ "\n",
+ "Our model is defined by the probabilistic expression \n",
+ "\n",
+ "\\begin{equation}\\label{model_disc}\n",
+ "P(Y=0 | \\text{do}(R=r)) = \\sum_x \\underbrace{P(Y=0|X=x, T=1)}_\\text{1} \n",
+ "\\overbrace{P(T=1|R=r, X=x)}^\\text{2} \n",
+ "\\underbrace{P(X=x)}_\\text{3}\n",
+ "\\end{equation}\n",
+ "\n",
+ "which is equal to \n",
+ "\n",
+ "\\begin{equation}\\label{model_cont}\n",
+ "P(Y=0 | \\text{do}(R=r)) = \\int_x P(Y=0|X=x, T=1)P(T=1|R=r, X=x)P(X=x)\n",
+ "\\end{equation}\n",
+ "\n",
+ "for continuous $x$. Model as a graph (Z is a latent variable, and can be excluded from the expression with do-calculus by showing that $X$ is admissible for adjustment):\n",
+ "\n",
+ "<!--- ![Model as picture](../figures/intervention_model.png \"Intervention model\") --->\n",
+ "\n",
+ "For predicting the probability of negative outcome the following should hold because by Pearl $P(Y=0 | \\text{do}(R=r), X=x) = P(Y=0 | R=r, X=x)$ when $X$ is an admissible set:\n",
+ "\n",
+ "\\begin{equation} \\label{model_pred}\n",
+ "P(Y=0 | \\text{do}(R=r), X=x) = P(Y=0|X=x, T=1)P(T=1|R=r, X=x).\n",
+ "\\end{equation}\n",
+ "\n",
+ "Still it should be noted that this prediction takes into account the probability of the individual to be given a positive decision ($T=1$), see second term in \\ref{model_pred}.\n",
+ "\n",
+ "----\n",
+ "\n",
+ "### Notes\n",
+ "\n",
+ "* Equations \\ref{model_disc} and \\ref{model_cont} describe the whole causal effect in the population (the causal effect of changing $r$ over all strata $X$).\n",
+ "* Prediction should be possible with \\ref{model_pred}. Both terms can be learned from the data. NB: the probability $P(Y=0 | \\text{do}(R=r), X=x)$ is lowest when the individual $x$ is the most dangerous or the least dangerous. How could we infer/predict the counterfactual \"what is the probability of $Y=0$ if we were to let this individual go?\" has yet to be calculated.\n",
+ "* Is the effect of R learned/estimated correctly if it is just plugged in to a predictive model (e.g. logistic regression)?\n",
+ "* $P(Y=0 | do(R=0)) = 0$ only in this application. My predictive models say that when $r=0$ the probability $P(Y=0) \\approx 0.027$ which would be a natural estimate in another application/scenario (e.g. in medicine the probability of an adverse event when a stronger medicine is distributed to everyone. Then the probability will be close to zero but not exactly zero.)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 89,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "# Imports\n",
+ "\n",
+ "import numpy as np\n",
+ "import pandas as pd\n",
+ "from datetime import datetime\n",
+ "import matplotlib.pyplot as plt\n",
+ "import scipy.stats as scs\n",
+ "import scipy.integrate as si\n",
+ "import seaborn as sns\n",
+ "import numpy.random as npr\n",
+ "from sklearn.preprocessing import OneHotEncoder\n",
+ "from sklearn.linear_model import LogisticRegression\n",
+ "from sklearn.ensemble import RandomForestClassifier\n",
+ "\n",
+ "# Settings\n",
+ "\n",
+ "%matplotlib inline\n",
+ "\n",
+ "plt.rcParams.update({'font.size': 16})\n",
+ "plt.rcParams.update({'figure.figsize': (14, 7)})\n",
+ "\n",
+ "# Suppress deprecation warnings.\n",
+ "\n",
+ "import warnings\n",
+ "\n",
+ "def fxn():\n",
+ " warnings.warn(\"deprecated\", DeprecationWarning)\n",
+ "\n",
+ "with warnings.catch_warnings():\n",
+ " warnings.simplefilter(\"ignore\")\n",
+ " fxn()"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "## Synthetic data\n",
+ "\n",
+ "In the chunk below, we generate the synthetic data as described by Lakkaraju et al. The default values and definitions of $Y$ and $T$ values follow their description.\n",
+ "\n",
+ "**Parameters**\n",
+ "\n",
+ "* M = `nJudges_M`, number of judges\n",
+ "* N = `nSubjects_N`, number of subjects assigned to each judge\n",
+ "* betas $\\beta_i$ = `beta_i`, where $i \\in \\{X, Z, W\\}$ are coefficients for the respected variables\n",
+ "\n",
+ "**Columns of the data:**\n",
+ "\n",
+ "* `judgeID_J` = judge IDs as running numbering from 0 to `nJudges_M - 1`\n",
+ "* R = `acceptanceRate_R`, acceptance rates\n",
+ "* X = `X`, invidual's features observable to all (models and judges)\n",
+ "* Z = `Z`, information observable for judges only\n",
+ "* W = `W`, unobservable / inaccessible information\n",
+ "* T = `decision_T`, bail-or-jail decisions where $T=0$ represents jail decision and $T=1$ bail decision.\n",
+ "* Y = `result_Y`, result variable, if $Y=0$ person will or would recidivate and if $Y=1$ person will or would not commit a crime."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 90,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "# Set seed for reproducibility\n",
+ "#npr.seed(0)\n",
+ "\n",
+ "def generateData(nJudges_M=100,\n",
+ " nSubjects_N=500,\n",
+ " beta_X=1.0,\n",
+ " beta_Z=1.0,\n",
+ " beta_W=0.2):\n",
+ "\n",
+ " # Assign judge IDs as running numbering from 0 to nJudges_M - 1\n",
+ " judgeID_J = np.repeat(np.arange(0, nJudges_M, dtype=np.int32), nSubjects_N)\n",
+ "\n",
+ " # Sample acceptance rates uniformly from a closed interval\n",
+ " # from 0.1 to 0.9 and round to tenth decimal place.\n",
+ " acceptance_rates = np.round(npr.uniform(.1, .9, nJudges_M), 10)\n",
+ "\n",
+ " # Replicate the rates so they can be attached to the corresponding judge ID.\n",
+ " acceptanceRate_R = np.repeat(acceptance_rates, nSubjects_N)\n",
+ "\n",
+ " # Sample the variables from standard Gaussian distributions.\n",
+ " X = npr.normal(size=nJudges_M * nSubjects_N)\n",
+ " Z = npr.normal(size=nJudges_M * nSubjects_N)\n",
+ " W = npr.normal(size=nJudges_M * nSubjects_N)\n",
+ "\n",
+ " probabilities_Y = 1 / (1 + np.exp(-(beta_X * X + beta_Z * Z + beta_W * W)))\n",
+ "\n",
+ " # 0 if P(Y = 0| X = x; Z = z; W = w) >= 0.5 , 1 otherwise\n",
+ " result_Y = 1 - probabilities_Y.round()\n",
+ " \n",
+ " # For the conditional probabilities of T we add noise ~ N(0, 0.1)\n",
+ " probabilities_T = 1 / (1 + np.exp(-(beta_X * X + beta_Z * Z)))\n",
+ " probabilities_T += npr.normal(0, np.sqrt(0.1), nJudges_M * nSubjects_N)\n",
+ "\n",
+ " # Initialize decision values as 1\n",
+ " decision_T = np.ones(nJudges_M * nSubjects_N)\n",
+ "\n",
+ " # Initialize the dataframe\n",
+ " df_init = pd.DataFrame(np.column_stack(\n",
+ " (judgeID_J, acceptanceRate_R, X, Z, W, result_Y, probabilities_T,\n",
+ " decision_T)),\n",
+ " columns=[\n",
+ " \"judgeID_J\", \"acceptanceRate_R\", \"X\", \"Z\", \"W\",\n",
+ " \"result_Y\", \"probabilities_T\", \"decision_T\"\n",
+ " ])\n",
+ "\n",
+ " # Sort by judges then probabilities\n",
+ " data = df_init.sort_values(by=[\"judgeID_J\", \"probabilities_T\"],\n",
+ " ascending=False)\n",
+ "\n",
+ " # Iterate over the data. Subject is in the top (1-r)*100% if\n",
+ " # his within-judge-index is over acceptance threshold times\n",
+ " # the number of subjects assigned to each judge. If subject\n",
+ " # is over the limit they are assigned a zero, else one.\n",
+ " data.reset_index(drop=True, inplace=True)\n",
+ "\n",
+ " data['decision_T'] = np.where(\n",
+ " (data.index.values % nSubjects_N) <\n",
+ " ((1 - data['acceptanceRate_R']) * nSubjects_N), 0, 1)\n",
+ "\n",
+ " return data\n",
+ "\n",
+ "\n",
+ "df = generateData()"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 91,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "(25000, 8)\n",
+ "(25000, 8)\n",
+ "(25000, 8)\n",
+ "(25000, 8)\n"
+ ]
+ },
+ {
+ "data": {
+ "text/html": [
+ "<div>\n",
+ "<style scoped>\n",
+ " .dataframe tbody tr th:only-of-type {\n",
+ " vertical-align: middle;\n",
+ " }\n",
+ "\n",
+ " .dataframe tbody tr th {\n",
+ " vertical-align: top;\n",
+ " }\n",
+ "\n",
+ " .dataframe thead th {\n",
+ " text-align: right;\n",
+ " }\n",
+ "</style>\n",
+ "<table border=\"1\" class=\"dataframe\">\n",
+ " <thead>\n",
+ " <tr style=\"text-align: right;\">\n",
+ " <th>decision_T</th>\n",
+ " <th>1</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>result_Y</th>\n",
+ " <th></th>\n",
+ " </tr>\n",
+ " </thead>\n",
+ " <tbody>\n",
+ " <tr>\n",
+ " <th>0.0</th>\n",
+ " <td>3911</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>1.0</th>\n",
+ " <td>8759</td>\n",
+ " </tr>\n",
+ " </tbody>\n",
+ "</table>\n",
+ "</div>"
+ ],
+ "text/plain": [
+ "decision_T 1\n",
+ "result_Y \n",
+ "0.0 3911\n",
+ "1.0 8759"
+ ]
+ },
+ "execution_count": 91,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "# Split the data set to test and train\n",
+ "from sklearn.model_selection import train_test_split\n",
+ "train, test = train_test_split(df, test_size=0.5, random_state=0)\n",
+ "\n",
+ "print(train.shape)\n",
+ "print(test.shape)\n",
+ "\n",
+ "train_labeled = train.copy()\n",
+ "test_labeled = test.copy()\n",
+ "\n",
+ "# Set results as NA if decision is negative.\n",
+ "train_labeled.result_Y = np.where(train.decision_T == 0, np.nan, train.result_Y)\n",
+ "test_labeled.result_Y = np.where(test.decision_T == 0, np.nan, test.result_Y)\n",
+ "\n",
+ "print(train_labeled.shape)\n",
+ "print(test_labeled.shape)\n",
+ "\n",
+ "tab = train_labeled.groupby(['result_Y', 'decision_T']).size()\n",
+ "tab.unstack()"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "## Algorithms\n",
+ "\n",
+ "### Contraction algorithm\n",
+ "\n",
+ "Below is an implementation of Lakkaraju's team's algorithm presented in [their paper](https://helka.finna.fi/PrimoRecord/pci.acm3098066). Relevant parameters to be passed to the function are presented in the description."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 92,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "def contraction(df,\n",
+ " judgeIDJ_col,\n",
+ " decisionT_col,\n",
+ " resultY_col,\n",
+ " modelProbS_col,\n",
+ " accRateR_col,\n",
+ " r,\n",
+ " binning=False):\n",
+ " '''\n",
+ " This is an implementation of the algorithm presented by Lakkaraju\n",
+ " et al. in their paper \"The Selective Labels Problem: Evaluating \n",
+ " Algorithmic Predictions in the Presence of Unobservables\" (2017).\n",
+ " \n",
+ " Parameters:\n",
+ " df = The (Pandas) data frame containing the data, judge decisions,\n",
+ " judge IDs, results and probability scores.\n",
+ " judgeIDJ_col = String, the name of the column containing the judges' IDs\n",
+ " in df.\n",
+ " decisionT_col = String, the name of the column containing the judges' decisions\n",
+ " resultY_col = String, the name of the column containing the realization\n",
+ " modelProbS_col = String, the name of the column containing the probability\n",
+ " scores from the black-box model B.\n",
+ " accRateR_col = String, the name of the column containing the judges' \n",
+ " acceptance rates\n",
+ " r = Float between 0 and 1, the given acceptance rate.\n",
+ " binning = Boolean, should judges with same acceptance rate be binned\n",
+ " \n",
+ " Returns:\n",
+ " u = The estimated failure rate at acceptance rate r.\n",
+ " '''\n",
+ " # Sort first by acceptance rate and judge ID.\n",
+ " sorted_df = df.sort_values(by=[accRateR_col, judgeIDJ_col],\n",
+ " ascending=False)\n",
+ "\n",
+ " if binning:\n",
+ " # Get maximum leniency\n",
+ " max_leniency = sorted_df[accRateR_col].values[0].round(1)\n",
+ "\n",
+ " # Get list of judges that are the most lenient\n",
+ " most_lenient_list = sorted_df.loc[sorted_df[accRateR_col].round(1) ==\n",
+ " max_leniency, judgeIDJ_col]\n",
+ "\n",
+ " # Subset to obtain D_q\n",
+ " D_q = sorted_df[sorted_df[judgeIDJ_col].isin(\n",
+ " most_lenient_list.unique())].copy()\n",
+ " else:\n",
+ " # Get most lenient judge\n",
+ " most_lenient_ID = sorted_df[judgeIDJ_col].values[0]\n",
+ "\n",
+ " # Subset\n",
+ " D_q = sorted_df[sorted_df[judgeIDJ_col] == most_lenient_ID].copy()\n",
+ "\n",
+ " # All observations of R_q have observed outcome labels\n",
+ " R_q = D_q[D_q[decisionT_col] == 1]\n",
+ "\n",
+ " # \"Observations deemed as high risk by B are at the top of this list\"\n",
+ " R_sort_q = R_q.sort_values(by=modelProbS_col, ascending=False)\n",
+ "\n",
+ " number_to_remove = int(\n",
+ " round((1.0 - r) * D_q.shape[0] - (D_q.shape[0] - R_q.shape[0])))\n",
+ "\n",
+ " # \"R_B is the list of observations assigned to t = 1 by B\"\n",
+ " R_B = R_sort_q[number_to_remove:R_sort_q.shape[0]]\n",
+ "\n",
+ " return np.sum(R_B[resultY_col] == 0) / D_q.shape[0]"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "### Causal algorithm\n",
+ "\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 93,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "def f(x, model, class_value):\n",
+ " '''\n",
+ " Parameters:\n",
+ " x = individual features\n",
+ " model = a trained sklearn predictive model. Predicts probabilities for given x.\n",
+ " class_value = the result (class) to predict (usually 0 or 1).\n",
+ " \n",
+ " Returns:\n",
+ " The probabilities (as vector) of class value (class_value) given \n",
+ " individual features (x) and the trained, predictive model (model).\n",
+ " '''\n",
+ " if x.ndim == 1:\n",
+ " # if x is vector, transform to column matrix.\n",
+ " f_values = model.predict_proba(np.array(x).reshape(-1, 1))\n",
+ " else:\n",
+ " f_values = model.predict_proba(x)\n",
+ "\n",
+ " return f_values[:, model.classes_ == class_value].flatten()"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "## Performance comparison\n",
+ "\n",
+ "Below we try to replicate the results obtained by Lakkaraju and compare their model's performance to the one of ours.\n",
+ "\n",
+ "### Predictive models\n",
+ "\n",
+ "Lakkaraju says that they used logistic regression. We construct the models using only *observed observations*, i.e. observations for which labels are available. We then predict the probability of negative outcome for all observations in the test data and attach it to our data set."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 94,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "# instantiate the model (using the default parameters)\n",
+ "logreg = LogisticRegression(solver='lbfgs')\n",
+ "\n",
+ "# fit, reshape X to be of shape (n_samples, n_features)\n",
+ "logreg = logreg.fit(\n",
+ " train_labeled.X[train_labeled.decision_T == 1].values.reshape(-1, 1),\n",
+ " train_labeled.result_Y[train_labeled.decision_T == 1])\n",
+ "\n",
+ "# predict probabilities and attach to data\n",
+ "label_probs_logreg = logreg.predict_proba(test.X.values.reshape(-1, 1))\n",
+ "\n",
+ "test = test.assign(B_prob_0_logreg=label_probs_logreg[:, 0])\n",
+ "test_labeled = test_labeled.assign(B_prob_0_logreg=label_probs_logreg[:, 0])"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 95,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "# Train model for predicting the probability of positive decision with a given\n",
+ "# leniency r and indivual features x.\n",
+ "\n",
+ "# Instantiate the model (using the default parameters)\n",
+ "decision_model = LogisticRegression(solver='lbfgs')\n",
+ "\n",
+ "# fit, reshape X to be of shape (n_samples, n_features)\n",
+ "decision_model = decision_model.fit(train[['X', 'acceptanceRate_R']],\n",
+ " train.decision_T)"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "### Visual comparison\n",
+ "\n",
+ "Let's plot the failure rates against the acceptance rates using the difference."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 96,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 1008x576 with 1 Axes>"
+ ]
+ },
+ "metadata": {
+ "needs_background": "light"
+ },
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "failure_rates = np.zeros((8, 5))\n",
+ "\n",
+ "for r in np.arange(1, 9):\n",
+ " \n",
+ " #### True evaluation\n",
+ " # Sort by failure probabilities, subjects with the smallest risk are first. \n",
+ " df_sorted = test.sort_values(by='B_prob_0_logreg', inplace=False, \n",
+ " ascending=True)\n",
+ "\n",
+ " to_release = int(round(df_sorted.shape[0] * r / 10))\n",
+ "\n",
+ " # Failure was coded as zero.\n",
+ " failure_rates[r - 1, 0] = np.mean(df_sorted.result_Y[0:to_release] == 0)\n",
+ " \n",
+ " #### Labeled outcomes only\n",
+ " # Sort by failure probabilities, subjects with the smallest risk are first. \n",
+ " df_sorted = test_labeled.sort_values(by='B_prob_0_logreg', inplace=False,\n",
+ " ascending=True)\n",
+ " \n",
+ " # Ensure that only labeled outcomes are available\n",
+ " df_sorted = df_sorted[df_sorted.decision_T == 1]\n",
+ " \n",
+ " to_release = int(round(df_sorted.shape[0] * r / 10))\n",
+ "\n",
+ " failure_rates[r - 1, 1] = np.mean(df_sorted.result_Y[0:to_release] == 0)\n",
+ " \n",
+ " #### Human error rate\n",
+ " # Get judges with correct leniency as list\n",
+ " correct_leniency_list = test_labeled.judgeID_J[\n",
+ " test_labeled['acceptanceRate_R'].round(1) == r / 10].values\n",
+ "\n",
+ " # Released are the people they judged and released, T = 1\n",
+ " released = test_labeled[test_labeled.judgeID_J.isin(correct_leniency_list)\n",
+ " & (test_labeled.decision_T == 1)]\n",
+ "\n",
+ " # Get their failure rate, aka ratio of reoffenders to number of people judged in total\n",
+ " failure_rates[r - 1, 2] = np.sum(\n",
+ " released.result_Y == 0) / correct_leniency_list.shape[0]\n",
+ " # onko jakaja oikein\n",
+ " \n",
+ " #### Contraction, logistic regression\n",
+ " failure_rates[r - 1, 3] = contraction(\n",
+ " test_labeled, 'judgeID_J', 'decision_T', 'result_Y', 'B_prob_0_logreg',\n",
+ " 'acceptanceRate_R', r / 10, False)\n",
+ "\n",
+ " #### P(Y=0 | T=1, X=x)*P(T=1 | R=r, X=x)*P(X=x)\n",
+ " failure_rates[r - 1, 4] = si.quad(lambda x: f(np.array([x]), logreg, 0) * \n",
+ " f(np.array([[x, r/10]]), decision_model, 1) * \n",
+ " scs.norm.pdf(x), -np.inf, np.inf)[0]\n",
+ "\n",
+ "# Error bars TBA\n",
+ "\n",
+ "plt.figure(figsize=(14, 8))\n",
+ "plt.plot(np.arange(0.1, 0.9, .1), failure_rates[:, 0], label='True Evaluation', c='green')\n",
+ "plt.plot(np.arange(0.1, 0.9, .1), failure_rates[:, 1], label='Labeled outcomes', c='lime')\n",
+ "plt.plot(np.arange(0.1, 0.9, .1), failure_rates[:, 2], label='Human evaluation', c='red')\n",
+ "plt.plot(np.arange(0.1, 0.9, .1), failure_rates[:, 3], label='Contraction, log.', c='blue')\n",
+ "plt.plot(np.arange(0.1, 0.9, .1), failure_rates[:, 4], label='Causal effect', c='magenta')\n",
+ "\n",
+ "plt.title('Failure rate vs. Acceptance rate')\n",
+ "plt.xlabel('Acceptance rate')\n",
+ "plt.ylabel('Failure rate')\n",
+ "plt.legend()\n",
+ "plt.grid()\n",
+ "plt.show()"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 97,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "0.0 (0.018718463137853268, 7.749450073818988e-11)\n",
+ "1.0 (0.33301477999280144, 6.337618003666896e-09)\n"
+ ]
+ }
+ ],
+ "source": [
+ "# Below are estimates for P(Y=0 | do(R=0)) and P(Y=0 | do(R=1))\n",
+ "r = 0.0\n",
+ "print(r, si.quad(lambda x: f(np.array([[x, r]]), decision_model, 1) * \\\n",
+ " f(np.array([x]), logreg, 0) * scs.norm.pdf(x), -np.inf, np.inf))\n",
+ "\n",
+ "r = 1.0\n",
+ "print(r, si.quad(lambda x: f(np.array([[x, r]]), decision_model, 1) * \\\n",
+ " f(np.array([x]), logreg, 0) * scs.norm.pdf(x), -np.inf, np.inf))"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "So it can be concluded that:\n",
+ "\n",
+ "\\begin{equation*}\n",
+ "P(Y=0 | \\text{do}(R=0)) \\approx 0.018 \\\\\n",
+ "P(Y=0 | \\text{do}(R=1)) \\approx 0.340 \\\\\n",
+ "\\end{equation*}"
+ ]
+ }
+ ],
+ "metadata": {
+ "kernelspec": {
+ "display_name": "Python 3",
+ "language": "python",
+ "name": "python3"
+ },
+ "language_info": {
+ "codemirror_mode": {
+ "name": "ipython",
+ "version": 3
+ },
+ "file_extension": ".py",
+ "mimetype": "text/x-python",
+ "name": "python",
+ "nbconvert_exporter": "python",
+ "pygments_lexer": "ipython3",
+ "version": "3.7.0"
+ },
+ "toc": {
+ "base_numbering": 1,
+ "nav_menu": {},
+ "number_sections": true,
+ "sideBar": true,
+ "skip_h1_title": true,
+ "title_cell": "Table of Contents",
+ "title_sidebar": "Contents",
+ "toc_cell": true,
+ "toc_position": {},
+ "toc_section_display": true,
+ "toc_window_display": true
+ },
+ "varInspector": {
+ "cols": {
+ "lenName": 16,
+ "lenType": 16,
+ "lenVar": 40
+ },
+ "kernels_config": {
+ "python": {
+ "delete_cmd_postfix": "",
+ "delete_cmd_prefix": "del ",
+ "library": "var_list.py",
+ "varRefreshCmd": "print(var_dic_list())"
+ },
+ "r": {
+ "delete_cmd_postfix": ") ",
+ "delete_cmd_prefix": "rm(",
+ "library": "var_list.r",
+ "varRefreshCmd": "cat(var_dic_list()) "
+ }
+ },
+ "types_to_exclude": [
+ "module",
+ "function",
+ "builtin_function_or_method",
+ "instance",
+ "_Feature"
+ ],
+ "window_display": false
+ }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}
diff --git a/analysis_and_scripts/Bachelors_thesis_analyses.ipynb b/analysis_and_scripts/Bachelors_thesis_analyses.ipynb
deleted file mode 100644
index 39665d65d8330aa188233997c17d551e4b1c4207..0000000000000000000000000000000000000000
--- a/analysis_and_scripts/Bachelors_thesis_analyses.ipynb
+++ /dev/null
@@ -1,2455 +0,0 @@
-{
- "cells": [
- {
- "cell_type": "markdown",
- "metadata": {
- "toc": true
- },
- "source": [
- "<h1>Table of Contents<span class=\"tocSkip\"></span></h1>\n",
- "<div class=\"toc\"><ul class=\"toc-item\"><li><span><a href=\"#Data-sets\" data-toc-modified-id=\"Data-sets-1\"><span class=\"toc-item-num\">1 </span>Data sets</a></span><ul class=\"toc-item\"><li><span><a href=\"#COMPAS-data\" data-toc-modified-id=\"COMPAS-data-1.1\"><span class=\"toc-item-num\">1.1 </span>COMPAS data</a></span></li><li><span><a href=\"#Synthetic-data\" data-toc-modified-id=\"Synthetic-data-1.2\"><span class=\"toc-item-num\">1.2 </span>Synthetic data</a></span></li></ul></li><li><span><a href=\"#Algorithms\" data-toc-modified-id=\"Algorithms-2\"><span class=\"toc-item-num\">2 </span>Algorithms</a></span><ul class=\"toc-item\"><li><span><a href=\"#Contraction-algorithm\" data-toc-modified-id=\"Contraction-algorithm-2.1\"><span class=\"toc-item-num\">2.1 </span>Contraction algorithm</a></span></li><li><span><a href=\"#Causal-model\" data-toc-modified-id=\"Causal-model-2.2\"><span class=\"toc-item-num\">2.2 </span>Causal model</a></span></li></ul></li><li><span><a href=\"#Performance-comparison\" data-toc-modified-id=\"Performance-comparison-3\"><span class=\"toc-item-num\">3 </span>Performance comparison</a></span><ul class=\"toc-item\"><li><span><a href=\"#On-synthetic-data\" data-toc-modified-id=\"On-synthetic-data-3.1\"><span class=\"toc-item-num\">3.1 </span>On synthetic data</a></span><ul class=\"toc-item\"><li><span><a href=\"#Predictive-models\" data-toc-modified-id=\"Predictive-models-3.1.1\"><span class=\"toc-item-num\">3.1.1 </span>Predictive models</a></span></li><li><span><a href=\"#Visual-comparison\" data-toc-modified-id=\"Visual-comparison-3.1.2\"><span class=\"toc-item-num\">3.1.2 </span>Visual comparison</a></span></li></ul></li><li><span><a href=\"#On-COMPAS-data\" data-toc-modified-id=\"On-COMPAS-data-3.2\"><span class=\"toc-item-num\">3.2 </span>On COMPAS data</a></span><ul class=\"toc-item\"><li><span><a href=\"#Predictive-models\" data-toc-modified-id=\"Predictive-models-3.2.1\"><span class=\"toc-item-num\">3.2.1 </span>Predictive models</a></span></li></ul></li></ul></li></ul></div>"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "# Bachelors thesis' analyses\n",
- "\n",
- "*This Jupyter notebook is for the analyses and model building for Riku Laine's bachelors thesis*\n",
- "\n",
- "Table of contents is provided above. First I will briefly present the COMPAS data set and then create the synthetic data set as done by Lakkaraju *et al.* ([link](https://helka.finna.fi/PrimoRecord/pci.acm3098066)). Then I will proceed to implement algorithms. Finally I will do the side-by-side comparisons of the results on the synthetic data. Finally I run the causal model on the COMPAS data.\n",
- "\n",
- "## Data sets\n",
- "\n",
- "*Below I load the COMPAS data set and generate the synthetic one.*\n",
- "\n",
- "### COMPAS data\n",
- "\n",
- "The following data filtering procedure follows the one described in the [ProPublica methodology](https://www.propublica.org/article/how-we-analyzed-the-compas-recidivism-algorithm)."
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 54,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "(7214, 53)\n",
- "['id' 'name' 'first' 'last' 'compas_screening_date' 'sex' 'dob' 'age'\n",
- " 'age_cat' 'race' 'juv_fel_count' 'decile_score' 'juv_misd_count'\n",
- " 'juv_other_count' 'priors_count' 'days_b_screening_arrest' 'c_jail_in'\n",
- " 'c_jail_out' 'c_case_number' 'c_offense_date' 'c_arrest_date'\n",
- " 'c_days_from_compas' 'c_charge_degree' 'c_charge_desc' 'is_recid'\n",
- " 'r_case_number' 'r_charge_degree' 'r_days_from_arrest' 'r_offense_date'\n",
- " 'r_charge_desc' 'r_jail_in' 'r_jail_out' 'violent_recid'\n",
- " 'is_violent_recid' 'vr_case_number' 'vr_charge_degree' 'vr_offense_date'\n",
- " 'vr_charge_desc' 'type_of_assessment' 'decile_score.1' 'score_text'\n",
- " 'screening_date' 'v_type_of_assessment' 'v_decile_score' 'v_score_text'\n",
- " 'v_screening_date' 'in_custody' 'out_custody' 'priors_count.1' 'start'\n",
- " 'end' 'event' 'two_year_recid']\n"
- ]
- }
- ],
- "source": [
- "import numpy as np\n",
- "import pandas as pd\n",
- "from datetime import datetime\n",
- "import matplotlib.pyplot as plt\n",
- "import scipy.stats as scs\n",
- "import seaborn as sns\n",
- "import numpy.random as npr\n",
- "from sklearn.preprocessing import OneHotEncoder\n",
- "from sklearn.linear_model import LogisticRegression\n",
- "from sklearn.ensemble import RandomForestClassifier\n",
- "\n",
- "%matplotlib inline\n",
- "\n",
- "plt.rcParams.update({'font.size': 16})\n",
- "plt.rcParams.update({'figure.figsize': (14, 7)})\n",
- "\n",
- "# Read file\n",
- "compas_raw = pd.read_csv(\"../data/compas-scores-two-years.csv\")\n",
- "\n",
- "# Check dimensions, number of rows should be 7214\n",
- "print(compas_raw.shape)\n",
- "print(compas_raw.columns.values)"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 55,
- "metadata": {},
- "outputs": [
- {
- "data": {
- "text/plain": [
- "(6172, 13)"
- ]
- },
- "execution_count": 55,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
- "source": [
- "# Select columns\n",
- "compas = compas_raw[[\n",
- " 'age', 'c_charge_degree', 'race', 'age_cat', 'score_text', 'sex',\n",
- " 'priors_count', 'days_b_screening_arrest', 'decile_score', 'is_recid',\n",
- " 'two_year_recid', 'c_jail_in', 'c_jail_out'\n",
- "]]\n",
- "\n",
- "# Subset values, see reasons in ProPublica methodology.\n",
- "compas = compas.query('days_b_screening_arrest <= 30 and \\\n",
- " days_b_screening_arrest >= -30 and \\\n",
- " is_recid != -1 and \\\n",
- " c_charge_degree != \"O\"')\n",
- "\n",
- "# Drop row if score_text is na\n",
- "compas = compas[compas.score_text.notnull()]\n",
- "\n",
- "compas.shape"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 56,
- "metadata": {},
- "outputs": [
- {
- "data": {
- "text/html": [
- "<div>\n",
- "<style scoped>\n",
- " .dataframe tbody tr th:only-of-type {\n",
- " vertical-align: middle;\n",
- " }\n",
- "\n",
- " .dataframe tbody tr th {\n",
- " vertical-align: top;\n",
- " }\n",
- "\n",
- " .dataframe thead th {\n",
- " text-align: right;\n",
- " }\n",
- "</style>\n",
- "<table border=\"1\" class=\"dataframe\">\n",
- " <thead>\n",
- " <tr style=\"text-align: right;\">\n",
- " <th></th>\n",
- " <th>0</th>\n",
- " <th>1</th>\n",
- " <th>2</th>\n",
- " <th>5</th>\n",
- " <th>6</th>\n",
- " </tr>\n",
- " </thead>\n",
- " <tbody>\n",
- " <tr>\n",
- " <th>age</th>\n",
- " <td>69</td>\n",
- " <td>34</td>\n",
- " <td>24</td>\n",
- " <td>44</td>\n",
- " <td>41</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>c_charge_degree</th>\n",
- " <td>F</td>\n",
- " <td>F</td>\n",
- " <td>F</td>\n",
- " <td>M</td>\n",
- " <td>F</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>race</th>\n",
- " <td>Other</td>\n",
- " <td>African-American</td>\n",
- " <td>African-American</td>\n",
- " <td>Other</td>\n",
- " <td>Caucasian</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>age_cat</th>\n",
- " <td>Greater than 45</td>\n",
- " <td>25 - 45</td>\n",
- " <td>Less than 25</td>\n",
- " <td>25 - 45</td>\n",
- " <td>25 - 45</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>score_text</th>\n",
- " <td>Low</td>\n",
- " <td>Low</td>\n",
- " <td>Low</td>\n",
- " <td>Low</td>\n",
- " <td>Medium</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>sex</th>\n",
- " <td>Male</td>\n",
- " <td>Male</td>\n",
- " <td>Male</td>\n",
- " <td>Male</td>\n",
- " <td>Male</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>priors_count</th>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>4</td>\n",
- " <td>0</td>\n",
- " <td>14</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>days_b_screening_arrest</th>\n",
- " <td>-1</td>\n",
- " <td>-1</td>\n",
- " <td>-1</td>\n",
- " <td>0</td>\n",
- " <td>-1</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>decile_score</th>\n",
- " <td>1</td>\n",
- " <td>3</td>\n",
- " <td>4</td>\n",
- " <td>1</td>\n",
- " <td>6</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>is_recid</th>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>1</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>two_year_recid</th>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>1</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>c_jail_in</th>\n",
- " <td>2013-08-13 06:03:42</td>\n",
- " <td>2013-01-26 03:45:27</td>\n",
- " <td>2013-04-13 04:58:34</td>\n",
- " <td>2013-11-30 04:50:18</td>\n",
- " <td>2014-02-18 05:08:24</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>c_jail_out</th>\n",
- " <td>2013-08-14 05:41:20</td>\n",
- " <td>2013-02-05 05:36:53</td>\n",
- " <td>2013-04-14 07:02:04</td>\n",
- " <td>2013-12-01 12:28:56</td>\n",
- " <td>2014-02-24 12:18:30</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>length_of_stay</th>\n",
- " <td>0</td>\n",
- " <td>10</td>\n",
- " <td>1</td>\n",
- " <td>1</td>\n",
- " <td>6</td>\n",
- " </tr>\n",
- " </tbody>\n",
- "</table>\n",
- "</div>"
- ],
- "text/plain": [
- " 0 1 \\\n",
- "age 69 34 \n",
- "c_charge_degree F F \n",
- "race Other African-American \n",
- "age_cat Greater than 45 25 - 45 \n",
- "score_text Low Low \n",
- "sex Male Male \n",
- "priors_count 0 0 \n",
- "days_b_screening_arrest -1 -1 \n",
- "decile_score 1 3 \n",
- "is_recid 0 1 \n",
- "two_year_recid 0 1 \n",
- "c_jail_in 2013-08-13 06:03:42 2013-01-26 03:45:27 \n",
- "c_jail_out 2013-08-14 05:41:20 2013-02-05 05:36:53 \n",
- "length_of_stay 0 10 \n",
- "\n",
- " 2 5 \\\n",
- "age 24 44 \n",
- "c_charge_degree F M \n",
- "race African-American Other \n",
- "age_cat Less than 25 25 - 45 \n",
- "score_text Low Low \n",
- "sex Male Male \n",
- "priors_count 4 0 \n",
- "days_b_screening_arrest -1 0 \n",
- "decile_score 4 1 \n",
- "is_recid 1 0 \n",
- "two_year_recid 1 0 \n",
- "c_jail_in 2013-04-13 04:58:34 2013-11-30 04:50:18 \n",
- "c_jail_out 2013-04-14 07:02:04 2013-12-01 12:28:56 \n",
- "length_of_stay 1 1 \n",
- "\n",
- " 6 \n",
- "age 41 \n",
- "c_charge_degree F \n",
- "race Caucasian \n",
- "age_cat 25 - 45 \n",
- "score_text Medium \n",
- "sex Male \n",
- "priors_count 14 \n",
- "days_b_screening_arrest -1 \n",
- "decile_score 6 \n",
- "is_recid 1 \n",
- "two_year_recid 1 \n",
- "c_jail_in 2014-02-18 05:08:24 \n",
- "c_jail_out 2014-02-24 12:18:30 \n",
- "length_of_stay 6 "
- ]
- },
- "metadata": {},
- "output_type": "display_data"
- }
- ],
- "source": [
- "# Calculate length of stay\n",
- "out = pd.to_datetime(compas.c_jail_out, format=\"%Y-%m-%d %H:%M:%S\")\n",
- "in_ = pd.to_datetime(compas.c_jail_in, format=\"%Y-%m-%d %H:%M:%S\")\n",
- "\n",
- "compas['length_of_stay'] = (out - in_).astype('timedelta64[D]')\n",
- "\n",
- "# Structure of the data\n",
- "display(compas.head(5).T)"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 57,
- "metadata": {
- "scrolled": false
- },
- "outputs": [
- {
- "data": {
- "text/html": [
- "<div>\n",
- "<style scoped>\n",
- " .dataframe tbody tr th:only-of-type {\n",
- " vertical-align: middle;\n",
- " }\n",
- "\n",
- " .dataframe tbody tr th {\n",
- " vertical-align: top;\n",
- " }\n",
- "\n",
- " .dataframe thead th {\n",
- " text-align: right;\n",
- " }\n",
- "</style>\n",
- "<table border=\"1\" class=\"dataframe\">\n",
- " <thead>\n",
- " <tr style=\"text-align: right;\">\n",
- " <th></th>\n",
- " <th>count</th>\n",
- " <th>unique</th>\n",
- " <th>top</th>\n",
- " <th>freq</th>\n",
- " <th>mean</th>\n",
- " <th>std</th>\n",
- " <th>min</th>\n",
- " <th>25%</th>\n",
- " <th>50%</th>\n",
- " <th>75%</th>\n",
- " <th>max</th>\n",
- " </tr>\n",
- " </thead>\n",
- " <tbody>\n",
- " <tr>\n",
- " <th>age</th>\n",
- " <td>6172</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>34.5345</td>\n",
- " <td>11.7309</td>\n",
- " <td>18</td>\n",
- " <td>25</td>\n",
- " <td>31</td>\n",
- " <td>42</td>\n",
- " <td>96</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>c_charge_degree</th>\n",
- " <td>6172</td>\n",
- " <td>2</td>\n",
- " <td>F</td>\n",
- " <td>3970</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>race</th>\n",
- " <td>6172</td>\n",
- " <td>6</td>\n",
- " <td>African-American</td>\n",
- " <td>3175</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>age_cat</th>\n",
- " <td>6172</td>\n",
- " <td>3</td>\n",
- " <td>25 - 45</td>\n",
- " <td>3532</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>score_text</th>\n",
- " <td>6172</td>\n",
- " <td>3</td>\n",
- " <td>Low</td>\n",
- " <td>3421</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>sex</th>\n",
- " <td>6172</td>\n",
- " <td>2</td>\n",
- " <td>Male</td>\n",
- " <td>4997</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>priors_count</th>\n",
- " <td>6172</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>3.24644</td>\n",
- " <td>4.74377</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>4</td>\n",
- " <td>38</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>days_b_screening_arrest</th>\n",
- " <td>6172</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>-1.74028</td>\n",
- " <td>5.08471</td>\n",
- " <td>-30</td>\n",
- " <td>-1</td>\n",
- " <td>-1</td>\n",
- " <td>-1</td>\n",
- " <td>30</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>decile_score</th>\n",
- " <td>6172</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>4.4185</td>\n",
- " <td>2.83946</td>\n",
- " <td>1</td>\n",
- " <td>2</td>\n",
- " <td>4</td>\n",
- " <td>7</td>\n",
- " <td>10</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>is_recid</th>\n",
- " <td>6172</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>0.484446</td>\n",
- " <td>0.499799</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>1</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>two_year_recid</th>\n",
- " <td>6172</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>0.45512</td>\n",
- " <td>0.498022</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>1</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>c_jail_in</th>\n",
- " <td>6172</td>\n",
- " <td>6172</td>\n",
- " <td>2013-03-21 04:40:57</td>\n",
- " <td>1</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>c_jail_out</th>\n",
- " <td>6172</td>\n",
- " <td>6161</td>\n",
- " <td>2013-09-14 05:58:00</td>\n",
- " <td>3</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>length_of_stay</th>\n",
- " <td>6172</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>14.6228</td>\n",
- " <td>46.6935</td>\n",
- " <td>-1</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>5</td>\n",
- " <td>799</td>\n",
- " </tr>\n",
- " </tbody>\n",
- "</table>\n",
- "</div>"
- ],
- "text/plain": [
- " count unique top freq mean \\\n",
- "age 6172 NaN NaN NaN 34.5345 \n",
- "c_charge_degree 6172 2 F 3970 NaN \n",
- "race 6172 6 African-American 3175 NaN \n",
- "age_cat 6172 3 25 - 45 3532 NaN \n",
- "score_text 6172 3 Low 3421 NaN \n",
- "sex 6172 2 Male 4997 NaN \n",
- "priors_count 6172 NaN NaN NaN 3.24644 \n",
- "days_b_screening_arrest 6172 NaN NaN NaN -1.74028 \n",
- "decile_score 6172 NaN NaN NaN 4.4185 \n",
- "is_recid 6172 NaN NaN NaN 0.484446 \n",
- "two_year_recid 6172 NaN NaN NaN 0.45512 \n",
- "c_jail_in 6172 6172 2013-03-21 04:40:57 1 NaN \n",
- "c_jail_out 6172 6161 2013-09-14 05:58:00 3 NaN \n",
- "length_of_stay 6172 NaN NaN NaN 14.6228 \n",
- "\n",
- " std min 25% 50% 75% max \n",
- "age 11.7309 18 25 31 42 96 \n",
- "c_charge_degree NaN NaN NaN NaN NaN NaN \n",
- "race NaN NaN NaN NaN NaN NaN \n",
- "age_cat NaN NaN NaN NaN NaN NaN \n",
- "score_text NaN NaN NaN NaN NaN NaN \n",
- "sex NaN NaN NaN NaN NaN NaN \n",
- "priors_count 4.74377 0 0 1 4 38 \n",
- "days_b_screening_arrest 5.08471 -30 -1 -1 -1 30 \n",
- "decile_score 2.83946 1 2 4 7 10 \n",
- "is_recid 0.499799 0 0 0 1 1 \n",
- "two_year_recid 0.498022 0 0 0 1 1 \n",
- "c_jail_in NaN NaN NaN NaN NaN NaN \n",
- "c_jail_out NaN NaN NaN NaN NaN NaN \n",
- "length_of_stay 46.6935 -1 0 1 5 799 "
- ]
- },
- "execution_count": 57,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
- "source": [
- "compas.describe(include='all').T"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "**Notes:**\n",
- "\n",
- "* Mean age is roughly 34.5 years ranging from 18 to 96\n",
- "* Defendants have an average of 3.2 priors (sd 4.7) and more than half have 1 or more prior.\n",
- "* 48.4% have recidivated in general and 45.5% recidivated within a two-year period following their arrest."
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 58,
- "metadata": {},
- "outputs": [
- {
- "data": {
- "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3EAAANtCAYAAAA+YPKsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzsnXt4FeW1/z/v7HsukBASvIQWREQjgiSKgO0pyjleqfwseCkEBS0XqdLT0yq2lmrL6SmIVKuWi7YFBVQU9Gjp8dKDYntEKgSUahQpohJFCCGBXPZ93t8fe89kz+zZsCEBEng/z5OHsPfsmdmZNe+8611rfZeQUqJQKBQKhUKhUCgUis6BdrxPQKFQKBQKhUKhUCgU2aOcOIVCoVAoFAqFQqHoRCgnTqFQKBQKhUKhUCg6EcqJUygUCoVCoVAoFIpOhHLiFAqFQqFQKBQKhaIToZw4hUKhUCgUCoVCoehEKCdOoVAoFAqFQqFQKDoRx92JE0KUCiEeEUK8LYRoEUJIIUQvh+38Qoi5QohdQohgcvt/cdhOE0L8RAjxqRAiJIR4Twgx+lh8F4VCoVAoFAqFQqE42hx3Jw44E7geqAf+dpDt/gBMAn4OjAR2Aa8KIc63bTcLuA94FLgSWA88J4S4qn1PW6FQKBQKhUKhUCiOPR3BifurlLKHlPIq4DmnDYQQA4GxwA+llI9LKdeQcPw+B36Zsl0J8GNgtpTyASnlG1LKKcAbwOxsTuaKK66QgPrp2D8KG8puO/yPwgFltx3+R+GAstsO/6NwQNlth/85bI67Eyel1LPY7BogCqxI+VwMeAa4XAjhS758OeAFltk+vww4TwjR+1AH2rt3bzanrVB0KJTdKjojym4VnRFlt4rOiLLbE4/j7sRlybnADilli+31D0g4bWembBcG/umwHUDZUTtDhUKhUCgUCoVCoTgGdBYnrhuJmjk7+1LeN/5tkFLaw5L27RQKhUKhUCgUCoWiU9JZnDiBc76oOMLtrG8KMVkIsVEIsbG2tvYIT1GhOLYou1V0RpTdKjojym4VnRFltyc2ncWJ24dzFK0w5X3j30IhhN1ps29nQUr5mJTyAinlBcXFxW0+WYXiWHAkdqvrktrGMF/Ut1DbGEbXVQ244thyIoy36j46+Wgvu1W2oziWHMpulT12btzH+wSy5APgWiFEjq0urgyI0FoD9wHgA/pgrYszauGqj/aJAvS6+8+H3ObT2VcfgzNRKFrRdcnW3Y1MenIjNfVBSgsDPH7TBfTrkY+mHTRYrVAokqj7SHGkKNtRdCSUPXZ+Oksk7iXAA1xnvCCEcAM3AK9JKcPJl18h4dSNs32+EnhfSrnjGJyrQtEhqWuOmIM1QE19kElPbqSuOXKcz0yh6Dyo+0hxpCjbUXQklD12fjpEJE4IMSb5a0Xy3yuFELVArZTyTSnlu0KIFcBDQggPsAO4DehNisMmpdwjhHgQ+IkQohHYRMLRuxQYdYy+jkLRIYnE4uZgbVBTHyQSix+nM1IoOh/qPlIcKcp2FB0JZY+dnw7hxJHe5Ht+8t83geHJ3ycCvwL+EygA3gOukFJusn32HqAJ+AFwCrAVuF5K+af2P22FovPgdbsoLQxYBu3SwgBet+s4npVC0blQ95HiSFG2o+hIKHvs/HSIdEoppcjwMzxlm6CU8j+klKdIKf1SyouklGsd9hWXUv6nlPLrUkqflHKAlHLlsfw+CkVHpCjXy+M3XUBpYQDAzH8vyvUe5zNTKDoP6j5SHCnKdhQdCWWPnZ+OEolTKBRHGU0T9OuRzwvTLiYSi+N1uyjK9aoCZoXiMFD3keJIUbaj6Egoe+z8KCdOoTiJ0DRBcb7veJ+GQtGpUfeR4khRtqPoSCh77Nx0iHRKhUKhUCgUCoVCoVBkh3LiFAqFQqFQKBQKhaIToZw4hUKhUCgUCoVCoehEKCdOoVAoFAqFQqFQKDoRyolTKBQKhUKhUCgUik6EcuIUCoVCoVAoFAqFohOhnDiFQqFQKBQKhUKh6ESoPnGHSa+7/3y8T0GhUCgUCoVCoVCcxCgnTqE4gdF1SV1zhEgsjtftoijXi6aJ431aCsUJhbrPFMoGFJ0RZbedG+XEKRQnKLou2bq7kUlPbqSmPkhpYYDHb7qAfj3y1SCtULQT6j5TKBtQdEaU3XZ+VE2cQnGCUtccMQdngJr6IJOe3Ehdc+Q4n5lCceKg7jOFsgFFZ0TZbedHOXEKxQlKJBY3B2eDmvogkVj8OJ2RQnHioe4zhbIBRWdE2W3nR6VTKhQnKF63i8vKShhd0ZOCgIeGYJRVVTvxul3H+9QUihOGgNfF4gkXkuN10RCMsnDtdmqbwuo+O4lQY62iM6LstvOjnDiF4gSlMOBh+oizmLqsiuI8H9NH9OUnV52DlBJdlyrnXXFCcDwL83VdsvtAmJkvvm/WlMwdM4AeXfwU5XqPyTkojj+FAQ93XXE2O/clohpel8ZdV5xNYcBznM9MociMMUd4eM3HjK7oSVGul3uuLqPAr1yDzoK6UgrFCUp9MGo6cD++vB8zVm1RxcuKE4rjXZjvVFNy58otPD9tmLq3TiIOhKPUNqY7891yvXRz+4736SkUjtQHozy85mNuHtbbMj9YNL6Cc07posawToBy4hSKE5RILE5xno/7xwxg7qsfMXNkmZky8eBftvKf155HSb7/eJ+mQnHEZCrMf2HaxRTnH/3JcyQWZ9gZRUwZ3geXELg0wb7mMFKXR/3Yio5DMBJn8Vs7LGPs4rd2cO+3z4Xc4312CoUzkVicm4b24ol1Vtv97f9+zK+uHXBMxlBF21BOnEJxghLwurjvmjL8Hi1tpW3O6AHE1URT0ck5noX5ui7xuTXGD+vFzX98x7y35o8rJxzXVcrySYSm4TjGako6TtGB8bg0enXPcbRdXdeP9+kpskANMQrFCYpE0hKJA8IcoCExyZ2xagu6cuIUnRyv20VpYcDyWmlh4KgX5htpnA3BGLctq7LcW9OWbyISk0qm+yRC18kwxh7nE1MoDkIsriOl8/wgpuYHnQLlxCkUJyihiM6dK7cQ03XHaIU9EqfrktrGMF/Ut1DbGFZOnqLDU5Tr5fGbLjAdOaMm7miLihhpnJrA8d7SBEqm+yQipsusxliFoiMRkxJdKtvtzKh0SoXiBCUuJcPOKMLj0lg5dSh1zRHWVO9mRFkPinK9uDRhpnwdb4EIheJI0DRBvx75vDDt4mOqTmmkccZ1yZRv9mLMBV/DpQniumTlxs/RJUqm+yTCrQlHO3CpsVPRgfF7XERjurLdToxy4hSKE5R8n4vKoV/nxsfWU1Mf5LKyEm6/tC/Tlm9Kc9SOt0CEQnGkaJo45jZqpHF+ureJkeeXMnHJBvOeWlBZQWGum64+JS9/spAf0BztID+gkp0UHZcCn5s6PeJou3k+ZbudAXWVFIoTlJaIbjpsAKMrelr+bzhqu/YHTSXLVGrqg+i6rlIsFR2WY5kCnHoslwaP33QBZ5/aNa0m7rZlVYQiOvXB6FE7F0XHojGoO9pBY/DoF8WpNHjFkVLbHKE57Gy7TWFV0NkZUJE4heIExV6nURDwOOa+h2M6n9W1cN81Zdz3UjWbdzYAcFlZCXubI0xZWtXaP6ayglML/BQEDp6ydjwbMCtOHA5mR8cyBdjpWE/eMjhjPUlMl6Bq4k4aMtXEOYlDtOfYqNLgFW0hpktcGep6Y7pUCrudABWJUyhOUNyasCj3NQSjjkp+2/Y0MfPF92mJxPnpVeeYr//s6jLTgYPEwD5lWRXv7dzP1t2NGVd8jYnFtfPf4uI5b3Dt/LcOur1C4cSh7ChTCvDRUIV0OtZNf3wn7R6DxL3j1oSqiTuJOJgdpNLeY+OxvAcUJx5+t4bbpWW0XfXc7vgoJ06hOEHJ82ssqKwwB+hVVTst/zf6wSxcu52a+iB3rtzCqQV+3ppxCS9MuxiXJhxX6HK8LstEwZ7Os7c5rCYWijZzqAlqW3vExWI6XzYE+ayumS8bgsRimdOHMh3L79GYP67cck/NH1eO36MddYVMRcfhYHaQSns7XcezT6Ki8xOXkmVv70iz3QWVFfjcmnpudwJUOqVCcYIScLnp0UXyzOQhxHWJWxPk+TWenjQEXUo++qqRB17daqZPptbESSSe5Apd6iShtDBAQzBqThSc0nmW3XrRUZlYqBTNk4tDTVANcRG7fWYTAYvFdD7a3cjUZa2pwgsrKzi7Rz5ud+vE27C5uJQsnnAhL/9jFyPKevD1bgFyfR6awnHius6TtwwGIK5LHv/rJ9wx4kyK8pRtniw0R+Ise/szFk+40FT4e/yvn3D7iDMpStnucJyubMa7w7kHYjGdPU1honEdj0ujJM9nsXXFyUcsLln0t0/J83l4atIQpJQIIcj1aoRiOsV5PrUg0MFRTpxCcYJyIBxj9/4wt6WoUS6orGD1uzWU9ypi1upq8+E/qGcBd13Rz1SyLC0MsGBcOc9NHUJzWKcxFKWhJUphrodfvFRtThScVpZ37G0+4sl1JlTtx8nHoSaoRo84u01kEwHb0xQ2HThI2O3UZVU8O2UopxUkVqSdbG7BuHI0IYjq8N3HW++VOaMHmAsipYUB7hjRl2g0jsejUipPBjyaxrpP6ni2qsZ8rbQwwA//7Szrdm6Ny8pKGF3Rk4KAh4ZglFVVO/HYnKlsx7uiXC9PT7qIcEyiCdAl+Nwi7R7IdtHicFCLap0fjythj+W9ujE2ZTxbVFnBKQU+po/oq9LCOzjKiVMoTlDCMZ1P9zbyzOQhaAKkhJZInHFDe/Pomm3MGT2AGau2UFMfZPqIvty5cotVoWr5Jp68ZTATFr9jmcT+9KpzKMrzUpTrZdf+YNrK8sNrtrGosoIpKROGtjZgVi0QTj4O5aS1pUdcNK47F/PHW1MqnWzOuCdu+uM75uuJ1Wqdh248n2hc4nbBgWAUlyZMh1BxYqMJeOKWC5FSmM6UEAnHKhWvS6S1eZk/rhyvy7phXXOEB/+ylZkjy0xn78G/bOVX1w6wjHe6LgnHdHbuS6S5t0Ti9OwWSBOkyGbR4nBQi2onBpqAX4w6l492NTHvuoE0BKMsXLudKcuqWDF5CH2Kc5EogZOOjHLiFIoTlByvxtmnFVBTH6Q438fndS08vGYbtU1h5owewIubv2DmyDLOPiU/46R2X3MkbRI7a1R/8vyJocPrdjmuLJ9a4G/XBszHsvZDrTB3DLJx0o6kR5yeTC1eOXUodc0RFq7dbkbQXJrgi/oWvG5XRptLrRUd1LOAe68po745ylf7Q7RE4pQW+skLuBEoQYCTBa9boIehZl+L6UyVdgvgdVvHjeZwnEdf32Zxzh59fRv3fvtcCnJat9N1nZuH9TYX2Yxor65b6zb3tUSobQwz88X3ze3mjhlAF7+Hki5+c7tsFi1aj33o8S9T3fPz04ZRku8/7P0pjg9ul2DX/ghPv/MZoyt6UpTrZd71A1m4dju79ocYs/Dtw1KlVhx7lBOnUJygRGKSvY1hM8KWmvY1Y9UWZo4sY9bqapZ/7yK+2h9yTF0LRa1OUqqwyQvTLqYw4GH6iLPS0nS6+DztWm/Rlvqnw0GtMHcs2ruRt9P1nTN6AE+s28EdI87iyXU7WPS3TyktDPDU9y5ytDlXUomwpj7IXVf0IxiJp02iTy8MoETdTh6MsTbNmfJZp1iahqNzZh9b4hJzG0iMuzNWbeHZKUOtx43raRkUd67cwjOTh1i2y1Tf7HYdWRpnKOq8wBGK6ke0P8XxIRTVeXjNx2k2OX9cOc3hGNCqSj1rVH9O6epX166D0W6zLCFEnhDi60IIT3vtU6FQHDlRXaY94Ges2sLU4X2oqQ9SlOtl7pgBCAEuTTB3zACLQtXcMQPw2Zwku7BJfTDqmKbT3o2OjdS61PNra4qmE0qy+8TBqQmy0/WdsWoLs0b1Jx7XufScU1g0voLiPB//+edqFo2vSLsnpJTMGZ24V07p4necRAsS4haKkwOnsfbOlVuI2mxASuHonEmbqcR158iZ3abiGfrT2WXhS/J8LLQpEy+srKAkz7pAku345xLOLRVsWaFqPO3gxHTJ6IqeaTY5bfkmi605qVIrOgZtjsQJIUYCvwQGAhIYDGwSQvweeF1K+VRbj6FQKA6fTA/4goCH0sIA3fN8/HDFu/zmhoF4XBq/+vOHljSf+1/ZygPXD6S0MEBxXqLI+WtFOexqCHJZWclBU84isThfNgSPWAnNKQXnSOufDgcl2d2xONJUrEwRgC5+d9r1Lc7zUdsU4fanN5vbzrtuILqUFOZ4eHbKUKSUaELQHIkSjUueWLeDmSPL0DK04YhLic+llP9OFrJ1pnQpKc7zWcbZhWu3o9u8OMNJSosC20w/4HFOZ/fbBHU0TZDjdTFrVH8z3TPH60q7lyKxuOP52ce/gNfFg9cP5IfPvmfeMw9eP5CA15W2PzWedlzcWkIEx+ka5aZEke2Lt7WNYZUe20FokxMnhPh/wCpgDTADuD/l7R3AzYBy4hSK44DP7ZxC0xKJ8+D1A/F5NIrzvXxRH6RHFz+1TWGmLK2ybOvWBE9Puoj9wZgZcbusrIR7ri4jHI0jBFxWVsJr1Xssn4vpknG/b1W7OhwltIOl4BipdUerzuJYpW0qDo2THSwaX0H3XC+aph30mmeKADw7ZWja9Z0+om9aNPlHz73HrFH9GTb7DdP+zuyeS11zhIfXfMit3ziDHz33ntlfyW4vHk2kiVooTly8GdIVPTZH3qMJ7rqinyXFfe6YAXhsxuJ2acwdMyBtO3v6Y2HAwx0jzuK2lHT2BZUVFAasCVF1zRGLGI9xfnZhKL9X475rytjXHDW/133XlOH3Wo/bxeehMNdrcQoLc7108VmPezjjqaqdO/bkeDWK832O16ggx2P+bpRhlBYGiOuSa+e/dVjpseraHj3aulR4L7BYSnkZ8JDtvfeB/m3cv0KhOEI0QVqK5IJx5eT53HjcGvuaI9xxaV9eevdLHv/rJ44NP3/3+j+p3tUqTT2oZwE3D+vNuN//nW/c/wY3PLae2y/ty2VlJebnFlVW8Ks/V1smxVOXVbGnKZzVeWeagDcEI2Z6XE19C/e8sIWL57zBtfPfYuvuxrRV7yPhWKVtKg6Nkx1MWVrFuzX7D3nNM4qSCNKub+/uuY7b5iSjCob91TZHmLqsiteq96BLyaxR/Tm1q58F9vtmXDkBr4ZQc5STBpcLRztw2XyVWIa0y5jNjsOxOPe/klCnXDF5CDNHlnH/K1sJ2xrS722JmA6csb/bllWxt8Wa8pZtRCweTygYz3zxfW54bD0zX3yflkicuC1wVh+MMmHxBiYu2cANj61n4pINTFi8IS2NvijXy+PjbePp+PTx1FiwuXb+W+0+pisyE9ehi9+Vlmo7PzlPeOPH32LWqP488OpWapvCLBpfwX/anu2HSrFU1/bo0tZ0ynOAu5K/269IPVj6XCoUimNIKKZz/ytbWTzhQiJxHa9LoykcY19LhFVVO7nz8rOZuHQDi8ZXkO/3ENd1npo0hJZwlM/2BXlkzcd8d/DXyfG6zBSbPsW57NwXpDjPR019or3Ao69v42cjz+XuK8/B59bwuIQZmRvUs4Cpw/tQEPCgCdjVECRyiBRLp5SeNdW72dUQsrQtmDN6ALWNETbvbODBv2zlvmv6I6Vs00pfW2TrFe1LpolnQcBjTh5emHYxRbnetFXeTBEATdMs11eIRO8Np22jKcp9NfVBYil1SpoQTFyygf+bcQmr3/uC5d+7CE0k6uAaWiJEYon0S8XJQTQmWf3eF5Zm3ys3fs5Nw3pbtotlSLu017ppQlCcb3V0ivO9adHdSMy5di5qc/ayjYhF4jqL39phGXsXv7WDn3/7XNtxs0u71HVJrt/FkomDzdYLXrdIk6zPtqWCon2J6ZKYDrG4zrJbL8KlCXQpeWr9p4wf1pu9jRHOLMnj0bGD8Lpd6Lru+GyPxOIZ2xCkLsYZn2kOx/jqQIhTuvjVs7WNtNWJOwB0z/BeL6C2jftXKBRHiEsIapvChKJxDoRiaYpocV2nOM+HlFgafc4fV86a6t28Vr2HGVeeQ3M4lpYCZKRXQEJtzf75y8pKqG2M8OPL+zFj1RaK83xp+8iUYhnwutK2nT+unN+u+ThNEGDmyDIWrt3OzcN6c/2it9tFAa29FREVR0amiWdDcrU/UXOkO6be9i3Oy9hjTkvWgRifK87zpdX3zB0zgIDXxaCeBWze2cBlZSVIiXk+DcFowikUcOWA0/iyIWix10WVFXTPU9HbkwVNwL/068HEJRssY6R9+PFnsGm7gFSOV3NMk8yxpTW6tQy1c7YDH6rnooErg3qmvbzTaYw27plU9rVE+KI+mLad3+2ytEDItqWCon1xaYI9B8KWeuC5YwZw5YDTcAn44bPvMnfMAM46JY+iXB+1jWGzRt54th/qmWssxg3qWZD1ZxTZI6RdFulwPizEcuA84F+ARiAKVADVwN+Ad6WUk9vhPI8ZF1xwgdy4cWPG93vd/ed2Oc6ns69ul/2cpKg73oaT3X7Z0MKexjBFuT6+m3SyDEoLAyyecCE19UFTFtv+3sQlG5j9nfPo1T2XGx9L//zMkWUAzFpdnfbe05MuIhxLRCXqmiN08bstaUTGdkazWSNnXtd14hLTIbMfL7VmD2DF5CE0BKOO59DBGoErm3XgYOOtU03c78aWown4cn+IVVU7ue+a/o62YrS/2NMUtojraJqgrjlCMBpj9/4w4Vgcj0vjlK5+PqltJsfrMqMKtU1h5o4ZQCiqc0ZxLrv2h/B7NG5/arO5KPG1bjl8vLvJ8R56ZvIQSgtznL5aZ0LZrQN2u/2ivoUbHMbIFZOHcHqKDezeH+STvc1pTs0Z3XPp0TVwRPv7fF9LmsDI17rlWPYH2dUlfVHfwi/+9EGaUMq93z7Xctw9jSG+M39d2vnZ+8TV1Lc4Pjvs98aXDUHH+/hIm5Gj7NaRbO32gesG0rMwwMVz3rBcL2NM/mp/yHHMc3rm1jaGueeFLYnMn+Qix6E+k4mToLbusL9MWyNx9wDvAFuB/yGRUnk3MADoCvy/Nu5foVAcIa5ketfuAyHHlJuWSJxe3XMc3/O6NX43dhD3vVTNvOsHZkxrS218bFCc56OhJcptyze1riKPKzdTMFP3EYvr7N4fZG9zhN/+b6Jfjc+tOabq2FeNjahMJnUtpYDWubGntsZ0ya/+XE1tY4TpI/ryk6vOIZahiXEoGmNXNM7uAyHqmhPpw3dfeQ4An9W1UJDjoSjPy+yXP+S16j2snDqUiUs2WPYzqGcB+X4Pd65sjYY8dMP5LP/eRcR1aabN5XhdWaXIHS9OgonPcScuM6RJ2hbJQzGdFzZZ0y4f/+sn3H7pmZbtYrqziqW9dk6XEo9bswiMeNxamtpltgiRoY+dzVyiWaZxZqvaKTP8/doSZFAcmkzpvcX5PjQNrq8oZURZD+K65Iv6FgJeF32L88j1Wcc8I02yJRKjthHzWV3XHCEci3PP1WU0hWNtek6rnoPOtMmJk1J+KoQoB34BXA7ESUTlXgF+LqX8su2neGJyqIieitQp2ooO/OCZd5k5sswx5SbH62L3gbDje5/UNlOc72PR+HKicekoY316YQABaZ+fPqKv6cBBsth++SZmjepvmSiXFgaIS8m2PU3c/fw/mDmyjBmrtjB3zADHVJ3TCvzmsYz0olO7+AjHdKUoeYJipLbWNoa5Yf5bljSe4jwf85ItMNLr2SQ3/fHvlgheTNepa4pYGjIbdZV1zZGsVCv/fcW7ph0bkZGWSNzZ/jpAiwE18Tk2ZG4JYP0b+90a44Z8zRTOaYnEGTfka/hsKeV+t+Y4Bvpt2+kS5r/xT0ZX9CQHF5G4zvw3/sm9tho2XZds/aqRSUtT7GD8BfQ7xWoHMkOT8RW25uEiw/cVtu+brWqnUgU+PmRqAv95XQtn9chj6vA+pqqpUdbQLTeGW2v9nGOa5PgL8LgFExZvsIzBTkrWnixbD2USPOtgGTfHnDb3iZNS1gC3tsO5KBSKdiQSS9S85XpdPHHLYD6va+HhNduobQqzsLICr1tjwyd1LKysMCerqfVutU1hnp40hPqWCNNHnGXZZv64crwuwc/++33mjB5gGcC/VtQa3Ustfj6tIGAO4qZ6myb4enL7goCH4jwfPbr4LXLYNcmaike/O8iyMv3Imo/5+chz2bU/xLzrBvKj596zTFQLA55kk+dEimZbRU8Ux55YTDdTIp+4ZTABj8aXDSFmjiwj1+ti9ssfptnfwsoKZr/8ocV+vv/UJp6ZPIQH3rKKJzyxbgdTh/dh4drtB7VjA2PyDYmIs8cl6FOSy5O3DGZfc8SM+k365hlpdUnHAzXxOTYIAY+OHUR9czRFct/j2GbCUH9Mdc7sSHBUsVw5dahlO02DaZecSX1KS4Bpl5yJZpsX720Kmw6csb9JSzfy/G3DLLVpmSKK9sieO6nGac+2sPtcIqmQbHdG7Zo/2dbspY4HR9J/VGHFrQkWVlbw8JqPGV3Rk1O6+OmW66U5HEUCwWjcYjPTlm9i5sgyVlXtZEFlBbctq2Lq8D5pjv+kpRuZNap/2hj85C2Dqd7VaOk7G4/LjKIoqaieg8602YlTKBQdE5/Dau6CceWU5PuY99rHrPukjkWVFZzS1cczk4bwRUNCsOGBV7eyeWcDQFr0AloH86W3Dua16j3UNkaYObKMolwvp3b1IyFj8fOCceXMHFnGP/c08/MXP6C2KcyCpBCKLiV3XdGPfc0R59Qa2/erbYzwRUOip9fiCRfw5C2DaQzFKM73UZLnY1ttEw/+ZWtaelBqJEKlmnVcYjGdj3Y3WhYPFowr59kNO2kIRrjn6jJeq95DQcBrSU/rGnBbVnshYT8C51Sxr3ULMHV4H/66dbe5n217mtjVEHRcpe6W6+VPt1+MLhORi/3BmEWAYv64cpa9/VlaitzRxsmW1cTn2OASgmhMtzhnD14/ME2hNJqhxcAztkhXOEO6YsSWrqghCDo4hZqttCYUdbaDkM0ONCEcsy7sEbZYXFL1aR1PTRqCLhNKrK9X7+Ky/qdZtoskFZJTF07uf2Urv709Q2gMAAAgAElEQVTx/LS/YZ7PqmLpc9uO6TAeHE7/UUU6wWicHl183H5pX6alOORzRg/gN3/5mOkjzuL6ilKeraoBEjZTku8zx9fFEy7E7UovqUhd7Ep9bX8wytwxA8jzuS0LANlkBxyLaG1nnA+0yfKFEH88yM/vhRDzhBBjhRD+Q+9NoVC0J1Kmr+betnwTkbhk1KDTKc7z8ds1H/NFQ4hte5r40XPvMWVplenAlRYG2H0gfMian807G5iytIoxC98mHNP55Z8+YM7oAUwf0Tdthe625Zv4555mJi7ZwOadDeZrPxtZhhCCO1duMVPbUiktDNA14GHW6mpueGw9s1ZXc9cV/dCTK8cTl2zks7oWmsIxXCLRx2jSkxsZXdEzfZUw2ddG9a/p2OxpCqelM962fBO3XdKHm4f15pPaZi4rK2HUoNOZuGQDl857k4lLNlDbGDH7FhqUFgbQM6SKhaI6s1ZXc/XA03n8r5+wbU8Ts1ZXM++1j5kz2tpncf64cua++hFf7g/x/ac2EdVlWp+uacs3ceV5px7TSFwmWw54XY73kkpTa190MMVFIGEHP3z2PezainqmGjFbpMvjEo7Xze3KzimM2sYwl+a8P6d0z9sv7WsZZ2+/tG9aGqdbE5T3KmLs4+sZPnctYx9fT3mvIuy+lFtLKCRPWVrFDY+tZ8rSKmqbwrht90ZDMLFwN2HxO1w6700mLE5kYjQEW/uPOY0Hh9N/VJFOF7+LSEw3HThoHRdHV/RkajLSNqhnAdD6HB7Us4DXqvewrzlCLC4dbaslEk97bU9jmFBUTyu3OFSvOTj6PVw763ygrcsXl5AQL5kAVAJXJP+dAHwHmAgsA/4hhCht47EUCsVhEMkg+qBLyYxVW3joxvOZceU5TFu+iYfXbEubsM4dMwAppSmnnkppYYC9TZG01z6ra+G16j088OpWenYLOB7faYVOSji1q5+a+qCZ2pZ6Lk4pcneu3GI6ksZ+Z6zaQly29jHqU+zcyDkSi2dMNdvbHDabiifSMTv2IH6iEs1gv5oQzFi1hYfXbOPuK89Jc8ymLKvinqvLTPu5rKyEJ28ZnFFkwaUJ5l03kLqmCNeWn27aX21TmBc3f8GSiYN5887hPD1pCI++vo3XqvdQku8zFzKc9tmre65jKt3RIpMtx3SpmtcfA7Lt15atM+USgrlj0sdj+3aZ7M8uqiMEPHTD+Zb9PXTD+Wk2GnaY0E9bvimtyXg0Lh23i8ZtY2UyzXTxhAtZMXkIiydcyKNjB6WlUwYjcUdnNJjiCGQaD2Jx1YbgSGmJ6ASjzn9Xox9nYyjG/WMGsHLqUJ68ZTAr3vmMqcP7mMJioWg87Xk977qBlBb6057h553ehZ7dnNPUdV0/6HM3VejqrRmX8MK0i9u1tjfTGHoo5/J409Z0yrHAUyRq4l6UUupCCA24FpiXfD8KPA/8GhjfxuMpFIosMSYM9vQDlyYYdkYRB4JRM22npj7IA6+2pr2U5Pv4j2ffY/bo8wCH+ofKCnp0SaSxGTUgpd0CvLJlF4vGV1AQ8Fj6aqUe32mFThMCTSR+37yzwTyX1hRNya3fOIPRFT1ZuHa7GcUzCuSNB0rCIZRmH6Od+5xT4rxuV8ZUs5ZwnMo/tIpiKCGI48PBemAZNrs/ec1TMRYFnrxlMMFonFhcctMf38ko8PNZXYspVLJgXDkAT6zbwfLvXUQkFmfnvhZ6FeXQEIpS2xhh2a2DKcz18tyUoRnP0ecS2OezR5NMthyN6ap5/TEg01hr/zsLQVrt5ZzR6TVioZjOzrpmnpk8xFRCXbetNk1uP5NwiF1Ux+PS6BpwW1QsuwbcuG3bZVIrtKtiZrudllRITk33/O2N56elZ+oZa/Gs38Hpu9q/gyJ7Yrrkq/3Oz0hj8Tbf77aIm8wZPYCCHA+Pjh1EUyhGwOsiFNWZ/Z3z8Lg0GoJR/vB/nzDpm33MZ3i3XC8L1243U8ztx7usrIS9zRGmLK066HP3aPZw7ayp5221/geBOVLKF6SUOoCUUpdSrgLmAA9JKTeQcOD+rY3HUigUh4FHS1/NnTN6AL/80wdUDv06j7y+zZK6aKRF/ui59/h4TxPF+V5i8US6zs9f/IBZo/rzxo+Hs/x7FxHwaOw5kKiVu+Gx9cx88X0OBKNcft4pzFpdzeyXP6KhJcrCygrL8ReNr6BbridthflAKFGYn7q916XRPc9HJK7zyz+1pvf8+PJ+DOpZYHnQzBk9gIVrt5sOWiyZZuQUYVxYWUFhwGPm2KdSWhhgx97mjKtxui5VlO4YoOuSuJTMH1duuXYLxpWbzj7AnmTz2VSMtLOFa7fjc7v4/lObMkZ4544ZwMNrtgGt6ZrTR/Tl+5eciSZgb7Ie9D+efY9uuR7uuqIfdz//D374zLvEdJ1fJFOHU/f54PUD0TRB4BjW6WSyZa/bZU58Ti/MSUqHKweuvQm4tTRbnT+uPM0GpEwsEMwcWcaKyUOYObKMJ9btwK6kn+93cc7pBdz42Hq+NXctNz62nnNOLyDfb81i8LoFCxzuEa+tniwa05m4ZCMTl2zghsfWM3HJBiYu2ZgWKfS5NUc7sqtnujNEFO1pkrpMKCSnjqc/eOZd7MOmS3M+bmpKco7X+W9sb4CuyB63Jvjbx3tYYHtOzxk9gFVVOx0zYGas2kJhjhdNCO5+/h/862/+yg+ffReXJpj98kfMWl3N9BFn4fNo9CnOpSjXy+4DIRqCESJxyZ3PvZc2Zv7s6jLTgTOOc6yjYAcbQzsybY3EDQS2Z3hvO9A/+Xs1UNiWAwkhhgNvOLy1X0pZkLJdITCXRJpnAHgb+KGU8h9tOb5C0dlwuwTd87wsmTjYbLptiJZU72pk5sgyR1W+OaMH8MS6Hfz0qjIzIpWoO0tEK2aN6s/Xi3LS6hN+8ExCft0QNPnhs+9SnOdj1qj+9Oqei8clWLpuB9dWlLJk4uBk4bogLiWagF/86QNqGyM8eP35BLyuNMXM2sYIm3c2MGPVFmaN6k9JFx8FAQ+zRvU31TSNVLFd+4MZI4y/fvlDfnXtAEdFtEWVFfzsv9+3/B2N1Tgl137sqGuOMPbxvzPsjCJTbESXkO9LyKgvqqxgyrIqFq7d7qh+t6shxHUXlFqK7lMjvOecko9LE3y+r4W7rzzb7MO1eWcDvbvnEo3HAWHut6Y+SH1L1Pz/zJFl5u+pwj4FOV50qbP4/z7h5ovPyEp1rT3IVt1PcXQIx3QefX2bRcDj0de3cZ9N6t/v0dJEJOaPK8fvsToiwYjOI2s+TlPjvffb50Ku9biP2I77yOvbuO8a63FjGSJdMZkeOXvw+oFpzcPtAi0elzDVCVOzMzy2mr1YXM/Q786eAin5/c0V7GoIm5HCUwt8iBQ5q+ZwnGVvf5bWY+8H/9qXghwUR4DfozF2SC9+9edqZo4soyTfR57PTSSu8/Nvn4tLw1EkKhpPT7s1BHoiMZ09B4Lk+r2WFgMLKit4ecuXlnG4IOChZ2GASIbI7rGMgnXWMbStTtxXwBjgLw7vXQfsTv7eBahv47EMpgOpXVljxi8iEaN/CegN3JE85k+AN4QQ5yfbISgUJwXhmI7P40JKGLPwbct7NfVBTuniZ/POBl7c3Np81ufWaInEmDnyXBpDUccHcI7XlVFBMsfrskgOpzp/S28dzBXnnUo0JvnN6x+mKQXOu24gq6pqKMjxMHHJhrTVv5kjW1fr+hTnclrXRLqS1+3i0bGDLKliqUpWRoSxtDDAzJEJRcN7vx1PaybtdbtwaVBrK5Q3VuOUXPuxQ9d1Zo4s47SuCU2s/cnai4Anh5guaQzHmDtmAB6XxmkFfuaPKycS0820ndEVpfzh/z7h3m+fa0nd2byzgVmrq3n+tmHsaQxbnD9j8cLjEjSHdfYHQ+bnBvUsINfnNv9v1IukEtclfo/Gum11LPrbp4wb0outuxuzcvLboopmfLaL382zU4biEqBpmkqbPIbEdMlr1XvSJrw/u7rM8v+mcJw3P9rDU5OGIKVECMGLm2q4ZtDpFOW1bpep6bY97TIaz+642faxi+k6+QG3RSUypseJ25wulxD43cKSnul3i7T9ed0a911Txr6UFgj3XVOWlu6Z69PY14wl7XJhZQU5Pi1lXy7WfVJnKiUa3+FO99kojoymcJyWcJzaxkTEKxLT+aSxmYVrt/PQjeezbXezc5qwcFakjMR0Pqtrod8peVy/aL3lWXnbsipmjUrEdVKfyc9OGcqO2qaMZQ/Z0lZlSaf5QGcYQ9vqxD0EPCiEOA1YCewBSkg4cFcB/57c7pvA5jYey+BDKeX6DO9dA3wDuFRK+QaAEOJtYAdwFwkHUKE4KfC5NPY2hahrSm9kXFqYkEpPVfdLjWS4tRglXXyODWd1KWlojmasd3Oa4Bp1Sg0tMWa++L7Z2Dt1kP/Rc++x9JbB1GVwEAsCHvM4X+4P0RyJ069HvqMD5bSqZvS/S3042HPs9aQQhNNqnBHds59XR8+Z72zoumRvst/azcN6c9vyTRTnJWzxhsfWW67nr/78IbVNYdPBNyLFupTcPKy3me5o7yMX12VaJHnGqi0s/95FPLluB8PP7kFcl6aNTx3eh8/rWsz/G2m8Tm005o8rZ8o3e+HWRFZOflsivCo63DHIVK9lb2rtc2tc0LsbYx9fbxlT7emK2Tbdbmstnt1EXEJQ3xzlzpVVlvMrDFijEaFkeqb9uE7n59QXz67c2hTSHZUnV0weQtdkhltnjZR0ZAKexMJlpsbyD6/ZlpbpsLCygq/2hxztzlD2XTCunOI8n+X9mvogvbvnmp8zrp+U0ix7SLXPRZUVWV/b9hoHj2bN3dGiTU6clPK3Qohm4OfA1Slv1QCTpJR/SP7/d0DQ/vmjwDXAl4YDlzzH/UKIPwGjUE6c4iQilFQaK87zpTXDnnfdQJrDUe664hxTzhla0yKeuGUwQjg3nJ39nfNYVbWT340tN+uNjEE3puvsy+Dg7WkMm4qVmRy9umTD5IMVWhuqmZOe3Mjz04YhEGkrZ8aq2vPThtESjrNjb7Ml5dJoBJ7pc06rcceiT40ikUo5ZWmVxdGf/Z3z0mzRiM4uXLuds0ryWDF5CA3BKGeW5PJZXQt3P/+PtHTHrgEPc1/9iHuuLnO0v4aWKOW9ilj70W4qh/Zm+fcuIhqXSKlz18p/mBMNI40zFNXTJtvTlm/i6UlD0JICLIdy8g83wpu64iyEOOzocGfshdTR0TI0tbb/WXVdsvitHZbshsVv7UhLu9SldMyCSGtFkKx7th/XYzuwlFj6IMZ1ycqNn9OrqLdlu2z72GUrbBKO6e22v84aKenIxOI6NfUhx2v0bPIa3f/KVp6ZPIQvkotXXQJuZv73+2lO14Jx5fz8xQ8SUbflm5g1qj8Tl7QmzZUWBsjxunh2ylCicR23Jsj1uYjEJbVNYUuKZSKd1p/VtdV1yVcHQidtlkybm31LKX8vhPgDUAqcCuwCaqRsHW2klJ+29TgpLBdCdAcagFeBu6WUnyffOxd43+EzHwA3CSHypJRN7XguCkWHJaZLM4LRLTdRO9Y9z0vA68atCTyuhNMy77qBlpqgmvogTaEYXTM4Wn6Pi4kX92b5+s/MejeBxO0S3PHUu9x1RT9HNcu4rptqkYZDZneI6pojjnV6vxubELSYObKM+1/Zyt1Xnk1NfauSZHGej+kj+tK7ey45PhfdcxMCDiX5fvRcSa7PzaNjBxHwuojFJTUNQXbsbebhNdtMx85Ytcu0GqdWgo8NhkqY4egP6lnAqQXO7SrOKslj3vUD+fXLH/Ja9R4zElaY67WkUE5ZWgXAislDeK16DzNHnutof18dCHFG9xxO6+rnuynRkgWVFQzuVcCLm79g2a2DkQj2NoY5ozjgONmO6xJdyqyc/MNRRbOvOK+cOvSwosMqcnd0CGVoav2Qram10GDaJWdSn5JeOO2SMxE2bQ6fW+OnV52dVptmj9jpSSXe1LTGgNeFjs3ZcwuuHmjNuJg/rhyPTQBF1zM4jzbnLJMqq13YJJ7l/jwZ9md3RhXtS1SX+NyaOc5OHd7HvE4IuO+ac/mfLV8gJayqquGa809Dl/CbGwbS0BJJLlQkoq1N4ZjZY7amPkiv7jmWqNui8RU0hWMWpUtDeM14rhrZFI/fdAEFgUM/V43xrDkcO2mzZNrsxAEkHbadyZ+jxX4SbQveBA4Ag4CfAm8LIQZJKfcA3YBPHT67L/lvIaCcOMVJgd+tcdcV/QhFE6kvRupXqnM1d8wA7n8lEaEy0g1rmxLF5bGUdDKD0sIAp3b1U9cc4bZLzuTTvc0seOOfXHneqZxRnMuMK8/GWL9ZPOFCWiJxuud5iemScb9POFtG7ZHdUZs/rpxHX99mKXw+sziPz/e1cN9LH1iakBtO4I69zY4pbU5OmdME1vjO2azaqZXgY4MR8TSusT2V0cBQr9te28y0S86kIOBlRFkPIjGdU7r6uaysxFIrlLpPt0swf1y5RWDCsMuZI89l4pLWeo7iPB97G8PcfHFv9jVH2bG3xUwPe+enIxxTkQIejbiEReMTSqjZfN9sIrz2qF2mqDVAbWM4zT5VXefRwZPS1NrAyQnREAQd0gs17M6Pc/Pw56YMtWynS3j3s31cWnYqupRoQvB69S7+7dxTLdtFY8593ezpj8Yzwym1zrKdR0u7f5wEWgJZ7s/t0hwjiqntA9QCRPvjd2t0DXi4rKwkrQZzYWUFv3tjG/dcXYbPI7h5WC+mLLOm2Qa8Lua/8U8mXtyb+1/Zau7XSCV+4pbBNIViFOR4yPO7GPXourSI36xR/RnYs2vG5+rBMgeM8SxT+5iTIUumXbRZhRADhRDXCyFusv+0x/4BpJSbpZQ/llL+SUr5ppTyIRLNxXvQmiYpACfN74Pe4UKIyUKIjUKIjbW1te11ygrFUeVQdquTSIfM8brMuh576tedK7cwdXgfMz1t+oi+zLtuID6PRiweTzbttDbxlECez834P/ydh9dsY9Sg05n54vt8a+5afvzcewD8/MUPmLhkAzk+F0IIpExE0QAeeHUroyt60sXv5plJQ3hh2jBmjepP14Cb6SPOorQwYApQ6FKnIMdDbVOYQT0LWDzhQp68ZTBdAx6WTLyQh9dsc/xeRtNugFhM58uGIF/uD6ZNYGesav3+2azaKbn2tnMouy3K9ZrX+IlbBtOnOM+xVcSCceV80RBk5ovvE4zEuWPEmcxaXc2YhW9z42PruWPEWVxWVmJub8hmzxk9gCX/9wn5fjdP3jKYlVOHmlLvNw/rTTxFyW9QzwJ+fHk/Zr74Pv/44gBTl1WR43VRnOdj0fgKYnHndLG4LvG5BL/934/ZVtt00FYURoQ39btlivDao3aZ2ibc/tRmrp3/Flt3N1qO3Vl7IXUEDma3Hrfm2JzbY3NWMqUrRm32EcnQ2Dpia2zdNaBR0bs7Yx9fz/C5axn7+Hoqenena8B63HgyPXPR+ApWTB7CovEVFOf5iNvSM+PSOYXe3vOwKUUp8vUffYvFEy5k2duf0Ry22lFMOn9fuypmOBY3I5lG64X7X9lKOMUu9zaHefAv1m0e/MtWc5xXOHMwu43pktkvf8jdV56T9gyduqyKGVeeg88tiMQkwWicmSPLGNSzwLyO9c1R7v32uXTP85qCYIbt3/HUZm7+4ztoIuHMN4Wcx56CHA/BSNzxuWo47tfOf4uL57xhjmmxWKIxeEskEYFzGgdPliyZNkXihBAFwJ8BYznHmNGk3qFPtuUYB0NKuUkI8TFwYfKlfSSicXaM9gaOCplSyseAxwAuuOAC1fhJ0Sk4lN1Gko28jehDpjo0QzCkpj7I14tyaAxFaWiJUpjjQQgsSmVxPY5bE9Qmmyw7CZTcubK1VmlvY5iJK1tTeIzIl5E2sWLyELrlesn3e/j79joG9ymyHM+lSdya4MHrz8fv0SxRxEWVFRTnezN+r2AkTl1ziF0NYaYuq2LedQMzfv+TZdWuI5DNeBuO6vw4WcO5eMKFjjUTTeGYqZJ258otLL11sMUOb0sKI/zimnOJ6hDXde68/Gxe+ccuynsVmcc6vTBAUZ6XmSPPBRLRjMvKShhd0ZM+xbns3BekOM9n2pkupRldWP69ixxtKqpL3DKhHFi9q/Ggka7DifDao3abdzbwxLodPDtlKDFdZ/ueZu5/ZasZtbZH2VRd55FzMLsNRuOO6ZQPf9eaThnPUPtld/IzCZbYBUH2B3VT5t/Yl2H3ef7W7bwu54iYXSUypjs7j/aWAC5NUJhjnT4W5rjTbDYaz1DrZvMK3S7NMZKZGomL67qjYqddOVNh5WB2a6iq3jb8TMfr1BSK0RwmLWPBaFWU43Wxa3+ILgG32Zg+rkuawzGmDu/DwrXbzfq4SFx3tOl8vyfj+OOUOfDgX7byg389y6ybNhZ9jedDUa6X0woCnNIlu5q6zk5bI3H/BRQB/0LCgbsWuBRYDnwCDG7j/rMhNfr2AYm6ODtlwOeqHk5xMmHULRirVC2RuLlSZWCkmBm/x3XJ95/aTBe/m/qWKPuao0xY/A6XznuTCYvfMaWiG1oO7RhOHd7HUYxi6vA+Zp3Rk+t28O/PvEttY4jyXkVU/v4d/vU3b3LpvDf519+8yad7E0XS+1oipgNn7GvKsip+elVZxu/10VeNbNl5wFQ9M5xZ+3YtkfhJs2rXGdjbHGbS0tYH98NrtrGwssKc5P3ouffwujUWv7XDtN2a+iBx20TYuOb7mqOMfXw9lzzwJnNf/YhvnV3CrNXVXPLAm9z0x3eoa4rw6//5kG/e/wbfvH8tXrfgjhFnMWt1Nf/6m78y88X3+fHl/cwat3hKNOVgTY+NiWo2ka5sI7xOUbsf/ls/SvJ86DpMXLLBdOCcjn04UT9F9rhEazrlDY+tZ8rSKmqbwo791RztxdZfzetyjuylO13ZCYzoGSJidqEUTTifn/175Ps1Rp5fysQlG7h03ptMXLKBkeeXku+3np8rw/1ht28NHL9v6t503VmxU/lwR45Ri7inMex4nXK8rrQ03NRneEskTl1zhFuWbOTTvc182RDkpj++w7cffYtZq6v58eX9KM7zkeN1OUbL5oxOpJ5nGn+cMgdGV/Q0Ww2l7tPI3sn1uU8aBw7a7sRdTsKRMyT/a6SUa6WUNwH/C/ygjfs/KEKIC4CzgL8nX3oJOF0I8a2UbboA306+p1CcNBiy0kYUI8/nZmFlRdqDcuHa7eaAGoomBk0hBIU53oypYn1Kcll662CK8nxmypqB4RhmcvDOPiWfFZOH0MXvZtzQ3vxy1LnEdYknpTHzoJ4FLBpfwdeLcijO89GnONdxX5JEb64F48rTHg5GTzvjc04PkUWVFQzs2bVNdRW6LqltDPNFfQu1jeGDps4pDo1hgwabdzYQ13XH1MeFa7cDiWu5tyli2U9pYQCvSzPrOCAxAbBPSqYuq2LmyHNZPOFCVkweQigq06IbM1Ztwe/RWHbrRZQW5pjv+T0aC2z31ILKCvwezUyRMyJd7WEnqVG7t2ZcwgvTLqZfj3zqg1F27G12nIilrnJn+vzJMuE5WhhjrX0Msvd1cwnBb28837Ldb288P62/miagON/HrFH9WTF5CLNG9U8497b9HWwRIZVYlhGxbL9HU8g5AtgUsnpUXi3RPDx1fw9ePxCv7fxShWFS0ylDsdb9ZXJY7Ys3isNAJOp2jTRz+zhmH4uhtcfsg9cP5MySXEryE8I1XyvKcVy0nT6iLw3BqCVa9vqPvmWO4x63ljb+GGNlXEoWT7iQQT0LzPeKbKJVxj7/eufwk3I8a6uwyanAJ1LKuBAiBOSnvPc88Ewb928ihFhOot/bJhLKlININPL+AngkudlLwNvAMiHEnbQ2+xbA/e11LgpFZ0BK2PbVfp6ZPISYnkgT27prP4snXIjXrbFrfwgpJXdfeTYNwShPrNvB6IqeZkF+JMODPxKXjP/D3830isUTL+SWb5yBAFoicbrlerjvpWqmj+ibUXQhtd/XvOsG8uTbn/KzZGpEqlDJ3DEDuOuKfqaqpX1fsbjku4//ncvKSsw0zG17msx0j1QVTGPAnzWqP32Kcwl43W0WJlHF9u2PU2Pi373xT+664mwaQzFK8n3cfeU5zH75QzbvbDDr4/ye1l5dxuSzJWKdhGRaWIjp0hSbcFJ8LM7z4dY0Kv/wd0sRfTCqU98UZEXyHnNrgu17DtAtx4PfrZn2UBjwtJudOKmnRmLxrHstdcZeSB2dbCX8dSlxadYm2S5NpEXEdJlY1OrZLcdMLfe4SNsu15dYRLgtRXBiQWUFub70iJiRImyke66q2pmWniklSXGf1rTQJ9bt4F5bC4RsHSqvR1CQ67V834JcL16PLUKZhTBMJkVM+3dQZI+U0DXg5j/+7Szy/R6Wf+8iXJqgKRSlviXKgVDM8W9ekONBCA/ffbx1HrCwssLSG85Qu+xdnEvNvhYG9Swwo2UzR5axqmon91xdRiRZ32Y8i52eqakCbMX5vrSU8lmrq09acaa2OnFfAYaL/BkwFFib/P+Zbdy3nfeB7wJ3ADnJYz8P3Cul3AsgpdSFECOBB4D5gJ+EU3eJlPJoKmcqFB2OgFfjgt7duTHFYVpYWcHKjZ8z+oKeuDXBv694zzLpfWLdDhaMK8elCdw4q1N+urfZfM1Q7kuttVhUWcEjYwexqyGU1p9uYWVCZtiomdu8s4EfPfces79zHrv3h9N6b8V1yd3P/8NUtTRev6yshHuuLuNAKMbiCRfSs1uAwlw3oYjOrNXVluhbquqZ8RA4rWsAt7vtuk5K7a/9CXhdaUp1Ey/ujdsl6FOSh5QSv1vjvmv68/1LwuT73cx++UNqGyPJlhc5eF0aT67bQeXQ3hYbztTaItWmnRQfp4/oa6bzGhHdJ9btIODV6JYXsCxKLKyswO/VCEd0Xph2MUW53pLKIU0AACAASURBVIy1Hfdd0x8p5SGVTg+m0KbrEiEE91x9DqFo3CL7nW2vJUXbyPNrjB3ai2gs4cQIIRg7tBd5fntLALj9qc1p9mdXiZQSvmwIpdWwndE917JdMCJ5ZM3HFqfrkTUf84tr+lOQ07qdz61x+6V909Qk7S0L/F6NOy7ta20PM64cv9emJpmlQ9USkUxcvCFtu2enDLWcn9ul8eD1A9NaKqTWxBkZF2nn5mkXfb6TkoBXwyWgoSWWNobtrGvmfz/ak7ZIMGf0AH7152p+ctU5luf41GVVZm84QxDKXr/4xLodTB9xFj634PuX9GXc71udQGNRK3WsNBxBj0vj4e8Ooqa+hafWf8qi8RVmSuXJnhLeVifu/0g4bquBpcC9QoheQAy4mXZMYZRS/hr4dRbb7QNuSf4oFCctoag068GgNXVsxeQh/OJPHzDpm314etIQM/XR6xbMHHkuDcEo1bsaKcn3pj00F1ZWMPO/W1sxOtW9TVlWxUM3nE9Rnhe/R+PpSUNoCsfwuTVmp/TzenTsIJpCMTwujdMLA8R1nX9/5j3mXjfA3J/HlehhU1MfNNMmTuvqRwhhfQCMv4DCgA8CWHq51TaF6dHFz3NGg1GXRkmer10cOFBqf0eDLj6PmUpmrN4X5/toDsWZsuwdiy22RGJmw3lI1ISVFgZYdutFXDOolC4B68RvVdXOtEmJYdPGhKEk38eyWy9KpDxKSUskTu/urem8m3c28OLmL/j+JX05EIxlvMdEMiVO1yWRWNzSjxHg5mG9uX7R24eMzB0s2gs4ts1Y/NYn/PDf+mXVa0nRdqIx2N8STXOScj1WwYZU5VODmvpgWoQt26bbkbhObaM1jbi2MZKmYhmK6Y4tBuz7kzr4bX3n/F4X0lZ3pmmkRX3njB6AZhtWo3GdYWcUMelfzjAjlI//9ROitvOLS538gNsiahXT48RTDiwllHTxmgIaLk3gcSVeVxwZkVhCiCTTGHZx32L8Hs2ySGBkudz6jTOYtbraInTy9aJEbzgnxegZq7bw9KQhPLJmGyPKepiLrca42xyOsWt/os64OM/n2DpozugBvPNpA7d84wzV6idJW524XwCnJX+fS0Lk5AYSkbKXSETNFEdAr7v/fMhtPp199TE4E0VnJZpBptpQpLrn6jJcGjSFEqu3duWvuWMGmAN4Ua6XHl381NS3mFLCkDk9rSjPSywuORCKke9zU5Dj4bqFb1sieMFInLuf/4clsjYv2dD2b3cNJxTVaQpb0yGnLK1i8YQLzdQ343iTlm5kxeQhuDVBjy4+np82jGhMb1OEIxuU2l/7Ux+Mcv8rH/HTq8rYfSBEJK6jCcFvbRGHh9d8zE+uOieDjetMWVrFs5OH8Mjr2yyfW/1uDSsmDyGqy8T1b4pQnO91tH8jhWfpLYMt13lEWQ++/9SmjOqURmrlnsYQTSFrg9s5owegiXQp90wR3INFewHHthnPThmasbi/rTavSCecwUmyR9icUoWdhEOyVbHMtq+bnuX+QjHdMXLm5Ow5pV3eZ0u7zPG6qBz69bQm4zle6/ioIdL6QInk66nsaYympY6e2kVlPBwpsaSapJNtBKNxJizewPxx5ayq2unYc9MYb2aOLGPW6mp8bs1MLU/dp+GoxXSdSf9yhjmvWFO9m1GDTmfGqi0U5/mYPqIvXyvK4cEbzqe+JcwdT7+bNrbNGtUfTdNUpkuSNi1HSym3Syn/lvw9KqX8kZSyVErZTUo5VkpZ1z6nqVAoDpdMymCpr8eSK76jK3o6tgooCHiZsrSKMQvfNnv5pBZAZ1KGlDIRFbniob9xw2Pr2dsYpjivddBNjeAN6lnAzcN6M+73f+fSeW9yw2Pr+XxfkLmvfoRb01hUaRUt+VpRjuNDZ9f+EP+sbebuVVuoa4pwatfAQZX+MvWgORzBCaX21/5EYnFqGyPoUvKj595jytIqPC7BzcN6M2t1NTc8tp5Zq6u5eVhvfMm6s1RKCwN8tT9ETX1C6v+16j0W1cBFf/uUuC7ZUdvMTX98h//6nw/56VXOrTKMHoK/fvlDFo1vFTAxiutdGdT8XJpA0+A789ex+0DItH1jInJKV3/WEdyDRXszvSelzOjAtdXmFelkWyOWSTjEfqm8Lme79rjS0zOdVSet+8tWACWjs2dXsdQEEy+23o8TL+6dZnORDM5tJKan7S8ckxYl5HDMasORmLOYin1fiuzRRGZFUleyfcu05Zu4+8pzHIXDILlom+tl/rhymiMxapvCxHVpbm+kVhqKwBOXbODzfYmsCCNiZ0TdZr74PiPmvUnlH/6Ox+WyzBmMY/XunquerymoZGKF4gTFowlH2WZX8vVcr4toLPHQzhRRM5rBGqvFd195NpqA2d85jzfvHI7fo6Wpj/1ubDmzX/7Q+rBdvonpI/qa+049XqbUizsvP5twTKcgx8vKqUNNZcJdDUHHh05dc4QZqxIO6aQnN1LXbE0zspMpwnGoz6Wi1P7aH6/bxfQRfZn98oc8OnYQiydciC4Tkzi7MyQlFufKEMp56d0vTZEJJ1vRNGEql27e2YDEefJq9FB8rXoPRUmBhhWTh9A12VtQCGdpdJDoutUZhMSEZubIMjQh0lTXMkVwvW4Xl5WVWBo1X1ZWgtftMiPB9u+XTd8l41yawzG+OhBSjlwbyOQkHUw4xFBhfGLdjjSnSxM4qljah5Vsm4JrGVQi7eNUxoU/W6QwnEFN0u5QZdsCIRtnL5phX/ZG6YrskRK+2h9yHMO+OhACEn/j/cEoM0eW8eadw5k1qr+ZPmls3zXg4dHXt+HWNOqbo8x++UNzsSLT8310RU/2NUcS6ZsO20xdVmWZMxjHyvG51PM1hbamUyoUig6KjiRgq28IeBNFBBt37CPf76E22R8mk+DDV/tDZgF5aj2bkbJzIBRjTfVus07ho68a0QSW1AtIDMpfS+bLF+f5KMprVZjK5EDuD0YZs/Bt8/hdA27+fcW7DDujiPnjytPqT5a9/Rk19UFK8n0ZoxqptFc9m1L7a1+Kcr2c1SOP719yJi5NM1NnjRVgYwJhXLtzTunC87cNoyUaJxaXxOJxpg7vw+yXP2Tixb3TRFKMhQwjilxTn5Bad7L/1B6K4ZjOxCUbgIQzNmf0AL5sCJFju8dyvC527gtyekFiUmTY+KCeBdx7TRn1zVG+2p+YIN13TRn3vVRNbVM4YwS3MOBh+oizzLoVo46vMOBB04SlBvRQkWDD5p2EB5Sq6pGTbY2Y161x5+X9qKlPXH+vK/F/ry39MZ5BxTJui4hlEhhJi7BJicetWfbncWtI6RwptH8Pe4sBn1ujON9qY8X53rTvke35xXRJcZ7Pkp65cO12i7OXbQN0RfboUvJf//Mh915TljZP+MVL1UDib7ynMcys1dVmiYVRUmE8m/c2hc3azO55Xm4a2oue3QI8PWlIxjrQgoDHFJHKNAfo3T3XvObGGNU9Vz1rU1FOnEJxgiJ1mP//2Tvz+CjKbO//nqrqLelAQkhQFgGRxahhCULQGRcYHR1RXgVEJTDiCEHcrq+Deu8Mow7jHRG5jriweB2QxYXFeXVw1JlB0BmRUSLKKAqRRRNECCGBdKfXquf9o7qKro10QrYO5/v55KN0V9fS9dTT5zznnN/Z9A3GF/VCBkREZQXPb/oGN4/ojcsG5WPhxt2oqovqqlEWefIpRQCAV6YXY+6GL3XHTIsuvDx9pNqHKxFlOHgsrMsH2/3YAsAfbx0OgTEcCUSw4rYR+La6Xm+ibN5ei4hpkbzHb7gAc8edj3Py/Zi74UvDD/6z75VjfFEvbNlbDb9HSqku7VTq2aiuqGUJxWRkeV16LRlwYgV3ztgClK4s09PLqoNRhOMyBAA1wWhCjCSC//zZoESlDcPyaSMQisZxJBBFhlvEU3/djcnFZ+kO3gsf7LUsDGg1cZohy5PGqSZu8sBVA/WG4hocwLIPT8iy98xRm+I+cNVAhKKywSmdP6EQi0uGQRAExzFUE4rZCg9o9XNaJDgaVw1zSWA4eCxkOy59blFXc604GtIlwU9Wk0c0jOJQI2aW5ueJiHIy0bhiEefgKapYuhJNwc2LFOa0Szjsb42NKmYq1yEJDHePHoA7VifVp00usjhnXkmwXXAz1+yparMF+nPkFtV/J2+nZZZYrpXm3SYjJJrUP/rmTsy8rB+yJVVUyty+ZcPnBzBvvDofAsDccefj7LxM7K0K4jdvfImqQCQx7hjAmK4ofc+Y/uibl4llt16IhRvLDdE7rc3FkilFOHw8YvtbnOER007ApLVtA3LiCKKDIonMIiutRTKqAhE8fsMFKHnxY7yx/QBm/3QQvC4Br84oRk19DDkZLvz9y4Pof0Zn+FyibWSNcxgM0mdvGYrFJUVYuHG37WruvLe/wgNXDcKxUAy/NLUd+OOtw3Hb8m2W80w+nksUUPLiVqybOcpWkS0304154wsRlRUsmVIERTH2nzGj1bOlGsXQoN5wLUt1MIqaYAw5SU1dNbQVXM0YjMoyyg8FsXBjuW5IPPHOLozok428rJ6G6NWSkiLk+t14+I0vdWPinp+co0eRvYmifDkhSiIIDH+4aQg453h+0x7c9qM+ukGa5/fg+mE98O8Dx/HtkTpcNugMHAlE9IWSe8cM0PvWvTB1OLp18qA+KuvtPrRrmb1uB16bUYwenZ0dp4YixlokuKFxqSgch45HThrZJFXVpiEwWIRx7GrdFK6KPZkd+Vy/tU+cXWTKXJsWisl6WqO23RPv7MLTNw0xbJdqWmOWT8DdYwZYxEOyfEanKxJXdAdO29cdq8ssTmYwKmPVR98a+ue98MFe3DX6HOQmbccYQzSuGL6Xp24cDJYUAmQMyPUbe87l+t2WKCGROi6RYUlJEUpXlekLYy/+fDhmXXYOHrz6XHyXWGS9pbgP/uPVz/R5c9ryT7Bu5ig9MwFQazPXzhyFI3URPHPzUHT2uSzZO5pQ1KLJwxCOKXjgqkHonZOBMzp7sLikyDBfa1G3dPpNbQvbgJw4guigxGSOT/dXG5p9v/FppT4Rn9nZh6G9snFL8VmorAnpP4w9c7wQBYbRBWfAJTIoCgyrZDcW9cSMS9Xo28vTi7H1myp0zvQgLnPk+V349dgCCABem6G2FhAYQyASx/iiXgiE47j3VaPi1MxVZVj1i5F4cuJgdO+sKur99s8nDG3AmNoWkxVbRbYeOT4cDUbRLcuDQERGMCpD4TLCsTh6ZGdYJtHkerbGrJpRb7iWRVEUeF0CKo7W267O9shR03RWfbQPS/6xXzeY39h+AOGYgvkTC+GVRPx2w5eGe1S6qgyvTC9G/3w/HrhqILpne3E0eEIWXlNIVTjHHpNjOO1HfZDpkfC3Tw+oz5PMUfLiv7Bo8jD0zMmDwjm6+j3o1smLwePOg8AYwrETfeIEgSEQCdoa0nIDJT1OEWOXKZpRHYziqb8ZDfqn/rYLj11fiLwsj+24NUc2SVW1aSgcKP/hGF6eXgyFq3PtezsPok9uhmG7uMKx7ENjpGvZh/vwG1Okyy0K+K+fDbL0TXOL1nRFu7RGc4qh4JCKaJ7rjocUSEwxNK8/HorgeEhBJ++J7ZwUDc1CLqLAsGVvNdaUVRqOe+9PjLVOMVnRr1Xb131rPjc4hTGZY/22CkwYfpahofrUi4wN1YnUickcZ3T26PdbYAw19dFE+waGhRvLcc+Y/sjOcOP+KwdgwV936xE6c+14nt+D6kDUskhUVRfF9opazF6nthj45nBAj949deNgdPW7oSgMeX431pSOSqlvZnulLWwDcuIIooPiEhmG9ck1NPt+fvIwVNSEsGVvNbwuAf/1s3NtU7xq6+PwuE7UTGgRiIvOzkXJqN64ddkJyfRFJUXY8FmlblBrK26/u/581NbHDKvTK38xwvHHX1Y4JJHhztXb8cufDsTOg3W2kTnGmB7J0z4/e90OPDlxMF4vq0TJqN6W9J0sbxTZGdZJtCn1bNQbrmWROXBHItpljujOn1CIu1/ejqpARO8ZtL2iFi9t2Yc7L++v94wzGxCAeo8OHQ+jZFRviIxBVqCPk2SFVPOYm71uRyKNV8Lwvl1QfiiAbp08ifpLNw7VRTFr9TbDeOuR7YVbYoYx53XZO2MNNSu2ixjPn1CIQDiOrpknFPwURbGNBimKmr7nNG61yKZWZ0c0Hq9bQFHfrrjlha2GedHcJFt0iNiJlogdbJ2ataWjDNul2sQ71VREn0tATb1gaPy8qKQIPlfTat08DumU5vNzqptKrgEUBOCSgd0M7Qrs6g6J1PG6BfxwLGJonD1/QiEy3CIyXKKlbnbBxMF48Z97cdfo/nj2vXLDvu4Z09+S9p28SKTNv8nRu/vWfI7Vt4+0afptXWBIB9rCNqDhTxAdlJjMbRW/ZlzaD/MnFEISGM7M9tpKVHf1uzFr9afI9XtQF47jjM4evDqjGHeOPkftq5WkEnjHqjJMGH4WAHU1LhxT8OTEwejsc+GlLfsM+95/pN5W/Uxg0NUGqwIRPPnuLswddz42/fJSrL59JF7asg8AsOzWCx3l2RmA6ZecbXvNwUjzTaKNVQQkGgdPpJKpjbfV5r7vz74My6eNwJ8+PaCn/j24/oTq4/iiXoam3+b3gRN1lrNWf4rsDJeujAY4K6RqLQbUvlYMs9ftwMKN5eiUcHziCUcwz+/BkilFWDBxMKoDUcRljmDEWPvUNdNj246ioUJ9IdH7UFPG1JQAp/7xY8NquMxhew1apM9p3OZnqWl7CzfuRk0oBqLxhKP28vfhqHEMKA73yCyw6KQ6aW6S7dTEO2yqu1PAkZPpwvJpI/De/Zdi+bQRyMl0QYGpT1zM4TpiVrXL5xLKsa/NKMayWy/Ec7cMtRje4biCZxN9GrWx++x75Zbzc2zVkZQrqSgO3x11GGgy4aiiO1jAid//o8EYInHFoqT64j/34jfXngdZUTD7pwMNc1mfrvatfzSFX7voXWVNCFV1EUvkKnk7ReGoqovgQE09quoi7VpFty1sA4rEEUQHxakOQhIZnnhnF569ZSgAYMHEwXrNhWYgayujssKx4qP9tqvHybU0LknAn++6GArHSaMhCzeWY9HkYbgjaWV20eRhakTM74GscLw6oxiMqUX2daE4qoMRPDruPATCcVTWhMEVZyXBvIQypfma4wpHPK5Akk593aqptXREanjdagPjZR/uS4w7Y61k+eGAPu40AyE3021bQ6Tdk+Qxm+f3gHP1M9o4clJH06JU9VEZSuKZULfjWDBxMOIKx0Vn52JycW/DuF9cUoR8U5pbU9N3ASAUlQ0r2BrJK7wig+X6t1fU6tF0u3E7b3wh/u+az/Xn8+FrKZrcFFKtOXNSYbRLQ0wl/THVpuAi1JYaDAJkzuFiDO7E66nsz3x+nHOEY8YatgUTB1vULpVEn0ZzTfWvrykw/DsVVUzHczOrwhApo41brRm3Nia7JmoN7X73a4JR3P3KZ3h+8jDMn1AIgTHUhmJ6b0O73+XkjJ1knBw7bV5Lt/rztrANyIlLY/o89FaD2+x//JpWOBOiPeKU8hKLK8jLciMUk1FxVI0yuEUBD19XgEcTcue19erEKwrMthG4uZZGZAx+j4QpJ1ETBNQoW67fjScnDkZelgffVdcb1K3ckoD91fXwugRdTa1njg/Lp12I2voY5rzxBfL8HtvUoCfe2YUnEv1uzNe8tyqI4+E4euV40cl7arn2p2KMEw0TjakN6OeMtTbgNo+7mKxg2a0XomeOz7ZOsmeOD+tmjkJ1MKqn4z5yXQHKDweQneHCittG4PG3v3JssVEflTF/QiG65/jgEtTebtkZLrgFAV6XAK8k4O4x5+Cbw0HDYsjMVWUW5T+g6e0oGlJSVRSOI8Eo5m7YaTC4XtqyT98medyGYjL2HA5Y+j1RNLlppJpe6KTCaE4vdDuoTppr4pwMZ7M6pcvFcCgg445VxjT4rByjCZiqjL/CgftNKe33r/3cImySclsADnyw65BBAGXdtu/QN7dvw/siZZMmIwkMVxbkY9bl56AmaUx6XALgEDVeduuFev3b2XmZiMYVfPV9LYb07GxZoH1+8jBkZ7gwZ2wBntm4Gw9dfS7e+uKQYbFr4cbdhnNKnofSrf68LWwDcuIIooOS6RGwbNqFqDx6QrSkR44X67Z9h4evPQ/fHa03rKQumjwMT988BHWhGGQOPeVyYLesk0YpFpUUYePOgzi3e7btdsnRkCUlRXjpw324cURv/DzJ4QNUdavHb7gAD73+b8yfUGiQP684GtLPtbImhCfe2aXLHEsCQygmq722bKTikxU5l08bgYPHIqe8kke94VoOLZWse7bPdjzlZ3l0AyHDLeLWZZ9gxW0jbNOCX5tRDK9L1J2bN++6GMGIUR3w+cnDkJfltqijLS4pQle/OnZD0TgOHDtRtL/lwctx58vb8cr0kTgWilsiEo+//XWzNCFOlqt++faR+N1bO3W1t+QV3upg1JIW9eD6HXj59pGGVeBkJctgJG7o90TR5KbjdTlI6ZtqyQTBXoXR2nRb7beVrMTY1e+GuXNAhpthUUmRRU0yw23cX31EwTMbdxsigM9s3I2Hrz0POUnaK4JDRMw8VaZSwwaoEbZnbxmKmmBMv46cTJdFUdLnFjB2SE9DvduikiL4kmoKXRLD8mkX6guP9VEZvbr44JLIiWsqfq+A399wAQ4ei1jq4rNyJdt7HI4rllq5xSVF+O5oCGu3VWDuuPNxVm4GDtaG8PAbX+Khqwfpi7h3XHYOHr/hApyZ7cN31fVYsWU/pl3c11D/njwPnazGrL22+Wlt24CcOILooIRjHMdDMcPkvGRKEW4p7gOFw2L03rH6U91o6Jnjw8NvfImnbx6CuEP6Yo8cH16dUYzt31bjzJxMx2hGt05evDajGD1zfODgGH3uGRAA3UnTqKxR2whoBnhyBC/DLRq23V5Ri2nLP8H7sy/DpKVbkedXa4b6dM1EJ6+IV2cU40BNCLWhmCHaIDC065U8Qu19dWVBvr5IYB5PXRPpaJGYtR4oGS2dbWHCeO3e2YvsDLdtDdGrM4qxYst+zBlbgPwsj0Uee/XtI3WhiaG9shFJOJoCY5Zi/vvXfo4nJw4+5SbEdqlES6YUYe648y195ZyMHTHRKiHZ4NGiNJluEa/NKE60U3DuU0c0TCByEil9/4ntovGGVRgBVcJ//ru7DD0+57+7C49cZ1SxDJzEOeuc5Jw5pcaZnSnZoU+cWT1Tq2FrKComMoYsrwS/xwWBqRE8xrhlu1DM/joeue485CS2cYvMksKpLrTQmG0q0ThQH5Utc5i2AGZ3j3MzXJZ7tXDjbtw8ojfGDe2hL5jOGVugZvUk6mx75vhQWx9DVFYMC7jlhwOYO+589Mv3w+cyOmMnU+Y9WZplPK6gKtHyRRQYfG4BOb70aleQKiRsQhAdlLjCLXL+pSvLsLcqqNf3JFNZo65w3r/2cyhclaqurY/hiXe+wrzxhYYiZk0l8KalW9G7axbO7pqBxZv32G5XWVOPF/+5F9XBKG5a+i9MWroVU/74MR64aiCG9srWj6/lz2vnkp2klKe2PrAWDGvXoTl1U178F3ZUHgegpvuUriwzpItpdRWkJNl+yfd78KtrCnA8HLOMp3njC1Fbr6YNds060UdOTiw0JKOlbf1152GUrizD98fCiMTtBSPiMseWvdUoXVmGw3URTP3jx4bm9snF9zMv66cL9DjVQp3Z2XvKTYjtUolKV5ZBEATkZRkNEpck2F6/SxJ0Z/D65z/ExfM24YZFW7CnKoB7X/0Mk5ZuRVUgihyfq0MaOK2FJqV/xVMfYPSC93HFUx9gy95qy3fqNF7MNWcxmevjdtLSrShdWYa/7jyMuGytsbPdzlLDZp8aZy4n09Qu527YiUlLt2Luhp24a3R/S7qnFrEzP5tmpUiXxBCMyLh12ccYveB93LrsYwQjsm307OcX9TUc9+cX9TVU7AUjim3Te7OAEJE60bhimNs0tDGptlAx/p5LkmB7r7r63QYxqNxMN+ZPKMTizXt0hzsn04VcU/9P7bdbZLDMa1qNmVkMShKYbZpldTCKeFzB14fqMHHJR7h0/mbctHQrdv8QwP7qYLsWRWkqFIkjiA6KbFNEv/vgcQzo5gcHbFe4akMxVNaoKmj/efW5eo1bVV0Uc8YWIDfTjewMN2avPSGGMGv1p3h5ejG2V9TqqpJn5Wbgu+p6+D2Suno69jzc/IK10fHccedj2vJPcGVBPh66+lxE4zL+dt8lqI/KyPRIGNorG1WBCLpne23T3RZt2mO4Zs0RXbllnyXN6PnJw/DCB3up9qedI0kCGFPv5fqyCtuowKpfjIRLVKMBeX4P6qNxy/1eXFIEt8D0mrjunb344VjIdtwfCUTw3C1DcTQYQ+/cDMtzE47J+rH65WVi9todqjR8irVQTeFkqURVdRE9jSjH5wI4LPUoWjq0nTOYHOmeuaoMa0pHoXu2z+40iBRwJdQajyalDXbJdFkc+VQFS8RErdL4ol76GFxfVmHZTnLYzjz+UhUFiSWpSWr7e/a9cjxiisQpDhG7h03bhaL26pmvzSgGMk9sx1PYn5MojNlhJVInrnB9bjOPSQ61ufor04shc7UF0JLNe3D3mP6OtXLJZRY5mW4EwjH8z6TBYGBwiwySyBCV7TN7kut7k9Mk++f5LTVmB4+FHOfGwwFFz75I7sV484je8HslMDDHFMz2mqJ5MsiJI4gOilcSLGIPi0qK8Oifv0RVXdRSOK/VjvXM8cEjCTictEK3vaJWT218bUaxoRG3umqnroZqwiU+l4A+XTMgMOBIoB4xWcHjN1ygpncl1KwWb96DfnmZ2PLQ5TgajOHxt7+ypPwsmjwMeVkeLPvnXny8v1ZPmXSLDLX1MWzZW224Zs0RXfKP/Zh+ST+sKR2FmKwgbOhuGQAAIABJREFUJnMsfX8PakNRvHz7SN0QTodJ+nTEJQhYX1ZhGQ+LS4pwLBRDlldCKCZj2bQLcaQugjtf3o6Lzs7F8mkj4BIZJIHhWDiGGxZ/ZHDi/7H7MBZMHKyLMmj7zPW7cDSgph4/d8tQLLhxMI4Go6gORrG+rAJ3Xn4Olt06HEcCUVQcDeltMF68tci2JkkQAJ+b4UBNvcEYcDIS7F53SiWSFY7rn//QcP4LN+5GVV3UUI+SrEBrZ/BokW41EknRjFOBMUBgzFJjbKn9cqidM/dh80oC7h7d36Li6zVFxPweAXePGWAZf36PSdjESQDFFDrTInsWNcmxRjVJgQF3j+6PIwFVWdAtqudrjsSlGnkUHNI9k6dmr2TfAN38nRCp45UEdPG7LbbA4pIiZPtE7D4UtHzfgP09DUTiuhjUvPGFWLJ5D8YN7YE7kwTKFpcU4c+fVVrqJHvnZkAUgAM19ZAVrtf+XlmQj19fU6A/X1pN6MmEnmKybDuWOnkl1EdklLxo7kmXpc/B6aSEqUFOHEF0UGS7urdVZfoKfLI4yN6qoJ7LvmhyEdySYJBg19BW6JLR1NA2z74MImNY9dE+XDO4Bzr7XJaavPkTCvH421/rapSMqb2JZibOy7zCd8fqT/HK9GJcXdgdH+9X0y565qhNw+OK4ihiop4nQ/dsr24g33/lABwJRnGLpbFo+56kT0fy/B7cO2YAnk6sqOZmupGX5cHLW/cbmspneiR9jK8pq8Saskr0zPFhzYxizFhhTL2alRhLczd8qe+zq9+DmvoIInEJh+siuOjsXDDGMDURgdYixHXhODLcEjZ/XYEfD8jHS7eNwHfV9QhF7Wt5HrjqXCgKcPG8Tfo465/nR3lVwGIknOx1s1z1kilF+N1bOy0pZdozrT0fWj2KatQojlF37f/NaoZE44jJXHe4gBNzl7nWLRrntpGuR68737BdXOF4xrTdM++V47fjjNsFIvZ93V6bUWyoiRMYbNUuzdOeo8qmyRvVxovZaTU7hU77M9eLOqV7Jiu8KnBogD7T2ACdSB2FAzNWlBkinPVRGaGojEDEvn7zlen2tXK19TEsmVKELpluhKMy7vlJfxwNRi2qvatvH4nj4biltjG5Bnne+EJk+9wYN7SH4fd6/oRCdOvkxVk5GY5S/gePhWzH0vJpI7DvSNCSgqnVx6ebEqYGOXEE0UGJOzSM1VbgtVz0D2Zfhqis4KGrB6F7tg/PbCzH9cN6QBLsZa57ZHv1SVxb+f3tn7/UJ+DFJUXwugTsrQrqE7V27OQ0rtnrduDVGcXQVvacenUdOh6GWxLw9M1D8NXBOizevAcCY3jkzZ347bjz8OTEweiZ40M4pqAuHMM9Y/qjVxcfsr2SIe3MJQkWBb90mKRPRyRJwIB8Px657jwoHIjLHAePhfHxfjUCrI0lrQmtuTdazCECEJUV3DyiN/p0zYDXJaA6EMPdr3ymj+UVt43QHbihvbLx84v6Ghy6u0b3NywarL59pG3k4qGrz9XTvLRxtrZ0lG4kaH2ZgpE4DtWFbY2H12ddBI8kGBQKs30uy7GSn2nt31otSW6mG1V1YYvioNaSQ4sEeV20iHEqpKrWGJMVh75p5qbg3DaaYO7Dlmpft3BcwZ8+PWArvJKMxyXY9vH0mCKFkbiSktPqkpht5NFcE5dKn72oQz1rNE5R5KYSU9TvtLImpGfaAGq2jdM9UTh3yGZQ0ycnJmU/aPNMVSCiL7ByDsvCg7YQ9dedhw3pmZpaqbadVoKR5XU5Svlz2GceSCLDwo3llte1+viTpa+fKi2ZpklOHEF0UJzqL7QVeO3f2qroi//ci1+PPQ/XD+uB2et2YMHEwXjinV0GI/mJd3bhN9cW6BE8xhh+t+FLgwjEzFVluuHZUBpXdSCKH46H9fOyO9/qRP+rOWMLMHfDTr2fXFUggjM7exGJK7qc+yv/+g5b9lZjSckwfFtTb5GjtlPEJJGT9oeicByPxlATjKE0KVXsuVuGQWDA98fC2LjzEGTO0S/PD4EBuX4Pfj/+fPzPX3c71x4BerRq1S9GWoQSjgaj+r9nXtYPD67foa9S98vLRMXRkKH1hVOkAYChLqmyRq0zzfN7DBLbj731FX51zbm2z0k4pugOpIbWE8/umdYcw9xMN87s7EW3LC+EhPJkcr1RTFbUBZqJhdhTFcSz75XjN9eeh5xMEE0kZbXGU4xMNbUPm1cScP2wHgYJ//kTCi2CJeGYYhsBNNe6pdrcPBbneOvzA5b+b1Mv6mvYTkihVjDlnnNEyjiN2/qorNdbJi84aPW+vbr49Fq5fVVBzPl/X2D+xMGYtnybIarHAfzhpiH44VgY9VEZ//WzcwF28vRubR5zSQLmjC3QF+a07TLcIqJx2VHK3+OQaukRBb2lSvLrWi1eQ704m0pLp2lSDgVBdFBcoroKqhmVWtRsfVmF4d+vl1XqKmSrP9oHv0dCnt+D2lAMVYGIQflMawSen+XBY2/txMHakG1kIMMt6k5ZMuY0rgy3qKtari+rsFU8W7x5jyFSN3vdDsRkBa9MH4kfjkcwaelWXDZ/M25d9jFKRvXGRWfnoj6qqm7NeeMLTFq6FXPe+AJVdRH1R8R0PiRy0r7QfvR2VBzXHThAHVd3vvwpvj8WxtwNO3Hflf1RWx8zKN9F4xyPjjsPnHO92TdwQlnth+NhAGq6Jrep7agORvXPZPtcyPN78MufDsTcDTvxk//5AHPe+AK//OkJVVWXyCwKbs9PHgZRZMhI6nGlGaQPXDUQD73+b4xZ8L6+L01YIBnVAbAaOws3ltse79P91fp5Tlj8ESYt3YryqgAUhSM30437rhioq8k99Pq/EZUVzF67Q1cz7Iiqba2JJDDb8WYWGNEiU+b7Z45MOUX2FFMkThBSU4mMK9y2j6LZ6XJSuzRH9rTFi2Q0A99wfgy4ZGA3TFv+CUYveB/Tln+CSwZ2s6Rxuh2+P3fShqlsQzQOxoA/TBpi+U57dvHhb18exK+uKTC8t7ikCDFFwZFAFDFZQcn//gvTln+C7RW1EBgM8+WkpVvxy7Wf4/vaEB5/+2vMeeML+NwiwtG4o10wtFe2/vnL5m/G3A07DfOt5mCe7DfbSdEyz++xfV3rSef0uVPtnemUplkdjJ7SfjUoEkcQHZSYbK2/2PBZJWb/dBBmXNIPZ3b24vWySiz4u5piMGv1p1h264U4Foph/sTBWPr+HkvaxKKSImT7JETiCn5+UV9d+t9uJU9zzpzSuOaNL0Q4JuuqljMv64e+XTPw6oxiPUKn9XhLdv4qa0IAB0RRsERStGuQBKYra2rvzV63A68m5fM31yRNNC/VwSie+tsu/OfV5+r1FBt3HsKYgm7I9rnQI9uLx2+4ANE4R3UgaoiMaT3fntn4DW4pPsuQipiX5UFVXQQ3FvXEjEv7gXM1srVwY7m+0ru+rAJLSopQuqoMtSE1NTc5IpLn9yAaV/QoVlTmEBjD8mkjIDDoaWpb9lbr9Ty6sSkKWPahVYFv2sV9Lc/JC1OHw+c+sTKcHGXL9EiJeiZVIGjVR9/i3p/0x6SlWw3n+cOxMDI9InwuSVd4C0Xj2JOof01uvUE1cadGVFZssxYW3jzEsF3MoSbOHOlyipAIpsieoqSmEpmqwEiq6pnaAqElTVI0nV+KKpYcaksb7TlST0sx1F/HOYffIxmeab9HQtzcJ4FIGc6BpR/swR8mDUFXvweiwMDBEY0reGTD17h00Bl4ZXoxFM4hCQz10ThK/vdjvDajGIJpkUlWuGW+NJdQzFxVhhW3jbDMd5o4k5b9YI5AJ2fhdOvkPelvtiAwx1TL/nl+XezMJQrI959oaXCyz50KLZmmCZATRxAdFielsSvOOxMelwhBAN7bVaW/XlkTwrFQDBMWf5RIXRsKgTHDj6ZXYnhm4zcYU9AN68sqMO3ivpYaiucnD0OXTBfuGdMfb2w/kFCUzMCx+hhCMRkPXT1I/zEfX9QLgFqfp6VMri+rwN2JXkXJ9SBPvrsLgGpUeF0iQjHnBscyt0/ZUMCbfZImmhdFURcINCdcG1PPvleOqrqoHs1Kfm/VR99iTVmlbpiuKatEbSiK/0xEXvcfqcfstTuQl+XGXaP749ZlJ/a9aPIwBCJxLPtwH+4dMwC5fhfmjC1AwZlZBuNXWyVONj5ev2MUfG4Roxe8b7mOmMLx2oxi1EdldOvkdVTg83skPPNeOdaUjgLnXB+XAPDC1OF46m+78IsfnW1YTFkwcTAef/trfYHjLn6O43mW/rgPplzUF7LC4RIF9MvP1NOKNAMqz081oaeCyJietaBhl04pp6j+yBgsCn45mS6L2iVjwIxL+uE/XjtR1/mHSUMs23kke3VKtymd0i0Ktq0S3CYnPyrbp0lOMadJOqlOmtYMGFPbEdyxepvhuezkPbEN58Az75UbGqDbpXoSqeMSGe4ZM8DQumf+hEJ0yXTjrXt+BJeg1pFt2VuNuePORzRRZx+Jq020S3/cB0v+sR8A8MIHezHr8nNsf3eTSygYA558d5dBSCUmy7h5RG+cnZdp+/lBZ2ThtRnF8LlFZPsa/s02p1oqCsfRYAQHa8OG9HxzWqNTiuap0FJpmhrkxBFEB8UtCij9cR9MGH6W/kP7Yflh5HfyoiYYRTCi4OHrCvDomzt1Y1AL8av1QTGDMAmgTj6ao/XAVYNQcTQEvwe6tLvCOX7/lxMqU5qBvGiTKjdsaB9QUoRnNu7W96vVO83+6SCs2/Yd5owtQH6WB519Ljz+9lf6Oaoyx87KZwp3Nlq8LpFETNo5sk090KzVn2JOwtA1p4WpfQpHovxwAFWBCCSB4f3ZlyGucBwPxXSJawBYMqXI0rfqjtWfYu6483HX6P7I6+RBLK7oCwqdEz2PKmtCllXiPL8HMZlDEhn+8cDlADjCMUWPxGmy7ppT9sPxsKNq2n1XDMQZnbwW42Rgtyw8et35mLjkI8Pn7l/7OVbfPhJ7q4LonZsBr0s0nKcW/ejdxYe4AtyUiNJpTtsbd16EYESGlFiNlkim/ZTI8Ai2fSwzTFL/jnVdJq9LEgRLjzmXwCAJVqfLHMFyibA4XQz26pRmU1gUAVEQLMqBosnedAkMPyvsrqfO10dl/Kywu+WcU43EpaLu6fcI+NXYAsTiXK+B/dXYAks7BSJ1OAeyMyS8PH0kRMYQVzhEgWH7t9Vwu1wo3VCGp28agsnFZ8HjEnCkLop1M0dBVjgef/sr3DNmAPZV1+OvOw9jy95q3POT/o41u9r/u0QBI/qo6ZGiwHB2Xiaee+8blB8OYMGNg20/n+GWmvy7raXn/3AsbBFaaw1hMy1N005JszkgJ44gOigZboaxQ3rqxexXFuTj7jEDcMsLWw0/5FovueRol/p5e2GSc/L8uP/KAQhEjDLBS0qK8PTG3QaRkzsSxveaskqUHw5gztgC9M/3o/xwABs+q8SDV5+LOy47B9k+F36fkBjWzlNTsNJk3u+47Bwcrovgv//yNZ6+aQgy3KIlpUdL98z22U+cXTPJgWvvcId6oOTVXPN7ssLxXz87F163iEdNSqlaCu/2ilpHBdQMt4hZqz/FkkTNx/wJhVj24T7cPbq/bvwmf1aLdk1Kco7mjS/ES1v24a7R/XHn6H6QBIb8TidqP5yuyyMJOLtrpu7AmZXMYg4qs3GFY84bX2DJlCJ08kpYMqUIpSvL0L2zV49+aGlIyYbLzIQEvc8tUSS6meAA8ju58eqMYsgJQ9gtMUs7Fq2GraHIFDi3yLDPn1CIPL9VnfL72rDFOTu7q1GlJlV1ynBUsaSoa+MluTk3E1RH0zz/M9N1CEJqkbhUhFJiMnC8PmZRzsx0UU1zU1G46kjVBGKWXoO5mRIqa0K499XP8OqMYhyoCRkyIOaNL8TCRDuVX/zobORneRCKySctoVhcUgS3xHDtkJ6GBY8FEwdD4apjaC7hWDKl6JQcHq0mbcHEwS2a1uhES6VpapATRxAdFHMPofFFvSzSvrPX7cAr04vx8vRi/G7Dl4Ym3k71bt8drUdelgfPb/rGsK/SJJlgjWTjOzllUks7Gn3uGZi0dKte8/Prawrgc0vI9kp67vreqiDuX/O5oYZHEgVwMNuUntsvOQeSJLToxEm0HE7pJ8mrueb39h+pR/9ufr3VBWBUSn3o6kF4/O2vT9ozrbImhAyPBI/EEIkruPNy1cDt1cWH12YUQ+Enjn2y2o1ZiQhCOGY0Dpyuy+sSDQ6cWcns1Rn2fZncoqA+dyvL9Oj4y7ePhCgw3dB1clojcQWP/vlL/PqagoTDQc/HqRCMKNj+bTWG9s7VX/vXniMY2jsXOUn92jhntpGpR0x94qIOQiSvmtQpYylul6o6ZVzhBnVBrW2HuXZOlmERHSpdVYa1pcaebYqSmsqmY3+6pPGYalsDInVispoW6dRrUPu3rHC8+M+9tvNddUAVF1sypQj98vy2SrhPTRqCfUeCkAQgEuOWhYL7136OlbeNQFVdFB6Xsa2KRxKgKDyluclOyl+rSXNSv24NYbOWSNPUICeug9Pnobca3Gb/49e0wpkQrY15ddPZoJPxxDtf4+4xA7DzYJ3+I5+T6bKsimnRuqpAxNZhM6+YaSIn2v+ba9s0w1xz8JJTG7pn+xCPK6gLxy01PFpB8v8Z1stgmCSnKbTkxEm0HHbpJ8k1cea0MG1M/eGmIY5Kqfev/Rx/mDQELpE5fl51BoOIymo65aLJw7AyUWu36ZeX4vd/+UpfZXZ6lrTX4wq3pJalklZjp2TmnArHDcf9687D2HmwDq9MH6l/3slw4VAjJNT4vnlwCQxds3yGtNX5EwotY0BkwLSL+1rupUkPxLH/m1lFNNXtnNQpzQ6QRxL0zIzk8zPXzkUdosMx2dizzbGPnUmMJMMtYFFJkSUalJyOmmpbAyJ1RIE1+L1qao3ji3rZ/t7XheMAgMWb92DhzUNsI6914RgWbizHb8edh0PHw/ZjlgP3jOmPu5LS37Xjrykdhe7ZRkVLM05S/t06eVS7wUZorSMIm5ETRxAdFPPqppNB53OJeOCqc/H2ju91EZJDxyN49M2dAICVt43A4boIakMxg6qdncOWl+UxqD8umDgYZ+Vm4M93/wiZbtFQ26YpUmmftZtQJUnAoG5ZWFM6CnFZsdTwULSt46Gln6yZUYz6mAyRMRwPxzB1VB90z/bB5xbx5MTBYIA+JqsCEcgKx5UF+Rhf1EuPIqwvq0BMVjBnbAHysjzYWxXE2/8+iMdvuAA9cjKw/0hQ/7zmzD109SB9lX/ZrRdiy95qXZCiqi6Kl6ePBINzD0YtgiAIDAdq6g3jMnm8uiQBksBw8FjIsGqsRULyszzweyRIAkOu321Ync5wi6hIHDt5MaSyJgQhSdnQznBZVFKELI+In5siJNT4vul0yXDjeFbccI+6ZnnQJcM4n4Xj9iqWf7jJqGLpFJky90TT6i4tESzRGmGzdaZMDpDi4OytSbE/nXnuTbUGsD6qoGzfEbycUEIUGMN7Ow+iS8aZeiRTSghpJNd4r9v2naWtAZE6jJ08CprsxGvOHKCmk98zpj/ysjzo6vfo/eQO1IZtI81TR/XBPWP66+UV9naIgD5dM+wdSllpsGG2k5T/67Mu0hfPnnx3F+aOOx99u2YiwyOia6Yn7e0FcuIIooPiNa1uri+rsKx2Li4pAmMcnHNcMjAfGW5RFyGpCkRQWRPC/up6W4ETs8O2qKQIm776wTCBv/jPvZj900GIxGR0y/Lgd9dfgIevVeCWROT4XHjs+kI8fO3JHTBJEhxX4Sja1jERBAaXJOLWJNl8AFg3cxTWbK5AyajellrItz7/HneN7m98PdGD66GXTox5zVl7vmQYAOhqqZozl+wQiQLDvPGFCITV3kbbK2px7yuf4Y/TivRWBMn7fWnLPiwqKYLXLaA2GMeY/3nfEuXKy/I4rhrnd3JbIiHzxhfig12HMGlEbxwLxfTIdnKrjiff3aUbVoIAvT5ue0UtXtqyD6tvHwmFA3FZwbpt3+HGEb1tjSVqfN80XC4RfXIykOESEVdUOfZ8vwcuU72WS7BXsTRH7CSB4akbB+O+NSeyIJ66cbDFYXGLzFZQxS2anT0HQ920XSzFaJdHEmxbDJjTMwXmUANoo7LZ/4zOhnrteeMLDSqbmR7BUOOtPfeZJGzSZETGoDBuGwX1utVm20+8swuPXHcexETzb00hOHmOWlxShLtH90cgEreopc6fUIgMt4hcv9txYempGwfD6xbBYvbj1OsST9owW1E4QrG4fXQ4rnToxV5y4giigxKOKtjwWaWhZuzD8sN4dUYxonEFHknAii37MKxPrt6D5ZE31bq48sMBzB13Pnp18aEmGLMYCk/fNASdvKJetyYwBoUrGNYn1/DDPm98Iea/+zXuu2Iguvqtq17kgBFO2KUf5mV5sGVvtS6So0lUB8IxDDizk6PypF0tx+HjYeT63ZbxmpzuK4mCvrKsPQPbK2pRcTQMztUWAsdCMfi9LjBwXVl16kV9cfCYNcqVm+lGdTCKUCyOH46FDT3upq/YhrWloyyREO18p/7xY6y8bQSCkTiicY6nbx4CcOB3b6kR82TD6sqCfL0+jjGGR978wpAKVdwvr83qQzoqLpeIHskFcDa4JQGLJhfhjtVJBvPkIku6oktiyM40Rl+zM92WpuCiAGT5JKM6pcRgbvsnMnun0KqKmVoE0K4H6bPvlVtq+7iDOuUj5j5xNoq05tq5oKnGu7LmRO1W9sm/dsIBhQMTl2zFpKKeeHl6MXgiChqJy6iPKChdWYaeOT78cDyMuRt2OvZg1WqPpy3/BK9MH6mPWy3SXBWI6D1atb6wc8YWIDfTjTM6exGNK8j2uaF4uO2iBADbKJs2p2rqk05zWkde7CUnjiA6KLLCseQf+/U+LhqXn3sGjodjEBnDkn/sx3sjeuPJiYOR63frtWdVgQhy/W54JAEy5/B7RYNBIQoM0ThH92xfUpqDOmm+PusixOIKGGMQGfDY9YUdauWLaB2S0w9D0Ti++qEOz733jV6nqRkYWr3cL350tu1KbIZbtLyW7XPhuU3f4L+uORcrbhuhG6mPvXWi3caikiK4ReC2H52NuMyx6atDekRLEoDjoRhkheNnC/9pOffJxX2w4K+7DceMxmXLarLmNG6vqFW3cag10mrtDtdF8NDr/8YLU4ejZ8Jyfez6QkTjsqHZt1Yfp60+m2sFF24st/R37Aj1Ie2dYFTGyo/2W1Qi7xx9DrombReOKZAE4Jx8v55eGFdkRGLGmrO6sIw3tx/AuGE9wTmHxBje+LQS44b1QE6SmmRUVvDff/na4Ez991++tjQjd4nMMi4W2TTxlhXFtt/dnLHG85NEhrtH97fszxwBVByUW3lS7RzVxDU/2nyz4O/lWPD3cv31jfdfimAkZpijKmtCOBqMQhSY7X3o0zVTTallDNOWf2I5VlzhhoWwuRt2YnFJEfweEZ06q6m4gsBsyycO1dnX0UXjsp5Gmef3dMiat4YgJ44gOihO9QiMMTy/6RuML+qlpvKIAo6FYni9rNKwOvbbP3+Jh689D71yfAjFVKXKQCSO2lAMz2/6Bo9dXwiAUhqJlkMbW1V10KXyyw8H8PgNF+DMzj54JHXsTh3VB2d09tqOdy39MPm1/E4ePHzteXj67+UYU9AN3Tt70TXLjd9cex5+dU0BJIEhwyPgeEjGTQnltdfKKlFRE8JdY/pj7oYvMe3ivsjOcDtGLpKVXrXnzryarEXZNIfUqcZJq7XrmePTV5+1RZG8LA8O1NQ7Gjl2qphVgQgCEbWGq1++Hz5Xx0oxaq9IAsOWvdVYU1apv9Yzx4d7f9LfsB3nwO//8rWlvtPcX00UGF4rqzQY4D1zfLihqKdhO8GhGbnArBG2DTaKv1MtTbwdauKaeX8saX+pKFgSjcPpO62qi6BbJ7U2V1tk0uZNt8McdaCmHnPHnY8eiXnK/P6uQ3XYuPOQ3lPW5dCj0q584mQNszX1ycqakKGJeM8cH87sbK3T7GhQMjFBdFCyfWpNnFaQrEUtVn+0Dz+/qK9eI/fMxnKUrizDmrJKlK4sw4TFH+F4KIb7rhiIMzurE2FM5pi2/BNc//wWzN2wE/ddMbDDr3AR7QcttVJLx3no9X/jcF0YCzeW43g4hode/zf+49XPMH9CoWG8L5lShN65GYbX5k8oBOeAzy1iy95qlK4sw7XPfohRv9+Em5Zuxdc/1OHieZtw7TNbdKdq8eY9WDBxMLbsrca9r2zH9B+fjdnrdiAmy5ZnbFFJEVwSM7z2wtThEJl9jzvN4Hhh6nDk+z36dWqfnTe+EOvLKvDC1OE4s7OaUmo2TDQjJxnNyEn+7pL3uezDfarjm22/T6L5yfW5bcdLrs84l2b7BNw9ZgDmbtiJSUu3Yu6Gnbh7zABk+4wmm99rneMXlRTB77U2GZ83vtAyBsxpktm+E3Vnoxe8j2nLP8HYIT0tx7X7bVF7dFq3S2V/nRz21ylpuyyHbbJ8ZMY2lWyf2qTePD/m+t0IhOOYu+FEZoI2b3bL8trOJys+2o8zOntxhs378ycUYvHmPdiytxrhmIwenX3onu2zOHBO2M1hWpQtee7bXlGL0pVluH/t53oaZUeHcU6h6GSGDx/Ot23b5vh+KpL96UYathjo+E9mI7Ebt+FwHDLiqA0perG9S2SIyRwCU1d7PS4B457bYlnhWlM6Cmd08jo2IKZV+0ZDX5YNDc23ySSPQU3ZMRSV4XOLiMscoZgMlhjXcZnDLQk4MzGGjwQjCMcUiEx13rITRrM5vVFrTFsViOCFqcPRP8+P8qqAnq7z4NWDcGZnL3wuEVFZgaxwdM0UDc9YJ58AAUAwBsTiiv68VAejuP75Dy3P2mszig3PVPJ1ainJgiCc9JlzEkpJLvxv7D4T0LhRXxY3AAAgAElEQVS1oTHj1kw4HEd1KKqPl1yfG16vMSlKUThC0YhhXGX7BPjcRmc7HI7jeCyGaJwbmox3crkM+6yqi+BXf9phiew9dn2hIYsi1ePG4wrC8SiOJW3X2SfAK7kNhnkkEkeEx1GXtF2WT4CHSfB4TpxfLCYjEIuhPnJiuwyPAL/LpYvDhMNxBOU4wtET23jdAjJFyfL9gcatLeZxq9kIx8IccVlRBaUEBklgqIvEE6njDKLA9HkzlflEUbg+50pMXUSIxK3q0o3ByQZpaO5LMxp9wpROSRAdFK9XQjgMAFH4XAJiMkcopkASGNySgGicg3NYColfmDrc4MABlDJJtD22YzCp7if5Rz7DLRkclPwsr+0+7ST/n71lqGNbgOTXYzEZhwMRmLI1IQDwuT3I9KbWJ86c8tOUZ+1k59nUfRItg9croYfV6TAgCAw+twf1sSiQuJ8+t9Xh1pyX6ngUgGoBmh04QB17910x8KQ9ChtzXEkS4IUbx1kEDByMMYsDB0B11CJAHaInXjM5cIAqCuMHUB+J6K8lO3D6tYaBcNK+HBw4IkU0G8HriiPE1Z5x4ZiCpe+rUbMlJUUYkO+3qKw2NJ8IAnOcc5uK0zEbmvs6OjT6CaIDk4rB0CXTc9pOgETHoanOz8kcw5PtN1mNMCfT8rbtsVrS2CBHrWOR6v1M1SlMdeyletyTtX5JxuOR0MPTsKmZirpnKtdKNA6vV4IXEpB5YiHs3p/0x2xpUNrYAqfz3EdPA5FSimgaplwSKXI6T4AE0ZrQs0a0FTT2iIagMZJ+UEUoQRAEQRAEQRBEGkFOHEEQBEEQBEEQRBpB6ZRESlDKJUEQBEEQBEG0D6jFgAnGWBWAb9v6PBpBVwBH2vokWhjzNR7hnF/VVifTHklh3J4O4wRov9dJY9aGU5xv2+O97mjnROPWhjS0E5Jpj2O0sTR0DTRubWhg3HaEcXEy0uH6Gj1uyYlLcxhj2zjnw9v6PFqS0+EaW5rT5Ts8Xa6TaJ/3ms6JaO90hPHQEa6hvdHRv9OOen1UE0cQBEEQBEEQBJFGkBNHEARBEARBEASRRpATl/4sbesTaAVOh2tsaU6X7/B0uU6ifd5rOieivdMRxkNHuIb2Rkf/Tjvk9VFNHEEQBEEQBEEQRBpBkTiCIAiCIAiCIIg0gpw4giAIgiAIgiCINIKcOIIgCIIgCIIgiDSCnDiCIAiCIAiCIIg0gpw4giAIgiAIgiCINIKcOIIgCIIgCIIgiDSCnDiCIAiCIAiCIIg0gpw4giAIgiAIgiCINIKcOIIgCIIgCIIgiDSCnDiCIAiCIAiCIIg0gpw4giAIgiAIgiCINIKcOIIgCIIgCIIgiDSCnDiCIAiCIAiCIIg0gpw4giAIgiAIgiCINIKcOIIgCIIgCIIgiDSCnDiCIAiCIAiCIIg0gpw4giAIgiAIgiCINIKcOIIgCIIgCIIgiDSCnDiCIAiCIAiCIIg0gpw4giAIgiAIgiCINIKcOIIgCIIgCIIgiDSCnDiCIAiCIAiCIIg0gpw4giAIgiAIgiCINIKcOIIgCIIgCIIgiDSCnDiCIAiCIAiCIIg0gpw4giAIgiAIgiCINIKcOIIgCIIgCIIgiDSCnDiCIAiCIAiCIIg0gpw4giAIgiAIgiCINIKcOIIgCIIgCIIgiDSCnDiCIAiCIAiCIIg0gpw4giAIgiAIgiCINIKcOIIgCIIgCIIgiDSCnDiCIAiCIAiCIIg0gpw4giAIgiAIgiCINIKcOIIgCIIgCIIgiDSCnDiCIAiCIAiCIIg0gpw4giAIgiAIgiCINIKcOIIgCIIgCIIgiDSCnDiCIAiCIAiCIIg0gpw4giAIgiAIgiCINIKcOIIgCIIgCIIgiDSCnDiCIAiCIAiCIIg0gpw4giAIgiAIgiCINIKcOIIgCIIgCIIgiDSCnDiCIAiCIAiCIIg0gpw4giAIgiAIgiCINIKcOIIgCIIgCIIgiDSCnDiCIAiCIAiCIIg0gpw4giAIgiAIgiCINIKcOIIgCIIgCIIgiDSCnDiCIAiCIAiCIIg0gpw4giAIgiAIgiCINIKcOIIgCIIgCIIgiDSCnDiCIAiCIAiCIIg0gpw4giAIgiAIgiCINIKcOIIgCIIgCIIgiDSCnDiCIAiCIAiCIIg0gpw4giAIgiAIgiCINCKtnTjG2DuMMc4Y+53p9RzG2P8yxo4wxoKMsb8zxi5oq/MkCIIgCIIgCIJoLtLWiWOM3QxgsM3rDMCbAK4CcDeA8QBcADYxxnq26kkSBEEQBEEQBEE0M2npxDHGsgE8BeD/2rx9HYAfAZjCOX+Fc/5O4jUBwAMN7fuqq67iAOivff8RJmjctvs/wgYat+3+j7CBxm27/yNsoHHb7v8aTVo6cQCeAPAl5/wVm/euA/A953yT9gLn/BiAPwMY19COjxw50mwnSRCtBY1bIh2hcUukIzRuiXSExm3HI+2cOMbYjwBMBTDLYZPzAHxh8/qXAM5ijPlb6twIgiAIgiAIgiBamrRy4hhjLgBLADzJOd/lsFkXADU2rx9N/DfHZr8zGGPbGGPbqqqqmudkCaKFoXFLpCM0bol0hMYtkY7QuO3YpJUTB+BBAD4Aj51kGwb73FLm9AHO+VLO+XDO+fC8vLxTPEWCaB1o3BLpCI1bIh2hcUukIzRuOzZSW59AqjDGzgLwKwC3A/AwxjxJb3sSYid1UCNuXWx2oUXg7KJ07RpF4agORhGNy3BLInIz3RAER5+UICz0eeitBrfZ//g1rXAmBNGxoPmZIJyh56NjQve1fZA2ThyAswF4Aayyee+Xib+hUGvfrrTZpgDAd5zzQIudYQugKBy7DtVh+optqKwJoWeODy9MHY6B3bLogSEIgmhDaH4mCGfo+eiY0H1tP6RTOuVnAC63+QNUx+5yAN9A7RHXgzF2qfZBxlgnANcm3ksrqoNR/UEBgMqaEKav2IbqYLSNz4wgCOL0huZngnCGno+OCd3X9kOLR+IYY5cA+NQuApZQihzGOf+gof1wzmsBbLbZBwB8yznfnPj3mwA+ArCKMTYbavrkf0KtiXuiyRfSgpwsLB2Ny/qDolFZE0I0LrfFqRIEQRAJaH4mTldSSaej5yP9oPuaXrRGOuUmAKMAfGzz3sDE+2JzHYxzrjDGxgJ4EsDzUFMwPwJwOee8ormO01w0FJZ2SyJ65vgMD0zPHB/cUrN9ZQRBEEQTcEmC7fzsktIpyYUgGkeq6XRkv6QXdF/Tj9b4pTlZgqwHwCm57pxzxjn/tem1o5zz2zjnXTjnGZzzMZzzz0/lOC1FQ2Hp3Ew3Xpg6HD1zfACgP1S5mW7DfhSFo6ouggM19aiqi0BRmtT8nSAIgkjQ0LwqCQzzJxQa5uf5EwohUV0IkaakYkukmk6Xqv1CtA/ovqYfLRKJY4z1gSpEojHcpsm2D8BtAL5riXNIFxoKSwsCw8BuWfjTrIsdw9tUZEoQBNG8pDKvhqIynnhnF+aMLUC2z4XaUAxPvLMLz94yFMhs4wsgiEaSqi2RajpdKvYL0X6g+5p+tFQ65c8BPAy1XxsH8AyMETme+HccwJ0tdA5pQSphaUFgyMvy2H0cgPPqyZ9mXXzSzxEEQRD2pDKvuiURVYEISleW6Z+jtCIiXUnVlmhMOl1D9gvRfmCM2d7XhPaEAbqv7YOWSqdcDlUtcgxUZ+0uGBUlRwO4CMAZnPMXWugc0oLmCEtTkSlBEETzksq8SmlFREciVVuCxn3HRGTAvPHG9PB54wshUoCt3dIikTjO+bcAvgUAxtjlUNUp61riWOlOc4SlqciUIAiieUk1S4LSioiOQqq2BI37jokgCHhpyz5DevhLW/bhsesL2/rUCAdaXNiEc/4+OXAnRwtL98jJQF6Wp9ETIa2KEQRBNC+pzqunOn8TRHuhMbYEjfuOR26mG/ddMRBzN+zEpKVbMXfDTtx3xUCyJdsxrdEnzg21T9vNAM6CqkiZDOect0argw4LrYoRBEE0LzSvEqcbNOZPb+j+px+t4TzNhype8jaA1wFEWuGYpx1UZEoQBNG80LxKnG7QmD+9ofufXrSGEzcBwMOc88da4VinPYrCUR2MNmoVpSmfIQiCINT580gwgnBMhsgYfG4R2T6aQ4n0pK3sAbJD2pbW+P7pHjc/reHE+QF81ArHOe1pSr846jFHEATRNOzmz/kTCtGtkxd9cjNpDiXSirayB8gOaVta4/une9wytLiwCYA/A7ikFY5z2uPU46U6GG3WzxAEQRD28+fsdTvwbXU9zaFE2tFW9gDZIW1La3z/dI9bhtaIxD0DYAVjTAHwFwBHzRtwzve2wnl0eJrSL456zBEEQTQNp/kzwy3SHEqkHW1lD5Ad0ra0xvdP97hlaA0nTkulfATAww7bUEOzBKeSM6z1eMnzezDzsn7I9rlQH5Xhczt/vdRjjiCIdKYt6yyc5lyBMZpDibTDLYm4siAf44t66X3C1pdVtPhYbqvjEur8yRjDupmjUB2MYvHmPdheUdvsdiDd45ahNZy42wDwVjhO2nOqOcO5mW6suG0EDh0PY/a6HYZ9OBXaa31hzMekviAEQbR32rrOwmnOXVxShByfq8WPTxDNSY7PhXvGDMDMVWWtOpbb6rinO3bz57zxhXhpy75m7w9H97hlaI1m38s55y+d7K+lzyFdONWcYUFg8Hsl3ZhIZR/JfUE+fPBy/GnWxVRoShBEWtDWdRZOc+7MVWWoCcVa5RwIormoCcV0IxtovbHcVsc93bGbPx9cvwOPXHd+s9uBdI9bBmqy3Y5oTM6wUwpRLK40Ou+Y+oIQBJGOtIc6i6bMuQTRHqGauNMLp++dc97sC/l0j1uGFnfiGGN/bGATzjn/RUufRzqQan3ayVKIqMaNIIjThbae7xSFQ1Y4zblEh8AlCbZj2SW1bNJWWz/HpytO37uscChK8zpydI9bhtZoMTAawOWmv/EAbgXwfxL/7nAoCkdVXQQHaupRVReBojRcFqjVp/XM8QGAY31abSiKH46FsWDiYCyZUoQ8v0dPIUp1HwRBEOlOW8931cEo3vvqB6z6xUismzkKS6YU4cqCfCyZUkRzLpF2SALDUzcONjxPT904GNIpGPOp2EJt/RyfruRmurFkSpHhe39+8jCs3rq/2VPS6R63DC0eieOc97F7nTF2CYDFACa39Dm0Nk0ttk+uT3NSWlMUjoO1Ycx54wtDIeqT7+5CNC6ntA+CIIiOQFvPdwwcw/rkouTFfxmK9c/s7KE5l0g7YnEFLknA3HHnI8Mtoj4qwyUJiMWVJu0vVVuorZ/j0xmP6X5zznF1YXcoStPuuRN0j1uGNquJ45x/wBh7CmofuR+11Xm0BE7F9n+adXGDtWcN1adVB6MoNRWHPrh+B+aOO18PS1ONG0EQpwttOd9F4gpmrf7UUqz/2oxiILNNTokgmozMgbte3m5JeVtTOqpJ+2uMLUR2S+tTHYzi1mWfWO733HHnIz/L2+zHo3vc/LS1sMleAEPb+ByanZMVcFbVRZq8ChGPK6iPxm333bdrJnJ8Ln3/PreIuMIRiyu04kEQBNECxBWOypoQhvbK1vvE1YZioKmWSEc4545CF02hMWIWsZiMw4EI4gqHJDDk+z1wuax6AG3VE7Ijot0f8/zVLz8TMVlBVV3E9jum+9B+aDMnjjEmQa2Lq2yrc2gpTlYsev3zHzapn1E8ruDrQ3WoqovY7jvDLaK8KoDpK7Yhz+/BA1cNtPSKo9YBBEEQzYdbFHBlQT5+flFfPLj+xHy7pKQI3To1v8IbQbQkzS0+kapQSiwm4+vDAdyR1ENsUUkRBuX7dUeurXtCdkS0Btzm+ev5ycPw8BtfoioQsXzHdB/aFy0ubMIYe8/m758AvgdwC4AnW/ocWgtF4ThcF0YoFser04uxtnQUhvbKVn/UpxThd2/txEVn5+Jv912CFbeNgEsUcDycWvHo4UAEM1eVYeHGcswbX2gpDpVEpjtwT0wobFSvOIIgCEKlMaJUAgN+PbZAN4AAdb4tXVVG8y3RrmhugZFU9icJDPMnGO2V+RMKLUIphwMRPLNxN+aMLcBrM4oxZ2wBntm4G4cDEX2btu4J2RHJzXTjV9dY569Zqz/F/VcOsP2One7DkWDE9hhEy9IakTgBgPnprgPwOoBXOeebW+EcWhy71Yn5Ewoxf2IhumS6EYsryPa5UTKqN6Yt/8Sw2uR3uyA1IOEbk9VeRJU1ITz57i7MGVuAbJ8LPXJ86N7Zh4PHQsjze/DLnw7EsVCM+nEQBEE0ksauMjMGyIp9ChrNt0R7obkFRlLdXygq44l3TtgrtaEYnnhnF569ZaihZpQxWKJB88YXgiUdNkJ9xpodzfG2+1575PgwtFc2tlfUGr5jpxTZ+ogMJZOyD1qbFo/Ecc4v45xfbvq7mnM+s6M4cID96sTsdTtQcTQEWQEEQcCMS/tZiuDvWFVmWG1ywiUK+mrW9opalK4sw/1rP4fAGASBwS2JuGdMfzy4fgeqg1F9W42eOT4wRg8XQRCEE41d7Y/JHPuP1NN8S7RrGjOuNfGJHjkZyMuyV1lNdX9uSURVIILSlWWYtHQrSleWoSoQsaRncg5LNOjB9TuQXIrHANvnjGg6hwMRxGRu+73KCjDzsn6WdFot5da8/b4jQYqKtgFtLWzSYXBanchwi4jGZZzZ2WcQJUkuJFU4R219BJ28zsWheZluLC4pwsyknPHFJUXI96tKP7mZbvTtmonKmhAWb96DeeMLLataItkUBEEQjjRGiAFQhU20FPfk+XZxSRHNt0S7obHjOpX95fk9hgjb4s17LPvLzXRjxW0j8G11vS5h3zs3w5KeKTsIqshJXhxjwIKJg3H/2s/152zBxMEkInQKxGQFoWjc1l4MReN6em3y/crNdGP5tAtRcTSk39OcTBcefXMnFpcMOyXxPqLxtIoTxxi7AMDDAC4FkAPgKIDNAOZyzv/dGufQ0jgVBNdH1cEsCEyPpmlpj+ZC0pyMOHpkZ9imLnxzJIiFiZzx3Ew38rI86N7Jq6dhCgJDhkc9h+0VtXrKZW6mG519Lsx/92s8dn1hq34nBEEQ6URjhR0kgaEqEDGkuNdHZXT2SRCEFk90IYiUaG7BEp9btIinzZ9QCJ/bur9IXDH0tX1h6nDLNpJgL4AiJT1DkiDA6zL2NPO6BIj0nDUZlyjgSCCKVz7+1uCQv7RlH24e0RsDz8jCGZ28FpvUfE8XTByMEX2ycSQYRenKMsO9JsGTlqU1hE0uBPAvAJcD2ABgPoC3AIwGsJUxVtTS59Aa2BUEz59QiF5dfMj2Sqiqi8AtMiwqKdLTHs2FpJE4R20oaikW1lIX/rrzMEpXluGxt77C3qogvj8exuG6sJ7X3DXTo5/D9opazN2wE5G4gvnvfo37rhiotyBIpWCfIAjidKMxwg6A2ih32a3Dcc+Y/roD1zPHC48kOH7mVGmM8EpLfJ5IPxo7rhsirnAs+3CfQYhk2Yf7EDeNpVTTLvP9HiwuKTKcX3KmEaDWnj636RtEZbUJdVRW8NymbyDT+G0yGW4BPXK8+I8rBsAtqu6AWxRw5+XnoFcXH0Sm3sPkOaI6yVED1Ht6/9rPMeWivpbXp6/Yhh+Oh2mOaUFaIxL3ewBfABjDOa/TXmSMZQH4e+L9K1vhPFoUQWDon+fHK9OLEY0rEBhwJBBFJCbj25p6vaFi6Y/74JbiPrapA6LAcLA2rDfz1ibaTl7JkIZpjuIlr3YkFyUzxiAy4LHrC5Hjc+ktCGiVhCAIwkqqwg4aCueIxLlhVXpxScutS56qvDfJg5+eNHZcNwRXuK0QCTcZ66mKkQgCQ4ZbNETZMtyi8fyY/THByEFoKsGIjM++rUFBj2zDHLakpAj/3H0Yj2z42jJHOKXmOgk8fV8bwrFQjOaYFqI14tDFAH6f7MABQOLf8wCMaoVzaBVqQjHc/MJWjPmf93H5gvcxcclHKF31KSqOhvTBveQf+yEwZlsY6hKZ7sABJ1YyWNL2My/rZ4ni/X/23jy+qurc/3+vfeack5AQEgRBBYvQiEEIs/dalUq1ov6UQQuBAsqgVVuvdfi2l6u9tPciam2tZdBWkEFFod5WrUOr1Q6oYKDSGkWKoGEyISSQ4Yx7r98fJ2fn7OHAISYxwP68Xr4wZ6+z1j57P/vZa63n83ye9J2t9KTk3vkBenYLUJTroy4cd+R5HThw4OAYyEbYIYW4KvU8ZUj61flrKoirskN86xeVWXdk2k9dHI9dHwtqBiES1bSeylaMpLYpxownNjFr5Waue+wdZq3czIwnNhnsUtPsx9S0Nv+MUx5CCIadVWjxYfPWVHDBgGL973QfITLMX92K/ecpn+P4mI5BZyzijrVNclJso2iaRKKxYuYI3rjja/zh9guZUtZHFzcZ2jef5dPLWDd3NG4XLJk2zEAdWDJtGF5XsqbKn+74Gn/8jwtZc8NIikI+XAKdCpEf8BxzZyudLrOvPsznh8OE44msdsQcOHDgwEF2SGTYfY6rskN869EEKrKhSba3wIWDkw/Z2JHMIEQipbGtEPDo1KGGuc6jU4daxEiysctMz5pDp2w7XAI8LmF7Xb1uheXTyxjaN589dWE0TaOmIQpIfjnVOH+9f2IpPrdioezeP7GUZW/udHxMB6Iz6JTvAj8QQvzRRKcMAncD73TCOXQoNE2yt76Z+nCCm9KokEmxEjcSDBTIFTNH8M7OGlbMHIFLEaiaZP17n/HtC/oBMP2JTXofD08ZgsetMLBnLs/OG0M0odomAKua1J2tmS5z/8RSXC27JO2V2OzAgQMHpzrcGfyqSxH4Pe2/R5pJoCLgdWVFk2xvgQsHJxeypdt63PZCJB5TvdugVyHocxtokkGfmxyvsV02dul1ZRjT5QibtBVut0BR7X0YwMIXK7l/Yil/3v65QbRkfEkxq2aP5HA4TnVDlCc37uIn15Tq89R99WFqm2I8+Op2tlbVOz6mAyHMOyftPoAQI0kqUUZICpvsB04DrgACwEVSys0dehLHgeHDh8v33nvvuL5T0xDlcDjOzBWbLA/C03NG41JgyvJ3DHltP7r6XB59YwcTy/rqapOaprG7Nhm506RE1SR+j4ve+QFOy/Ojqho1TVEONcUNpQbun1iqP0QA1yz5G0Uhn17CoDmm0j2Y/DddTWp5eRm98v3kB044GdgT6mQ7A8ey27PueemYfexedEV7npIDIxybtUFb/G1XwsGGCHXhOHvS5Lb7dg/w14+rGT+4N73z27eOVaZJdmHIy7VLNlreP8/ffAFFuT7b7xeFfNw2bgD9egTJ8bnoEbSl2Dl2a4OuZLcp8bP2yHWraYjyw+e3MbGsr65UuKGiip9cU2qwo0NNUfbXh6luiOl2X5zrpVd+gO7B1nZ765r5wwf7uaSkF5qUKELwRuV+Lj23F6cX5Bh+w/YDDcxZnWbX04cz8LTWxePnh8N8crDJoojZv0eQnt0sz5ljtzYw2+3++jCv/nMfw/v1MMwpl5aX0TPXy+FwgvXvfcb0sf24/rHkHDZVHitd+fz2SwfqC/32zru1s2+g3Wy+i+G4f0SHR+KklJuEEKOB/wK+AXQnWWLgDU6SEgOxhIoi7Kvea1KiqsZjW6vq6RZwc9u4cywPztObPqWmIcZdlw3knt/8w+DQPG7BzBWbWTptmEEONrXbce+VyXC1XQmDpdOG8fyWvTw7dzQxVbLrYBP/+X//pKYx6iS2O3DgwEEbEFM1onGj3PbDU4Zw/pnd6YhMgUwCFfsPh23fP3YCEgN75vK7Wy6wFdFy3gMnFtp/wqzZiodopsQzTZNoEoPdL502zEK99HsUyvr1YOrj7xjmOeYotapquN3CELFzuwWqqqEoyQhOJKGx+JXthrnP4le287Przz/u3+kgCYlkwGndLOWr/vThAV3UZMm0YQikvoAzzy2XTy9jQFFIt7f2FNGxs+9Vs0cSTWiOOFMLOiUOLaXcJqWcJKXsKaX0tPw75WRYwEGSCqBJ+wRetyLYVdNkOaYIxZJMetOaCiaW9WX+RWfru02pY3NWv6cLpOw7HGHhi5Vc99g7zFtdYQhXe90u2xIGN63dwuXn9QIhKP/1u8xauZmtVfVOYrsDBw4ctBFSYvHjtz/7PnVNcTqK5GInUJGio6UjE4VJUQSqhq2IlvMeOLHQ3kI12QqWRBMaN63dYpljRBPGxV4krukpJnq7NRVE4sZ21Y1RZq3YbBA2mbViM9WNUb1NqibjvNUV+tynpjGK+xScuLcXZMv9TpWvmrTsbab96l16FQSB1vJXkiTl0k5Yb97qCurCcUO/7SWiY2ffn9Y2O+JMaejwSJwQoggokFJ+bHPsHOCQlPJgR59HR6Iw6CUST7Bk2jBubnFsKalpn1uhR8irHxvbv5CbLj4bLUNicH7Ao/+/+ViPXB9/uP1C3C7BmhtG8fS7uxl2ViGFQS/FuT4KAh4URXBOz5Bht+r1ys8ZV9KT/kVBXAIWXXseHpdCfTjOsjd3srWqXt+xbU9qhgMHDhy0FSeCL8okq52ixHcWUnXALDTLDHXAHIGTkwPHcx8TCY3qxihxVcPjUigO+XCbctiyFSzJVmQkUztzPbmEJrl3wlcp6d2NhCZxK4LKfYcN7YI+hbVzRhFPSBQBmgSPWxD0OTlxbYWa4X6f0b2V6lrUUqtv7Y2jkNJ+btpRfsPOvnO8Lsd3paEzhE2WkKRPzrM5djtQCEzphPPoMCiK4PT8HHL9MdbNHU1Ck3hcCqqm8f8t2UhRyMd9V5WwvLwMCUz/9SYWTCixTSatb9nRsDuW53cz7VfvGmgJv3j9Y16rrNZf2gOKQtSH4yx8sdIgsPLoGztsaZqpfDqv2+XUEHLgwEGXwInii7wZBB6aY2qnCi4cL4XJETg5OZDtfUwkND76vMGQvrGsvIxBPXMNCw10W8sAACAASURBVLlsBUsy2b25nSeD8I/HZJcFOS5Oy8/huseMtMuCnNbfoSiCIybxuGSbzpjGnpzwKPb3Mcfr4o6vD+CN7TXcddlAPR9uxcwRneo37Oy7OWYv7neq+q7OeMv8G/BqhmOvARdk04kQYpIQYoMQ4lMhRFgIsV0I8b8tRcPT2xUIIX4lhDgohGgSQvxRCHHeF/wNx4SmSZpjGkKAx6WQUDUk8Nz8Mfzs+vPJC3jIz/HoTnTZmzu5f2KpQY51aXkZGyqqWPbmTh6YZDy2rLyMn7xUaaEl3PmNQaybO5oFE0r4vy1VfN4QIRxTWTChRJeGvXntlow0zbs3bOM/r0hyoZ0aQg4cOOgK6EhflI2EerbtBfDwlCEGX/3ApFJOL/B3urLC8VCYUpG79PM+WuTOQddEtvexujGq5z2l5guPvP6xga4IScri8vJhrJg5gnVzR7Ni5giWlw+zUBZdQljmKA9MKsUljO28HsXyfDw8ZQheU05cQ8SedtkQaaVdNoRV+zbhUzMC0x7wegTLy8ssZQF+8lIl15T14YHJQwxzxkde32G578unl1HQwiBrb9jZ95mFObY2XxDwHJdfP1nQGVsYBcDhDMeOkIzEZYPvA58BPwD2AEOB+4CLhRBjpZSaEEIAvwP6AbcCdcD/A/4khDhfSrmnzb/iKEjtcj3y+sfc8G/9ueO59w3qSYtf2U5NY5TVN4zUH4atVfU8+GoySXdAcQiPS2HN27uYWNaXAcUhPj8SYdG15xH0uekW8BCOq7xWWW0Yd09dmMPhONc99g7jS4q55ZIBhp2s+yeW6qInR6NpuhSBogiHYuPAgYMugY7yRccb4TtW+2hC439+/xGLrj2P07r5cQnBgSMRonGNiLvrViFuT/EBB18esr+P0lawRJjEd7IVLIkmVFuRkUe+ZRQZUVWJx60YBEs8bgVVPX56ZrbUTAfZQwAFQY+tUN4PryihvjlmEeVb/Mp2npk7mgOHI9Q2xfj5Hz82qFO2JzLZN2D4rCDgYUdNY5dnbnQEOmMRtwcYBbxuc2wUyZID2eBKKWVN2t9vCSEOAU8CF5FUu7yKZOTvEinlnwCEEG8Du4C7gNva8gOOherGKPPXVLBgQom+gIOkg7lz/TYWTChh3uoKdh9sNoSBt1bVs/DFShZePRiA5X/ZDX/ZzfLpZTodcvn0Mm59emtG+mVqZ3piWV89Hy819t0bkmMvfLHyqDTNVBjaodg4cOCgK6CjfFGmCJ9Zij/b9q4WsYXyX28ynOfCqwfTvYtHtVKROwcnNrK5jzKDYMm6uaMN7TIJlpjbKaJVZCSFPgUBFFMkTpVwy1NbLc/xs/PGGNodrd7isdo4wiZtRzimEY5r+nwzhT4FAYQQ1DbFLNe8pjFKJK4xadnb+meV+xsy+tAvikz2nf5ZTUP0uPz6yYTOWMStJ1ns+30ppV6sSghxBXAPsDSbTkwLuBRS9eVOb/n3KmBfagHX8r3DQogXgKvpoEVcXNXYUxemONfHnrrWOhqpnY2v9splaN98Hnl9B8vKywy89IenDCE/x8Nd6/+hf68418eq2SNZ9PKH5Ac8BvqleSftwVe3A9C7m9+wm5ISLCkMelkxawS1jTGCPjdPzxnNkUicPXVhNlRU8d1x5+BSkjtwx5sc78CBAwcdgY7yRccb4cvUXtM0ahqiJDTJM3NHcbAhRsjvxqUIPC6RFF1wZT+5PBFEXBy0D7K91/G4SnVjVBf6KA758HjatokhSQpUmOcIZmQb7RICHpo8xMA6emjyEExrOKSUtuOahVKUDP0paaxLv0dhaXmZJSfOXK7AQfZIaJKXt+3jqTmjiCUkLgEHG2MUBD00R5OieOZ559LyMh57y2g7dj60rT6tLd87lVlknbGI+2/gQuB3QogDwF6Si67TgHeAH32Bvr/W8u+HLf+eC/zTpt0HwAwhREhK2fgFxrOFx6UwvqSYbgEP40uKLbSFJdOGce9VJby8bR8el7EWSkHQS8jnoijXa/nesvIyCkNe+hQEDPTLwqCX4jw/P37xA7ZW1TO0bz6axCBmkhIsKc7zcTic4PtpzvH+iaVsqKjilksG8OTG3Wz8pFYPPTsUGwcOHHzZ6Ci63/FG+Ozajy8p5mBTjHmrKygK+fjBNwdx+7NGCn1Rrg9/llHDE0XExcEXR7b3Oh5X+ai60bJgGVQcatNCLsfr4q7LBloKZQe8xr6yjXa5hMDvMdIk/R7FkhPn9yi245oXXi5FydBfazspIc/vYuWska3qlC46rJTHqQC/W2HsgB5Mffxdw/0BeH7LHq4eejp/3v45K2aO4HA4TlGuj4BHYeMntYZ+zD60rT6trd87lVlkwrwj0iGDCOEBpgOXksyBO0hS1GSNlDLRxj5PB7YC70spL2357GNgi5TyelPbG4HHgTOklFVH69dc0T4bJBIaew+H+clLldxz+VeZ8cQmQ0SuMOilZ54fj0swadnbFkNbdO159O2eoytPph97bt4Ydtc2GZzg/RNL+fP2z5lwfh9uaqFx2oXD1944Crci9Dy59GOp76Sonn0KAidS6NmZ2ZhwLLs9656XMh5LYfeiK9rzlBwY4disDdrib48X6Tu7Aa+Lz49Ev1BO3NobR+m+OkV9Lwr5dPZFc0wl5HNzer6f3gU5NmdkRE1DlGuW/M3iozvKHx/nTrdjtzZoq91me6/31jXbvrfXzR3N6VnYlBnVDRGuXbLR0t9vbh5Lca5f/yzbxeP++jCTl1vnMs/NG0Ov/NZ6hfvqw0yxaffsvDH0Tm9X18wUm9/77NzR+jO0r66Z+174gIllffWo3oaKKu678ly758yxWxuY7fbA4TCTlr1t67/yczzMWrmZVbNHcsez7+v1iBddex4uRRjmpOk+VNMkB45EbO/7M3NH43Ur9AjaCy+11ReeRBthx32ynaLNKqWMA0+0/JcRLcIkC4DHpJQHjtIuBPwWSACz0g8BdqvSo14YIcRcYC7AGWeccbSmtnC7FQTwWmU1N130FX0BZ65sv3TaMIpCPoOB7qkLUxjyZawbF0lovLfrEAsmlFCc6yPkcxOJq8wY249wXGXh1YM5szDH9rv1zXEicfswc4qmmS54ciqEnk8mfFG7deDgy0Bn2q3dy33V7JH85uaxxBPaMRcx5oigqknqmuO6T80PeCgK+Wx9vZlalgmdSQU6iSY7nY72sNts73W2ddiyRTyh2fYXT1jFd/ICbmO0y221i5iaoT/V2J+q2bdTNWO7RIb5j5oeZBDYirM4y7Wj42h2G1e1jP7L71HYU5cUz9taVQ+01pDzuAQPTh6CICn572spLZHyL03RhO393FsX5o7n3s/oc9rqC09loaauRiZWgHuB3pkaCCH8JBUo+wPfMClOHgK623ytoOXfOrs+pZSPSSmHSymHFxUVtenE3a5kvY3qhih9Cuwr29+0dgu3jRtg+F6fggC5frcuemI+tvtgE5eX9mbe6gquWbKRSx/+Mzet3YIqYeaKzcxauZkd1Y223z1wJEJ9OG57LPV5uuDJqRB6PpnQHnbrwEFnozPt1k6YZMYTmxCIrKT4oTWx3ut2MfVX73LgSET3qfXhOLeNG2Dr67Odb6eoQOnoKH/slJFpO9rDbrO91ylao7mdq42TUiHs+xOmnYbqxijTHn+Xr//0LS556C2+/tO3mPb4u7alCLI5PyXDuGYBFFcW7TKJszh0yqPjaHarCJHRf0mEPqdNoU9BALdLYfLyd7j+sXe47rF3mLVyMzOe2ERtU0z3LylBlHSk5ptH8zlfxBceT4mVkwldbREHR9lXaaFlbgBGAt+UUv7D1OQDknlxZpQAn3VEPlwKxSEfy1rqvP3suvMpDHptdxT6FwVZP38My6eXMb6kmPsnlqIIdNGT8SXFLJ9exvr5Y1g1eyQv/2M/Hlerg0vtPqXvcNnVnLt/YinL3tyZ8diGiioemJRs4wiYOHDgoKvjeOu7QftGuVJ9pfvUZW/u5Kwe9kyIbKMmnVmz7VQWAOgKKAx6eXy66V5Pt95rlyJs39t2SozZPBcugW1/Zu2dhJYUIlk+vYx1c0ezfHoZRSGfRdhEUez7U0wzSpFhXHOUWgj42XXnG9r97LrzDe3UTNFJZxXXZghBRiZXYyTOEzOH0y3g0WsGrpw1ApewL1UVS6gGH/nQ5CG2c9JU+3AsYbBXTZNIJGtuGMWKmSMY2jffmZtmgROm1L0QQgHWAuOAK6SU79g0+x0wSwjxNSnlWy3fywOuBJ7qyPNTFEGOx8Wcfz+bbgF3xkTL/YcjXP/YO7pwyQt/38OZhf2oaYzSPejhlksG6KUCUkmmXreiJ5bWNsV48NXtLG4puLinLqyLniy8ejB9uwcAweJXPtRD4KljZxbmoGqSSFxl9r/1p3+PII9OHXpKhZ4dOHBw4qErJLyn+koXmSo9PY+EZl+6xePKbo+0M6lAp7IAQFeApkk8bmGqm5bMI0q/3y5F4cmNuwyqjk9u3MVPrim19JfNc6Fk2Z/fnUGIxG2yZSls+/vRVYONzSS27e690rjX7nMp5AXchuuSF3DjS3uGvG7F3nazfM4c2EDCkRZWlvm65ud42VcfNojiLZ9eRvdghvvgdiGR+sLN1yJU07d7gKpDYb3+XKr9hwcaWPhiJY/PGM6AopClztvy8jJ65fvJDzhz06PhRLL+XwKTgQeBJiHE6LT/+rS0+R3wNrBGCHG9EOIbLZ8JYHFHnVgqkXPGik0kNI09dRG8bsVS2f6BSaX43ILl08v4xbeG4ve4+PYF/RDAhvlj0CSWWm93rt8GEh549SOiiWQ9j61V9ax/7zOWlZfRpyDA0L753DZuAGcW5nDgcITH3trJrAv66WPXNEYpDHlZ9PKHOh1TSonblT2lyIEDBw6+LGSiAR44EjlqRK6tUS676EZ6X1ur6tlQUYUQggOHI7a+/nhcamdRgToz6ufAiurGKItf+YhYS+5YTNVY/MpHFrpiYdDL7ZcOZOGLlVz32DssfLGS2y8daLlPtU0xHv5DckNh3dzRLJhQwsN/2G6hqmXbnxCtghXQOgcx0y69HsHdlw/i7KIQRbk+zi4Kcfflg/B6jO3yAi7uumyQvtDyuhTuumwQeQHjpkEkofHgq9sN1+XBV7cTScvZE2D7nDkzl7bD7VIozvPx8+uNUdDl5WUoAr637u8GW5i3uoKEJlk1eyQrZo7g+ZvH8sf/+Brr5o5GInEpyXt027gB3PLUVmat3Mydz23D61aoabHxPgXJ8hHL3typ+/HqRmudt3lrKlA1nLnpMXDCROKAy1v+/WHLf+n4EXCflFITQkwgudBbAvhJLuouPpYqZVuRnshZFEoKj9y5fgsPTR7C4le2G3agFr+ynQenDOHWp/9uSMx9cuMuZl3QD7/HlVHc5K7LBhGOqaycNRK3S+BRBOG4ysNTzsfvUfQCnak+n9+yl2fmjkbVJC5F8MgfdzCxrC83/Ft//VwenToUgh1xVRw4cOCg/ZCJBrivPpl4nyki15YoV6boxoCiED53cne5R8iLJCkMoElp6+t//q3z2/syfGGcygIAXQPSVpxDYKYrZnefNE2z7U8zCYdk218mwZKYSbDEJaA5pllULAtN8wlNSiJxjQW//afebll5GZq01p2zvS5ppxdNaPbP2fVd7zk7URCOqwQ8Cnl+YxQ0oWnsPxzJKIYTTRjvaWoe+4MrSlj8ynYemFyqfzeduTDotFyqDjWz6OWPDGIp8Ux259C8j4kTZhEnpTwry3aHgNkt/3U4UjvECyaUcNu4Afpiqj4cp6YxyrzVFXrbPgUBPqtttiTmPjNnFJFEMjK2YuYIHnl9hyHsvPtgE+f0DNEQSbCzppENFVUsmHAuM1dstpQXSPW5+oaR+rgCqA/HLOfiUGgcOHBwIiATDbC2Kcb31v39qBLUqShXtsgU9Xt23hgWvfwhE8v6EvK5mf7EJtbNHU1zTLX19eaaWV0BTlHxLxeZxDnWzR1taZuN3SY0mXV/qqoRS6jJ/LaEiqpqKIpxDiCwpwabLaQp2rqAS41705oK1s0dTX6a2n9jROOFv+9hxcwRuBSBqknWv/cZM8b2o1uafoWU6PXI0tudVdhPb+NShP1z5thvm+FWBHFVMmtlq79LMbv69QjazkcBi3+8e8M2FkwoIaFKahqj7KxpYnxJsaUcxI+uGkxcldxz+SC98HtNYxSPKzNFs604VXzdCbOI66owJHJOGWIRGzEX717wf8Za5EUhH3XNcUMk7YFJpSx+ZTs1jVHun1jKg69u54dXfJVJy95uTUhWkg9Pca7PdgcD4PqWuispyVhIlkFIhcsdCo0DBw5OBKRogOnRsZRvbO8d20xRv0Ra1OOhyUlfrwjo2z3ps9PziB6aPCTrEgOdBae8wJcPib0oRFuRUezDRDHOtv6b0iJEYo6ImQVLMpVAMAugeFyCK4aczqyVm/X+lkwbhsekqOJxZ2iXVt4g23NzkD1CPoX6sGpYwJnLDaTPR5dOK8tYDis/4CEcS7B02jBefH+vRd9haXkZCc0YwXtgUik98/wUh3wW//5FaN6nkq/raos4CXwKRI/VsKsgPdl9X304o9hI1aEwRSGvzgtOIT16B60c9NWzR/JxdSMPvpp8eFIcdz16N3c0fQoChHxu2x0MVcPQ501rt7By1kjuumwQVYfC9Mr3n3TG7MCBg5MTKTrYs/PGsK8+rAs8pQrQtierICXHbvapLiH0yU2qRIsmk4ILbkVh9eyRqFJy4HCEX//1E4t4w5eNTBHGjioq7sAKXzsLy7gzRDDcJrGP6sZoxshZevFwLUshklSJAcu4pjlFXJWWPP+b126xRArjiWO3y/bcHGSPxqhGXJX6vbQrjXXn+m08PSd5HzQp+Vd1k+29rw/H9ULhk0ecycwVmyz2tvDqwZa+f3PzWNxupV1p3qeSr+vwPQwhhCKEcJs++4YQ4g4hxND0z6WUmpSyn5Tyg44+r/ZCYdCrJ3n2CHlZOWsEK2aO4IVbLuCByUPo1yOI163wleIg0YTGqtkjGV9SDCQN/4wM8q61TUn6Y01jlF9OHcbrlZ8bjrsUeHrOaHK8LlsJ33AsYenT5xb43C769QgSjWskbAp9OnDgwEFXhKIITsvzE/S5dYGn9hTmSImZJFSrn07l56Tk14tzfayaPRKPS+D3KPTK9yNJ0sL+8nE1t447B7+na4UInPICHYtspP6PR1gmm/5SpY3S+1tWXkZxyDhRzTZyJgTc+Y2BBsGSO78x0BJV9rgFS6cNM4y71BQ5O55xs2knBNx6yQCDSMqtlww4LgEhB0YkNMmW3bX6vHVAccj2PmhSEvIpBNwKA3qGWHvjKP70/a8xpayP7h83VFRxZmEOhSEvIkMZghyvy/JZquB8urhTYdBLbVPsuMrJpONU8nWdEYl7mmRkbQaAEGI+SdERgLgQ4gop5R874Tw6DNGWxN2ikI8ffHMQT2/6lG+P7ZeRIvnLqcNYMKGEf1U3sT8tepdCSt719f/4Gv+qaeSXf9rBt8f2Y0d1I1ur6hlfUsyhxjjz1lSwYEJJS46ccXfqWyPPNJzj+JJiDjXF+c5Tree0rLyMQT1zcZvlgx04cOCgC6KjhDns6De/nDqMWy8ZwL7DEZ7cuIsfXlFikV9/bv5oDjbELTS1swp9NEe7Vv0qp7xAxyF7qf9sBUuy7y/H6zKIUuR4XZb+so2cBTwKtU2Sm9ZsMthzj5C1xIDXLVg5aySKSEbJVE0F2bZxvZlyotIiir6WeUo6HW/ptGF4nflLmxHyuhh6ZgF1TTEW/PafLJhQYnsfPqlpoijXh6ppfOeprYY55Pe/cQ6Hw3F9znlGQQ41jVHbfppjxkWUnf9pDyrkqeTrhOzgQolCiE+Bu6WUz7T8vRN4HbgDeAw4TUp5cYeexHFg+PDh8r333su6fU1DlGuW/I2ikI/Fk0qZtXKzvrAyJ3VOLOvLvNUV9CkI8ItvnU+3gJdwXCWhSsPi6uEpQ/C6Fe77XaUhofSBSaVE4hr9i4J8UtPEI6/vALBwmJOOTbCvPqo79gHFQa5//F2LUT87bwy98wO2v60Lw9l7M+FYdnvWPS8ds4/di65oz1NyYIRjszY4lt12VnJ6dUOEa5dstPjHlHDU/RNL8bkVg+Q2wF/uupiFL37ArAv6cVqeH1VKDjbG6FeYgyahVyf71kzXS9Mk9eEY++sjzEtbcC6fXsbA4qNu5Dl2awOz3abmAWb7saNvZWPTNQ1Rfvj8Nssc4ifXlBr6q2mI8qs//4tJw88wCILceOFXDO1qGyPsq48YNpaXThtG73w/hSG/3m5vXTM/euEDy7j3XnmugXa5r66ZJzfusoz77bH96J3WrrYxwt76iCE3asm0YZxuGvfzw2E+OdhkqU/Xv0eQni0KKHvrmlllM+aMsf0M59YCx25tYLbbvXXN1DbFqG2MkeN14XULugd91DQkU3g2VFRxw7/1Z9HLH1HTGGXh1YOZtXKz/v0+BQFWzhqJ161wuDnGwcYYw87sRjQuqW6IMj/N1zw+Yzg+t8KMJzYddXF2PM9SJpzAOXHHfXKdEYkrBvYCCCG+AvQDHpVSNgghVtDBRbg7GrGESlHIx/e/MZDD4Th76sL07ua3lcvN8ycvd1HIh8fl0o15fEkxa28cxZFIgoMNUQqCXh576xN9AZf6Tq7fw53rKwx9Pvjqdl2+dUBxCEUIXvj7Xr42qNi4Y1VeRlHIKIKypy5MQnUolQ4cOOh66KwXsaZJmqP29JtBp+WyYEIJD766nXsuH2Rp41Lg5ou/QjimMj1tcrK8vIyivM4VjjpaaYRUId2ikI+FVw/mzMIc9tWH+fkfP+b2SweeCJObLo1s6VvZ2nS2pQME0lYQxFyyIJrQ8JgiZwlNJWZKqchG6h/AnUGwxG0SLGmMqqx5+1OD6uTjf/6EWy75CoWh1naRDOUDfpZWPiBbkRQH2cPjEihC6Eyyuy4byLRfvWu4vt4WimwmOqQioPpIhGhC49ODDfTI9XHTmgrd15zVI0jQ66JHC8X3mGUu2oEKeSqVUumMOPQRoLDl/y8CDkopt7X8rZKs5XbCQgjBbeMGcPeGbdQ2xehTEMDvcdnK/vpbVKBuGzdA36GApGLktF+9y776MLNWbmbmis1cfl4vwzjm76T6nH/R2Wytqmfhi5V8WttMXNU4p1eeJUn4pjUV3DZugKFPR57XgQMHXRWZktPNhYzbY5xdB5v0/J4U+hQE8LgUPf8uJWaSDimhriluKZA8b00F8UTn0ikzXa/0Qrpbq+qZtXIzM57YRFNM5bXK6g65pqcaUvStdNjRt7K1aTVDKQLVZFLRhGYrCBI1Lc6khBufrODrP32LSx56i6//9C1ufLICc6pRphIIZsJWJsGSuOkE3Ypg4ye1XPrwn7nkobe49OE/s/GTWsu8w51WPuC6x97R9QDSaZfZjukge8RVqc8r5190tsWP3bx2C/vro8y/6OyMdEhNJu367g3buKSkl04tT/ma6b9+l4QmURRhyHsryvXZLqqyfZaOhWzGOhnQGZG4jcA9QogE8D3g92nHvgLs6YRzaHekKBEgdXGSVFmBTIULG6MJ+hQEOKuHvZhJfsDD0L75zL/obPoVtdboqGmMZhRAyQ94WvI3htIt4EUiMyan9i8Ksn7+GD1M/u2x/RAC9tWHKQ75DJSaU6XGhgMHDromOjo5PZHQqG6MElc1/B6FhyYP4Y7n3jfke7gEPHXjKH78UiXL3txpKSWgapIcrysr8YaORqbrlel9lB/w6P9/Mib8dyYKg14enz6cOavTImzTrYIl2dq0zCDjbk5/ySQIYi4xIEmyedIjXcve3Gn5HRlLFmQ5rtnmFQGPTh1KXVNcT+0oCHosZQFCPoWl5WWW3NKQr7VhtmM6yB7p1zQ/4LG9vjleF7mKm59ddz5uV2uOY4qS61IkQa+LhyYPQWkRf8rE+Er3uR6XYpl3gn05meMVr8pmnJMFnbGIuxt4Efgd8AlwX9qx64C3O+Ec2hXplIgFE0romevTywz8dutebr74KxnESjw8M3e0bpjm43FVs81vK871cbAxZvud3vkBnpkzikPNccp/nQyDr5g5ImNy6qyVm/Wk/d9v20tjtJCFL1YaRE5OYD6xAwcOThJ43PZiB552eBknEhoffd5gyNl4dOpQFl17Hh6XQnNMJRxTGfW/b+i5Yz2CXtxuwXPzxhBXNSTJAsTNMTUr8YaORqZk/kyFdOvDcf3/T8aE/86Epkk8bmEQGPG4k7mI6e/MbEsCZLpnHlM7X4ZnxCz2EfAqFlGeByaV4vca22UUIjHxKbMRIgFwKQrxhLE22MNThuASxnaKAoUhD0/PGY0mJYoQuFwYFnvZiqQ4yB7p1zTFNDBf3+aYyhmFOXz/2fcpyvXyzNzRaJrEpQg8bsFntWHu+c0/DHa1+JXtBj0Ht0ux9bl24npflAqZ7TgnCzr8F0kpP5ZSngMUSSm/IqXcnXb4u8BdHX0O7Y3aphgP/yHJ3y7O9XF6QYDVN4xk/fwxzLmwP0+/u9si+79k2jDiqsbqjbsIeBTW3jiK9fPHsHx6GeNLilkybRguRVioDDet3YKqSQpDXh6YZO0z4FFACAPN4JHXd1jaPjCpVBdC2VMX5jtPbeG6kWey7M2dyVD6mgqqW2rYdRaNyYEDBw4ywa0IW5+nahq1TRH214f5tDap8FvbFDkuOerqxqiFnn7LU1tpiqnc8dz7+D0K//P7D/Vj81ZXoCgKhUE/vfIDBLxufvJSJT63wukFfn45dSgrZo5g3dzRrJg5gpWzRuhqeh2NlBR9LKHy1I2jDKURHp8xXC+ka34fLHtzZ5t2uR1YUd0Y5bnNn9GnIEBRy6buc5s/09+pOqS0fTeb+YoyQztzJE6AbTvzdDcWlxaq3J3rtxGLm+iPLsViy7+cOtSyoiOIggAAIABJREFUyFSE/bjmeXZC1bj92fcN497+7PuWXPwjYY0n/7qLSFxF1SSRuMqTf93FkXBrO79bsS1r4D8JJ+adBZ9b0eegeX635fo+MKmU7kEP33/2fWoao3x7bD/++4UPqG6IcstTWzkSTkaQF117Hk/PGcVDk4cQiWv84Jtf1ftIlbyobozyyOsfs2BCCevmjmbBhBIeef1j6zPCF6NC2vn29PntyYYOjcQJIbzAAWCmlPJ35uNSyn905PgdhfSk41QyaPoO1/0TS/nt1r0G6kJ+jodoXOXy0t40x1S+89S7hl2C4lwv4bh9OLumMUZ+jofnt+xl1eyRHGqKUdsU49E3djDrgn74PUY6z9aqeha/sp2n54zmUFOM7kEvtz291SCUsqcuzOFwXP8sPeR9KtXYcODAQddEOKbqYgfFuT66BTwsevlDahpitlGFVAmXbFgDmSiGg07LZd3c0dzylNVfpvu/1DugKaby6j8OcPFXexqiDcvLy3AF2/+amGHHmlg+vYyFVw9uWXR6LTvbHreCWxE8OnWoQ5VvJ7gV7IU+TOuLTAIeP08T8GiPdj8ztUto9vaeMAmlqFIjEjdGzh6aPARVGttlP24GCqRpMZqNaElUtRdniTribG1Gc0ylMOjhlksGcPPaLWliJDl4XUqygHdU5cEpQ9hfH+bBV5MRtgUTSrjn8kF6Qe+UD06pWC4rL+Ptey5GItKojNJeNIf2pcNm8u0nq4hfh25hSCljQAKIdOQ4nY30pGO7ZNC7N2xjXElPPUl34YuVfLi/ASEU2yT4+WsqOBJRcbWEttPRpyBAjtdFXJVcfl4vZjyxiUnL3mbe6gpeq6zmzvXbCPnclu/VNEb5V3UjB45E+Fd1IzWmXYg+BQGqG6KGv1O7be2VWOrAgQMHbYXX7dLFDqobosx4YhOvVVbb+tw71ydFnrJlDaToaulI0dVS45qPpfu/hCa5e8M23IpgcJ98y87vvDUVNMc6ftJgx5pIRQ3Td7DTd7aLc/10D578Cf+dieMR+rAT8MhW6KMtgiAAirCfWygmmqSmoeeFpn7HHc+9j2mtl/W4meY0LtO42Vy/TOIsHVwl66SGSxEkNPRr3ypGsolwPLkYmrz8bb7dIoS0taqePgUBoglpsZN0Hzx/TQVul4ve+QGdwphJNKe9Uxoz+XZzNPlkQWfkxP0fMAl4rRPG6hSkJx3bJYMWhXyc0zPEurmjaY6p9O0eaCnCqXBGd3uBknBcxe8W3D+xVDf08SXF3HP5VxECahtjDOgZtE1OjsRVlk4bZqgBkyo/cM/lg1j08kcsKy+zcIRf+HtSU2Z8STE/vKIETSZpOfl+N8unlzFvtbHGh0O5ceDAQWchPcE93c9mSsDPRqgjJdjkdQmLkEKK9qMo4piJ9SkBCFVK+vUIdrjgQiahKYc10TWQrcCIIjC841PvavM6WlFg9Q0jcCsuVClxCUFCUy2CIEVBL0/NGUUsIfXolNctKDK9q0UGgRFz6YBsf4fHJXhi5nD21kX0/k4v8Fvk/kWG32s37tj+hcy5sL+hFEH6M6RmEHvRnFVcmyEEqKr9dW2IxPlKcYin54xCEYLiXB/jS4r59th+NETix/TB4ViCmgZ0v6lluH/tjeKQz3a+WxzynZSCfZ2xiHsZeEQIsZ7kgm4/GOOnUso3OuE82g3pCffmZNChffO567KBTP/1JoMBvfT+Xq4Ycjq1GQRKgl4XGvDkxl0smFBC725+NImhMOLy8jLuu6qE7zy11RDCbowmAPSk6vpwnAdfTVKL6sNxahqjFIY8BirmI69/zC2XDOCGf+9PTUPMUBsktcBbMKGEwqCX4lwfvbsFTnhjd+DAwYkDMw3wWAn4xxLqSKceFoV8LJo42EDN8rmFLoN9tMT6lFhFn4IAQZ+Lxoi9sImnnfzl0YSmMomZOKyJzkUmIRLz7r9E6O/41Ebskxt3cd9Vgw3t/B6Fg41xblrTSi9cWl5Gfo7HMvbhcMKi6nhayFgU2SVEBoGR7ARLzIIqQggaIglDfz+//nwKcoyLRymx/b33XnmuoV3Q66J8zJkWOmUwrS6ZW9gLm5h/g4PjgMSgOJlCn4IAeX43n9U24/coOk1yybRhrHn7Uy4/r9cxffCHBxpY+GKlXuT709rmTvFVbrfCoJ65PDtvDAlVw92iTqko4qQU7BPmRNl2H0CITJwSSTIvV0opu8wbx1zR3g6HmqLsqQvz0vt7uW7kmQgBuw8288jrO7ht3ADdsaXQpyDA03NG863H39ELg6fvTP1y6jB++acd3HvludQ1x5m/poIFE0pY+GKlpZ+FVw9m1srNhs8WXXseq97ezZ2XDWLPobBhZywS1wj63PjdCtc99o6lv3VzR9t+vmBCCfNWV+h/P3/zBRTlGl8MXyJO3Ceug3Asuz3rnpeO2cfuRVe05yk5MMKxWRscy27Td05VTfLjlyqPmhNXlOvlP68oaVFOS+Z+hWPJhZhEcu2SjeypCyfzxmz8a8rPpUtUu1tU2OIJSdDnojmmoUmNw80J8nM8/OiFDyy5Hsk8Zx/Fef6MvyebnWBNkxw4EmFffZjappguC3/buAGcXRwk5HPz+ZGovjC9bdwA+vUIkuNz0SPYLlRJx25tYLbb+uYonx0K67S01CLkjO4B8nNa35vZKuftrWvO+L4+vSDH0O5HL3zAxLK++iJpQ0UV9155rqHd/vokLc7c33PzxtArv5V69vnhMJ8cbLI8W/17BOnZrbXdvrpm7rMZ974rz6V32rgHGyIcao5ZInbdc7z0yG19NvbWNbNq4y4mDT9Dj8Stf+8zZoztp/+Omobkc3AoLZrYPeihd36AolxLuWHHbm1gttuDDRFUTWa85zuqG1n19m4mlvVl3uoK+hQEeHDyEAqDHiJxzcD+Ss9LXj69jG5+N5GExsHGGC4h+J/ff8g9lw8ylHHpzEVUTUOUa5b8LaPP7yI47gvRGZG4izthjE5FOKbSLeBmwpA+hkjZsvIycv3ujCH/PXVh9tQlk0NTO1PFuT4aowleq6zmhn/rz6//+gmrZo9ECGuoeU9dcoFm/qxv9xwWTCihtilm2BlbOm0Yj76xg9cqq1k/f4xtf9HE0WsIpf526DkOHDjoTGQS7TBL/XtcCl6PYFn5MA42xZiaxipIn1isuWHUMSmZsYRqO9FeMm0YW3bXMuysQp1V8egbO/jhFSW8VllNTUPMEG3oHvRYRCOOt3SLXftHpw4lGtcME6FVs0fy21vGcqA+yrw1FV/KBOlUR1NUZc3bn7Ji5ggDHfC7Xx9AfuuaBikluQG3IQLscYs2138TAnuxCNMtj2cpbJKtYAnZjqtqHAkbI3Y/u+58cv3GqWc2wiZx1V505WQVrOgMaMij3vN7fvMP7p9YSl7L/dpTF6Znni+pTtoyjy0MeukR8lHXHOWhKUPYVx/G6xJc/3irH15aXsaA4hA+j2Iow9FZCr5w8gr2dUaJgbeO9V9Hn0N7w+t2oQiFm9ZaZUzjqrRNqkz/fGtVPfNWV3DHc++zu7aZfYcjeij6tcpqZjyxCZdin5zZHFMtn31S00QkrnFLC80ydT43rd3CxLK+QDIB3q4/VbM/31RYPPW3Q89x4MBBZ+Jooh0pqf8zCoP0yg9QGPSjKIqex5tqn55sv+tgk+7rUpTMdKT8nJ1E9c1rt3BJSS9uXruFScPP4Oa1W3itslqvs5Ty6Skhq+aYZhFcON7SLXbt65riFkGBGU9sIhqX+gIum74dtC+8bhcbP6nl0of/zCUPvcWlD/+ZjZ/UWt6b1Y1Rpj3+rkGcY9rj71rkz1N2lY4+BQGLsEkmsQiz7bmyFDbJVrAkW5EKTcL31v3d0O576/5uEUrJVtjEVnTFSYlrM+IJSVyVtvdcCKHfV78nacd9CgLsPtjMJzXNvFZZzbzVFUxa9jblv36X+uYEO6obiauSWSuNfuumNRXMubA/tzy1lVkrN3PdY+8wa+VmZjyxqdN81Mkq2NcZkTgAhBDdgTFAd6AWeEdKeaizxm9PFAa9VNU1267qw7GERWRkaYvU8NobkwmikXiCWELSLeBBlRJVk6yYOZy71v9D78ejYEnOXDptGIUhL+NLinmtstogYPLwdecfNaK27M2dlgTjpdOG0RiN8+jUofoCMBVRfOT1j4HWWkOOqIkDBw46E0fbOU2nJQohcImkBLld+wHFIdbcMJKgz83aG0fxk5cqWfbmTh6YVMqKv+1iYllfPfe3IOBh7+GwbT9SSopCPlyK0I97XIKVs0ZQlUZj71Pg57G3dnLLuK9Q0xBtswiJXfscr8u2j0yy2if6LvOJgsKgl6fnjCKaJjDicwvLezOhSVtxMrMITrYCKGqG/lRzPTkBP7vufH1BlYqImSNnfo9iEfxZWl6G32Pc71czCJGYI4UJTXLbxWczdkARakuB6I07amzbHeu6ZPytziquzUhokpe37bPONcvL+O2WPQztm8/8i87G61b4y10XI4Rk98Fmeub5WD69zHAfzijM4f6XP2TBhBLLPdpaVY/XrXxhH2VHRwey+ixdKCudrdAZc9uOFFTplEWcEOLHwB2Al1bOZ1QI8aCUckFnnEN7QlFExkTmg40x/GkhY687+XPL04ROVswcjgZc//g7hofmkoFFuoRrOK7xwt/38NScUSRUyae1zfzXbz/Qk0tvvWQA+w5HePDVZB6IEBw10XRrVT1PbtzFurmjiSY0Q38PTxnCb24aS1zV8LpdFAQ8/OSaUu698uRR8HHgwMGJhUw+1uNSLDTD+yeW6nLm5vYA9/zmHxZKps+j8N2vn2NR4c3P8dj280lNE3ddNhBJq6/1uYWF4rW0vIyCHDcel8I1S/7WZhESu/bJRaL9NXEETr48qKpGvZ3ASK6GorTeA79bsc3nNBes1rIUBPF7XBn6M953r0shL+A2UNnyAm68JsGSaEJSsesgT80ZjZQSIQRvVO6ncHBvQ7ugL4MQic84bn7AxVdPz+f6x4xznW4BY7tsrosvQ5vOpOSdbAh4XAzv110vwl0Y9FKU6yPkT94fs37D/RNLWfX2bm4ddw4bKqr0YMIDk0rJ8Sr88IoS6prjer5x6jtPbtylR4MtvivL+2dHL181eyTRhHbMz1I++GiCVR2F46XRHy86Q9jke8BPgV8Da0gW/z4NKAdmA7dLKR/p0JM4DmSTaF8fjgGShohKTUOU2qYYGyqq+M7FAwj6XCx+5SO+PbYfT27cxX9OOJepjxsTlJ+bN4bGaEJXklz25k5qGqOsmzsaTSalWN2KSMq/avCtx60JzimBkz4FAdbeOIqn3tnNN0tP5ztPtUYAV84aQTSuEVM1cv0eQj4FVYP9hyN6knxq0bjhprEoQujGXRDwUBeOd1Up1i5zIl0FjrBJl4djszY4mt1WH4mw/7BVyKBnnp+qQ2GiCZWeeT5Cfg8JVaJJiRDw4xcr9clFSjTqtcpqvd8+BQGenTcGKaWteMSG+WPYUx/mu88YoxY/eelDinK9ejmW3QebOadniOseSwpWzb/obPIDHppjKoN65aIAo/73DT15vjDo/cI5cZkmKAOKQuyoacyq79SusKZpqDKZp3UUH+/YrQ3MdputEEn1kQgf7a/n7OI8ElryPb+z+giDeuUbRHAONkTYfzhiYfT06uY3CILsq2tmis24z84dbRAY2VvXzB8+2M8lJb3QpERpWZxdem4vw/l9fjjMgSMRyzN3Wp7fIGySraBKttdl/+Ewk5fZCK/MH0OvlnH31YeZYiPO8uy8MfTON9LkcOzWFma7zWQ/T88ZjSKwvccTy/qy8MVKi/jd6htGsrcurG+YpfeXZKEllTDDMa0l11Ggaho+t4ueef6MPjCdcXHf7/5p8OUrZo6wCAnaffZlCpgcp6BKlxQ2mQ/8XEp5e9pn24G3hBCNwM1Al1nEHQ2aJtld20RDJE5Ck4aX/LLyMtwuWPzKR0ws60vPPB/fHXcO1Ucihps3tG8+fo/C7c/+07BT8eCr20lo0iD1f//EUkI+e6GUMwtzeP2Or/FZbTPhWIILB/bkl3/aoe+m9MzzcTicsCTnp4RO0sfdWlVPc0zlf3/fOvlJUSpTfztJ8g4cOOgsaJokodkLGdQ3x0hoGn/5uJqJw/uy26Sstnx6mc5UUASGlz4k/ee++rBe6818LBzXyPVboxaXDCxi2FndDT567Y2jbBWHl5WX0T1orFt3rNIFZmRqD9j2kU3fqYXhw3/YbhGmcHx825FtDTOfBwpCAX1hk4pM+UyVA4SAvByPRQDFUl8tw7hmOqXfo1DWr4e+oZyRJiml7TNnR8/MRtgkk0CLmT6ayEAHThctydjGnGDnIGvEj3J/PC77e5zXIt5nFr9ThOD0/IBtf/XNcda+8ynlY87k0Td2ZOV77Dax7p9YSk1DjK1V9YA9vTwT5fzLopZ3tKBKZ8ShzwIyhQFeajl+QqC2Kcantc0caorrCzhoFTXZXx/Vkz131jQxb02FRVBk/kVn67trqe/evWEbt40bQDyt6GLq8xyvfTKmqkm+/cQmZq3cjEtRuHvDNkOi6b+qm2yT81NCJ6n+5190dkuyapPh2Pw1FYa/nSR5Bw4cdBZqm2Komr2QQY7Pw53rtzFp+BnsrYvoC7hUm3mrKzjYGGPe6gpdNCodfQoC1DbFMoqbuF2C2SvfMyTgz175HlcP62MRc3ArgtvGDbB8Pn9NhS4ukU5rVBRBUa6P0wtyKMo9dgkAu/aZ+sim75RYysSyvpZzdnx825GtcEhjRNMpl9Aq+tAYMalExjVbAZRI3Ngu23Ejcftxzf1lEg8xE7ayFVTJJNBiFkpRsvgdrgx9OXXi2o5M92f3wSakFLb32O9xGVJ1Ut9xKQJVYtvfgSMR5lzYX5+DZuN77ISdUnPWFFL08nTYffZlUss7WlClMyJxtcBg4I82x85tOX5CIJZQdYl/88q6KOTjrMIc1s0dTVzVOLs4yIIJJRTn+nh6zij21kcQQGHIZ7sqP6tHkKV/+pfl80hc1TnFqQT8olwfQa+L39w8lmhcQ5OSBRNKeL3yc8aV9CQ/4KEw5LUdx7x7Uhj06hG5/7qyxJCsemb3gOFvzdnxcuDAQSdA07SMO+9qi1y6SxEZd137FQVZMXMEPUJelkwbZqjflfJ3YBWPWDptGIJW/55K7M8PeFAEjO1fqPvY+nAcv0fhrB45GaMhXS15PrUrfLQSCw6OH0LA0mnDONgY06O3PULetkemMgiCmNuJDAIoduNmI6iiSnvBEnNEUcsy8uhxCcvzZy4dkPodj04dSl0ajbMg6DH8jmx/q4Ps4XEJlpWXsavmCEPPLERtofhW7jtMIkNZiuaYyrLyMsIxlXVzR9McU+nbPUAkruJxKayaPZJ1mz7lokE9Oa2bH0UkI8juFkGobH1PpghWypf2KQhwZmGORazE7rNsfXBHCJB0tKBKZyzingcWCiFqgWeklHEhhBuYDPw38GQnnEO7IFkwNlknLj1Bc2jffO66bCDTW2rGjS8p5tZx57DwxUqKQj5DMu6KmSPsE9Bdgo2fGNezfQoC7DscYcvuWr477hxDDaBfTh2KIoSBM59Ol8w0jnn3pFvAw13rt1HTGKVbwMOtTxtVKtOTV5dPL6Mo15677MCBAwftAU2THGyKUX0kauvDEi3lWlRNZhT6EKBTwsaXFLP2xlFAUuDh3t/9U6fjPPjqdh6cPISeeX52H2ziv377AT+9bgh9CgIWmuT4kmJuuWSAYUL63PzRejkY8zm4FUXPhesqyfOpXeFUFNLyHnKEUNqElECIuU6rWTjEm0mAxtQukyBIvkkQRGYpgBLIIAoSMIlKBL32giXm+rRHEx1KR1yVvPT+XkP9vFQR73S4hCCeMNI4H54yxBRlE7a/9b6rBptvh4MsISX07e5DkmextUz3uDDkRZPSUrA+Pf1mxawR1DfFmJ4m6Ld02jDGlxRn7XsyCUH1zg/wt7svPiq93O6zY/ngjhIgOV4a/fGiM4RNcoHfAxcAKnCIZJkBF/BX4JtSysYOPYnjwNES7TVN8vmRMIeaYsRUqS94zImUy6eX6eo86f8PrQs+szM9p2eIfYejOuVhfEkxP/hmMoHe61b0ByyFTMmbqWRTu3HMD1p6Idz0Y3b9pf7uItXtnVWkCY6wSZeHY7M2sLPbVCJ4plyz/Bw3moRIXEURgpqGKHeu30ZRyMdt4wZwVo8cDjfHCbccTyXkL5hwLhIJEn78UiU1DTFuGzeAfkVBdtU08fI/9jOupCeDe+fx2aFmQ24QYPHlAH+562IWvviBJcdjeXkZPUJegxjE0fBFd4BrGqL88PltugiB1qIs2DvfT8DjNghVaRKOhONoUho2ATNMWBy7tUFbhU0+Pxxm3+GIYbP0F98aSu9uVuGQbPo71BRhb33Uoop5er6P7sE0AZQsRUGO53d8YspFfWBSKf17BA2/42BDhKZYgrhKa26fC4Jet1GgJYvzqz4cZqfNmGf3CFJsfc4cu7WBRdikPnxUgSfzPX54yhB65fuJxCV3Pvc+kEwRKgx6Kc7zc7g5xr7DEbrneLn92b9b+lw1eySLXv7Q6i+nlzGwOOl7Un4w4HXx+ZFoh6k6mnGcAiQdha4nbCKlbBBCXAhcAfw7yQXcIeAt4GXZ0avIdoSiCBQhcLkEBX43q28YSW1jjPwcY3g4PVxsDh1vrapn8SvbeXrOaOrDcfbVh1n8ynbuu+pcenXzsWLmCAJehcPhBOW/TibQr58/xhJWzkQjStElU+Osnj2S6oYo9eE4bgX+68pz+c8rSnApguqGCPdcPoj6cByfR7EVAChOM16HbuPAgYOORopGs6cuzIOvbmfBhBJ65wfI87v5yUuVhk2o57fsZfLwvjw3fwy1jTHD7vADk0pZ9PJH1DRGuX9iKSC5cPGb9CloUe5NaIbyAikmQ36gP4tf2c4Dk0sz+vUUVCl5rbKamoaYIUJQEPQQSWRHP2+PHWBN0/SJUYr98f2W3KbxJcXcNu4cy7V5fsteFl49mH49guT4XPQIHjtHz4E9MtEkzTXM3C5ByCSaE/K7cZvohZlEdyz11VRJnt9lFEBxJT83nl92oiDZ0j0jCY3Fr2w32PziV7bzs+vPN7RzKXAkoloWmXl+49QzU53DeJqwSTjDmD83jekgewhkRnGTSEKjb0FSqTKmauyvD/M/v/+Iey4fxB3Pvc+jU4cSjWt6DmWK3rqhoopbLxlAkSl1aE9dmKaYyoIJ5+JS4Jm5ozkSjlNVF+bnf/yY2y8diM+tMOOJ1ujdqtkj+c3NY4kntA5XSe9oAZKOQqfUiWtZqL3Y8t8JDVXC/voo0EqdWD69zBD2TQ8X24WOaxqj/Ku6kT4FAeatrtATP7sHvcxauZkFE0oMO74pcZT0PjLRiNLpkjWNUT6ubtTHWHj1YM7xe5j6q3ctY/zh9gtt+wv53Ia/HbqNAwcOOhLpNJqtVfXMW13BH//ja7oqJCRfrneu38aCCSVMXv62hZmQfnze6gru3rCNp+aM1o9VHQpb2t+8dgsLJpRQH45T0xhlZ01TRr+eQkpYInWe0Bq5MIs3ZIJdAv+cVe8d1w6wmiY0sWBCiUHsZWJZX4vIVerapMrUPH/zBc4C7gsgJRBhfn+6TNc0EteYtWKzbaQrHZlqHpr7i6uSab/adOz+MtToMouCZPodZlt2K4Kaxqhu85naNcfsBVXWzR1NQfD4xs00pvmaOMgemjz6tbeL0NWH4+ypC1PXFLf40Ls3JP3KTWu36GWw0r8b8Lio3H8Er0uxMMkq9zew8OrBhv5mPLGJ52++wBAF7igcbx3ProJOWcSdTHAJOLs4SCzRunO07M2dPDR5iL4jsaGiiqXThnHT2i0se3MnD0wqNYSkl0wbRlGuF02Dt+68CLciaIzGcbuSibv5OR4WXXsep3Xz4xKC+nBM7y/VR/egx/LZ0mnD+MUbOwB06lFcVZn372cxdfRZunLU2P6FBL0unpw9ks9qm3n5H/vxuhWenD2SQ40xogkVv8dF96BXT1RO7Q67lCTlogvWjnPwBXEsyqVDt3TQGSgMelleXsbPX/+YWRf005PjF0wo0WtbQivzYGjffPp2t5e2TjEJ9tSFaYom9GNHYzJsqNjDmhtGoSiwrLxMXwCl+/WUz3W7YMWsEew5FNYjK326B8jxKsQ1K8nEjjaZzQ5wIqFR3Rglrmq4FUHQ5yLP3+p/ZZrQRH7AYxCxKM712e6Kp1gbJ8Juc1eHIpJ02+ojUd0OivN8mF+P2Ua6hIBffbsMt+LSI2wJTbXtLxvBkmyEQyApdGE3rzALkeR4FZaWl1kibDleY05cQrMXSjGfnz9Df/60/hQBq28YgVtxoUqJSwjba+IgeyQ0ScjvMlz78SXFei3MtTeOMrAf0kWh0n1ougBU7/ykzeTneBlfUqx/96HJQ/B7FAadlovXLVgxcwSN0QTVDVFer/yca4adTp+CHP5810XEVcmRcJzqhminCeoVBr2smj2ST2ub9WfkzMKcThGl+iLokEWcEEIFxkgpNwkhNOBolEkppTwhFpOaJmmIJmiOqdQ2GqNjPo9ioEh43IJF156Hx6WgCMFTN46itilGfsDDM5s+5cKBPS2qaHFVsuPAYUb076EXTExRX/L8Cs/MHUVCBUlSQcjtEqy5YRSqJvG4BVJK7rn8q9xz+Vf5tLaZBf/3T4pyvdw67hxDbaN0AZSUCMu0X72r03DSx15WXsbb/+8SXIqgMZLgqkf/1in8ZAcOHJyaUBRBn+5+7rpsEDUNUUNyfHptyz4FAeKqxve/MZCqQ2HbXdRuLYu8msYofrfC0L75bK2qx+e2T9rXpGRiWR+dyj6+pJhVs0ciBPjcLuKaxoOTh1Cc6+PT2mY0KYmZ6mr94ltDEaDXiUshE22yZ57vqDvAiYTGR583WGp+FuRlMnFIAAAgAElEQVQkOD0/B0URhl1kTUrbvOvFr2zXF8DprI0+BQE87s6oNnTywu9RkNIobLLMpg6bN4PdeU3X3+dWiCckN641CkOY2wU8LnvBEo8xeuBWFFvhELdiuu8yuVhKn8soAtsZXLeA20Dj9Lqt84BMQilBk1BK8nlx8/Sc0XoxckWRhgQhn0fhYFOcm9ZsNiz0ugVMRfYcZI2Qz0VjRKWuMcy6uaMRAuqa4pb54n1Xncv2A42674VWNphd7vL9E0v5xRs7uG3cOfznhBI+q23m13/9hBljzmLV27stAlFLpw1Dk9I+v7gTBfWipmfk8RnDO3zML4oOETYRQtwLPC6l3CeEuI+jL+KQUv6o3U+ijcgkEKFpkgNHIiRUjaktC56U4S669jzbKvWpzx+YVMpXikPEEhrRhIbXreBSIBrXULWk0zzYGOP0Aj9SYhvCXnj1YL7aK5eEJnVefDShEvC4iSZUDjbG8LkFuX4vi1/5UE9wT1E099SF9d2SVBLqo6/vYFxJz4wiLKmxf3PzWASiKyR9puCsGk1oD2GTY8GJxH0hODZrg0x2W30kwpFIgvrmGLVNMX2n9rRufgSCA0ci5PpcHIkkdFET80TioclDUISgIOhFEbBu06f8+znFxFXJV3vlUtMQ5Rdv7NBLt5zRPUAsIYmpmh4xeLZiD30KAqy+YSQeV1JgasGEEjZUVDGxrC+De+fxoxc+0P1tSkTlWyPP5JyeIU4vyDFE3+za/uiqwdSH45bF3YCiEHXhONGEahG2Sub1jaRbwENRrs+wQMz0PkrRm8yiVg9MKmXgabl0D9r6ccdubWAnbGJ3b++98lyLIEhUVREoejRJouFzuSzCJtn0t6+umSk2tvHs3NH0NrW7z6a/+64819iuPsx9v/untd1Vgy0CKKs27mLS8DMsqpPp55etUEo2/WU7Zgscu7WBnd1u+bSW4Wf1INYS5f/RCx9YBO5SG1m7DzbzyOs7qGmMJvOK4xrVDf8/e/ceH1V5J37885y55wIJIUEh8QKlaLQoBCng1utqbUVdC9ZW8AItF1nXXdfrbpetXdpdEa2ttQiyFS/gKoVe7a7aYq27XqoGVrrFIj8FJVYhQAK5TOZ2nt8fk3MyZ+ZMMpNMSALf9+vFC8icnHMm8+Q5z+X7fJ+Ia5K9e770GWIJTc2IEHsOhhlTHuTOjX9g8XnjXNuZy644nWjCdH3N6mDqZF4qAl6PI1lTISLDJLFJitROmdb67v64xpGU+oC8/6ozaGjqWnS/YvZEjhsedA2RGJOygPza6Sdx4/p6x8jDDzpnw6yHans04QjTTD3XyNIAjS0RO8zBSnc9/7GutLCr59YB2jGSYSVFmVRTltHIWTlnMn6PcoThuF071rlAv6eQHyGE6KtYLMG+log983RxbRW3fX4CB1qdKatXza1jXGWxoz62wspGl4U40Bbhpqe6sgAunzWREyuK+Mojr/PwnMkMC3ntEeEZYyuYO/3EjP2sADbUN6A1NLVFaWgKc+KIkF3H/njxtIyR4+WzJlI1LEDC1I5nx8NzJrsee7A9SsjncSzgLw/52NnY6njmpGpoCmMo7Po3NY11ezTuevy4qhJevv08PjzYDmAntbr3uR08dM0kKEb0kqFw/WzTJ7r8XsW+1gQ3rkuZYZtbR3WZsymmspyvt/vOZT1f2v0ppbNcN3P/t0vPGJMxw5Yedpk1QYvLfnI9nS/rMS4zgCI3QZ/B2KphXNWZGdT6vBtboo6w9YNtUWaves1uu1aU+GmLxPn+5ne545JTXD/j48tCXP+os74eX1VCVan7XslFfg9FuIe572+N0hqJZ2QqTs223tfIsKGa2KRfYyiUUn6l1EGl1OX9eZ3+lrrw3FrcDskMkIc74uze325/zVJdHmL3/nYOd8T5wmeOtztw0Lm4t3Pneuv/t2/cxp6DYUqCPtdzlYV8dgcOkovVl6T8v6EpzKJ19fi9HrugW/deXR5i8XnjHF9vaEou5C8KdF0v9b2lXtvv9fT7rvNCCAGwrzXiSMQxq66Gj5o6HMk6GprCLF5XT6xzzzjATi5y64/fRoPdgbOOv3PTNjpiyUGyoM/D+43tdh264JyxGfXpkvVbWHDOWKrLQ3xwoJ2gL1kHFgV8dl1qmmTUq3du2kbI58HTmS7benYEfR7XY/2eZEY2hWJMeRGVpQGaUmbmstXLyRC2rvrXMBSVpQGK/F7X40M+DyG/l7t+8ge+uub3XP3I6yx6sp7G1ojU431kavdykL6cJ1uij/ao80Cd5XzpgVNWApRUbsk+cr2/bOU5/bhYQrv+vsTSsmJmvb+03mgu58t6THzIJDgfdDpiXRl6oevzXnzeOPuY6vIQB9qi9us3rt9Ce9Rk3mNv8cL2fXYCqFTV5SE+PNCeUV8vOGcsJQH3+qk9msha1xX5M+vOxevqHW3oBU+8Zd9nbwzVNm6/duK01lEgDnT053X6W2oPfdVL77F6bh0X11ax+trkyMKDm3fycMrXNi6ezhPzp/J/Dc3UHj+MsZ2jxZNqylh9bR3PLJzG0pm1nHp8KZNqyoCukQiPkUxuYhUma2TEY0BlSYDV19bxy5vOpvb4Ydx/1RmsvrbOcY7UBe7W/S6fNZGKYr/7iJipkyOB5SE7CUvqtVfNrcM0TQxD88T8qY7XCrnrvBBCQObsQlnIlzURSdzUrO6svwB7dL49EnM93mMkE1B0xE3HOT2GynK8YvmsiTy4eSetkTjLZ00kkZKu3UosYdXrq6+to7IkGd5ozZRZx7ZG4q7Hej3JpC2pC/jTnznpz4SH59ZRGvCgteZQOMJHTe18eKCNj5ra0ZgZP5PV1ybrcY3U4/0h1y0G4qbm5vPH8T93ns/vbj+P/7nzfG4+f1zGzFmuM1hKwfeuPtPxeX7v6jNzTqjidn99Oc5tBtCtPdObGcVcrylyZ/1M09umJ1Ukw1Otz2vVS+/Z39PQFMbXWWc9s3Aaw4Je7r/qDGd9M7eOcVXFrPvaVEf71O81CPiMjDKR/P4gm+r3uL7WEXOfJStLWQ+Zy6yZaWoaW5L1ZWNLBDOl7FQU+1lz3ZQhVzceiYQiPwNmAy8cgWv1i/TUo6UpYTj3fOkzTD2pjFGl/ozFmivnTGbZs8k49ItrqzLCFFbOmcw3L6/lW7/YTmNrhPZognDU5PFXdzmyTVn/v+OSCax9ZRfXzzjZkT3KWuzf2BrBSEslvHVPc+f3n+a6oNprKJ58dTePzZtKc3uUWMLke1efSXmxnw87k6NY6yZGDQvyi5vOJhwt/K7zQggBmSmvm8Mx/B73hBBxU/P9ze/yxPyptHTEKSvyMTzkoTViuh7/7t5Wlj27nSfnT2V/S8Q+JmFq9/rRY9h1676WCKteeo8HvzrJPjZbYomg1yChnYksOmIJ12M/OdTBsme3Oxbwp2+zcN/zO1h2xemMrSzGYyiefHUXq/97N4s+dxIzz6x2ZPVbPmsiL+/YyxPzp+IxFIrk5uapYUdSjxeWL0v59Hqc4+TDQx5OHVNmr3G0OuTDQ87R/qxbDKT1fvweg2Eh575zw0JefGnXzXULhGwJf9ITquS6FYHWuLZnvnnZaRnvw/W6Ke8j12uK3Pk9hmvbdNXcOt74xwuIJjT/8ss/2qGVgN3JsdauVZeHeOiaSTz01UmUFfsxTc09//WOIyultV8nwN5DkYwy8aP/eZ9vXFrLP36xFr9X8cT8qRwKx2hujxH0GRxy2d6luty5pVZPs2Y97ceZGpJeqHV2R0K/JDZxXECpK4EHgd+T7NB9TFqiE631i/16E3lwW2hvmpp3Pj7MonX1/OCrZ1Ia9NPcHsXnUVQUBzC1RinF4Y7k6K+VBvvumadw0WnHEzc1HkO5Lk5fdsXpQDI2ucjvYf3rH3LFpDHOzJVz60gkTG76j62ORfXpC+lHlgZ49n8bMjJfrppbx+iyAB83R1iU8rBfNbcOpeDxV3Zz+ZmjGdNZqRsq814vrq3ijktOxe9VhHzefincbum3s1xjcP9WDQBJbDLoSZl14VZuY7EEO/a1OtbE3X7JKexviTg6QKvn1jGyJFlHtEcTeJTiYHuUUaUBNLBrfxu3b9zG1XXV/NXkakytk1n/EiZKJRMjmFrzb//5DmUhv+uauHWvfcCr7x9wbCxeMyLE/tYoD25+l29edhofH+rgQFvUrvery0NsWDQdaznP/rYoi56s7zYB1twfvUF1edcieqvB8cCvdzCrrobjhgWpKPHjUdAWNXnkd++xob6BX99yjr1GKPWca284ixXP/4k7LjmVG9Zm7iO2YdF0tNa5NFSk3LpIL7f7DoU7O/rRri0GSv1UlgSoSktYkkuij4NtHRxsi7EnZeuKmhHJrYVGFAcd58sl2ceBlg4+OtSRUb7HDA9SUdp1vr2Hwnx4sJ1bNnRt4PzAl8/ghBFFjsQrB9s6+Kg5krElwJiygOP+cj0ul+sebOvgo6aOjO0PxpQHHefqJOXWRXq53Xsouan3d361PaNN+c+XndaZeEfz5+ZkHbepfg83X/hpwtEE//qf7ziy3T69cBo797Y6kpxMqinj5gvHUzOiyJFcSimVkYGyPZIgEk+41pEPfPlMogmz12virMSEf24OZ9TVA5ScL5vBkdgkzabOv7/U+ceiSd6wBgZ10GlyrYGfn//1DD4+FOGGtW/YC+G/sqZrRM3arf62z0/go4NtnDqmzK6wrQQjqawQyuOGB/F7DZas28LWPc3s3Ndqj1KMGhakNGjQGkmG8IweHnRdeHzyyCISpmb1f+/mjd3NjlEOn0fxcXOEN3ft554vfYbRZSF7C4LG1oidZKWxJcodl0zA5zEc9zqppozrZ5xsNwZ6+oXpjZ5GSYQQxwaPxyDkM7jvqjOo7EzohNZ4DYN1X/ssSsHHhzr4p876K72z1dKZvvze53bw5PyzaI2afHXN61SWBLj78lrao4mMfTtDfg/Dg16emD/Vvo+SgIclF3yKL9VVE/IZLDhnLPtbI8xe9RozxlZw0wXj7fo9NSJi655mogkTv0cx457fcnFtFeu//lnMLOFg1qxJajiQYSjGV5bwt3/5aXvNinWNx1/dxU0XjE/+rLKEgR4Kx7h+xsl4DfeEVI0tEa744StSzxaKkVx3lpqe/OE5kzMWrOQaEug1kuuV0rcsSN8JItcEI23RBOte+4C1N5zl2K/tpgs/RUXKcQmt8XnTt0syMsI4O2ImQa9ybDGgtUkk5lw819KR4KV39vLUgmnozsHun29p4PJJYxiRkkjHzHJdM+W6kc4Ea6nHpH5d5K8jbuLzKNc2ZTiWYN7aNx2ZbFfOmcwTr+7m1fcPOOq7hqYw8YTO2DvObeuBYUEv//zzP3btYzkskIyo+M27/MMXT3X9/bD2XLS2oAj5PYwI+fnOlRP55mXdD/q7tS0ddfUgT1zSkyPRiTv/CFzDQSlVAzwAXESyo/gb4O+01h/29pyxhCae0Dy4+V1WzJ7ICSOKHCNqDU3JBaFPzp/K/tYoZ4+vdLxuJRhJH2FojybwGgqF5rtXn0HC1HxyqIP7X3iXylI/37zsNFoiJvHOBfxBn8eR4MS67jMLp6GB3/z9ubR0JDdJtKawn5g/la89/gbrv/5ZonGTlo441eUh/u1Lp+MxDKIJk29dcTqRmMn+1gjDQz57k0bANSnKgife4umF01CAYYDHMBhZHOh1QyA1AUDqNQbZKIkQop8daIuy/Lk/cfvnT+H6R99g6cxabvvx21SWBLh39sSMmacl67fw9MJpXNt2IiVBL4ZSeAzFj26oIxw1KSa5sWx7NE44ZrL2FWcoz0Mv7uTuy07j/f3tKCCWMPEYioMeg7IiPz+pb2DBOWNRStERM6ksCfDlszITS925aRtLZ9ay7NntxBMan6F48dZzSZiap17fzVVnndhtSFB1eQillL2xt6m1a9KBpTNrWbJ+C/+xIFnnu53zQFuUZc9u5+mF07i4tipjlH1kiZ8Xbz2XQ+EobZE4nxzu4LhhR2YvpqNRLK4znss3rt/CMwunOY7LNSSwpcN0JPdpaEomcnhm4TSGpeReyJbsI/26HkPRHI7y/v42uxw0h6MZ4ZlaOxMCWfeXfj6t4ZWdjVxQezym1niV4sXte7notOMz3u8z9Q3c/5udjvN9qa7acZyZw3VNE3tLkCI8RBMmP3hxZ0Zopsid11BZk+g8vXAaDU3JpHtLZ9ay6Ml6lqzfwtKZtQCMGhbkgavPJJYw2fjWh3ZEhFW+3dqNd27axtobzrKTUFWXh+w686kFyfak2+9HNG5y0QMv2///6ZKz8XoNu21oRXGZpklC44gycGtbptbVgz1xSU/6vROntf4dgFJqGHA6MAb4CPg/rfXhQl9PKVUEvAhEgOtJzvR9G/itUmqi1rqtN+cN+g2iMZMl538KRXIEw23EoDkcI26afHyow/G6tTg9dVRixeyJlAS8PPHqLi49Y4wj1OH7XzmT0qDX7ggu+txJPDy3jvao+wLPeGfn79Yfv+04f5HfQ8I0mTG2gpaOuGPDWGtUd8n5n6IhLZTBSq/9wvZ9WZOifNQU5tYfv22f55aLJvR6RHeopncVQhRWwjS5fsbJHAonw9PLQj57Hzjra6ms+i+W0I4tCB6eW8cPUsJt1s47ixHFhuuoc0JrbkurO7/zq3eo7FzrnDrT8e/X1xH0el3vo6LYz/JZE4knEnSgufC7L9v16YhiLyvnTHYN2bTuI+hT9sbe2bYWsLaC+XNzmB/9z/s8PLcuY03cfc/v6JzlU/zNBeMdIWgr50ymJRLjuy+8y7yzT7ZH2WVGrvcS2n2GzUybwSoNGhmf18Nz6ygNOqfYCp1gxOdVrmv209Pzmzm+j6DPoO7kkVyzxrm2L31z82Eh9/c7LOQ8Lmsil5T3YRi5beMgclcaMmhqc297WT/7hqYwVZ2dpYamMKPLQsydfqIjMuvhuXUEvXB8WcCu47JtWWXNoKbXVfsOd/CdX73DitkTMyIl1rz8vuMcqe3C1NDz9PKx5ropjChyvw8rkclgT1zSkyNS/JVS/wzsAf4beKbz7z1KqX/qh8stAMYCf6W1/pnW+ufA5cCJwKLenrQjamLq5G72B9tifHjAfVuBkoCX2zdus2feLFaCkWcWTmPz35/L0wunEfQZ/PPP/8jkkyoyRtP+9un/5aOmro7g5JMq+MHmdykv9rteN5bQdgfOOsftG7fR0hFn3mNvsfDccRkje3du2sasuhqa2mJ2B856bcn6Ldz++VPYuHg6xw0Pul6zubNBZZ2nLyleh2p6VyFEYSVMzZ2buurQ5nCMmy8c7/haquryEPGE5uannVsK3JiWgrrhYJhIXLuODkfiOqPuXHzeONetXLyGh7ipXe9jeMjH46/uwuvx4Ousu6z6NBw1eejFnXZWt6Uza3noxZ3ceN44ls6s5fFXdxGOds3AZEu3bX29ORzjhe37+MHmd3lm4TQ2Lp7O0pm1dphQdXmIlo5ExgzRkvVb+Lg5wqy6Gvt9WpEPfUnRfSzzKPdU+obKnGH7weZ3HWXgB5vfpaXDGRJozdilny89EUm249Jn9mLx3NLzGzm+j46Y+1YJHWnhlIfD7u/3cNh5XC5bJeS6/YHIXUvYdGzTYkn92VvtWuvfw0O+jLJ047p6oglobo/bdVxVacD1vOXFfte66kBbcm+6e59LJnF6+fbzeGbhNH719kdsqG9wnCO1XWjNtM2qq3GNGEtoXO9jdFnoqBi06veZOKXUt4ClwL8DTwN7gVHAV4FvKaW8Bd4Q/HLgda31/7O+oLXepZR6BbgC+G5vTmqNbFWXBynye1Eqmao6db3CQ9dMIuQzuP+qMzC15uE5k9nf2rXQuXpEiLjW/L/GVk45roSSgI/7v3wGPo/BjLEVbKhvYFJNGYvPG8fo4UEqSwP8z53nkzCToTmbt/t5aPNOVs2ty5hRC6dt8Gqdp2ZEESvnTKbY71znBl2jukV+jyO8yFr06fMalAQ9JLTmwa9MshtJqRmHUs/T0BQmHItzsI28M59ZoyLpa+KG+ijJ0SSX5CiS/ET0lTW7YEUvPP7qLu76wqmOr2Ws38iywXVqCuoiv4eWDveZvJaOmOv3utWNhoK2zu0G0hfaDwt6WT77M0RipmOPK2t25IXt++wwdcvX/mIsy57dzprrpjhmVrK918df3WWPYEMyWuKbl51GyOdxZIxbPmti1veburGu9TOSyIfeUwrW3jAFpQzHGjG3VPplIT9jRyazjI4o9lMW8mfMnBlZzpf+KPV5FA/PmZyR7CN9TVyuM3v5vI8ZYytYcM5Yxxq79PcRNzWfGT2cU48fhqk1x5eF+Mzo4ZkzhR7F2nln0ZCSyKV6RMjxPnJ9DyJ3cVOzZfeBjLbs6mvrSJimXY90xBJ2shlF9s9h1LAAjS3JgaDmcIwn5k91ZKpcPmsiD23eyRWTxth11aLPncTc6ScTT5j8+pZzWPPy+8x77E1evv08RpUGmTPtJC467Xg7scotF01wtAutKK5sM39aa9e2ZX+Fj+eRoK8gjsSauAXA/Vrr21O+9kfgRaXUIWAhcHcBr3ca8HOXr/8RuKq3J/UaioDX6JzZep0ZYyv4+jknOxbZFge8PPnablb/924urq3iby4Y71iY/P2vnMm3n30nmUwkLdRn5ZzJ1JSHmHzSCDvE0cqultpI+O07eylLSSncHI5x3/M7uPnC8XYssduC0lWd+9ilNiCSM3gmmsyH/+Ov7iIaSxCLa0aVeojGnWmMAylhE6mjw+/tayPoM/IO0Rmq6V2FEIVlzS5YqfUXnzfOHqm3vrZ0Zi0VxX6Gh3zcsXGbo/6zpK43A2iPJhxrNhzHtTs7cdXlIUytMTUZdaPfq/jgQJT/eOMDNi6eTjhmsnt/W1eiqLl1nDAigCfuPF+29VBjOtd4VBT7+eRwh31M6tYC4yqLUUrhMeCrU0+0R7Ctc/i9Ho4v83T7XEi9Zns0QTRhZqzJk8iH3gn5DA60aW5c5wwxG1niDHYq8XuYO/3EjEQkJX7nzz3gM9jvcr6KkvTgKYU/LcFIwkyQnuTOl+NavGCW95F+3Vzfx/CQh/NOHZURdpm+pULAq4i4JHIJlHTdX67bJIjclQQ8/MWnK2mLJhztO7/XoKLYa2e5/efLTuOpBZ8FQKnsn0NrOJ6xjcrKOZOT+2BqeOr13Wyob7CT942rLCYcSyaeSj2+vMhL0OdhZ2Oro/O1+to6xleWONqFVhRXc5ZtCPxeDxNGhY5I23IgEvQdiXDK4cDzWV57rvP1QhoBNLl8/SBQ3tuTBr0GkXhXqMuCc8Yy/7G3mPfYm1z9yOvMe+xN5q19k9lTTgBgVl1NRhjL3z79v3boSnqoz5L1W/irydWOEEfrF8E6ZvG6ev5qcjX/8ux2/F6DW3/8NouerKexNcLosiAPfDm54aLbgtLF6+q56wun2tPK1eUhfnjNZAJej+sC/bu+cCrDQj5uXL+FtqjJvLVvOt7rTU9tZfF54+yGjbVJ44Obd/Y6RCeZBTTAmPIiKkt7nyRFCDF0jSzy83DnZtVb9zSz7NntRBMmq9K+FkuY3LFxG42tEcaUB3nwK5Mc9dvDc+vYVL/H/n/1iBA1I5Lr3VKPu/+qMxhR7HN8bcXsiSRMzV8/lVk3KhQjin3MO/tkEqbm2h/9nnmPvWlnabtxXT2tHSbRzqx5VoO0sthvv4fUrx9XGrTru6qSgOOYxtYIlaUBvv2r7Xzu3t/y1+u34vca9p5LqRELZSE/xw0POp4LJ1YUsebaKRnvrbzYx6b6PayYndzIVyIf+qY96h5e2B51xvp1xE3XsMaOtAyL4SznC6edL54w+drj9fzld3/HBff/jr/87u/42uP1xBNp4ZkeI6Pcr5g9MWMfu1yvm+v7aIu4n68t4jyuNeKeyKU15biA12DlnMmO97ByzmQC6Sk7Rc46YiYJk4z23by1b3IobLLi+T/xNxd+mgOtEc659yXOufclnnx1l10/Q1dd+9P6Blo64hnt1iXrt/DBgXbm/PvvuXrqiY46POD1ZJSPJeu3cO2MkwEyEpIserKeprBzwM2K4nLbKNyq045U2zJbgr7+DFM/EjNxvwfOIpkhMt1Zna8Xmtv8evaNcJRaSHJGkBNOOMH1mNbOxZjWh5MttbM1KpRtajc1dCV9t3lrUXHqMenfb+pkSE5jS9QR5hONm/zrf/6JpxdOI5ol6cqhcMzxPSOKfUS6OdZ6j9kWHZ9yXClPLZiGoTT/8IVT+fsNb9ujwxKi0/9yKbdCDDY9ldtAwMunRxbzzMJpxE2N11CUhgxiMXhm4TQSpsbrMfAo+N5XzsRrKPxeg5KAyYZF04knTAxDUeQ3+OZlp/GNS2vxGoqgz0BrGBb08fTCaZid+3caKjm6vCHlel5DEc5SN0bjJnf/Yjt3XDKhx8QSv7v9PHweg6qSAF6vwSmjSu179KZ83eJ2jEdhR1CkzkSeelwpIb9zz063aAaAnyyZQUfMxKOSMxqm1nzr8tPx+xQPXTNJIh9y0F25zTXBSKGPi2U5LpZ2XDiW4N7ndjie//c+t4PvfeXMQfE+cjku2zYJf33BpxiJyKancqtxb2vGTc3tnz8F0Bxs6+pMr/7v3cyeUsPTnXWxoRTt0RiXTxqT9VzWtlWpbdCq0gDxhHsda5qaGNnqX2d70ori+s6VEzFNM589MAtuIBL0HYlO3M3AT5VSceDHdK2J+zIwH7hCKWU/xbTWfV2m2kRyNi5dOe4zdGitHwEegeRmiG7HWGEH1nRtonNhe/rUrRWfnW1qNzV0JX23eWuaujkcw+8xXL/fOsZK0Wp9fenMWhpbI5gaPuhMupL+vftaIhnfk/qeHPfZHqOyc2GqJ0sYQzRuMu+xN1l2xelEE6YjvEdCdPpfLuVWiMEml3IbCHgZE0h7PGXs55uD4p4PyeajJvd6NG5qGlZW4vkAACAASURBVFsjfHXN73nlzvO7DVNL3XAZkp200WXORfbp0o9pbIk4rmGNYrttv2KNOKerKu3mh9eHn9GxpLtym+vWAQN5XGNr1/N/MN5fT8d5DcWr7x/ISHLxt385HpFdT+XW1O5tQK+hmJPSvkt9raGpg5oRRWitueiBl3tuT3a2h602qHX8uMoS92t7DHxZ2sBu7cls9d6RZoV25nLPhXIk5qG3AeOAe4D3gNbOv/+t8+t/AGKdfwox5/hHkuvi0tUC23t70oqQn7KQYYe6rHn5/Yyp/Yfn1rHxreRWdJvq92S8nhq6kh7qs3LOZH6+pYEVs5OhieXFvozwh1Vz63hx+8cZU8ZWOOPKOZN5cfvHVLuEDD08Z7Ljetb/V730nmuYRfWIEFs/OMDDcybj96qM6fOVcyaz8a0PWTF7IjUjQo5zS4iOEGKoqwj5M+q9FbMnsubl9/nBV5Ohm2WdKdTTnwMlQYOKUGHqPStcyC1MSAwObmXl4bl1GWVAjuv9cbmeS+SuIuTH48H151oWMrK278aUB/F7FWtefr/H9qTVPk1tF1pfC3iVa4h5VUlgSNZ7A3HPSuv+HcBXSt2Ne3ijK631t/p4vb8D7gM+rbV+v/NrJwE7gbu01vd39/1TpkzRb731lutrHR1xEsQ51KGJJUyK/B6icdMOwSnyG7RHTUydnGL2eRSxhHaE/0TiXaE+4WjX9wZ9Bm2RBAGvQdzUGCqZ7jdmaszO7w94k5vNGkYy3a51HaWSm2/6vIpYXBP0GcQT2v5ev9fA1xkeZFrhQta9aU3Qa2BqiHbem89ILpZuj5oEvAbDAl7CiTitHV33a10r5PcwLOCjKRwjGk/g8xrJUKQ8s1PmSWJ+0nRXbiG3zJJHyjGawVLKrIueyu1A6+iIcyAcJW5q/B4Dj6EIxxKEfAYJDbG4SUWxh+ZwV904PGTgxUswWLhAlyOd8SyFlFsXbuU2tax4DUVFyO9aBuS43h+X67mQcusqW7kNJ+K0p7RHy0IGTWETn6HwGIq4qR3tQ68nuUl4ezSBpzPpn9UWDvqSW7DEE2YyXN0AdLKdGk2YeI1kW9gwDCqK/ZimZl9rxDXEfADrvV7r4z3n/eaOxGbfd/f3NdKsAW4Cft65D50GlpHcp251X06crCy8FHcTmVKeT2hK2rEVJb25q/6T+l78eBneTRRQxlS2hOgIIYa4YNDLmBw6Y909EwphsIQLiexyLStyXO+Py/VcInfBoJcg3oy2a3/XaRbDUFlDzIdivXek7/moS+ujtW4DLgDeBZ4E1gO7gAu01q0DeW9CCCGEEEII0VdH5ZCG1vpDYNZA34cQIlOhQjuP0bBMIYQQQoijsxMnhDj65dIZlI6eEEIIIY5G/Z7YZKhRSjUCHwz0feRhJLB/oG+in6W/x/1a60sG6mYGoxzK7bFQTmDwvk8psy76WN8Oxs/6aLsnKbcuhmA7IdVgLKP56uk9SLl10UO5PRrKRXeGwvvLu9xKJ26IU0q9pbWeMtD30Z+OhffY346Vn+Gx8j7F4Pys5Z7EYHc0lIej4T0MNkf7z/RofX9HXWITIYQQQgghhDiaSSdOCCGEEEIIIYYQ6cQNfY8M9A0cAcfCe+xvx8rP8Fh5n2JwftZyT2KwOxrKw9HwHgabo/1nelS+P1kTJ4QQQgghhBBDiMzECSGEEEIIIcQQIp04IYQQQgghhBhCpBMnhBBCCCGEEEOIdOKEEEIIIYQQYgiRTpwQQgghhBBCDCHSiRNCCCGEEEKIIUQ6cUIIIYQQQggxhEgnTgghhBBCCCGGEOnECSGEEEIIIcQQIp04IYQQQgghhBhCpBMnhBBCCCGEEEOIdOKEEEIIIYQQYgiRTpwQQgghhBBCDCHSiRNCCCGEEEKIIUQ6cUIIIYQQQggxhEgnTgghhBBCCCGGEOnECSGEEEIIIcQQIp04IYQQQgghhBhCpBMnhBBCCCGEEENIzp04pdSjSqmTs7x2olLq0cLdlhBCCCGEEEIIN/nMxN0AVGZ5bSRwfZ/vRgghhBBCCCFEt/INp9RZvn4cEO7jvQghhBBCCCGE6IG3uxeVUlcCV6Z86VtKqf1ph4WAzwH1Bb43IYQQQgghhBBpuu3EASeQ7KBBchbuTCCSdkwEeBX4h8LemhBCCCGEEEKIdErrbBGSaQcqtQv4K6312/17S0IIIYQQQgghssm5EyeEEEIIIYQQYuDls8XAFUqpeSn/P1Ep9ZpSqkUptVEpVdI/tyiEEEIIIYQQwpJPdsp/wrnFwHeBauAR4Bzg7sLdlhBCCCGEEEIIN/l04sYB2wCUUiHgi8Dfa61vBf4RZxZLIYQQQgghhBD9IJ9OXJCuveBmkMxs+ULn/3cAowt4X0IIIYQQQgghXOTTidsN/EXnv68A6rXWhzr/XwUccvsmIYQQQgghhBCF09M+calWA/d1bgB+JnBjymvTge2FvDEhhBBCCCGEEJly7sRprb+vlNoPTAMe1Fo/kfJyKfBYge9NCCGEEEIIIUQa2SdOCCGEEEIIIYaQfNbEoZIuV0rdp5Raq5Q6sfPr5yqlJLGJEEIIIYQQQvSznGfilFLlwH8CnwUOkwyhPEtrvUUptQ44qLW+ud/uVAghhBBCCCFEXjNxK4Aa4GxgJKBSXvsNcGEB70sIIYQQQgghhIt8slNeAdymtX5NKeVJe+1Dkh08IYQQQgghhBD9KJ+ZuBLgoyyvBXHOzAkhhBBCCCGE6Af5dOJ2ABdnee1c4A99vx0hhBBCCCGEEN3JJ5zyh8APlVKHgKc6v1amlJoH3AQsLPTNCSGEEEIIIYRwymufOKXUPcBtJEMnFaABE7hXa/2NfrlDIYQQQgghhBC2vDf77twb7iKgCjgA/Fpr/X4/3JsQQgghhBBCiDQ5deKUUn5gOfCU1vrNfr8rIYQQQgghhBCuckpsorWOAouAUP/ejhBCCCGEEEKI7uSTnXIr8Jn+uhEhhBBCCCGEED3LpxN3K3CbUmqmUkr2hBNCCCGEEEKIAZBzYhOl1B5gOFAMxIF9JLNTWrTW+sSC36EQQgghhBBCCFs++8RtxtlpE0IIIYQQQghxhOW9xYAQQgghhBBCiIGT05o4pZRfKbVFKXVxf9+QEEIIIYQQQojs8tli4GSSa+GEEEIIIYQQQgyQfLJT/hqQmTghhBBCCCGEGED5JDb5AbBOKeUFfgZ8TFqiE631+wW8NyGEEEIIIYQQafLZYsBM+a/rN2mtPYW4KSGEEEIIIYQQ7vKZiZvXb3chhBBCCCGEECInssWAEEIIIYQQQgwh+SQ2EUIIIYQQQggxwPIJp0QpVQV8FZgABNNe1lrrrxXqxoQQQgghhBBCZMq5E6eUmgC8DniAYmA/MKLz/03Aof64QSGEEEIIIYQQXfIJp1wBvAGMAhTwBSAEfB1oB64s+N0JIYQQQgghhHDIpxN3FrASiFjfq7WOa60fJbmH3PcKfXMD4ZJLLtEkt1CQP4P3j0gj5XbQ/xEupNwO+j/ChZTbQf9HuJByO+j/5C2fTlwJcFBrbZIMnRyZ8tpbJDt5Q97+/fsH+haEyJuUWzEUSbkVQ5GUWzEUSbk9+uTTidsNHNf57x3AVSmvzQSaC3RPQgghhBBCCCGyyKcT92vgos5/fxeYp5TaoZT6I/C3wKOFvjkhhBBCCCGEEE75bDHwD0AAQGu9QSkVBq4GioDvA2sKf3tCCCGEEEIIIVLl3InTWkfoSmqC1vqXwC+zHa+Uug74pda6qU93OAiZpuZAW5RoPIHf66Gi2I9hqD6fA+jzeYVw09cyW4gyX0iD7X7E0e+ku37V4zG777n0CNyJEF0KXRcOVN0qdfrg0NPn0JfPST7jwstrs+9cKaU8wFqSyU6Oqk6caWp27G1hwRNv0dAUpro8xJrrpjBhVGleBTn9HE/Mn0okbvbpvEK46WuZLUSZL6TBdj9CCDEQCl0XDlTdKnX64NDT59CXz0k+4/6Rz5q4fB2Vn8qBtqhdCAEamsIseOItDrRF7WPicZOPm8N8cKCNhqZ2mto67P//uTlMUziScY4PDrS7nnd/WyTzJkj+QjS2RPioqZ3GlgimKVl1hbtsZfaTwx2OcpNapg62RdjX0sFHTe18crijxzKfTb7lNJfju3s/8vsghBjM4nGTP6e0B+Jxs9fnyqU9YulL3drb8xX6utLuKbzUn2m2Z73VVsj2OWVrp6bKp2yJ3PXLTNzRLBpP2IXQ0tAUJhpPAMkK+k97W1i8rp6GpjAX11Zx0wXjWbJ+iz36sGpuHZUlAfs8k2rKOKmiiPuvOoPmcIxVL73H1j3NNDSFaY8kMIt1xnS2jGiIXEXjCSpLAiydWUtZyGeXsT83hzkUjjFhVCmAXaYqSwLccckEbt+4jYamMBsXT++2zGeTrZyOryyhKRzLCKnItVxn+x38c3OY2atek98HIcSglN4+sNoDp4wqxevNf0y9p/aIpa91a2/Pl8913Z5RqdeVdk/hpf9MNy6e3m1bYUSRz7V8uLVT0+VatkR++nMm7qhjmpqEqakuDzm+Xl0eImFqTFOzrzViV9AAs+pq7A4cJAvt4nX13HzheCDZgbvt8xO49tE3uPqR11n27HZu+/wEJtWUUV0eYtf+toyRChnREPkI+g3uuGQCy57dbpexOy6ZgMfALjepZWrxeePsDhwky5tbmVeq+wenWzl94Nc72LGvhStXvsLZy3/LlStfYcfelm5H+dLLtd/rcb0f6zj5fRBCDEbp7QOrPbCvteeZDDfZ6kK/1+P4Wl/r1vTz7W/LjCZym5HJ9bo+j/szyucx8j6XyF36zzSWMF0/h1giudQnoXEtH27t1HS5li2RnyHTiVNKzVZKbVJKfaCUCndub/BvSqnSlGNOUkrpLH/K+noPB9qirH99Nz+8ZrJdGKvLQ/zwmsl8+1fbOdAWJZYwHTNs4yqLqSwJsPraOp5ZOI3V1yZn4U4aWUx1eYjF543jzk3bHBXTnZu2cfOF41k+ayIPbt6ZMVIhIxoiH9GYZu0ru1g6s5ZnFk5j6cxa1r6yi+Ehv11uUstUWcjnKLPFfg8PXTPJUeaXz5qIp4fBT2t0NbXsXzf9JBY9We/6IM61XFcU+1lz3ZSM+1n10nvdfp8QQgyk1PaBpaEpTDzRu5BKt7pwzXVT7ERplnzq1ifmT2XtDWfxzMJprL3hLJ6YPzXjfB0x9/N1xJzvw+0ZUFkSyLhuPGE6Bg4bmsLcvnGb4+ci7Z7CS/2ZTqopY9SwoOvnUDUsSGVJAK01q+fWZTx73dqp6XItqyI/Qymc8jbgQ+AfgQZgEnA3cL5SaobWOrX2+DfgF2nf39LXGzBNk3MmjOKHv93J0pm1VBT7GVHsJ+hT3P75U2iPxvEaikWfO4k3djdz2+cn0NQWc4SmVZeHWDF7IsV+Dz9dcjbt0bhrxVQzIsTtP95GY2skY6TCGtFI/b7uRjQkI9CxTaNZcv6naGqLAeD3GCw5/1MUBzw8vXAaCa3xGgYX11bxwvZ9mFpnlNkHvnwG67/+WT451EFzOMbjr+7iO1dO7Pa6Ib8n4zyPz5+a9UGca7k2DMWEUaX8dMnZROMJlFLc/Yv/Y+ue5m6/TwghBpLPY7jWcV5P5nh6Ls/t9Low23G51q2mqTG1c52ZqZNRRqnn9Cjler70gb2Q38Pdl9dyMOXZc/fltYT8zuvGTO36XIilrHnLt90jeub3eri4torrpp/E8WUhTI3r59DUFuWOSyYQ8nsI+T0su+J0ivwemsMx7nt+h2s7NV2uZVXkpz87cYVecXqZ1rox5f+/U0odBB4HzgNeTHntfa316wW+PgmNPWv2wvZ9ACz63EnMPLOaG1Ni3B+eW8flZ1azaF0993zpM9z1kz9kjGz85MYZVJYGaGzBtWJ6r7GNxtaI60iFNaKRHhvuNqIhceTCoxThaIKlP/8/R6espSPObT9+27E2AyBh6owye8uGt7nvqjP4yiOv5zyCFjd1xqjehwfasz6I8ynXhqGoLA0AyTJ+y0UT2P5xS4/fJ4QQA6WqJMCquXUZa+KqSgKO4/J5bqfWhdnkWrcebI/S2BJxPCtWzJ7IsKCPqmFB+7iQ38OK2RMzBqfTO2caTXvas2fF7InotOZhts5tajhlPs8HkZvykI+/ufDTdvt17Q1nuX4OB9qiLHt2Oz9ZMoORxQGOGx7s1eeQS1kV+enPTlxBewhpHTjLm51/jynktbq5h4xRitlTTmDeY286Gqo3rqvnmYXTaGgK4/MY7iNMnWECbhXT6mvrGFns56dLzu7T6BtkjyP/6ZKz5ZfpGOHWmbplw9ssu+L0jLUZGxZNzxryM7osxCt3np/zCFosnnmeBzfvZPXcOhalNGKsB0BvR+pkhE8IMRR4vQanjCplw6LpxBMmXo9BVUkgI6lJoZ/budaR0SxhjU8vnOY4rizkZ9SwoD0j0x5NMGpYkLJQWthl1P18zyycBsVdx+XSuZV6vvCawjG7AwfJ53N653z5rInc9/yOZLs1bsrnMMj0SydOa53gyKy3O7fz73fSvv5vSqlVQBvwO+AbWus/9PVi6dP5k2rK8HvdO2nxzgQozeFYtyEAfWm45lKZSxy5SLgMPjQ0hSlKGzW1Yt6VgrU3nMWDm3faIYrV5SFCPk9eDQi38JfG1gjHlwWzlvfejtTJCJ8QYijweg1Gl4W6PaY/ntu51JGJLGGN6an8DUNxQnkRQZ+HWMLE19kZTW+3JLR2zXaYSIvTyrVzK/V8YaWXs617mrn3uR38x4Jp7D3cwYG2KPc9v4Ote5oz2q3yOQwOOXfilFKPdvOyCRwC6oGfaK07+npjOdzPGOBfgN9ord/q/HIEWA28ADQCp5BcQ/eqUmqq1jq9s5eX1FkzKw37+41t7rHhhuKJ+VO557/eYfmsiXYYptvUc3/+QkgcuQj63MtAe7SrQTCppow7LpnA1Y+87gh7ufe5ZLz7qrl1lId8eV03W/hLWUhG7YQQIpuBem6Hsjwrgr7MtXM7G1t7DPcsDmSui14xeyLFgcwx/lw6t6Kwsg20JjvSimXPbpfQ1UFOaZ3b0jWl1C5gOFAGxIH9wEiSHUEro0AZ8B5wvta6oeB323UvJcBLwGhganfXUkrVAH8EfqG1npvlmIXAQoATTjih7oMPPsh67dTFxlc/8jqVJQFu+/wERydt5ZzJ/Ortj/jaOeNQKEzTJKGT4ZjWzANgn8fnNfAayXVLPc3E5Zuk5ChdEzdkb7yQci23pqnZ8UkLC57sKgMPfmUSHgP++qmtdiy8tW7BUl0e4sn5U3l3Xyub6vfwnSsn5j3YEI+b7LMfCu6jq8cIKbOd8qlvB5OT7vpVj8fsvufSI3AnR5SU2049ldtCJhDL57ld8OumPSvWXDuFCcc5r9vYEuHKla9kPC/Swz33tXTwpZWvZhz3kyUzqCrtWmNX6PeBlFtbd+XWNDXvfHLYzhhthU8+/uourpt+ErGE5qSRxRT7PYx0mWm1ziGJ8wom7x9cPuGU1wBPAV8Dfq61NpVSBnAlcH/n6zHgJySzQ16b783kQikVJJl5cixwbk+dRa31HqXU/wBndXPMI8AjAFOmTOm2V2vNmn3U1E5DU5iGpjD3Pb/DDhcYXRbiB5t3sqG+getmnMyY8qKMc7hV0KmzHt1V1vl2yCR++eiVT7kN+AzH+oVhIS/Dgl7uu+oMFFCRsvm8paEpzL6WCIuerAfgm5flF8qT62itOLbkU26FGCy6K7eFHizN9bndL9c9Loe1czmGe7qti7bWVvXn+xBduiu3hqGoKPY7wl2t8Mm7vnAqfq/BcaXBrAOv8rkNvHyGxB8Almutf2ql89dam1rrTcBy4Hta6zdJduAuKvytglLKB2wCpgJfzGOdm6IA2TJNU9PYEuGjpnZUZ4pdSMYRL3qynlt//DbbPz7MhvqGZGalzoKf+n0H2yJ8criDtkicpTNrmVRTZi/2XXzeOHsBc/rGiaap+eRwR06bXaZer7EluflmZWmAMeVFVJa6j6aIo9eBtijXPfoG8x57k3v+609EEybN7THaogmsmXit3Texbw7H7H8rpRxrI9LLWfq6iSOxOWtP9yCEEP1toDai7o/rWgPV3bUXct24OZ/NyH+2ZQ9rbziLF289l7U3nMXPtuzpsW0j9X3faY29ufeiJ+vt9W/7WiL4vUa3kTOyAfvAy2cm7gySoZJu3gNO7/z3dqC8LzflpnPWbz1wIXBprlsIKKVOAM4GftqX66ePOFxcW5WRTcmaTbP+3doRx2N00BJOsGt/G//1h4+5cvIY18w/W/c0U9a55ii5kNiksSVih1u2dsQ52BZ1jGpNqilj8XnjaI/GaWzBDtOUkRGRyho1nVRT5gj9vbi2iru+cCqHwjGa22M8PLfOsVVGanleOWcyj7/yPn81uYYJo0oB93I2vrKEpnDMHpGtTJvhK2RSHRkFFEIMBvkkIskl/Mw0NbsPtPHBgXY7euLEiiJOqih2HDtQictyTfdfHvK5Zp1MX1+t0Fx6xhg707f1zFE4Bw2lvi+seNykNRrnB1+dxN/8x1bHs7+ixE/CNDP2CEyV2rZYfN44ezbPNHu3eb3IXz6duE+A2cCvXV67Ctjb+e9hQFMf78vNDzuv8x2gTSmVmvO2QWvdoJS6n+Ts4mskE5tMAP6BZOKVf+3LxdNHHKx94jYsmo7WGkMpGprC3PWFU2gOx+zQyGVXnM68x96kujzEE/Onct2jbzhGLe7ctI2lM2tZ9ux2e9bj4toq9rdFHXHKK2YnN1a2FqGmN8itCm3UsIBsKSAcrNHQxeeNs8vLpJoyrp9xsl0eq8tD/PCaScnsYKbmz81htNZ2eX7oxZ3MqquxyxLgWs6e+vpnuebff5/REUzNclmoxfmyfYYQYjBQWTa/Vqp34Y/N4Sh7D3dk7K9WVuRjRHFX3TZQCVByDfc8HImRME1HKH/CNDkciTHC2/U+InGTJeu3OOryJeu3JLci6CT1feHta42w4rk/JdugKRt4W+3XJ+ZPpTkcdZS5VNZm4dfPONnRFl19bR2VpUHpXB8B+XTivgc8oJQaDWwE9gFVJDtWXwT+rvO4zwFbC3mTnb7Q+fc3Ov+k+hZwN8kEJjcCNwClJJOvvAh8S2u9oy8XdxvxemH7Pv7pUpOQ30s0nuCq1a9lfJ+Vxr2hKZwxk2Z9vaLY75j1+KdLa+2GsHXM7Ru3cc+XPmNnukxtkFvHLHjiLXt/uvRryJYCxy5r1LQtErfLhlv5+euntvL0wml4FHzlkcyJ7q/9xVhHWcq2hi693KYOZBQyw5VsnyGEGAw8iows1MtnTcST1obNtSMSjiZy2l8tnw2wC52AIpes2uFowk6eZakuD2W8j3iWrQ0SKeGSUt8XXixhMquuhr2HI8x77M2M1w+2RQl4Dcdnlaqi2O/aXl30ZL1dpiXxSf/KuROntf6+UqoN+GcgNQVXA7BAa/2jzv//EAinf39faa1PyuGYR4HutkLotWwjXu980sKyZ7fz1Nc/6/q6NbsGyQrc7ZjRZSGCPoOHrpmE3+vJWln5PAb3/Fdy1GR8VYl7pacZkJE5MXhZo6afHO6wy0ZZyOdafkxTE/J7s5bl1LLkdkx6LHxDU5hxVSV5bRKeK9k+QwgxGBiGweOv7nIkiHj81V1858qJjuNy7Yhk29szfX+1gUqAkqtc34fPY7jW5V5P13osqe8Lz+cxqCj2Z22bHmiLUjUsmPX7DUPhMVTWMi0hsP0vr1zfWut/B07s/DPN+ndKBw6t9W6t9d4spxiyrBEva5GuNdK26qX3aGgK8+1fbWf1tXWO11fMTr5u2VS/h5VzJjuOWXPdFI4bFmREcddC4myLgdujCbbuaWbZs9vxGsr1mKDPyLhP2d9DGIbiuGFBu2xYHbJU1kMzW1nfVL/HLktux6y+to5N9XsyzhnyefolqY7bPUhZF0IcaRXFfm65aIKdIGLZs9u55aIJGXVRrok+rL09048L+jKbbLkkIhmoBBS5vo+qkgCr5jrbT6vm1lFV0jXTJ/V94VWVBKgsDbCpfg/LZ010/Gx/eM1kNtXvcS1zqbor05L4pP/lvE/csWLKlCn6rbfecn3NmhYOR+O880kLm7fv5cLaUfbIW90JZZhAR8zEqyCuNe/ta7NjwWtGhBhZ4iea0MTiZt6jZqOGBey95MqCXt5tbHWsm7NGOICjffr6qHozhdBduQVnKI1SCq8B+1qijgXnD8+t4/hhAUaWBjOO96jkaHNqWUrfA66y2M//2992REfdcrnPQWJQ3cxg0VO5HUxknzhhcSu3hUxYUugZjI+a2jl7+W8zvv7KnednbINk1euxhImvj3t75vM+ctlTNI99R6XcunArt9FonJ2NbXx/87vMqquxB2mffuMDLj+zmuPLgpSFut+7ONtn/PGhcEa5m1RTxkPXTAI4WtunfdGv+8ShlBpGcv3bCUD6HKvWWi/L9waGEmvEq7ElOavmtpgz5PNw3aNvUFkS4B+/eIpjYfKa66YwLJgssFaF//GhcEZB7jZEorjrl+b7v3mXpTNrqSj2U1UaYPTwkH0OWegrLG6V7Opr6ygOGDw2byqGguTSAxNTJx/4uWw677YH3PjKkiO6J6G1z42EbAghBlIua8QgmcQjvV3gdq7xlSXJRFMpHZbe1me5hiLG4yZ/2tuSkU3ylFGlverI5bNPrddrMLos5HKWJNl3tPBMU/PegXZHB25EsR+PR/GXtcfzTz/7v273LobuP+P0cjeppow7LpnA1Y+8Lp9hgeQ8E6eUOhv4JVCW5RCttR7ywcm5jAybpqahqd2xmBOSlaKVxGH1tXUse3Z7xus/XXJ2nxudjS0Rrlz5iuu5j5HOm/y2p+mu3GYrL1ZZdftaT2VyMJXBwXQv3ZAy60Jm4gY9Kbcueltuc62rCj0Tl+v5/twc5surX8u4vw2L1rHzmwAAIABJREFUpnfbwToS8qznpdy6SC+33bUNogmTRU/W21/rzfM0vdytveEsewAj9XqD7Fk9kPp1Ju57wG5gAfAHrfWxHdSqkntgLZ1Zy+jhQYaFfGgNSsG6r01leJbEEdF4Imuc8C9uOpuEib03nNdQdvhk6uiVZGkSuTJNTTgWd4yEWfu5jC4L8eW6ajbUNwAwY2wFn6oq4cVbzyVhaja+9SFfP+dTrpVrtjIYjiW63VemP8jvgxBiKMi1rsonnX4uYZy5zojFEqbr/cUTmft+FTrrYE/ni8YTzBhbwYJzxuIxFAlTs+bl96We74NoPGG3Y61lQateeo+aESECXoP/WPBZEqbG5zGIxBPsPRTGMAzKQz57P9juPvv0cpct0Y18hr2XTyfuVODLWuv6/rqZocAaWWjtiHPHJRNY+8oulpz/Kf7cHHZs4r1yzmQurq2y95ODrvAFt4q8siTAx80dLHLZbDl9OluyNIlcWGX1k0PJrJSVJYGMvQVXzplsHz93+ol8dc3rjtdSN1tNla0MvrevlbZI/IiGR8jvgxBiKMi1rsq1s5fPjF0u4Z7+LFkifR5nKOVAzBQWBzzMnX5ixobgxQGp53sr5PdwxyUTHG3XFbMn0tQWI+AzKPJ77C0irARnL+/Yy2VnVjtCbnsKt7TKXWNLRJ7VBZZPkPOHwDE/32mNkEXiyX1cZtXV0NQWy9jTZcn6LXzj0lo7a8/FtVU89fXP2gkYLq6tcpz35gvH2x046xy3b0zuB5ee0UeyNIlcWGX1wc07WT5rIjdfON5Og/3MwmksnVnLQy/uZME5Y1l47jjXzVYj8cwRWHAvg/dfdQYPbt55xLNPye+DEGIoyLWuyjWLZaGz//l9ihWznVkKV8yeiN/nbJwfaIvysy17WHvDWbx467msveEsfrZlT6+vm8v7aI+6bwjeHnV/RomexU3N2lecbYK1r+wiEk+wZP0WDrbFHD/vOzdtY/aUE+wOnPX1XMucPKsLL5+ZuG8BdymlNmutD/fXDQ12pmkmQyjLQiydWUtVaYBo3D0E4WBblGVXnM6nR5XQHI7Za+isxcKQ3DC8ujzEySOLXc9RFvLZ/7ZG4fJZLCyOXdZobkNTmPue38EP50zKSMazfNZESgIeoomeN1tNF/AaLLvidDvLWqAzFfGRDo+Q3wchxFCQa12V6ybehQ4l74ia3PvcDkd43b3P7UhmE0zZ8FmhufSMMRmzYtkiN3qSy/uIJUzX0D+3UE+RG21q1zbBsKCXhqYwRX7noEFDU7jbfeF60pdntWwa7i6fTtxMYBSwSyn1GnAw7XWttb6+YHc2CJmmZn9b1E5YUl0e4on5U/ngQLvrFPG+lgiLnqzPWMzZ0BRm8bp6Niyazjcv0/i9HjTa9RzWZuHpo3C5ZsISx67U0J2te5qJm9iVNXSNrD29cBrvN7a5lj+vx32y/kBblOsefSPj+KUza1n27PYjHh4hvw9CiKEgl7oq18ZuoUPJ/V4Pja0RO6FFtvNF4u6zYs8snNbr6/b0PkI+99C/oE9C8Xorod3bBGtvOMvemzhVdXmIhOneVs21zPXmWS2bhmeXTzjlXwAaOAycBnzO5c9R7UBb1E7rb009P/PGB4wuC2aEINx/1Rn2Rt9Ffo/ryIXW2t6gc2RxIGOa2dosXKacRW+khi5MqikjYbrPGB9ojdohl91ttpoq28ipdU0pq0II0Xu5bOJd6PC0XM8XN/OP3CjEddOXrdy+cVuvrieSdJZEI+3RBCvnTOZTVcWsveEsJtWU2bN0G9/6MGNj9v5+5sum4dnlPBOntT65P29kKDBN03XqeXjIy/CQj2cWTiNhav58qIN7/utPbN3TDEB7NNHjyEX6yJuVnfKhaybJ1LHoFatM/fymGXzcHGH3fvcZ408Od7B1TzP3Pd8VRjOmPMRxpcGsewNlGzkdXRbiuGFBKatCCNHPCh1KnvMMYI4JUAp53WyZM2MSTtlr2Z7jFSV+/uWXf7SX+6yaW0dVaQCtNZ8e9SnKQ74junxBMlBn17vfuGNUtqnnlo4Es1e91vkLUcTwkI/G1giQ/IU4saIop1Gm1JG3qtIgI4q7H4UToieGoYjENIvX1bvOtq2+to5N9XsA2LqnmUVP1nPrj98m4PV0u7lrtpFT6cAJIcSRk8uMXaHP5zHcE6B4+nDtnq6ba7IXkTu35/jKOZPtDhx0Lf+Jm5pRw0NUlgbweo2ClrmeyGefXbczcUqpE4CPtdaxzn93S2v9YcHubBDKNvXcGonbowLWiNIvbjqbcDS5L0bQ5+GEkL/gIxey0FPkwhrBtBKcpM+23XLRBLZ/3OKINS8P+WhsiWQtW/2dTETKthBiKDmW6qxwLOGaAOX7Xzmz365ZUey3cxBYybROrCiS0P0+sJ7jP1kyg0jMxFDJvY5Tt8aCrr0Cj/QesJZcE/0ci3oKp9wFTAfeILnRd08Bz0d1tzjb1PO+zr0vUkcF9h6O9OsiTFnoKXLlSwl9sWbbqstDbFg0Ha/XyOiMlYd87Gxs7bFs9VcyESnbQoih5Firs3wewzUBSrZEWIUSiZt2kjjrZyz6xjAUI4sDdvldOrPWfQ/YxjbaookBKdOSgTq7nn7j5gPvpfy7pz9HNbep5+WzJrKpfo9jVOBILMKUhZ4iV1UlgYyFyKlJS9LDWJrCsQEtW1K2hRBDybFWZ/X0TOkPx9rP+EhK/dmueum9jGUXy2dNHJA9YFMVOmz4aNHtTJzW+vGUfz/W73czyBmGYnxlCRsWTSeeMPEYCq+h+M6VEx2jAoVYhNlTaIYs9BS58noNThlV6ii3Aa/B4UiMeFgTi5uOMjbQZWugry+EEPnIp846GsIu058pXo9BVUmg23XUfSXPhf6T+rO1kpw9OX8q+1oiNIdj3Pf8DjtRX64/76OhnA8F+ewTd8wzTd1jmJlpapRSfdpHI5fQjELvDyOObl6vwXHDgna5qiwJZOy5s3puHceXBfF53TOPHamyJWVbCDGU5FpnHU1hl16vweiyUI/HFaoxn+255OvHjuOxwDR1xt5vW/c0s/tAu2N/Y8j9OXw0lfPBLq/Sr5Q6Vym1Sin1n0qpF9P+bO6vmxwseprOtwru3b/4v4zp6HwWYeYSNlDo/WHE0S+1XC0+b1zGnjuL1tXz9p5DtHbEB7RsSdkWQgwludZZx1pIoNUmunLlK5y9/LdcufIVduxtwezFfnLeLBkxvdIp6JMDbVG+/avtGW3WmhEhHvjyGb16Dh9r5Xwg5TwTp5RaBDwMHAB2ApH0Qwp4X4OOaWqi8QT3X3UGzeEYq156j617mu3pfNPU7D3cgc9j8A9fOJXDHTHu+dJnCPo8HD88yKjS7lOvm6Zmf1uEjlgCQyUrq3uf65rCTg8bkIWeIh/xuEkkpfyOHh6ksiTA0pm1VJUGKAl4SZgmxQEfhztijCjy2xlWs5Wt/gqXkLIthBhKUpdaxBImvs7wwvQ6ayBDAmOxBPtaI8RNjddQVJUE8Pl6H92QS/2frTH/0yVnZyTF6ul84ah7RsyHrpkExb1+G8c80zSZVVdDkd/D+q9/FsAuw6+/d4C1N5yFx1CYGkaW+Hp8Dqe3NdLbyqKw8gmnvBV4CpivtT6mutNuU8PLZ03kvud30NgaIeT3ZLy+YvZESoJevvWL7Xzj0lM53BHPOpXsdv4Vsyfyzctr+dYvtrN1T7PrNHZ/ZQcUR5d43ORPe1tYvK7eLl8/un4K//jFU7hlw9uOMv3dX7/L9TNO5vu/eZdbLpqQV5ktZLiElG0hxFCRy1ILGLhQ8VgswZ/2tXJjyjPg4bl1nFJV0quOXK71f66d1lyXkLhlxJQw+96Lx00aW6Mse3Z7xhKLi2uruOmC8cx77E3HkothwewDqm5tjdS2snxWhZdPOOUYYO2x1oED99GkOzdt4+YLx7PmuinETZ3x+u0bt9HUFuPmC8fb359tKtnt/Nb3Lz5vnL0ps2maNLZEehWKII5d+1ojdqUKyfL15+YOuwNnfe3OTduYVVdj/51vmZVwCSHEsSjX+nCgQsX3tUbsDpx1fzeuq2dfa3pAVW5yfb+5btIsS0gGRmrbIH2Jxay6Gpas35Kx5KK7Z7xbWyO1rSyfVeHlMxNXD4wFjvq1b+myjSaNqyqhuizEx4fCrq8X+T2MLA1wW2djOdtUcrbzF/k9jBtezI8XTeObv/gjL2zfJwtERd6szb5TFfk9rmWuLORz/J1vmZVwCSHEsSbX+nCgQsXjpna9v7jLgHAuYZK5vt9cN2nO5XwSZl94qW0D65lvSf8/9PyMd2trNDSFGVdZTHV5kXxW/SCfmbibgb9TSp3TXzczWGUbTQr5PBiGyvp6ezTBx83hrOGQPZ2/PZrgvcY2dh9op7ElOfohMx4iX9Zm36naownXMtccjjn+zrfMSriEEOJYk099OBD7XXkN5Xp/6UlBck1Ekuv7Te14vXLn+fx0ydmuA9BWRu/08ynlPE72Cissf0rbwHrmW9L/Dz0/493aGtYm8PJZ9Y98OnG/BKqB3yqlWpRSH6b9+aCf7nHA9TSN7/b6itkTGVHs4/4X3u1x2j/b95cX+3hw805u37iNxeeNs4+XGQ+RD7eNWWtGhLj/KmfmKWvjercN7NNJaIsQQiQN9vqwqiTAw2nPgIddNufuj7DQXDpeHoXrBtMeaff3K7+vK+Pnqpfec2T/3FS/h5VzJudVpgdiE/hjndI6t/VVSqnHgG4P1lrPK8A9DagpU6bot956K+PrVoiBaZokNGitHdP5XdklTbwKAj6DcMwkYWr8HoPKHjbCtL8/miBhaj453OHITvnMwmlc/cjrQPIXwy270zFEqvY02cqtJRqN09gWtTOTBf0G7+9rp7I0QFN71M5O6TEMOmIJRhT7OX54qMeMqrKZZ87kB+Oip3I7mJx01696PGb3PZcegTs5oqTcunArt7nWh4U+Llfpz4DKYj9+v3NFzUdN7Zy9/LcZ3/vKneczprzo/7N37uFRVef+/6y955oLJIQEhaAgIhotl0RpgF4QTrH1UusBxQooaAVEtHqq1R5r7Smn56hItV642RZvKCAcf7Z4vLSotUekaqBQjVyK3AJIQkhCMpnr3uv3x2TvzN6zByYQNNj9fR4fmZm1116Tefda613v+36/J2x8dc1R7nkpWYttME+uqtrDL68YfKz7HNduHWC3270Nrfz6T9u4ecyZKEKgS4kuJV5F8PH+ZtZUH+A7XzmVASV5BL3Z/caJhJ5kQf2cROC/ZOiw3WZdEyelnNrRzr9MUBRBUa7viAxKJfkBdF2ysz7ErkOtFiHlo9WxGdfXNUe5Yv67aexVrTHN/HdXOuFz0fWRSOhsrQtZGKPmTyrnT9X7mTj8dG55YUOavb00a9RRJ2uXQdKFCxcukshmPsyW1bGz2X91XbK9vrVT2TM7c/4vyvVx+7cGHbV2zkXnIuhTmVR5Gvsaw5b96qLJFaypPsDaT+uZ9rX+lBYc+UA3FdmKwLvoHGQdiTuZIIToCzwMfIukZ/sn4DYp5e6jXXukk+FMDlZqVKyuOcpHe5scle6ziZ4lEjr7msLUNkepD8VYVbWH2781iF7d/IRjGkGfSkKXxBP6P3P045/uCx8NR7LbfY1hrlr0Xpo9Lpl6AXNf38wPx57FjBQH72ibBTcCd0xw/0AOcCNxXR6u3TrgWCNx2ewhOtKus+97oqVjjoROXldcu3WA3W5rmyN8vPew4351+fTKzzVS7AI4kZE4ACHEMOBe4BtAATBcSrleCPFfwDtSytc6OoDOhhAiB3iTpBj5dSRTQP+TZC3fYCll6Fj7zoZBKZbQMjL/Ha2OzUlrZtGUCgYW5+HxKOhB5wl2YHEeDeG4+zC5cEQmxihVEbxRXcucy8/LmvHri1zkXbhw4aIrwszAqW8lx6fSGtM4vSiHfkW5x6Sb1pn6ah3p74tkgHQzOz5/xBN6xv2qpkuK8/3ouqSuOWqxB8DdB3QRZJ2oKoT4GvAecDZJ0e/Ua3VgZucO7ZhxI0kphO9JKf+flPJl4LvA6cCM4+k4G0Ymn0fNyPx3NOY+p6LiGc9W0RCOZ/z84T9uYUvt0dmkXPzzIhMzmaZLSguDKIqSNeOXqw/nwoULF1Y0hmMcOBzh3pc/YuLiddz78kccOByhMXxsummdqa/Wkf7AZYD8Z4IQIuN+1aMqGdlKG8PuPqCroCPEJv8H1APfA1QgBpzfFon7V+ARKeVpJ2ykWUIIsQYISClH2d7/M4CU8ptHuv5I6T2JhM7W2mYe+dNWrh3Rj1O6B/F7FHJ8CtG4jhCgS9DaCoeDPoXWmG4WEucHFMKx5N87rks0PVlA6vMoqApoOigKhFOuUUTyQRNAJKGb70mJ2b+mS1RF4PUI4gmJJiUeIVAUQTShE/QmUzCFkOh6cnyKIvAqoo3CVyJlcuxxTUdVBElm32Txs9erWkLnR0rp/JxC7O6qYkMmu41EEtSHUwhNvEmb0aXE71EQgCaTv7+iYNqH2ma/CU0S1ySizeY0XeJRFVTRfp0uJYpI2oxHUUhoOvEUm2+O6Jb7h2IaqhDkpDwfAY+CLkG22aImJWpbn0jwqAqQHIvRXpK0V+PeUoKqCFRFMU8Lj0RGBM72alzXyTbs2qwD3HTKLg/Xbh3gRBBREFRoDLfPdcbrVEIQXZeEY9G0dkGf1VnKtl1H7nugOUxCa59bPSr0yk+vdbKvGUVBH4FAetLWF9Eu275w7dYRdrs90BQmzy8IxyGaaLehXL9CLCHR2/aFidR9oQSvR0EhuSdN6JKgRzH3tAGvatkDeFWBlMl9bCyR3Cd7VUFcS671fq9KQcCTzCbTdPS2/Yexphfn+miKasQSGl6PgkcRhGPZZQ19Tmt7Z+KEplOWA/8qpZRCCLvndxAo7ujNTxDOBV52eP9j4Mpj7VTXJdFEjIJcr6WGaFxZCbPHDOTxN7dx3cj+3LWqvTh0weQKHluz1RTpXjC5goKgh5oGaxHpgknldM/xEvAq7G+McVNKfdID4wfz9NodTBvVnwdf20JdS5QHxg/mnS0HuHRoqdnWGMespevNa+dOGMxL6/dyRXkflry7I218cycMplvAQ37QS31LzCS4SL3vrWPP4qziXLMoujjPz4+/PciRtAXcEHtXQiSSYFt9yGJPdpt8+KoheD0K89/6h6P9dguo/PKVT9I+M66b/Xy7zTx+zTDiCZ3b28Ttne43f1I5z723i8Icj2m/hk052ahhhzO+OYD8gIfrn3K2QaPddSP78/TaHdz9nXOIJnQe/uOWtD6PZq9+j8K1v3vftWEXLlxkhYKgwqf10bS59owia3pgLKY5thtY5LE4I9m268h9D7bE09oV+q1OUKY1Y2BR7hfeLtu+XGSPPL/gYKtGU2vcsndcMLmC3gV+dtWGLOu5sc7eMvYs8gMqk3/zvmU9dlqbF0wqx+cR3PB05r3qkmkX0BiKZbzX6r/VsOgvO819q7EXzrQ2Z0oz/jKu7R2JxB0CfiCl/B8hhArEaY/ETQR+LaU85QSONSsIIWLAr6SUd9ve/0/gbinlEZ/2TCfD7TnBCuGYhgRyfCqxlNOLZ9buYNFfdprXzPh6PyaN6E9CS0bL3t1Wy7fOPZVwXOezpjDz3thqCoG/OGMEUkquWrzOfBD6FgaRtNO+dg96aI3pBL0Kkbg1WqdJeLbt/ldVlHLzmDNRFYGmSz6tCxHwKuaDZaC0MMiyGysJeBUOHI6S6/egKsn7vfXJZ4wpO5WGUIyeeT5y/QqhaPKeyWhHMmKo6ZKVH+7mulFnkNB1tteGeHTNNlMa4QTJIZy8T9wJgpPd7m1o5XA4Sreg37SVaDyOx+OlviVKJK4hhKB39wCqIlDaosHhuMandSFWVe3hv8d/hUhMZ+LidWm288KNlUTiGht2HWLY6T0IeFW+/2R6u3svLWPGs1Xm62XTKxFgPi+LplQwZ3U1915axpzV1RkJWK7/2hk0heOcVZLHf7/6iYWOev3OevNZ83kU/B6FpnCcgFflaoexG2P4jz98zBvVtZbPHrpyCE3h+BGprrONTKfAtVkHuJG4Lg/Xbh3gFIl7Zu0OJpx/mrnurvxwN9eO7G+JiHWknaJIdF20r/Ntr+3t1u+qZ9jpRWYGxYZd9ZSfXpTW7o8f72dM2alm5sSb1fv51rmnprX7jz98nEb1f99l5x7z9zjQ1Eqv7jnm9zBed/T77m1odVyHlk+vTJM/wLVbRzjZrdej4FWwZI0Z+wanv/f9//oV/rK1lmtH9jfbG+2MtdxpvZUSHluzjbFlvVhVtce0MV1KenULmM5V6nXP3jAcVUlG8hQhOBSK8sRb/2DaqP5ouuSU7kHy/Gqb9plE6pDQk1lG8bY1OaFLXvv7fs4rLWBASS7RuE5LNEFtczRtbTfkEeKajvcY5RGOMwJ4QiNx/wfcJoRIjXIZHuANJMlEugqcPNOMfxwhxHRgOsBppzlnhOZ4JX6vwu76KDOfq2LkGUX84Bv92dsQMQuZx5/fl4bWBCuqariqopRLhvThmrZN7biyEm4Ze5Zp7KWFSbHl+1/dzIY9jUQSGgFvUk/uvu+WIYA9tojdwskVHGoJU5gXdIzW3Tr2LMaf35ccn8qhkO1kZVI5xXl+y0NS0xDmcCROQ6vkJlsEb9RZJSx9L7nJnvH1fpaoX2pEZe2n9cyfVE5rLMG//OodczwPvZ7UuHOFyU8cjma33YIKh1oVi80tmXYBiq4D0KtbgPtf/cSMkqVGs1ZV7WHmN8/kUChOU2vcsfB5X2OYH724kQWTK5j7+mZu+NoZju0Kgl7L67rmKLe8sIH5k8ppaE1Qku832zld3xSOc93I/vQpCHDHixtZMKk8Lbo2f1I5/7n6Y/O7LJxcgd8j+Kwp4tjn3obk2B8YP5i65ph56FDTEKY4388dL1pPBPW2vxlYT/mOFJk+mU/3TiSymW9duOhqOJLdelXBJUP6MO2pDyxzktemVu31OLfzeaztso2w5fgVzu5dwLYDLeY+5OzeBeT4rRvPgFehon9Pcz9i9BfwWtspCo7ZEIptH5vx+9q+R35A4VCr17IGLZhcQX7A2mE23zehS8e5POHW/x8RR7Lb/IBCJCE5GEpQcyhs2lBpjyDdgorj37tvjxwuHdLH/E1Xzhxhtsu0hhvr7cLJFeQHPaaNGevnoVDM8br6lhhxTTcjb/MnlfOTi88mFNWZ2ZbF8+8Xn82Tf/mUG752Bj96caNzNHByBVU7DuLzKGm2baztiYTO5gPNFimmhZMrOLtXftaO3BcRAeyIi3kvyZTKjW3/lsB1Qoi3gErgP457NJ2DBqCHw/uFbZ+lQUq5WEp5vpTy/OJi56zQxrBOc1g3f+Cbx5xJfUvMUshc3xLj5jFnAnDjN84wnSiA8RV9zQkKkgb6oxc3MnP0AEoLg+w82EpCg1vHDqQhFOdQKG6JnNU0hJn5XBUDSrql9XPXqqRI5sznqtjbEOHjfc2We9c0hLlp6XpuHTvQ8p1KC4PkBbymA2e0vXPlJmoOhZlwfvKBn3D+aWn3nLV0PTd+4wzz30aBtDGemaMHmPc4GqGLi2PD0ez2cFi3/G7FeX4ONkeZ/Nv3uWL+Wq793ftcN7I/w/oWWOzI+P+tyzZQcyhMfSjmWPjcGE46dzc9V8X4ir40huMZ26W+zvN7LDaU5/eY7Zyurw/FuGvVJqKJ5CIe8KrmRAzt9ji+oq/5OvmcRo469lRbNT7bXd+a9nxpKfuEVDKBmaMHpD2nboH3kZHNfOvCRVfDkew2rsm0NXfW0vXENauDEU84t4slrO0abXO3Mc82hnVLu1hccrA5atmHHGyOEotb+4vEnfuLxK396Tppc+tdqzahW5tl/r6279Eccb5vc8TaYTbfNxNBl8c9LDsijmS3zREdXcfRhlqjMuPfO3XPmLrGHmkPYKzLHiFMGzPWz0zrdH0oxp0rk2u0YWMeRTX34TNHD+D2FRsZX9GXH7UdvDqtyTc9V8WYslMdbdt4RGtboma/xuczn6uitiWa9d86E9HQLtueojP3CFk7cVLKjSSlBQ4A95CMbM1u+/ibUsotnTKi48fHJOvi7CgDqo+104QuLSdBihBphnLnyk0oSUYQVEVYThYynVAU5fpYMKmcR9dsI6Hp9OuZQ45PzUj7muk0yug/x6dmvNdpRTnmg2KcQgic+8vxqfg8CsP6FqR9F6ON2jZ51jSE0TTd8llB0GueOGQr2GlQ2e5taKWuOeoybB4n7LbiNLmlOjGpdpRqTwvf3s4D4wen2c7Ct7dbrnNqN39SOauq9liui8Q18zpVEUTiGg+MH8yqqj0Z71PTEKY5knQGW6KJrCJ+2Y7dsM/SwqTI6aNrtqX1nZp2nkrXnelZc6PPLlz88yDbKFFnt4vr0nEfEu/k+2on+Htk086rCuZPKk9bX+zRThfZw9jXZrIhp7VTCCy/Veoau/Dt7cydcOT1NvW3NtbPI63TqWt7TUMYXaZfn7oOZ1qTpXS2MWNtzyTFlNBsJxhHQCYpjxyfmvZeZ+0ROlQNKqVcD4wVQgRIRrsapZStnTKSzsPvgYeEEGdIKT8FEEL0A0YBdx/huiPCOO0pLQyak1omg1g5cwQeVTHbQvsJReo1pYVBCnJ81DVHqGuJosmkV90a0yz3Sm1vnEbZ3zf6N6IeTm3qmqPce2mZmev+9Nod3HnR2Y5tW2PJuqg7LhqUsT9jYi8tDFom29LCoFkLly37j6s/1vmw20qmyc2YIO12ZNjBhj2NPPT6Fu69tIwzi/PYfajVTJdNvc5oN+fy8+hfnIsiBEvf28H4ir7c8LUzTJszImalhUl2V6/q47E3tzG+oi/dAh6WTL2A1lhSb/HHKzeZdaPRRHIyrW2OZnwGUl/bx16U66NXtwC3vrDBMvaSbgFWzhxB74IgAa9Cne3kzR5NNui6axrCGZ+rZ/DhAAAgAElEQVRrN/rswsU/DzKty/YoUWe3y7QPsR+AHu991RP8PbJpF9ckr2zcy5KpF6TV4bk4NngUkdGB1nVprud9ewTZXhfi6bU7+Nll51p+qw17Gnl67Q6WTa/ks6YIvbr5WTa9kmhcd9wrpP7WxvppX6e7B72WtT91X6uI9OtT1+FMa7IQzjZmrNVe257dHK+afcJi6t4gtQ9jT+903+NFxyr22uEBvCTJTboangR2Ai8LIS4XQnyXJFvlHmDRsXaa51coCCosnFxBaWGQgy1Rx/BvJJ4kP1n63g7LqdGqqj0saLvWaPvrq4fSEo2z5N1kW9BZ+PZ2euR66ZHrTTvRWDi5gu21h9P6MaIYcyckTy6cTjUevmoIPXK9zFldzcTF65izuprrRvZn5Ye7WWA73Zo7YTCFuV4eXbONu1ZtoiWaSLvn/EnlPPnOp5QWJvONV3642/zsyWvP59TuwQ5pzLj6Y52PgqBi+d0y6cEYk15qNGxV1R4WTa6gtEfQnGTnrK5GR1KS7zcdHeP3N6JtdS1Reub7SWg6S9/bwSVD+lhsbtqo/ix8e7t53R8/3s+tL2xg2qj+zFldzWWPv8u0pz5ACJj7+mZzEp87YTClBYGMp32pYzCeldLCgGXsfo9CayxuGfv8SeUsfW8HuX4Pp3QLUBD08eS151v6tkeTi3Lb2ziNpSPRZxcuXJz8yPEpaWvkgskV5PjSa7+c2hUEj62d36M4zuk+z4m97xfRriCocOnQUqY99QFj5v2ZaU99wKVDS9P6cpE9CoIKXtU5TVVVBHUt0SQp3oubmLO6mtljBgIybc94y9iz+NPH+wl4VeasruaW5zfQEo1bDkWNdXn9rnpzb5y6fhrrtM+jpK39xp5h/qRyVBWenFJhXv/wVUNYVbWHeVcOOeL+4M3q/Wn74tS1uiTPb+7vU8dbkpc9KV/q3iD1HqfbsuA6c4+QNTslgBDiUuAXwBCSNXHD29gpfwO8KaV8vlNGdZwQQpwGPAx8i2Ta5xrgNinlzqNdm4ktLR7X2N8coTBHpTkiAUlDa5wZz7YXQS6aUkFRrg+PIogkdHxtWhgGg4+h66brEo8qUIUwtbY0KdnbEMarKvRp+7F1vf0zjyLweQTRhDTZKQ02KkWARNAaSzB1SbLQeFxZCfdckiRIURSBTxGINt04vU0nxmC19HsU9jdF6B70ouuSzw5HePC19tOTd+4cTUGOaup9pbJT+jwK3YIqkfhR2fmOiL0NrYx64K2099+960KXeSoLONnt7kMhinJUU0Mo4FGoD8WYnmqzkyvoketFT9GJS/3/639PMppJKRFtjGYXf6U3cb3dru1acEGfQks0qQXnTXkGvG36cpGEjhCCjbuTLGqalPjVdt03IQSP/inJYmVEjRe+vZ1fXz3UPIH1t+nExRJJ/ZqAjbHV71HQpbQ8f4ZeTaoeYrIf0WGtQ5edsnPgslN2ebh26wC73dYeDqNJSUJr1830qKAKQUm39g3yZ01h8v0iTdetOZpk2jOw+1CI7kGVlpR5NS+gcDis0bdHrtluX2Mrdc0xbn6+nZjsiWvKKc730bugfd3c19BKlQOLZcXpRfROWV931Yfomaumje9gSOP0ovb71jaHyfWmf49QXFKS3/49Gloj+BTS2sV0KMwJmO1qGlopdNC7awjrlKawU2ajidcG124dYLfb3fUhinJVdh2KWvYFi6dUUJDjNTVaw3HNXFcN5sZUXbmAN6mV7E9Zh4PeZKTJ0H7zqgp5AcHhsG7ZFxgascaabHymSYlXUVBEu0ay36PQPeBFVRUOhqJE4jp+VZDM0Wlnpwx4lOSa3DY+Q2vuSJqx0M5OmdB0PF82dkohxPeAVSQdoruAB1M+3gFcB3QJJ05KuRsY35l9er0qXkXhw52HWbe9jqlfO4OCoJfl0yuJJnS21bbw05c+Mh0fgJdmjeSK+WsBWDlzBBMWvsewvgXMHD2Aolwfp3YPIESyfq6uKcqpBUG2fNbML1/5BEjWMJXk++mR62NfY4TbV/wtLUxr0LevnDmCX77yiSVd8rZlf2PeVUPQEjoNcY3LHn/XvPYPs0exrylCQdBLj1wfs5auz0jxLoTg4kfX8uS15zOwOC8pypjQyA94TWPs1j4fHxMyhaHdtLRjh0cILvr1Wu69tMyk9O3dPWCmKxbm+tjb0GpqHhow7Koo18fPV2/m56s3W/o9p3cBExevM2069bqVM0dwKJTApwpoE+v0eVS0WIKR96cT2P7p375BfUuMknw/LdEElz3+LoumVLD203pWVNVY+lYV8bkIdCqKOKokRjZtXLhw8c+BmCZ5NoVyP6brrHxvN1NsqX7RhM5Nz/2NmaMHWA6oHrl6qKWdR1G45NG1afPyihkjLO2khCfe2mZZ9594axv3XWalBYjrkluWbUob95/vHG27b3LNsN93+fRK6/dNSH77zqft3zeh88zaHWnfNxyT/P1AEwNKupnvVe1q5Mxe3Un1uzRdcu7P16SN752U8SWyaOOiY1AVwf7DCZ56d6clTfXJdz5l9pgz+cbctxlXVmLak0TQq1u6QHxHkH+ce0UDJR3sqNh3dHfH41HoXRA8arsjIdPe4ETtFzriYt4HLJFSjgMesX32EXBep42qC0LXJUJA3x5BJg4/nfte/oid9UntFVURzFldbXHgSguDdA96Gda3gNLCID1yfZaQcTSh8x9/+JhIXOdXb2wlHNf4rCli9rNhTyMznq2isTXOpN/8lf/630/SQsGpYebithS3Gc9WMXHxOmY8W2WGsee+vpl9TRHL2A62xMy2P165ibkTnIklFkwq52BLxExvbAjHKc7306cwp0PpkkdDpjC0m5Z27PCoivm7XjfSmq6oKIKX19ew5N0dybRJh/Rcw2ZTkZp+WZzvT0uxNSiBSwuTOkC92tJqjXxze1/b60L86MWNbK1t4WBLzEyHcLL12c9v4Ir577LlQLNLeuPChYsuA5+qMPrsXpZUv9Fn98Jnq6fxtqWo2ddpr71GTJCWEjZ3wmBsDP7k+RVuGXuWJWX9lrFnkWeTGMiW2THgURzJQwK2aETA4/x97e2Kc30U5gWZuHgd35z7NhMXr6MwL0ixbV3PtD6k1iN5XXbKTkeuX8HnUVj7aT3fevgdxsz7M996+B3WflqPty1V99axZ9ErP9Dpez4XnYOOiH1HgMuklH90EPv+BvCGlLKTfOwvDkcS+97XGOadLbV8d1gfvjn3bTOq9pU++TSEEmZEw9gIP712B98ffjo5PpU3P/mMicNPpykcp7Y5ysK3t7NhTyMvzRrJLS9s4KlpFwAQjeuWfp69YTgXPvRnAPN+BUEvp3YPsLcxmX5ZnO/nr9sPMujU7mlpFf+7aS+XDOnDKxv3Ut6viKJcH0W5PqKaxg+ebr/PU9MuIOBVzbQ5TZdt6Z46ty3baDqoGdIbOwXZpLG1wZ1FbHBMp6wP8fib/2D22IFmCi9IwjGdlR/upnJAMT6PQu+CALqkTTheEIrG2d8U5YySXJrDCYtuyrwrh/Db//uU2WMGckp3P9G4TIrZy6QGUtCj0jMvfaKvD0XY+lmLRbslVZfu6bU7uPs75xBN6Kb+2q1jB3JGcS6f1n0uAvJZoQM2aodrsw5w0ym7PFy7dYDdbhtaIzRHNGIJiSKSKds+jyA/oFrSBhtCEfY2RtO0qPoU+CnMbW+3qz7EbcucI3apaY01Da2WCKBB9jFlZH8zDRGgORJhp4MOW78iP/mB9vvurg/x8oa9XF5eaqaFvry+hsuH9eG0lPseaArz2eEIh0JxU1usR66XU7oF6JWSFlrXHOWK+e+mRfbs8/f+xlZ2HwqbNPHGWnNaUZBTuye/x/6mVnbXO7XJ4dTuadET124dYLfbUCRCTIP6UJw9KTpxJfk+eub50WTSEfd63YyozwknVOz7MNAzw2f9gLqO3vxkQiyh8dnhCMurahhxZk8zqjbj2SoAXrn1a5aUBoOR565vn4NEMu1rZ/Czlz/ijepas8/SwiAFOV6WTa+kKRynNapR2iPAsumV7fVutDNDGvcrLQzyzPXDUYSgPhSjd0GAD3c18sIHNdx7aRkDS/LQJSQ0jQnnn4bPI/ju0FKLczjvyiE8fNVQenXz41EVinN9hOIJ9jZE0tqljvdEpje6KWqdC79H4YryPhaB11Qh9isvOI0HX9vMtFH9ORxJim7n+T3ENcnAXnnk+BRiCZ05l59nTu5FeT5+cfl5vPb3fby88TNuHTuQfj1zCHoUYpqGlAqKItKcHV2XPPhakn2qd0EySu1V4b7LzkWTkp9/9zxK2py/l2aNMq+LJTSmPfWB5Xt9URT+LoOqCxcunBCJ6RwKxZj9/AZzbnj8mmF4FT+knHk2RzTe+uQAz99YaR6Yvry+hsvL+1DY7iPhSYnYGXCKOmm6ZNFfdrLoLzst719T2c/yuqFVw6/C8umVZh1TKBqnoVW3pLf5PAqVA4osa8bDVw1JI0qJJHSWrtvNjd84w5KCZ+jkGoglNIrz/Ja90cK3t6fN37qE3/7fp5Z2v/2/Ty1poboOn+xr5PkbK00H883q/WnRORfZ42BIw6cKU/M4dd/XI1cSTegcDEUpzgt0uDbMxeeDjjhxfwR+IoR4FWhue08KIfwk9eJe7ezBdSX4PKrJgLP4nWS6V6rye57f41hP9o+6FuasruahK4cwe8xAqvc3WybHw+E4P/99NXUtURZMKmflBzWU9+th9j2urIQFk8pNcUUjreJwJM79r26mriXKnMvPY8qI03n2vV3MWV3N49cMIxrXzROrJVMvMB9QaBcaf+b64fxw2d+oa4kmUxfzfJb6KKOdUSvnpjeeXNAljrpwxu+551CYOy4aRFNr3LRdw74icZ2e+T6mtRHlGCgtDDLn8vMYObCY80oLuW3538zr5k8qp1vA6+jsLJpSQXG+j4Vvb+eOiwbx2JqtXDeyv+UZMhyiVEe+LoOcwBdRK5mJQfWLigq6cOGia0BKTAcOknPD7Oc3sMJWS+ZRBMurapj3p3YtytLCIOMrSi3tjLTGWSnrvlNao5FiaJ8f7emZ+QGVxtY41z+9ztJfcTdrf1LC7Ss2Wr7H7Ss28qKtFi/QdkA47akPLOtGWtqlV+HH3x5kycCYO2EwAa/te6gK00b1T2vnTUmnzPEpVPTvaXEwnRhAXWQPr5okADP2itC+71s+vZIpv32fuRMGE4ppnNEzzz2s7ILoiBN3D/A+sAX4X5LslHcDg4HuwPc6fXRdCEW5Pm7/1iAe/uMWfnLxObTGNJ6+fjiCZPwz6FXSHDsj6lHTEMbvUXj8zW0smXoBh0KxJN17jyDba0Pc/Z2zaQzHeezNbdx50dnmxAiYkbunpg2nviVKYzjOg69tMZ03n0fhodeTr1+4sZKxZb34j98nNc2NqFwmse6mNm0vgBuf+ZDnb/yqY7tzTsnvkOabi66BWAbxSkNgviWa4HA4YS7aRrquV1UozPVxOOwsqp3jU9nbEEk7GJi1dD3Lp1c6Ojsznq3i+R98le11IdORNJ4Vo42TQ2TUStqjX9kcJhxH6qPz3zODkKcr7O3CxT834lmKWiuK4IlrhqWlIdrnpZaYxnPv7Uojm7h5zJkUpbTzepKyR/b0TK/NmQrHpOkQGmObtXQ9K2aMsBCMZFoz4jbBY0n6AeGdKzexcqbV2YvGdZa8u8MSYVvy7g5+biNe0aRO0Kdasj6CPhUtpdynNaab6aDGPW96rorl0ystUUwX2cOrClpjzr+5oUF458pNzLn8PLoHfe5hZRdE1k6clHKnEKIc+A/gIkADvgG8BvxMSrnvxAyxa0BRBIN65fOzy84loUl++tJH3HHRIHMj+tYd3+TptTtYMvUCmsJx6kMxM23NIDmpa45xKBRj4uJ1DOtbwC8uP5e7/+fvFqfPq6Y7XG9U13LD185g4uJ1lvf79ghy54ubTEcsrumW9Is5q6uZc/l5lBYGHU/rapvbRY1rGsKoGcQQgz6P+/CehMgkoFqQ46MlmjwMuOeSc0wHLtWejZRdp+sbw/GMwuGJZB6v42eqIhhQkktNQzjj9XaHyHjuUlMss3HGTkTqo8ug6sKFCyeoGebaNGkSKYnE9bTUNTs3gUcRjgy9P/yXgZZ2cU1HSmlxfqRM1inb2znO17Z22X6PWMK5v1jC2h+CtIyLB8YPTqv80XWY/9Y/GF/RlxxUYprO/Lf+YUmnzCRKbXeUXWSPuKYfVeDdOLh1Dyu7JjoiMdAdqJNS3nACx9PlkdRdg0e/P4y4prNk6gU8+c6nfNYUYdqo/sx9fTM3XzjQkp72wPjB3P/qJ9w6dqCpPP/vF5/DwZYY864cYuaJ37VqE8umVzKurITxFX3Nk6tVVXscFd+314UsZA+pivNGOkKPXB8rP9ydFiVcOLmCZ9butPTnUQSLplRYtO/cFMqTF0VBHwsmV1iK2R8YP5g7X9xIXUuUp6YNx6sKZny9H5NG9CeRYs8rqmq4/9VP0k55jejyrWMHOqfxqIrFDlM/C/pUInGdlTNH0D3oZVxZSVqNqJNDdCy1kici9fF4ooIuXLj48sKnCOZOGJyWDuizOT9SkjF1LRUBr5JWRrFgUnlaGiIIs40BJymCTHOyx4E9c9HkcmqbYxaSC3t6ZiZnT3X4vk+vtUbinl67I00CIehTufnCMzkUSu6PfKrCzReeSdDXvh5kcjZcdsrjgcDvESyZej41DRHzNy8tDOBvi+aWFgZpjWnuYWUXRVZOnBDCA9QDVwB/OKEj6qIwTvaRGgmpWFmeJpXz9uZavj6omO8PP50euV5HkpO7v3MOP1qxkXFlJQR9qqn7lro59qmCW8aelcYi1T3osThoCyZX8NiarUC7Iv17/6hjzuXn0a9nDl5VQUpJJK5x8eA+/O+mvTxz/XAOhWLUh2I8umYrN194Jt8d2puAV6U438+8N7bSGI7x/A+++rlpcrk4cQgEPAwsymXZ9ErqmqPk+T1E4hozRw9g4dvbaWyN0bvAz6VDSy11BvMnlQOwoqqGn11WxkNXDqE438/u+lYzdbe0R9AxjccgJzGcHYNl8qxeeexvilgOCIz7vFFd2+kO0YlIfTzWqKALFy6+3Mj1C/oUBnlq2nALO2Wu30ZEIp2jSZotEqcqOKYX2nwu9AzRKbsES3Fu+oHegskVaVT/AZ9AIiyRwoWTKwj4bBIIinAsH7E7VKoiHCNxdmevm9+Lqihp9+3m95ptgj7F8TsE3Zq4Y4YiIK7rRBLS8rdfMLkCiTQPI3p1C7iHlV0UWTlxUsqEEOIAyRTKf0oYJ/vLp1cycfE6a1720vU8ff1wHnj1E+677Fx0SQbRbHjoyiHoUlrq3moakoQTcy4/j5gmHfO+H7pyiCnAXJzvp1tQ5WeXncs9l5SR0CUr3t/FZUNL6ZHrRdMl9S1Rbm4rtB5XVsK9l57L959sH/ewvgW0xrS0dM6HXt/CNb/5q0vW8CVBIODBF40TiWvc8kI7c9rcCYOJazqtsXR7m7V0PUumXsDaT+vZebCVUEwjP+BhYK88Hrl6KLqUtEaTTJXL25hUPapCSZ7fZLAa1Cuf388exf7GJNupXUg+tYbuvsvO7XSH6ESlProMqi5cuLAjFEuSMP1wWfvB7K+vHkrAk0OKckDGkgVVWOe91qjOVAdSqRXTKylIqWFTBI792afR5lgCn4rFydR0jeZYgh4p9PEtEd08mIPkPD2zre4slcVfKMIxwvafV3zFcl9Nl2m1z3et2pQWeWwIxx3vm7oPCcd0Hluz1XLPx9ZsTUb13Jq4Y4KmS3RJxlrD388eCQgKgu5hZVdFR44wngN+cKIG0tVhUOUaednD+hawaEoFy6dXcu+lZQS9CteN7I+UElXBUUD535Zv5GBLlEOhmOPpWb+eORlP1gQw49kqJix8j0m/+SvNEZ1f/OFjNn/WzMHmKOX9inh0zVaEMKhWBPdeWsawvgW8UV3LgcMRS78zRw9wZC6cOXqAS9bwJUNCl45F6Jqe1DTKVL/2wPjBzHtjKzOereKSR/+PvQ1hJv3mr3zWFKGmMVn/kHSWcuhdELRQECuKQNMx2U4z1cABJ0RE1BWPd+HCxeeFmKabDhwk57YfLvsbMVvNmSJg3pVDLPPSvCuHpDldmYhS4g5EKU6i4Pa5NBzTmPfGVrbXtVDXHGV7XQvz3thK2FamkW3dmUcRTBvV3yIyPm1Uf0cJhEykGakw9lfGnmrRlAqK8/yWfYimS96orrUIpb9RXZvWl4vskdBlxt8ooSc1ZXvkugLfXRkdYafcCVwjhPgAeBnYT5KkyISU8nedN7SuhaBP5cffHoSqCMaVlVhSBMaVlXD2KWUU5fkQQtAa0/h124lRUa6PHrlJavW6liitMY2Ypmc4PRMZc82NWjpIPmBSSnMMRsraTy4+h1hC55evVJspak9cU87SdbuIxDVLv5k21QVBLzO+3g9ICo56bREWFycfMm0IDPpmx4iVqqAITObUVVV7aAzHKc5Lasml1n48ee359OrmJxxrTzHUdUk0oZk1n/EMNn+i8uzd1EcXLlx8XsjWWRFCUJTntUTEQG87fG1HtvVf0YRu6m8a0akHX9vCr68eammnKM4EI/b5MNv7hmOa430fv2aYJSqmZEmUEvBlkCJISZX0ZFnX5yJ7eBWBJp33AKoiLH9/F10THfmFngD6ABXAL4Angd+k/Pdkp4+uC8GIZqgK3HNJOz36sL4FXDeyP5N+81e+/chfmLh4HXXNUeqaY2bk7Nrfvc93vnIqiyZX0CPXy/qd9Sx0iNT98pVqPIpg/qTytJO1hW9vN8diPGCGA3fHRYO49+WPuPChPzPpN3/lupH9Gda3gJqGMDc/v56bRg/glO5+y4ldsnjVKpJZWhjE71G4dGgpExev45tz3+aqRe+x+UAzCTvrlIuTBoaWUCpKC4NIIM+vptnb/Enl+DyCO1duMk9ZZ48ZyJrqA9w6dqClkN4gDNm4p4lRD7zFFfPfZW9jK5sPNHP14nXm9XkBT9oJ9ImOjBmpjyci0ufChQsXBgzikFQYRE+pUBRBXXOMqUveZ8y8PzN1yfvUNcfSnRqv4jgvO+mrGaLgRnSqriWa5tjoOo5pjfbaOb/H+b5+2yGuz6M63td+KOdTFcdIoc82vljcOVskFm8fX0meP23fZNRhuzg2+H0Kfk/6nnP+pHKaWuNEYu6+r6ujI5G4/idsFCcB4m2UugebY3RLiWLNHD3AdKaMU6nWmMaPvz2I7z/5VyA5IZ1RnIvPI9jbEOGayn74PAr3/+tX8KqKhfzknkt0iz4MQCiaoDjfx6IpFWZNnEdJ9uukt2XocBkkEvWhGM2RBI+u2WaOUREijf1q4eQKivN8afpfM5+rYsWMEfQuCDr8ZVx0daRqExXkeMkPePF7BKoQNEUSjnpEV13QN61+zSDNcTpxzmljEatpCBNNyLT6htnPb+Dhq4YeE2lOZ+u9uXDhwkVnQhU4slOqtmkqlnDWTfuZja0xHNd5/M1tlnaPv7ktrZ0QkgWTyjnY0s4m2TPPhy2wR0KXlj2KwYhtjxSGjqBP1zOlXbZMvT1yfBzO91sIWorz/fTIsbbLRp/O41EYVJLH8umVJHSJRxFultBxIhzTeXbtDqaM7McLN1aiS4mqiKTu8atbuHnMmV/0EF0cBR3Ridt1IgfS1WEQJexripDrb2eKLAh6zWhYaqrCgknlDOtbYOrEHQpF8aoqty3/m2WSv//VzRaZANVBH2bG1/vxw7FnmfVFpYVBFk2uYFxZSca0yN7dA6bT1yPXxynd/GZqnHHPP8wexYrplcQ0yY6DIe79fx9R1xLlgfGDqWuOmeOqaUjXk3Fx8sBJm+iB8YN5eu0OZo8ZSGM4xrcefsdsX1oYZGxZL0sfNQ1hTuuRw96264+U7pupzq5XNz+lhTkdcsBOhN6bCxcuXHQmIgmdl9bvTXN+Zts2wUoG3TT7VGbUf6VKsEAyCygVKoJIXLPM7Q9fNQR7knrA45yuaI+wZatPpyiCgcV5rJgxgrimm2UXaemZHoV+PXLJ8XlIaHoaAZYBX4ZUydRIpq5L/nEw5K4FnQhNl7y/s5FvnRtN25tOqjyNgOsgd3lk/QsJISqFEFdl+OxKIcRXO29YXQ/GydP6nfUoCmb4uTEc59axA9OiYTctXc/M0QPMCFdhjj8tOnHnyk3cOjY5ORohbCllWvrBxOGnmw6cce2M56q4+zvnOKZFjisrMRkyjXTOPQ1h7n91M3NWV3PHRYMYV1ZCKKaBEEz+7V+Z9tQHbNjTaCE4MeDmnZ/ccNImumvVJsZX9GXW0vX85OJz0lIpVlXtsfRRWhjkH23F8A+MT0+PSU331dty7O3Xe1Slw4ttJr23+lCsY38EFy4+B/S7+5Wj/ufiy4egT+WK8j5Me+oDxsz7M9Oe+oAryvsQ8FndKSmd0xptCgMZU+Dtem2ahNtXWOf221dsRLP1J8ExXdFOCZLTRuOfOr8vmFxBjs+eninZVtfCVYveM8suttW1pKVnQtKR610Q5LSi3DQCLAM+rzNBi8/b/n0PhqKOa8HBUDStPxfZwasq3Dp2oOnAQbttHArF0+zDRddDR9Ip/xt4J8Nn5wA3AWOOe0RdFAZRwtRRZ3Df7z/i5gvP5IUbK1GUZIqEU+Th7FPyee6Gr+L3JtMondr07RHkzR99E0hqwzy25h9cU3maJf1AEcLxWkUI8vyetLTIey4pY9Jv/pr2UBoplnet2sQz1w8n6FXQM+jWGGkRbt75yY9MRfdGFFcVwmJv3YMebhkzkOr9zaZNzbtyiBnBfej1ZEH7Oafk41UVapsj1LUkF9LSwiCarqXp+RyrDZ0IvTcXLly46EwkErqjk7TCRqWfkQDF5sX1zPU56nD2zO14GuIR72tzunQgz6+YqXWKECR0DXseTqbDtWOVJoU2WtQAACAASURBVIrEnAlaUolSInHntSASd7OEjhUeVRyxRCLmciF0eXTEiRsCPJjhs/eBW49/OF0f0YSelubw3A3DHVMB9hxq5e7/+TvP3jCc+lAsQ5swZ5bk4VEFOw+GWFFVw7baFmaOHsDp+Tn4PQpej3OqQU1DK3FNckr3XJ65fjgeNVnnlMkxKwh6zX83heP4Pf6Melqndg/wzp2jM6Y/uDh5kIkhrDEctzCFGYvv4XCC4nw/j39/GNGETmtMsxTUb9jTyJzV1Sbz489/X21ZfO9e9RFPXlvBihkjjphCkw1OlN6bCxcuXHQWspUEyMQ+bdeJUxSFkm4+lrXpcCbriAWKkp7+mIlZMBWZ9hBe25wcien82/JkJk5q7ZyddbKzD9e8nnaClkzjy6yxd0y3dAFE2oIETn/X1piWZkcuuh46sqsKHKG9yj+B3GJ9KEZdc5QlUy8wtUyG9S3gmfd2prEmzbtyCPPe2EpNQ5idB1tZVbUnjZ0vmT4QACSqgAdf22Leq0eOD6+qJPVZdMniKen9//5v+4hpOgcOR/F5FHp3C3Jq9yCq4syUZdQtlRYGaWyNmyQRTnpap3Y/cvqDi5MHPgctoQfGD2ZV1R7mThhMUzjOtKc+4Nrfvc+22hZuWrqeT/Y38735a5m4eB3TnvqAm5/fYKb+jisr4fkffJVYQkMIQXG+L42lTCKOmkLjBF2X1DVH2dvQSl1zlMKg19V7c+HCRZeGJ0P6o52av1tQdUxX7Ba0HkrVtkT56Usf8fG+w3zWFOHjfYf56UsfUdtiTR30ZNCJs98323ZZs056VMaVlVh03caVlRzz4Vo24wv6VMc2QZ97oHesUBVB7eEoD1+Vvjft2yOYVjPpoutBSHsydqaGQlQBH0opZzh8tgj4qpRyaPqVJxfOP/98+eGHHzp+dqApzM76VrO+yDD24nw/Xo/Cp7UhMyWttDBASzRBQdCHLsHnUXjuvR18/awSTu0exOdRaGiNcd/LH1Oc7+OeS8qQMklCoUuduuaYmZ4xrqyEOy8aRE1DxMJA5fUI9jdGzffOLMmlJarx8B+3pBVPz50wmAdf20JdS5S5EwbTq1uA0wpzaAjH0XUdTYKU8mRh/+vSg/sicCS7DUUi7D8cZ8+hsIWdcm9jhKI8H/XNEZoiGiX5fnrm+RFCsvNgK/Pe2GqS2wC8c+docnwqB5qjJvOpsQl5bM1WU5vwWIvNM5GYDCzOoyEcP9nZKU+6AX8eOJLddjV0Vj3bzvsv6ZR+Pie4dusAu91GIgm21YcsKeQLJlcwsCiXQKA94SkaTdAUixNLSEuErbvPi9/f3m5fYyv7GiMWsolHJg6lT0GAUwtyzHa760P8cNnf0iJnv756KKcVtZ+r721o5d1tdYwcWGzed+22OkYNLKZPYXt/2RJJxeMam2tb0r7v2SV5eL0dd6r2NrQy+/kNjhFAY3y6LtlZH2JXfau55zm9KId+RblO64Frtw6w2+2hUIRDoTjRuEZzNLmvVBWB36MgRHLfWpQb+AJH/E+HDtttR9IpFwKLhBCHSWrC1ZDUjZsO3ADM6ujNTzZoDgQRRt77VYvXme8P61vAz79bRiSuM+WF9y2MlZB01Oas/pg3qmstOnNGu/mTylny7g6zv/EVfZn21Idp4e5nrh9uYaV65vrh5uRb1xyziI23RBM8cvVQpIS65uRp3u6GVq793fu2yTpdiNPFyY2ohlmgHEvo7DnUSmmPIEGvwtzXNnPfZedyR5v9GlG6Z97byR0XDTKlL0oLg2yvCzGgONd04KCNxOe5Kp6aNpx/v7iMHL9Kz9xj02Tr7DoLFy5cuPg8cCgco2rHQZ6/sRIpJUII3qzeT1GOl94pTlxzLMFnTVFmpdSwz59UjtJdWJw4gUgjm7ht+d94ccYIy31TdeIM2FkdAfICKuf0KeDqlHl+weQK8gJWh8uo/TdS5TMdmtWFYqYDZ4zvpuOQIhJCOH6PVBF0RRH0K8olP+A92Q/0ugwicUk0rpMf9BL0eTgUilEfirGqag83fv0MTu+Rc/ROXHyh6IjEwJNCiEHA7cC/pX4EPCylXNzZg+tqkBlqzeJtGizGZzNHD+BQKJ6mt3bT0vVm7ZBRU2fozKW2m9XWzmiTSUbgUChmuS719YY9jeaE+NKskdQ2R4kldLoHvfzX/35CXUuUOZef526Y/wkQjulMW/JB2iHAkqkX8EZ1LfdcUpbGlnb/v36FWEJn3lVD8KoKigKRuE5LNOFoi/UtUX704kZemjXqmBdVl8TEhQsXJyPiuuTnqzfz89WbLe9feM4pltfRhG46cNC+3i+3EaBkS1ji8woevmqIyVBpSAyksjoCtEQ0R6dr+fRKuh+D/Gs8y/FBMmpX2xK1aLvZo3WqgAfGD06TXrDXuymKcPcnnYi4ppPjU9leG7LsVwGq9zezzGaXLroeOhKJQ0p5hxBiAfAvQBFwEPiTlPLTEzG4rgYjD3x8RV8Kgl7imo7foxLXJL+aOIQFb21nRVUNJfn+jIyVRqpAqs7ckdghAUt7A6WFwTSa9UzkKd2DXm55YYNlcnzo9S2mQHPqfd0N85cPiQxF90aRvbAV1SdZU3P45SvVZoqkoSt3zyVlpo0N61vAzNEDKMr1UZjrozjPf1z245KYuHDh4mSENwPBiF0SINNcnMiSAMV+QJZISLwexcIu7PUoJBLW/jKJfdvvq+uSvY2tRBMSRSRFxyPxBH0KrPqemQhV7DV22aZdKorC02utIuhPr93BL68YnP7HdtFpUBVBXJPk+FRHu9R1SSKhu7wIXRgdcuIApJTbge1HbfglRGHQyy1jz7JMSHMnDObOFzdS1xJlflu6ZPegl131rRkZfxa+vZ35k8qZtXR9RgetR67PfH9V1Z40GYGFkyt4dM1Wy/hWVe1h0ZQKS73SE9eUc/+rn6RFWuZcfh6tMeuG290wfzmRacFVhGDB5ArerN5vaV9aGOTTuhDXjexvir7ftSopUfHLV6qZP6mcx9/c5lh3eTxF5gbJjr0ewyUxceHCRVeGR1WSewGbmLZdXzVb5yfoVcw9QmraZdBr7S+hS2Y/vyGtP3tkL5PYt13M+XAkRkNrPO2++YEYBTntETC/x3l8diKM2pZoxghgai1eUa6P2781yJ37P2f41KTMlKE3bLcjCTRFYhTluXVxXRVZO3FCiJFADynl6rbXPYAngPOA14G7pJRf6jDOoXB6Hniq/tqspet54cZK5qz+mFkXnpk2qT8ycShnFOfw2DXD0HRpTrTP3fBV/ut/26MeCydXoEvJs9cPR1EEXkUQ9Ck8dOUQ/B6FPL8HBNxzSRmAed0Px55FcZ6PZ64fTiimketTCcc1ixyCMe7+PXPRpTQfXHfS/PLC71H43dTz2ZtCjNOnMECeX2V3fTMDT+lusYMHxg/m5Q17+c5XTmXulYPZXhdi4dvbzTTgW8YM5M6LzmbaUx+kPQv/M2vkMY8z23oMFy5cuOhKiCQ09tSHLJIAa7fV0cfGWOlVRdqB7IJJ5XhteYOaDo+/uc0SmXr8zW385/e+YmuXvf6bk47dizOtNXahqJYx3TOFT4VIQufPm2t5PkVP7uX1NVxRUWrpL9vIozv3fzGQUhLwKJzVK48XbqwkmtD5rCnMM+/tZPaYgfzylWruu+zcL3qYLo6AjkTi7gfWAKvbXj8EXAz8iaTQdxMwp1NH14Wg68nTikwpksa/dSmpa47x/Lrd3DL2THNS9yiC7bWH2duoWiJ5RprarWPP4uffPZfWmI6ma7REE7y6aR8Tzj+NxphGUZ6P04uC1LfEzc2z4fD94vLziGs6z723k/J+RZQWBinM8aLpku5BL+PKSkwSFSP9zedJancdz6Sp65L6UMyddLs4YprO4XDCQoLzyMSh5Pk9DDylG1LCsumVtETi5Pi9KEJy7ch+FqHZX189lJ55flbOHEHQ5yEcc66Niyd06pqjx2wTn3fNg2vDLly4OF7k+1XKbMQhCydXkO+3ZiZICd1zPBYxbYmOnSQ8piUZqlNR1xwjZqs586iKpcSjMRxnVdWetAhgpvKOuE3MOdu0y6BP4fz+Pbjmyfbvm8zEOLbII7j1bl8EdCmJa5KEJi2kJreMGUjAq/BGdS0/bQsWuOia6IgTdw7wAIAQwgtMAG6TUv5OCHEbMIMvsRNXH4ohUyJXBuz6a5/Whfj3i88mP+gxT7uMSW7Z9Epzkof21MZ7Ly1j5nNV3P+vX2Hyb99vK04eymVDSy0O2zPXDzc31sb1M5+rYs7l55HnV/nO4N7Mf+sfXDeyfxoFfP+iHL4xqJcl/e1YqeAheypiF188pMSR6Wz59EoOtsQ4pZsfTUpimuQHT67j3kvLmLO62tL+h8v+xpzLz2PaUx+YqTPG4YCB0sIgmi65Yv67J4VNuDbswoWLzkA4Lh3X5hUzRlCY0k5RBHsPRtLSGs8sybP051WEY/qjvcYuxyfSSjwWTK4gx2fTf1MziH3bnL2gV3W8b9BGRBKLS+fIno09M+hTWDC5Im18dmfPxRcDj6pQ1xKxpMU+MH4wj725jbu+c07S4Vbd36oroyO/Th5wuO3fw0mKextRufXAaZ04LguEEGcJIX4thNgkhGgRQuwXQvxeCDHEoe3bQgjp8N9txzOGWELDqwgW2YQ6504YzMK3t5tpEf175lCcHyDf76VbwEtxXvJkKUkakpnspKYhTO+CIIumVFCc56dnni9tUUhln0y9PsencvuKjXQPernzorPT2C5veq6Ka1Pql4z3b3zmwzRylFTYhZf1lNO4THTwR+rPxReDI6XcPP7mNmKaRNcxJ/JMZDsGEY6RYmOQnEDyWVg0pYL/fKX6pLEJ14ZduHDRGciWrTGa0B2dn6hDRMypnT0iForq7Kw7zLLplfz5ztEsm17JzrrDhKLW/hSBo1C2/awqoTmPL2H7HtmyZzZHND7Z22gZ3yd7G2mOHFvlTSKhs68xzK76EPsawyQS6WyYLrKHE1vqXas2Mb6iL4KkzdgZQl10LXQkErcXGAL8BfgO8JGU0jiGLwRaO3lsqRgHXAg8TdJhLAB+DPxVCDFKSllla7+JZGQwFTuPZwA+j0pTOE5BjtdMNfB5BN2DPn41cQg+VUEi2XYgxKNrtplEJ/89/jx+suojNuxpRNMzR/JKC4PsawwzZ3U1cycMRkLaJFkfijGurIRrR/TjlO4BVCE42BIjoScn1NrDUVRFOE6u0QwOZCY2waNFKVw6+JMHSoaUFk1KrhvZH5CE4+2OXiayHSPiDO22+dS04XhVQdCnIvVkKvGiKRWWVJyuahOuDbtw4aIzkIlNUlXstW7Z1bBp0jmtUbPlXeb4FM7uXcC2Ay1mvfPZvQvIsUW6IgmdB1/bYunvwde28MjVQy3tsq1h82aI7NmjNgGPQt+iXEuaqROhChw9tT2R0Nl8oNmS5r9wcgVn98p32ROPEZns0Si5eWn9Xm4ec+YXNDoX2aAjlv8C8F9CiJUkdeKeS/msHNjWmQOzYRkwREo5T0r5lpTyJeDbQBj4oUP7ZinlOtt/nx3PAIpyffg9Co2tceasrub+VzfTGtO59nfvc+FDf2bi4nXsqg/zwvu7uOOiQRTn+Zm1dD0eRWXm6AEArPxwNwsmWSN5D4wfzKqqPcydMJi8gIfiPD93rtyEz6OY7Qys31nPrWPP4u7/+Tv/8qt3mPK790noOnl+D+PKSixpnakw0tyc3s/ERnm0KIVBB59tfy6+OOS0pbSk2t2CyRXEEhp3rdqElCIpAN72+cK3t/PA+PRT24Vvt5PSGuyWU5e8z77GMB5F4PEkU4DmrK5m4uJ1zFldzY+/PYhAF02dcW3YhQsXnQGfIhwjXT6bE2dIEaTCSYrAryqOc6k/rdZNcrA5yr0vf8TExeu49+WPONgcJWaTGPAq7WLaExevY8azVdS1RNPu68kwPnsNW67feU3J9VvHJ4RwjOzZZW2MQ+Mr5r/LqAfe4or577LlQLMl+6e2JeqYslrbEsXFscFIs02FwY7+iz98zBXlfdJSaV10LXRkd/VzkjVxfpIkJ79K+WwI8GLnDcsKKeVBKa1HUFLKJmAr0OdE3TcViiLID6gU5np54pphzL1yiCU9MamRpfPvF5ehtuWz1zSEUURSrLu0MMg3BvXi7c0HWDL1At6645s8f2MlpxcFGV/Rlwdf28Ls5zcwc/QAahrCqAosmmKdJCcOPz1tErtz5SYOtsT494vL6N09QJ+2xcPuKD75zqdpG/MjsVEeLUph0MFn25+LLw7hmM5ja7Zy76Vl/GH2KJZMvQBdl+T6vYw8owhdSh5ds820mw17Gnl67Q6euX44L80aydIffJWgT6WubbE0bEpR2m0wHNMy1knE4tJxXEdK1/084NqwCxcuOgUCeub5mHP5eSyfXsmcy8+jZ54PbKlonjZ2SovzM6kcjy1nLdt0yniGdnEH3bknrhnGkqkXsHx6JUumXsAT1wxLixQqGZxRe41wS0Qz15Tl0yu599IyHluzlRZbmmRc0ynO87NoSgXLp1ea5SL2tMtsUtszpazaUz1dZA+nNNuFkyt4ZeM+3qiudUylddG1kHU6ZZt8wC8zfPa91NciecxyL7D4eCNgmdAmcXAesMTh42FCiCYgB/gE+LWU8rfHe89QVKM5EicS14lr7fVpw/oWcMdFgyykIQsmVzCurARdQu+CIE9fP5wHXv2EN6prmfenZNCytDBoyhMYMBw+j6JyzilBVswYwb7GMPWhGE3heMZapaZwnMufeJeVM0dY0iZ65PqY+/pm3qiuZVttC/deWkZRro/eBUFO6RbISOBwNOFllxL45EFCl7xRXUtdc4w7LhpkobeeP6mcoE9lw55GHnxti8mmuvmzZn60YiMb9jSyaEoFq6r2pAmx3ntpknq4piGMJjHTelPhVCcBXYNUxLVhFy5cdAbimmTlh3uYcP5pqIpA0yUrP9zNtSP7W9rFNMnqjXtZMvUCS7sptnZZp11maGc/ENOkJBLXLQzF864cgm5Lz4xmmXapyeSaYpcv+umlVjr6oC8DUYpNTzSb1PZsUzhdZI9IQuel9Xt55vrhJjvlo2u2ct3I/gzrW8CGPY1pBwIuuhZOlPUrwH1A7xPUP8BjJM+5HrG9/w5wG/Bdkgya24DfCCF+erw39HlUfB6VH724kfpQzDy9mDl6gCOZyD2XlOFVoaahlRXv72L2mIHmNePKSnjm+uGU5CdPqYb1LTBT1Jb+4KtE4xr7m8IIJMX5flZV7aG2OUppYZBhfQtYNKWCP8wexZ/+7ZuUdPOT41MZ1reA+lDMkjbx45WbmDaqvxlhmbO6mrim41EE+5vCGSMg2UQpDErgPoU5FOf73c1vF4VHEYwrK+HBCYPT7HTW0vVIXfLOjy/kVxOHIABd11lVtYeZowewfHoluT6Vmy8805LaM3vMQA6F2iNzAa9iLrKpyLTI2k9ei/P8fNYUoabx843KuTbswoWL40XQpzB99AByfCqqIsjxqUwfPSCNhdGrCC4e3JuahuTaW9MQ5uLBvdPSGtUMaY32yJlT2UVpYRCvrUZMSvjRixstc/+PXtyIfZr1qgrF+dZMhOJ8XxqLZcDrnIoecBAjzyaimE1qe0men4W2FM6FkysoyXNlCY4VHkUwecTp7KpvNQ8I6ppj3LVqEw9OGMy4shJHOQgXXQcdITbpKDL+8kKIfwH+mEUff5ZSjna4/ifANcANUsp/pH4mpfyZrfnLQoiXgHuEEI9IKVsc+psOTAc47bTMJJuFQS+haFIfy6gbumvVpoxsfgJ4/M3trP20ngWTK/issZXl0yuRwKFQjGt/977ldKp3QYCmcIJJv/mrhe716bU7+OHYs+jV3c+SaRdwsDnK25sPcMrgPkxd8r6l7csb9jLvyiHmhF3XEqUoz8fzN36V+pYYtc1RHnxtC7+aOIR/W76RupaoYwTEjVJ0fWRrtwGfwi1jz3KM5Bbn+TnYEmOGrVj8x98+m6lLrHqEy6Z/lVhCokvwqDD/ze2mc98z14+m6Y500sUO6YmpJ69OkWyX6v/Li2zt1oWLroQj2a1HgV2Homlz3+k9rA5Gd7+XZn+CQ6F2kqhcv4fufq/tXpj7i9T13VZKRsAjWDSlwiIptGhKBQGPjVBFZojspRGlZCdZ0DPXz5NTzufGZ1MyKaYk14FUxLPUpzMOje2ZGamHxh6Pwtm98lkxYwQJTcejJrVuXVKTI+NIdpvjV2hoFZYI7RPXlPP/2Tv3+Diq8+7/zszs7K4uti62DFgyGGLsuKkMWpEmkBIHcoEXkpRXAhokTEhrsE1uvNRA3zRpkjZtQKEkJbUEakMAmQSw35QkpJAW6rgvTkMkgd03BOP4hmSoZetiS9rL7Myc94/ZGc11tWvvrnV5vp+PPtqdObeZ+Z3nnLNzznMEZkz93XzVKlRF6f7OZM7U09kFY9+56f7WuSMyxjYA+BsAf8E5/16O+f0AQATA7/ud5Jw/wjlv5pw3L168ODCR0UQaB49PWm+1vvXCXjz66Utw1sKI769Ib59I4srVS6w3cxeetQBpjWPf0QmPW9fN2/ZA5wh093p7Tx9SaR3DEwoeffkgbnzvubjjSW/YK1cvgcCYNTf/y9euxj3b/gs3df8KQ+Mpa0HzoeNxa/1dkFt1eksxs8lVt0lFx8aePsfbY5N7rl5lDeCAqcXixycUzzGdMxwbT2H/sQn8zXO/xRc+vAI/2nSZNdgaS6q+6yTGkqqnTPZfXv3eZJOr/7lLrroliJlENt2OJ3Vr4ANMzcYZTzoHKydSaRxzOSI5Np7CiVTaEY6BZaasT9nSx3YdBHP9Nh5XdHzn35w29zv/9ibiimtTcCFgloTg7ALGFe57HXHFOzMiHBIcawDDIW93MpTjm0IACEuu9HzCSJKAc6qiWFZbjnOqojSAy4Fsuo2ndI+fhTue7MfbJ5K4+Z9ewfHxFCZSNJ1yJlPMN3GBcM7jAN7INx5j7GYAWwA8wDn3XZ8XFNXMOt887Siqhr9/cZ/1C9mrA2PoeOENfPUTv4eO1kbPvO+zF0ZQbttbayKlYmE0hHfVlePRT1+CiZSKofEUunbsx6sDY5BFweNS+NWBMetNX1rnYABaYg2Be8bVlstYsiCMkUmGmvIwliyI4Fs3rMHDO/Zb6+3MN3YbM9PlxhJp6LrT6E/n7peYPZhuo+1vj02dnrUw4qujsxZErDnxgPHGznz65bKIWy9bDsaMOvH2iQTCkghd1z3rJC5uqEJK1XBkNO7Qkf2X16A32eTqnyCI2UCurvmDHJH88Lb3OcKJAsOtly339Cnc0ynVzLYudo6NK558TQcW7vTcTXqu+90NTyr45r/8Fi2xBpRBhKLp+Oa//BbfuK4Riyun3sZJAsPD7U0YGlesLRDqKmXPFL1h28wkk/rqKH606TJHetQvKSxu3V7cUIUNay/AiroKfPna1Xj05YP4ysd/L0sKxJnmjAziTgXG2HUwnJj8I+f8z/KMfhOM7Qj+63TKIEuGh75nXz2CrX/6Bzg2nsLwpIKxeNp3MfCXrnk3QqKAixuq8N7zqsA58PWf/Aa32DbeNgdVO/cetbYvcE+lNPftEhhDXNFQWy5bb1XcRu+chREIAiAIAj7VPbU3S2dbExZXyvjytavx7KtHcF3TUtxsm8758M0xLK40HJ3MBKcTROEw3Uabb49N5zYLoyGAw1dHHBx/9rGV+NYLewEAd1+10qGnB29Yg6HxFD775KsODX10dZ01iLu4oQp3X7XSsUeQXUf26brZnOiYUANOEMRMxLSxHqcbOe4T514DnEhrvn2K77gcjEQkAf/7f6zCnU/vdthm9z5sQQ5L3OmFRP/rCLm8Z+q67tuP8fsxWOdwTNfrbGvyXK+iar774tl/yKN+SeGx69ZvWcN9LcZAX9c53eMZyqx4F80YuxzGlMg9AL7PGHuf7e9iW7g/ZIw9xxj7E8bYlYyx/8kYexaGk5Ovcc4nT6cc5tuDq3//bLT946/Q2vVL3P5EH0ZczkTMKYvDkwrueLIf91y9Cje97zxs6OlDS6zBM3Xsnu170P7+5Z5pbfds34N7r343tvcNoKs9hpdefwf11RHUlMvY3jfg2TJgS1sTjowlMZ7UvFMitvYjrui4/Yk+XP37Z3t+Dbz9iT5r+lou7n6J2UNEErAl49badG6TUnXcvW0Pjo2nfF1K//eJJO7Zvgcb1l6Az1+5wqOXO5/ejdHJtEdDf3HNaistv3h2HZnTdc9eGJ3WiU4u+wgRBEGcCSKhKRsLTLXHbkcf4YDphbJr0BUSBd8+hdtJlA5YAzhgyja7/QEH7RPnHmSKzH+LAZG519jBtx+jucxxStUtb8hmuI1b+5FyrYkzvVi698Wze7GkfknhiYQEy1mM37KGe7bvQTKtW9sLETOP2fIm7goY+9NdDOBl17nDAM7LfH4HxsD06wAWAUjDGPjdxDn/wekWwnx7UB4WHb9UmXuwjEymrSkD1eUhfO3Hr1tT00xPVEFTx/SAhceiwLD5Y6vw6uERXLn6LAiMQZYEfOHKC/GdzPqj2nIZNeUyunbsR0usHqLAAtN6+Z4PBS5yNn/1ysXdLzF70AEsrY7gqdveh5Sq4/BwHN96YS+OTaQQCQmIyiH81SffY2m3IizhK8/+BoOjCaw6qxIAfPVQ5nITbWrsqdveZ0ztrZBz0lEuTnSCGnD3dBuCIIhSM5HS0PPLw46tA7p3HsAdV7wLtRVT4QTG8OANazxvzgTXIMn0xLjB5XDK7YlRCXAcorgGSWDAt2+8CF986jUrvW/feJHH/Vwyxy0GeEAfwrWdb+A0U/dWCarO8ejLBx35PvryQfz1dVNuDKhfUng0naO6LIQvX7saF9ZV+N7fE4k0eaicwRRrEMdhDK4KMnznnH8Vxmbj04X7HYCrC5FnEILAEA1JjikHTBElKgAAIABJREFUFWEJKdW7Bwtgulhn1vRHc2qkZ/pawLS2N/57HF079uMvP7Hamj8/OJrAR1fX4c//x7sxFk9jaDyFbzz3WxybSOETF50DTecBUyKMRcHHMlsVBE1fm26POGJ2IQnA4FgKx8dT+MErh9ESa8C9V6/CWCKNn+4+gpbmBkd402OZOYX3d0MTvnqIK87G066Ru57ZjS9fuzpnHZlv5YKgBpwgiJlKSBSw68Awnu4btI7VV0fxxY9c6AiXTGv4m5+94Ris/M3P3vBMaxQEhoVRCd+/9b3GdDYOhCXmmdImBkzjdK+dExlDSGSOH+tCIvO8YZNsb+zs6bk78bn2EeSAvd3cWxZwnftOz+S2wR71SwqPonHDEd5PXw9sr8fiadTRD6UzlrynUzLGFjDGLmWMXZ/5v8AdhnOuc86Xc85/U5hizizce6hFZcn6hQuY2oPl81euwMPtMQgM1vRHv2mQHa2NOJFQPNMYHm6PWft1jU6mHVPTfv76EG7+p1cwFk9bUyPMhc/dOw94pnbYf8Wbbg+4XPaII2YP8Yx3yr9/cR9uuXS5Y8rKje89F5/5fi9u/f6vceMj/4lbv/9rfPbJV/H5K1dY0ypNZz52PTx4wxpUl4d8NWLqx0/rp6qjXPYRIgiCOBOIGcch3mmIrnAB0xrdg67hSQWf6v4VPvx3v8AVD/wCH/67X+BT3b/yTB2URcE3X9ln2uUdT77qsPN3PPmqZ9plWPKfFur2FJlrH0EO+U/PlEP5T8+kfknhCQkMUka7QX3T+uoIvYmbweT1Jo4x9hUAdwGogPEingOYYIx1cM7/ugjlm5G4p38FTYVsqImiMiyhpky2pj+2xBpQVRbCD9YbXiHfHkvg/uf34t6rV+H+5/fiic+8F8OTCmrKZTAG3Hv1u8EADGWmY/rlYXqYNJ2pPN03iOoyCT+87X3QdI6Qaz+V6aav0R5xcwtzSsvgaMJybFIVDaGuMowxn73jBkcTWFZTBg6Ozc/scThEqYqGcHZVBLIogHOOpzJe1dwaWbmkEt+4zljo/vTt7wfn/LR0lMs+QgRBEGeCXKchCgH7v7lNYq4zD2rKZJysDDvesC2uDKOmzGkX9RwdqiRV3Xda6Oc/vMJ5HTn2EZKK/3357k0XA+VT4XKZnkn9ksIjCgyTiuFEZ8PaC7AgIlme0xdXhnFsPIlkWkdYcg/3iZlCzoM4xtjXAHwZwD8C+CGAowCWAPgUgK8xxqTMtMd5gX36V9D0xIGRBN6zdCFCIRErl1Tia594D9Kajv3HJvEXP/p/lvt2c5qluXh0eFLBIzv3Y/PHVuHubYZziaBpCfuPTVpTH+qrozinKoqX7/nQtAZuuulr050nZg92D1SvDozh9if6UF8dxZevXR2oq0hIsBbXA3DEe/r296Mu48k0iELrhxpwgiBmKrlOQ+S2/d/MQc1juw7iq594jyNcrlMHJUnAeTXlKJOlrBtg55peWBJ9p4VullZ5rjkXG2969HbfF8+0yxzLR/2SwsLBArX76Kcvwed+8Br+6pPvQW0F3fOZSj7TKdfD2J/tNs75S5zz32T+rwfwIDI7ws9H/F7zd7Q24tzaMutNgSQJOLsqinMWRo1fODKdY/OXuO19A+hobcT/eno3/uqnr+PzV16IjhfewKsDY9jeN4CGmqhnWkJXZrql+b17XTPOWhChzbkJB4vKZHRmPFABTs2dW1Nmeacyz3Vmpt4uqgj7Tl85a0H2AVyxoM3nCYKYidRVhD02ttPHEUldRRifv/JCx5T2z195oSdcPlMHc9kAO9f0Cj1l8UzlS+RGbbmMkMQ82t3S1oRtvW+ho7URDTVRjz6JmQNzexMKDMjYJIBPcs7/zefchwE8yzkv98acXTQ3N/Pe3t684+k6x/HJFJJpHSIzXOZWRf3fFKiqjqGJFFRNhygYv4SwzP+EYrxlqI6GMJpIW28dqqMhnEylkVA0aNxwDVsTlR1h5tGbiXlxkfkwnW5TKRXH4wo0nUMUWOZPQG25DE3L6FHnkASGuoowQiHjF1Dam61g0E3z4VTt7ZngvHufK0g6h755TUHSKRGkWx/8dJtOa4F21I69/Q96cwYU3vbmmt4cyZd064OfbnWdY0JJYTyhW9oNiQyKxiGLAmrKZF99EkUhb93msybuVwAuAeAZxGWO/yrfzOcSgsBQVxnJKaz5y5kvtmGwe9pAjRR2nPcLQxB+hMMSlob9q7sgiFhaXRZwjqavEARBTEcoFGxH7WRt/20UY0p6LunNlXyJ3BAEhgWRCBbk1n0lZhhZB3GMMfvw+/MAfsQYUwE8g6k1cTcA+AyATxarkARBEAThJpe3Y7m8+SrUWzaCIAiCKBXTvYlTYXigNGEAvpn5g+v4nhzSIwiCIIicKMTgigZoBEEQxFxkukHX1+EcxBEEQRAEcYaZp2v0CIIgiAw5OzbJO2HGlgF4m3OuFiWDIsEYOwbg8JkuRx4sAnD8TBeiyLiv8Tjn/KozVZiZSA66nQ86AWbudZJmfThNezsTn/VcKxPp1odZ2E+wMxM1mi/TXQPp1odpdDsXdJGN2XB9eeu2KIM4xpgIQAFwCee8v+AZEBaMsV7OefOZLkcxmQ/XWGzmyz2cL9dJzMxnTWUiZjpzQQ9z4RpmGnP9ns7V6yum31By8UoQBEEQBEEQBFFgaPMHgiAIgiAIgiCIWQQN4mY/j5zpApSA+XCNxWa+3MP5cp3EzHzWVCZipjMX9DAXrmGmMdfv6Zy8vmKuiUsDaKY1cQRBEARBEARBEIWD3sQRBEEQBEEQBEHMImgQRxAEQRAEQRAEMYugQRxBEARBEARBEMQs4rQHcYyxWvcxzrkG4EMA9p5u+gRBEARBEARBEMQUOQ/iGGPrGWObbd9/nzE2CGCIMdbLGDvLHp5z/gvO+WQBy0oQBEEQBEEQBDHvyedN3OcAJGzf/w7AGIAvAlgI4OsFLBdBEARBEARBEAThg5RH2GUA3gAAxthCAB8E8Eec858xxoYB/G0RykcQBEEQBEEQBEHYyOdNnAhAz3z+AAAOYEfm+wCAusIViyAIgiAIgiAIgvAjn0HcPgDXZD7/MYBdnPN45vs5AEYKWTCCIAiCIAiCIAjCSz7TKb8F4AnG2C0AqgFcbzv3IQB7ClkwgiAIgiAIgiAIwkvOgzjO+ZOMsbcA/AGAX3POd9pOHwXw40IXjiAIgiAIgiAIgnDCOOe5BWSM8VwDEwRBEARBEARBEEUhnzVxhxljX2aMnVO00hAEQRAEQRAEQRBZyWcQ9xKAewEcYoz9H8bYR4tUJoIgCIIgCIIgCCKAnKdTAtb+cLcAuA3AagAHATwC4FHO+VBRSkgQBEEQBEEQBEFY5DWIc0Rk7A9hDOZaATAA/wygi3O+o2ClIwiCIAiCIAiCIBzkM53SzcsAfgTgNQAygGsBvMgYe4Ux9u5CFI4gCIIgCIIgCIJwkvcgjjHWwBj7OoABAE8DGAPwSQALAFwFIArgsUIWkiAIgiAIgiAIgjDIZ4uBjwO4HcDHAJwA8CiATs75AVe4jwB4jnMuF7isBEEQBEEQBEEQ856cN/sG8CyAXwP4UwA/5JynAsLtB7D1dAtGEARBEARBEARBeMnnTVwT57y/yOUhCIIgCIIgCIIgsnDK3ikJgiAIgiAIgiCI0pPPdEowxuoAfArASgAR12nOOf+TQhWMIAiCIAiCIAiC8JLzII4xthLAfwIQAZQDOA6gJvN9FIazE4IgCIIgCIIgCKKI5LPFQAeAVwAsgbG599UwthP4UwBxANcVvHQEQRAEQRAEQRCEg3ymU14CYAMA0yulwDlXAXyPMbYIwLcBfKjA5SMIgiAIgiAIgiBs5PMmrgLACOdchzF1cpHtXC+MQR5BEARBEARBEARRRPIZxB0CcFbm814A19vOXQtgrEBlIgiCIAiCIAiCIALIZxD3rwA+kvn8dwBuZYztZYz9BsAXAHyv0IUjCIIgCIIgCIIgnOSz2XcYQJhzfjLz/eMA/hiGc5PnAXRz2nSOIAiCIAiCIAiiqNBm3wRBEARBEARBELOIrN4pGWMv5ZEW55xfeZrlIQiCIAiCIAiCILIw3RYDAgD7q7qVMJybHAJwFMaececBeAeGsxOCIAiCIAiCIAiiiGQdxHHO15qfGWN/BOA7AN7HOX/FdvwPADyVOUcQBEEQBEEQBEEUkXwcm/wXgA7O+eM+5z4N4C7O+e8XtngEQRAEQRAEQRCEnXy2GFgB4FjAuSEA7zr94hAEQRAEQRAEQRDZyGcQdxDA7QHnboexTo4gCIIgCIIgCIIoItM5NrHzNQBbGWP/D8A2TDk2aQWwCkBb4YtHEARBEARBEARB2MlrnzjG2IdhDOaaAYQApAH8GsBfcs5fLEoJCYIgCIIgCIIgCItT2uybMSYAWATgOOdcL3ipCIIgCIIgCIIgCF9OaRBHEARBEARBEARBnBnycWxCEARBEARBEARBnGFoEEcQBEEQBEEQBDGLoEEcQRAEQRAEQRDELIIGcQRBEARBEARBELMIGsQRBEEQBEEQBEHMImgQRxAEQRAEQRAEMYugQRxBEARBEARBEMQsggZxBEEQBEEQBEEQswgaxBEEQRAEQRAEQcwiaBBHEARBEARBEAQxi6BBHEEQBEEQBEEQxCyCBnEEQRAEQRAEQRCzCBrEEQRBEARBEARBzCJoEEcQBEEQBEEQBDGLoEEcQRAEQRAEQRDELIIGcQRBEARBEARBELMIGsQRBEEQBEEQBEHMImgQRxAEQRAEQRAEMYugQRxBEARBEARBEMQsggZxBEEQBEEQBEEQswgaxLm46qqrOAD6m9l/hAvS7Yz/I3wg3c74P8IH0u2M/yN8IN3O+L+8oUGci+PHj5/pIhBE3pBuidkI6ZaYjZBuidkI6XbuMWsGcYyxesbYQ4yxXzLG4owxzhg7zydchDHWwRh7hzGWyIS/vPQlJgiCIAiCIAiCKDyzZhAH4F0AbgAwCuA/soT7JwDrAXwFwLUA3gHwAmPsoqKXkCAIgiAIgiAIoshIZ7oAebCTc74EABhjfwrgo+4AjLE1AG4C8BnO+aOZY78A8BsAXwfwidIVlyAIgiAIgiAIovDMmkEc51zPIdgnAKQBPGWLpzLGfgjgXsZYmHOeOtUyJJMqFKgYT+hQdQ5JYAhLAiYVDZLAEBIZAEDVOMAAiTEoOoemc8iSAAFAUtUhCQwVEQETSSOdkChAEhiSqoawaLwcTWfihQSGcEiAIABpjSOtGsdFgUEQAFEQsKg8DEEw8tZ1juFJBQwcKXUq/bqKMCRpNr14JQpBMqliOKEgJDKkNZt2GMA5EJIEcBi6MjVtho3KAlJpHToHBAboHNB0jrAkQOOAqulWWjoHGANCooC0qkPlHCJjiIQEJNP+9cU8p+kcEUmAygGAg3NA5xwCY2AMAAdkSQBjcKRVGRWh6xyTKR0scz1a5pwkCUgqGkTb9XAAYUlEbbls1RdiZmPaM0XVIPs8O1XVMTSRQlrTc7Jzus4xllCQUDRonCMSEi37aaalajoiIRGqpiOd0VNEEqDoOkRBgKLq0HUOSRQgMKAyzJBIY8reCgw1URmRSPbm1X5tUVmEqnOkVd26TgDW+VCmXUikjc9lsoDJlPOeTHevcr2nxOlx3r3PTRvm0DevKUFJCGJ6kkkVJxUViqZbfc6FUQGTitEnAGxtq8ggMgY5xBBPTbXFUVnAeFKzbN9JRUVaM86LAoMsCpk2miOt8az9Urd9qo6GMJpIQ9d1aNxIYzq7pao6jk2koGT6KOVhAWkNDvtaSttYTJs7awZxOfJ7AA5yzuOu478BIMOYkvmbU0k4mVQxqal4eyyFjT19GBxNoL46ii1tTej55WHsOjCMzrYmyBLDAz9/E5+7YgUmUio2b9tjhX3whjX4m5+9gWMTKXS2x/DQi2/i568Pob46io7WRvyo/wja3rcMcUVzxOtsj6GuMoShkwo2bu23jt/X0ojHdh3EnR9ZiZVLKgEAe4+O45/7B3DNmqXYZAvb1R7DqiWVNJCbRySTKvYNT+Knrw169GBq5wtXroDO4dDVlrYmPLf7CK69qB4VYQF/+7M3cMuly3HP9j1YXBHG3VetdOjTTOtPPnA+IiEBdzz5KgZHE7j9D8/DtRfVB9eXTB04Nq7g7qtW4tGXD1r5uNP+0jXvxsmk5kirsz2GBVEJ3/jp6554Ha2NuP/5vVhcKeOzV6xwXHv3umasXFJJHdcZjq5z7D06jvWP9/o+O1XV8cbRcWywaSKbndN1jkPDkzh6MunQb/e6ZrxrUTn2Dk1gQ0+fr8Y725qwsCyEQ8cnHMdf+MKlOB7XcCKedmissz2GFbXlgQM5+7X55de9rhlhScC6773i0fSxiZSjHnWva8aKxRXYd2wi8F7lek8Jgpg/JJMqBk8mcGw85bA/Xe0xREIC7n/+DU/b+si6GHQdDrvb2R7Djt8exVN9g/j+rZfgRCKNL/zwNYftWlwZBucct36/N9Be+9mnrvYYfvLaIC5fucRRjiC75dcubGlrwndf2mf1t0tpG4ttc+daj74Gxpo5NyO286fEcEJBUtGtTiQADI4msGlrP9Zffj4GRxPYuLUfoiCiJdaA4xOKVSnMsHc+vRsb1l5ghO3pQ0uswTq3edserL/8fIxMpj3xNvb0Ia1NdbTN4/ds34OWWAPWP96L4UkFw5MK1j/ei9bmZVaHwgy7oacPQxOn/BKSmIUMJxRs7Onz1YOpnaFxxaOrTVv70dq8DBt7+iBl9Gwazw1rL/Do00zrrmd2Y2QybZ0z0wisL5k6YKZpz8eddlqDJ62NPX1Iq9w33uZte7Bh7QVoiTV4rt2sL8TMxrRnQc9uaCJlNdTm+Wx2bnhSweHhuEe/6x/vdaTlp/GNW/vBwDzHxxI60ir3aGxjTx+GE8Eas1+bX37rH+/F4eG4r6bd9cgsf7Z7les9JQhi/jCcUDAwkvDYnw09fRgYSfi2rUdPeO3uxp4+fLKpHoOjCQyMJKwBnHl+87Y9GBhJYHA0mdVe+9mnDZk+jLscQXbLr13YtLXf0d8upW0sts0t2Zs4xtgiAO8DUAvgJ5zzEcZYBICS41TJnLKB/14LWYe7jLHbANwGAMuWLfMNY75WNh+EyeBoAmJmND04moDAgKpoKDCs/Zz52Z5OmSz6xtN0Hpje4GgCiqo50vELq2qFus3ETGA63aoZzQTpIZtOzTg655bGADg+u9MaHE2gTBat40H52uuLuz4EpS0w/3Ka9S3f6zPrC1F6crG3AKCoWtZnl9b0vOycomqB9lW12dcgPenca4NVnQdq02wzpru2oPzsdck8Zte0vR4F3Qu3zqe7p0QwueqWIGYS2XSr6jzQJpbJIsrgPRcUnnOe9bzbnpnH7fY6yD4F9SX87FaQLXT3t0tlG4ttc4v+Jo4ZdAAYBPBjAN8DcF7m9LMAvlTA7Ebg/7at2nbeA+f8Ec55M+e8efHixb4JSwKDJDDUV0cdx+uro9AyjXV9dRQ6B8YSacQVzTfsWCLt+WxPJyieGJD3WCKN+uooZEmELIlWOn5hJXGuvXid30ynW1OvQXrIplMzjsCYpTEAjs/utOqro4grU4YpKF97fRlLpK242dLWOXzPmfUtKF7QOVnyNihEacjF3gKw7Jkd+7MLiUJedk6WxEC92217kGYE5rXBksACtSllmSpjv7ag/Ox1yTxmbz/s9SjoXrh1Pt09JYLJVbcEMZPIpltJYIE2Ma5ovrYpKDxjLOv5uKL52jS7vQ6yT0F9CT+7FWQL3f3tUtnGYtvcUvTq/xzAZ2F4h/wDON+K/QTGNgCF4jcAljPGylzHVwNQAPzuVBOujcqIyAI622PWAzHn2nbvPGCtm9B0Ddv7BrCoQkZHa6Mj7IM3rEHXjv3WHOLtfQPWuY7WRnTvPICa8pAnXmd7DCER6Gxrchy/r6UR2/sG0L2uGbXlMmrLZXSva8a23rewxRW2qz2GuorwqV4+MQupjcrobI/56sHUTl2l7NHVlrYmbOt9C53tMagZPd/XYmiya8d+jz7NtB64fg1qykPWOTONwPqSqQNmmvZ83GmHRHjS6myPISQx33gdrY3o2rEf2/sGPNdu1hdiZmPas6BnV1cRRpdLE9nsXG25jHNryzz67V7X7EjLT+OdbU3g4J7jVVEBIYl5NNbZHkNtNFhj9mvzy697XTPOrS3z1bS7Hpnlz3avcr2nBEHMH2qjMhpqoh7709UeQ0NN1LdtXbLQa3c722N4tn8Q9dVRNNRE8Z0/vshjuxpqoqivjmS11372qSvTh3GXI8hu+bULW9qaHP3tUtrGYttcZr4CLRaMsQMAujnnf8sYE2F4j2zmnPczxq4C0MM5X5Rnmn8KoBvAcs75IdvxiwC8CuDTnPPHMsckAP8F4Hec849Pl3ZzczPv7e31PVdq75S6mYflndLwrpOPd0ot40VtjnmnpBX4LoJ0G+idUgC4HuCdUmJIqy7vlAKg6y7vlLoOkU15pxQYIGW8U2oZ75JZvVPKApJKgb1Tcg6J+XunBHAmvfGRZn3IZm+B3L1Tqpqek51zeqcEIiEhD++URt05k94pk2kNUmm9U5JufcimW/JOOSMg3frgp9vpvFMyTHmmztc7paZzCD7eKbP1SwvpnTKt6RBml3fKvDMrxZq4pQD+M+CcAqA814QYY62Zj7HM/6sZY8cAHOOc/4Jz/hpj7CkA32aMhQAcBLARwHIAbadUehuRiIQIJCyIOI/nNQK1sTA6fZh8EQSGxZX0xo0wiEQkLJ2mM1lqHPUl59ofTJX7vTsxZ5jOnkmSgHOqcjekgsBQUx721V2+adkpj0wfxq8s09nqbOfdus/V9lMbQRCESSQi+f7gNJ1Nq3bZnxqbTZ3uB6xs+NmnfO2VJAk4+xRsebFsYzFtbil6d0cAvAfAv/ucWwNjoJUrz7i+b8n8/wWAtZnPtwL4BoC/BlAFYDeAqzjn/XnkQxAEQRAEQRAEMSMpxSDuGQBfYYz1Y+qNHGeMXQjgLgCP5JoQ53zaV42c8wSA/5X5IwiCIAiCIAiCmFOUYhD3VQCXAtgJ4HDm2DMAGgDsAvDNEpShIKTTGpJaGpMpY72ZKDCEBGPdjigwx3qdqCxAVTkU29o2llk7xBjAYKwlSqo6IpIAncOavxsSGDRurA2SJQFc52ACg6rr0HXDLWxEEsABKKpurFHSubV+ozIq4GRian5zRBZRFXWusWCMQWSAIAhnao0QUQLMNXHWOrKIgPHk1NodSRSg6rq1nkxkU2vIorKhMXPdmmhqnTFwZPSaWbfGOax47nWiaY1jQdTQ+ETStn4tk2ZIZNB0WPVBEqbW0llr3EQBmm6sz7Nfy2jcyGdxuYwTKc0x51zXubHGKbN2z8y3riKMUMjwDOU3Vx1AUebFE7mj6xzHJ1NIpjWIzFg/ycCQUJzrJHJ5Rs51cBzlYRGabqyPiMgClDSHoumQRQFyiCGp6AhJxjrltKqDMZaxzYAIAYpurIkz68PCiIAx2zrpqqgAEd4pSu5yREKiYz3zdPdjOk2aa/rSmlGHysMiFkT81xCmNd2q/2kt+zoRgiDmLsmkirFU2tGnlSWGk8mpNlzVOFhmXbzADF8Mqs6R1ow4ssiMPkBmPX3a9AUhGmvudQ6EBIaKMMNYYqoPYK59D0sCFI2Dc8NHhCgAibTR74iGRCyqyG4nTfuYz7q5uULRB3Gc8wRjbC2AmwB8DIaHyGEAfwVgK+dcLXYZCkE6rWEincY7Yync9sTUTvAdrY1YVhvF0ZNpa9Pk+uooHr31EoxNKrjz6d3WsftaGvHYroO45dLleGzXQdx62XL8qP8Irmtaam22aKYZlUVs+fff4bNXrEBYMkQ4Fldx1zO7sbgijLuvWonN2/Y4PpvxO9tjeOjFN63d6R+8YQ3OWhjBREpz7BpvlufOj6ws2O7xxMwhmVSxb3jS2iTbTxv/cNPFSKZ13PXMlE63tDXhud1HsO6y8zAyqTriP3D9GkRCAu548lWPjj57xQo8t/sIrlmzFD2/PIxdB4attG75wHIMT6Tx0Itv4pZLl1sbd9rzW7tqCe5/fi8WV8r43JUXOvL1K2dnewz/PRbHM32DnvCPf+a9iCuatemnvZyfu/JCrKqrgCgK2Ht03FEnutc1IywJWPe9VxzHqH6UDl3nnufS0dqIMlnEV3/8OhZXyvj8lRc6nm3QM9J1jkPDkzh6Mumxl362s6O1Efc/vxfHJlJ48IY1qIxKGJ1M49GXD2LTh96FhKI5wr/whUtxYDjlqWPLa4FkcmptiLsc+WjL736446mqjjeOjjvuyZa2JlSXqVhaVWY5bnGHsV8v6Zwg5hfJpIr9I5O43dWnXVQZxvN73sFTfYNW+3zFu8/CP/3fA9j0oXdB07m1obcZ56wFYWgp4Nh4ymHjvn3jRXhk537c9dELMTSBwD7Ad1/aZ/VL3N+7b27GyrP8bZNpHx/8172edOeDTSuqu0LGWIgx9kkAyzjnT3DO2znnH+Wcf4pz/thsGcABxi7w8ZRuDeCAqZ3odZ1ZAzjz+OBIwhrAmcfu2b4HLbEG6//mbXuw/vLzLcHb0xydTKMl1oBNW/shCiJEQbQ6sBvWXmDFsX8242/s6XPsTn/n07uRUrln13izHIXcPZ6YOQwnFKtzCfhrY2QybenKPLZpaz9am5cZunbFv+uZ3RiZTPvqyIy3aWs/1l9+viMtTYOVt2lk3flt3rYHG9ZegJZYgydfv3Ju7OnD6nMW+oY/PBy3Oqvucm7s6cPQRArDk4qnTqx/vBeHh+OeY1Q/Soffc9m8bQ9GJtOWPtzPNugZDU8qODwc97WXfrbT1KBpNyVBxOZthm5GJ9Oe8GMJ3beOnUjoGE4ogeWYrtzT3Q93vKGJlOeebNraj5TKrXB+YezXSzoniPnFcEKxBnDAlE21vsnbAAAgAElEQVQYHEngk031jvb5rmd2W3bQHMDZ42icYWAk4bFxX3zqNbTEGiAKYtY+gL1f4v6+/olg22TaR79054NNK+qbOM55mjH2NICrkJ8DkxmHmtlY1W/ndZ3znHe1r4qGHP+DdqIvk0WUwUjD/BHBDGfGdX9252P/LjD/spvxC7V7PDFzUHWvLt3aCNKpmJnSG6RNvzTtehYzonWnFaRXM55Ztlzrk6pz3zSnq3+qzgFVy/n6qH6UDiXLcymDaH13n/d7RoqqObSQj+20280gXQbVMbO9CCrHdOV2x50uXlrTfcMIDFa4oDD2ayOdE8T8Ich+lckizO3HcmmfTVuTrd2129Jc+qzu70G2ybSPQenOdZtWio3DDgCoK0E+RUUSGCSB+e68LjDv8aBd68cSacf/oJ3o44pmhdG5sZbODGced39252P/bo/vV55C7R5PzByC9GrXRpBONd1YHxekTb807XrWMp1Yd1pBejXjjSXSvmGCyikJLK/wZlhJYJAlMefro/pROrI9lyB9BD0jWRIdWsjHdtrt5lgi7aupoDpmthdB5Ziu3LncD3u8kCj4htE5rHBBYezXSzoniPlDkP2KK4bfBPO7vX0OsmM6z97u2m1pLn1W9/cg22Tax3zahblEKQZx9wP4EmNscQnyKhp1FWGUhQU8crNzJ/iO1kYIAkdnW5PjeH2NsRbNfuy+lkZs7xuw/ne0NqJ75wF0tDZ60qwuD2F73wC2tDVB0zVouoYHrjfS69qx34pj/2zG72yPOXanf/CGNQhLzLNrvFmOQu4eT8wcaqMyOttjWbVRUx6ydGUe29LWhG29bxm6dsV/4Po1qCkP+erIjLelrQndOw840hJFWHnf19Lom19HayO6duzH9r4BT75+5exsj+H1t0/4hj+3tgxdrmNmOTvbY6irCKO2XPbUie51zTi3tsxzjOpH6fB7Lh2tjagpD1n6cD/boGdUWy7j3NoyX3vpZztNDZp2U9U1dLQauqkuD3nCV0UF3zq2MCqgNioHlmO6ck93P9zx6irCnnuypa0JYYlZ4fzC2K+XdE4Q84vaqIyHffq09TVRPNs/6GifH7h+jWUHv/PHF3niiIyjoSbqsXHfvvEibO8bgKZrWfsA9n6J+3v3zcG2ybSPfunOB5vGzFemRcuAsScAfAjGnm3/CeAdAPZMOef8lqIWIg/8drQ3sXunVNQpT5JMMDz25eqd0viB1sc7pW54+5vOO6WmGx4Ap/NOaeY7B71TzqrCloIg3RbMO2XmXEG9U9ryC/JOqWfCSKIAjU/pX5x93ilJsz4E6XbKO6UOkaGA3imB8rDg8U6Z1nSEpvVOySCCFcg7JRAJCbPBOyXp1ods/YTz7n1u2viHvnlNoYtEOCHd+uCnW9M7pb1PWwjvlLpueJoEOLQA75Scc7As3il1PePFd/54p8y7sKXYYuADANIAjgG4IPNnp7ijyAISCokIhURUTrOT/UxgQUAZi7VrPDEziUQkLHV1JhdEAwIHUZ5/vosCji/MN+9psF/LYtl5nYLAcE5V9gwFgfnWCaonZxZBYKjzM7Q2Leb6jASBoaY8nLuOT0Hv5Tm0CXmXwxV3uuuVJGFavecShiCI+UMkIuGsiHcoUH0KdioXcrGV+ZKLfZyrlGKLgeXFzoMgCIIgCIIgCGK+UIo1cQRBEARBEARBEESBKMV0SjDGygB8BsAHAdTA2Ox7B4Dvc87jpSgDQRAEQRAEQRDEXKDogzjG2FkwBmwXAjgM4L8BnA+gFcDnGGNrOedHi12OQpBKqVC5imQaSKm65ahEzSzglASGRFqDmFkcqnMO3bbY0+70RNcNJw5a5lhIYBBFhkRah5xxBmGGN508aDqHJAoIS8xa9CkKhoMUnQOCAHDdmB+cUo2F62CGcxRV49AzzlI0nUMQGGRRQE2ZsfCzCI4ciBmA27FJRBaQVHSUhQVwDiQU3dKX5XAkoyu7gxK7Q5QKWURSnXLkUBYWEE/pEJjhxtysG3bnJdZiaE2H5NL3lPMTETrnDucnOucQBUPzRhkNhzyiyKCkdcuZj3kNZbKAdGaHANNhg90JRlQWjQXZalZnDnmRi9MJIj/sC9VVnTs0KQjMeoamA5KEoiEkCuCcI6nqqIyISChTGrU7s/FDVXWMxBUo2pRDKGTsqpyxzcm07nCaM5lWMJHUURV1OjapjAqQfRyb+F1fITRjT8usHyFJcNwjUQAmUxpExjxOYkivBDF/SSZVTKRVpGxteiQkQFF1cBiOzFI2B3oaN/qhoYwdVm3O+6RMX9e01yHRSCckMsuOS6IAxjh0HYhmnO5lsz8OZ0yigLqKMCTJO4kwF5taira61P2BUryJux9ANYA/5Jy/bB5kjF0KYDuA+wB8ugTlOC1SKRUprmIsruFEPI2HXtqHWy5dbu0Qb7pZvf/5vTg2kUJHayOisoh/2fM2rlmzFJu29lvhtrQ1YWFUwjee+y1+/voQ6qujeLi9CToHfrr7iCd8Z3sMP31tEA//xyF8dHUdPnvFCsf5+1oa8diug7jl0uV4bNdB3HrZcqsc377xIoREhn/499/5ljdZHcVkUsP6J3qt493rmrFySSU17LOcZFLFvuFJbOzpc2ip7+BxfODCOiTTOnoPHkds+SI89OKbHn1saWtCzy8PY9eBYUuDo3EV7e8/16PPQ8dO4uyqciyISvjWC3s9aT1w/RpEQgJ+FlAfntt9BB+/qB6RkID7n38ja9367k0XI63quPPp3Z7ruuzCOsRTGu54cir9rvYY/v7FN3FsXMHdV63E5m17CqZ1XefYe3Qc6x+n+lMozHv64L96ddTZ1gSBAbf39Afa3YHhSbx7aZVH96vqKnwHcqqq49DIJI6NpxzaMO3qrZctR5ks4qs/fh3HJlLovrkZdQtlvD2Wwvm1YRwYTnnyqq8GkITvQK6QmvFLy69+uO+R43pIrwQxL0kmVRyNp3AynsZGV5tcJou4//k3sP4Pz8czvYO4rmmpwz5+79PNOBFPO9vhtiY89NI+q1/b0dqIH/Uf8cS129YlCyI4r7bc1/6oqo43jo5jg82+drXHsGpJpWMgl4tNLUVbfSb6A6VYE3c1gD+3D+AAgHO+C8BfAJgVvnaPxxWMJ3QoKsfGrf1oiTVYnQvA2Bl+87Y92LD2Auvz6GQarc3LrA6rGW7T1n6kNaAl1mAdGxpXsHFrv2/4jT19aG1eBsCI4z5/z/Y9VnlaYg2OcnzxqdcwMpkOLK+icmsAZx5f/3gvhieV0t1coigMJxSrcwlMaemK1WdjYCSBDZnPG3v6fPWxaWs/1l9+vkOD6y8/31efF59biy8+9RqOjCZ907rrmd0YyVIfWpuXYUNPHwZGEtPWrdHJqYbDfV2DIwlrAGee25C5vg1rL7AaEvPc6Wp9eFKxDHah0pzvmPfUTwcbt/ZjaFzJancvXbHYV/dDEynf/IYmUhgYSXi0YbenI5NpK4/1T/QiqejY2NOHsYTum1c8pWM44a+BQmrGLy2/+uG+R47rIb0SxLxkOKEgnenTuttkgKEl1oA7n96N9Zef77GPR0aT3nY40zc2v2/etsc3rt22Hh6OB9qfoYmUNYAz427wseW52NRStNVnoj9QijdxFQDeDjg3mDk/41F1YycEgRkPpioash6UiXnc/FwmixAF5htOYLDCAkCZLGJwNBEYXsyM4rPl6/5vL0cZxMBy+B1XVC2v+0PMPFSd+z5bzrmlN53zrHo2def+7A6nZfLKprVs9cE8ni2+qWmz7NmuKyhuobWuqBrVnwJj3tMgTZbJoueY/flqAbo3bbibtKZn1Y1dl/a07P9zzct+fe44p6IZv7RyqQPu6yG9EsT8Q9V5YB/Q7KMG9UunszPm96A2325bg+xPWtP97aumO47lYlNL0Vafif5AKd7E7QVwc8C5dgBvlKAMp40kMEgCg86NneDHEmlrZ3gT87j5Oa5o0HTuG07nsMICQFzRUF8dDQyvZToF2fJ1/7eXIyieeT3u47IUvH6EmB1IAvN9towxS28CY1n1bOrO/BykTzGTVzatZasP5vFs8U1Nm2XPdl1+cYPSPR2ty5JI9afAmPc0m47cx+z2TgzQvRQwnSUkCll1Y9elPS37f7+8gvIrpGb80sp2LeZn9/WQXgli/mHv09qx91GD+qXT2Rnze1Cbb7etQfYnJAr+9lV0Dl1ysamlaKvPRH+gFIO4bwH4FGPs3xhjn2GMXc0Yu5Ux9gKAmwB0lKAMp82iMtlYsC4xdLY1YXvfAO5rabQemDn/t2vHfutzdXkI23rfwpa2Jke4LW1NCInA9r4B61hdpYzOtibf8J3tMWzrfQuAEcd9/r6WRqs82/sGHOX49o0XoaY8FFheWWLovrnZcbx7XTNqy+XS3VyiKNRG5cz6HKeWXnr9HTTUGHPLX3r9HXS2x3z1saWtCd07Dzg02L3zgK8+Xz08jG/feBGWVkd803rg+jWoyVIftvW+ha72GBpqotPWreryEB68YY3vddXXRPEPNznT78pcX9eO/ehobSyo1mvLZXSvo/pTSMx76qeDzrYm1FXKWe3urn3HfHVfV+G/GWxdRRgNNVGPNuz2tKY8ZOXRfXMzIrKAzvYYqqKCb15lYQG1UX8NFFIzfmn51Q/3PXJcD+mVIOYltVEZoUyf1t0mAxzb+wbw4A1r0L3zgMc+Lq2OeNvhTN/Y/N7R2ugb125bz60tC7Q/dRVhdLnsa5ePLc/FppairT4T/QHGefC0j4JlwthtAL4OoM52+CiAr3DOu4tegDxobm7mvb29vueCvFOaHnckgSGZ1iAEeKc0PfbYvVOaHiYL7Z1SUXVIAgMzPZVpHDxTnjngnXJGF+5MEKTbU/JOmfEombd3SgEICU7vlDrnEBiDIAAim9feKUmzPgTpdiZ4p2QM0OaUd0odIkO+3ilJtz5k6yecd+9z08Y/9M1Z4QpgNkO69cFPt0X3TqnpVtgp75SG3crHO6XRd5jz3inzLkxJ9onjnD/CGPtHACth7BM3AmAv51zPHnNmEQ5LCENCeeRMl6TwLK70/5WamN1EIhKWujuT5QGffVhk+7wgGhgM1WX5liyYhVnyOVWKqW9BYFR/Ckze99RPx9No244kCahbkLthFwSGhVLE0mq+bUIhNZNrWrXu+5HH/SEIYm4SiWT/welMI0kCzqmavlOQix0sRVtd6v5AyZ5cZsD221LlRxAEQRAEQRAEMRcp+po4xtiDjLEnAs49wRj7VrHLQBAEQRAEQRAEMVcoxZu4TwD4asC5FzLn/qwE5ThtVFVHSlUwnuJQMnOEzTnBsiQ41rK517WFJQFpTTfWrtnW0pnr5ySBWeuMQiKDJAhIpDWIAkNEEqBzQNGMdXEhgaE8s0ZJ0zkqwiKS6an5zBVhARMp3bHmSOew8tA5R0QSIYlT6yKqoyGcTKWRUDRonCMSErGoPDzT18YR0+BeE2euP4uEBKRUDimzvsyuHzNMmWz8xhNXnGvUQqIABmNNpyQwa80nYwC4sR6Hc460xqfW2dnWNJnr7Mz1edZaooiAhGLUGUEAdB2ZuiRAYMZaPbMMYUmExrm19jMqC0imjXWf9nVS5hohQRAKOv+9FHPrCf/1XoIgONY5msfDIQFxRTfWQooCQhLDZErLuLDWLS3JIYak4lwTac8nKovgnCORWYcclqb0HpEEMMYy9cCoN+41cZGQgHLJf4pSLrox14CkNR3RkAhN51A03WpHONi0egvKR9c5xhIK2XmCIAAYfYRJVXX0AYy15YYPBZ0DqqZDzKwTZjD+p1TDzoZDzFp/HBIFSCKDlunr2vsN4ZCIqrCE4/E0AG6tv4+ERNREZd916/a2PMhe59P+zsV2uxSDuKUABgLODWbOz3hUVUdSVXB4JIXbn+jD4oow7r5qpWMX+ofbY1B1HT/b8zauWbPU2tTY9NoDAA+9tA+3XLrc2sS2vjqKB29Yg6qyEG79/tQu7x2tjbj/+b1YXCnjzz62EsMTiiOvrvYYfvLaIEbjKtrff66V10dX1+FzV15obUBrehrqPzSM2Hm11qaO9jyOTaTw/VsvweikYm3eaHrVKeZO80RxSSZV7Bue9Gjhud1HcO1F9eg7eBwfuHAxEmnuG6bt/efhRELFQy++6dGsXTtm+MtXLsFjuw7i9g9egMqIhG+9sNcTb0tbE3p+eRjVZRKuvajekW9newzRkID7n3/DE+8fbroYybSOu57Z7Vv3OttjeOjFN/Hz14c85buvpRGP7TqIOz+ysiB61nWOvUfHrU09qa4UB7/7fF9LI3buPYqPX1RvbQJrt69u+9Z7cAQfXFXnsMV2bXSva8aKxRXYd2wC6x/vxeKKML76idWIK5pDXw/esAbP9A7iuqalePTlg7j7qlWIKxouWBTBgeGUp/5UlYVQBzgGcrnoRlV1vHF0HBt6/NsYs679UVNDoN6C8lmxuAJvjcZx9GTSkSZplyDmJ8mkiqF4CmPxtLO/2h7DkgUy3hlL4nM/eNVhfx/bdRC3XrbcsqFd7TH8va3tNdvqf/q/BzzteGd7DD99bRCXr1ziOG6mcWxc8di8bPY6Hxs2V9vtUmwxMArgXQHn3gVgvARlOG2GJlI4kdBx+xNGY71h7QWeXehv7+nDyGQarc3LrAphntu4tR/HJxS0xBos8ZrnjIFT0nFs87Y92LD2ArTEGnBkNOnJa0NPH1qbl2H95ec78mqJNVgdCjPspq39uGL12VYHx53H4GgCAyMJawBnni/2TvNEcRlOKL5aaG1eho09fbhi9dkAhMAwimoM7vw0a9eOGf6e7XvQEmvAF374Go6MJn3jbdraj/WXn2+VwVFHevoAMN94I5Np3PXM7sC6Z5bTr3xmuQql5+FJxWoIzPyorhQev/t8z/Y9aG1eZg3gzOPHJxRf+/bJpnqPLbZrY/3jvRiaSFn5bFh7AUYm0x593fn0bqy//Hxs3rYnozOGTVv7MZbQfeuPonIMJ5Rpr8etm6GJlHVtfjo361o2vQXlMzSRwuHhuCdN0i5BzE+GEwoUlXv7qz19UFRuDeDM42ZbarehG1xtr9lW+7XjGzP9VvdxMw0/m5fNXpthcrFhc7XdLsWbuH8D8CXG2E8450fNg4yxJQD+N4B/LUEZTpt0Zod4UwDmbvN2zN3ng3aoL5NFlMF/l/syWfQcq4o6d753nxczvx7YzwWVi3Pue9zMo0z2L1cxd5oniouq+z9zU5+ccwhsem0FacrUjj1NM2w2rfvp1vwuMH8N2/U5XXn8ymfGKYSeFVWjulICgu6zn30Nsl/T2b3B0QRUTXdoyzwelG9VNGTVm6A6JjCj/uVyPXbdpF1lyVaOIL0F5aNqOtl5giAsVD24D6AF2DbTLrnbVxPTxkxnv/zSNT9nO2e31/Yw09mwudpul+JN3JcBVADYxxh7kjF2P2NsK4A3M8f/ogRlOG1CmX3gzE38zN3m7Zi7zwftUB9XtKzx3MfGEmmMJdKIK5pvHC2z74b9XFD6jDHf42OJNAAE5lHMneaJ4mLXq4mpG1MTOkdgGPNckKZM7djTNMNm07qfbs1zOvfXsF2f05XHr3xmnELoWZZEqislIOg++2knyH5NZ/fqq6OQRMGhrWz21oxr1o2gOmauQc7leuy6CbnKkq0cQXoLykcSBbLzBEFYSEJwH0AMsG2mXXK3ryamjZnOfvmlm0tfw26v7WGms2Fztd0u+iCOc34IwCUA/hnAhwB8MfP/RwAu4ZwfLHYZCkFdRRgLowIevtnYPb5rx37PLvQPt8dQUx7Ctt63sKWtyXGus60JiypkbO8bwH0tznjGrvcRx7GO1kZ07diP7X0DWFod8eTV1R7Dtt630L3zgCOv7X0D6HTtcL+lrQkvvf4OOl1lMvOor46iocYsR+l2mieKS21U9tXCtt630Nkew0uvvwNADwwjSwyd7TFfzdq1Y4a/r6UR2/sG8J0/vghLqyO+8ba0NaF75wGrDI460h4DwH3j1ZSH8MD1awLrnllOv/KZ5SqUnmvLZXSva6a6UmT87vN9LY3Y1vsWulzaWVQh+9q3Z/sHPbbYro3udc2oqwhb+XTt2I+a8pBHXw/esAbdOw+go7UxozNurH2LCr71R5YYaqPytNfj1k1dRdi6Nj+dm3Utm96C8qmrCOPc2jJPmqRdgpif1EZlyBLz9lfbY5Alhoc+dbHH/m7vG3DY0C5X22u21X7teGem3+o+bqbhZ/Oy2WszTC42bK6224xzPn2oeYTfjvYmDu+Umo6wKFi70Idy8U6p69B1WJ73/LxTGjvaG94pk2kNwjTeKXWdo7zg3imBSEiYyV7LZmShziRBui2ld0ohE8bXOyXnEJlRD+JKxutqDt4ptYzHqyDvlGnV8Jo1C7xTkmZ9yGZvgVPzTqlqOkIZ75TxlOHt7FS9U+oZz8NnwjulqumIZLxTpjPe4QrrnTInO0+69SGbbs+797lp4x/65jWFLhLhhHTrg59uc/JOqesQGYNhJgzvlIpq2Nls3ilNT+w6N+yz2zul6bWSvFNa5F2YM7JNO2NsNYB3A/gl5/ztM1GGU0GSBEhSBOWRM10Sg4VZNrFfWJZjIuVTH2uksOM7MfuJRCQs9elM5kN1MTXhSntBFk2fbtqFRBAYFleGi5cBASD7ffY7XuW2exUBCbu0cbrPM9c2IZd8JEnAOVWnVxGC8hEEhppysvNnGhroETOFSERC5HSHAnnYk3Nk/7yy2sUC2eu52G6XYrPv7zLGumzf/yeAPQCeAfA6Y+ySYpeBIAiCIAiCIAhirlAKxyZXA9hl+/41AD8BsAbAKwD+sgRlIAiCIAiCIAiCmBOUYhB3FoBDAMAYqwfwewD+lnP+XwD+HobTE4IgCIIgCIIgCCIHSrEmLoGplQkfBHASgLmycgJAZQnKUBB0nSOhpJBIAynV5ggi49hB07m1+NxyUpJZ6J6yO46QBGg6n3KKIgookxlOJDQsjIqYTE05kjCdnxgL6oFk2ulkQsgsNuUcYJlFpKZjCi1zPhoSoGkcasbZhJhxpqKDQ9MBWRQgivAs9idmP27HJmWy4fzB7qDE7jBEYAyCAHDOIDEADA4HPSGJQdPgcDRid1Bid8JjrxumUwmBTTlEMcui6RzhjMMIRdPBAEvjdRXG/PWhiZTDSZAgMGvxcz7OS9yOMmSRQdEMhyim9gEELn6eBQuj5wS6znF8IgWAW/qLSILhzEab0lJK1RGWDAdTpsZEBiQyGnM7ArE/v4gsQIDhuEfVOcpl0WHXI5KAlKZbdti0mwlVd4RdVC7iREJH2mb/K0NTjk0KoRl3uZW04VwrJAqoqwhDkoJ/jzXvZSJtOBSSRQE1ZUYZSMsEMb9JJlXENdVyTmL1WVXD9kVlAWmVI6VOOVoSBWe/wN7HlTN9ANMZCmMAA8PiCsOB0lhCQVo17KXpqC8ii1gQDmEkoSCZ1iAyhmjmmOnwJJ+2eD6106UYxPUDuIMx9haAOwD8K+dcz5xbDuCdEpThtDEHcMfjGk7G09iY2eHedPv83Zf24eevD3m+f3R1HT53xQor/EdX1+HPPrYSwxOKtTO96WJV1zWcTIp46MU3cculy61d7euro3j01kuQVDQ89NI+z7n7Whrx2K6DuOXS5di59yiuWbMUm2zl62xrgigy3PZ4n3Wso7URUVnEln//HW69bDnKZBFf/fHrODaRQve6ZqxcUjlnRT9fSCZV7BuexMaeqefe2R7DT18bxMP/cQj11VH8w00XI5nWcdczuz16+sKVK6BzeLT+3O4j+PDqsxCSBLx2eASx5YuwsacPiyvCuPuqlQ5du+tGR2sj7n9+L45NpKyyvHJozBPPLMOXrl2NkwnVeQ1tTRAYcHtPvyf8nR9ZGahdXefYe3Qc6x/vteriZ69Y4agr3euaEZYErPveK45jK5cYvzXZ49vPUV0pHLrOsfe/x/HPrw5YtsxPWx2tjfhR/xFc17TUc9zSWFsTfrr7CP6oqQErFldg37EJrH+8F4srwvi7GxsxkdKxsacPl55fi/b3n+uxm4BT/x2tjeg9OIIPrqrDpq39eOELl+LgcAobepy2dXFlGPWIQpbF09aMXbd+96GrPYZVSyp9B3LmvVz/RK+jfCcrwxAY89U5aZkg5gfJpIqheApj8bTH9kVlEfc//wY+d+WFqAgL6Pz3A7iuaSl2vOHTx8yhLf/8lReitiKEoZMpxBXNEebh9ia8jaTDjj54wxpUl8v49KO/zqstznZuLtq2Ukyn/BKA9wHYDWAlgL+ynfsjGOviZjzDkwrGEjrSKrcadcDY8X3T1n60xBp8v7fEGhzhW2INODKatARsxtnQ04faigg29vShJdZgDdLM84MjCWzMpOs+d8/2Pdbx1uZlVuUyz2/c2o+jJ1KOY5u37cHoZBotsQZs3rYHI5NpbFh7AQZHE1j/eC+GJ5US3VmiWAwnFGvwA2S00NOH1uZl1veRybQ1gDOPmXoaGld8td7avAx3Pr0bo5NpXLH6bCuPDWsv8OjaXTc2b9tj6cwsi188swxplXuvYWs/hsYV3/DZtDs8qViGHTDqoruurH+8F4eH455jw5OKJz7VleIwPKlg/RO9Dlvmp5HN2/Zg/eXn+x63NJbR6/rHezE0kbKe34a1F0ASREtb6y8/39duHp9QPGl/sqneCjuW0K2Ohz3MwEgCw4nCaMaeht992NDTh6GJVNZ76Ve+IJ0TBDE/GE4oUFTua/sAZvRfe/ogCaJla337mDm05Rt6+qBqwMhk2hNmaFzx2NE7n96NgZFE3m3xfGuni/4mjnP+a8bYMgCrAOzjnJ+0nX4EwL5il6EQKKpm7F/FYInDZHA0gapoyPd7VTTkCG8e90tD1bkV132+TBYDz9mPiwLzPV8mi77HyiA6PpvnFFXL+d4QMxNTT3ZMjZiYunKHyaZTU2NlsmAFFL4AACAASURBVAidT+WRTZtBaYsCyxovqL756dlMJ0i7iqp56mKuaZtp+oWnulJYzOdkt2VBzyrI3rk1NjiaQFrTHelpNu3mYze5LV5QHSvL7HUEl+bM8/loxq7boPugarpfVI/mg67rVMpFEMTsJlufVmBT9kbn3LKRQbZyurbcTMevzxHUDznVtng+tdOleBMHzvkk57zPNYAD5/w5zvmb5nfGmMgY0xhjTaUoVz7IkghJMDbNNnd8N6mvjmIskfb9PpZIO8KPJdKIK5pvGpLArLju82Ycv3P245rOfc/HFc33mBnP/GyekyVvI0/MLkw92TE1YhKkxWw6NTUWVzQIbCqPbNr0+26mlS1eUH3z07OZTpB2ZUn01MVc05Yl0RPffo4oHOZ9ttuyoGcVZO/cGquvjiIkCo70RJt287GbzBYvqI7FFQ2SwAqiGXsaQfdBEv2b8qD844oWqHOCIOYH2fq0Op+yNwJjlo0MspXTteVmOn79iqC+xqm0xfOtnS7JIC5PZuSk1dpyGVVRASGJobOtyRKJue5ne9+A7/ftfQOO8Nv7BrC0OoKO1kZHGl3tMQxPJNHZHsP2vgHc1+I8X19jzFP2O3dfS6N1fFvvW9jiKl9nWxOWLAw7jnW0NqK6PITtfQPoaG1ETXkIXTv2W/OHTQcPxOylNiqjsz3m1EJ7DNt637K+15SH8MD1a3z1VFcp+2p9W+9bmfnqIbz0+jtWHl079nt07a4bHa2Nls7MsvjFM8sQkpj3GtqaUFcp+4bPpt3achnd65odddFdV7rXNePc2jLPsdpy2ROf6kpxqC2X0X1zs8OW+Wmko7UR3TsP+B63NJbRa/e6ZtRVhK3n17VjP1Rds7TVvfOAr91cVCF70n62f9AKWxUV0OXSZ0drIxpqoqiNFkYz9jT87kNXe8xyABR0L/3KF6RzgiDmB7VRGbLEfG0fwI3+a3sMqq5Ztta3j5lDW97VHoMkAjXlIU+YukrZY0cfvGENGmqiebfF862dZpzz6UOVCMaYCCANoJlz3n8mytDc3Mx7e3t9z7m9U1oeywK8U+qZ75bnHs7/P3t3Hx9Hdd+L/3NmZle7kvwgy5ITLBsbAiZurg2WSAjcEgoN4VW4SakNSWNhQn5x/FBIQqlD2l5uaUnbOA6XklDb4JsQwE4TYpeSQkpIAoQUkoJkgksNxgWMLQdsWZZtPax2d2bO74/ZM5rdnVnto7Qrfd6vl1/WzrO033PmnN053wNDpGentG0JI5Wd8lTMwvTM7JSpjIHlyk5pWk62yxrPTlnVFzcRguJ2orJTJi07q2yEgrJTSok6fdJnp2TM+giK22rOTjli2qivguyUScuGUUB2ypGkBa3w7JSMWx+52gkLvvJ4Wc5x4GtXluU4UxTj1odf3AZlp0yYNuyg7JQ63KzTQdkpLdt226cIyE5pq/3TslPa0AWmanbKgi9yPLJTjishxCUAnvZZdVJKObOUY2uaQEMkgoZIKUcJNrM+/f9x1zBB56WKiUQMzI2kF/OmSrzPRR4z32uZ21SeQqFpAi3T/L+18AraJt/9qTSaJtA6vfwVrd/7V47ykOueUI6YKeUYuf6WjGWiqS0SMRCBMW7tv1kNwXVO67TseqqYe/FUuk9Puk6cxxcAvOh5bU7UhRAREREREZXLZO7EvSql/PVEXwQRERFRueXz6CYfyySavCZzJ66s1DO2lu2kcralBCTccRDR1DPBUGPZUsuN1BiKzLFCasxPNKxhJGnDcAYiIZka26HGViRT4y50TaSNdTM0AV04x/Y9f2r8naE7j9iatnPNdZ7ng/2eG7ZtiaODcSQtG6E8xltQdVNj4jQNsO3RuNQ0QIOTmUrXnHFv7rpUDBm6BtO23XjSxWjc1oc1nBqx3OfhR5LO2LaQpqWNl1Pj7AxNQAj1uLdMG2fX2liHUMjJHKVi0rZtWBKQUiJkaDA0J6uVGi8HiLzGAtXoc/FTnm1LHBuKI550MqCGdAHTcurHulRdalo2NE/dl/SMyUxYNsKGM+ZNjcFsjOgYHLEQTaX/T6bqTuHZRpWTxogzNlnVg/VhDUNxC9OjGk7FnDo5pAnomkB9nXCXGZpANKyhXjcwkLSyYi+ZtNLGd3pjv9i/U6ExznJRG8o1to4ol5ERE/3x5Gh9KABNCIQNgYQpkfAsB+C0VQFASHdcXMTQIKGmLBAQkLCkk/1SE8I35wLrofKYzJ24HUKI2QBOAPgJgK9IKQ8WcyDblth3ZAB3/XQfbrz0LNzz1H78f//zDHeSZJWF7/GXD+PiRXPcybhVJrBHdh/G1cvm+s5if9NlZ2N6RMfASBKnYiZuftg55uWLW3HjpWe5kyqqc9zz1H48ufeoe+yvP7EPvYPxwPPf/5kOnBox8cXv/8Zdtm1VB85qacT+3sG0We0f/OwHMZyw3EkXVeazc+ZMY0euBo2MmNjfN4Rv/fx1XH/hwrS4uPOapYiENPx4z29x5dK5aXGmYnP1756BkKHhxu+95Bu3z7x6BP/525O46bKz3UmTVVapoP1uufxsJEzpTiKuMlud09oIXdfccpZ5vd5YV8f6wmVnB8amKrPe+N62qgOL5kzjjaLKZb53ly9uxU2XnoV1O3ajpbEOX75iUVpdqurF3oFE1jpv3GzpbMerh09gXnND4DYbly/B/ndPon3h7LSY3rxyGSKGwPFhpC3fufYCHOhLpC3b0tmO5kYbT/7nO7j9sdfc2DuzuR77eoeytj2ntbGojlwxMc5yQUTKyIiJN44PYc1D3Vn36i/+/tm4+2evu+1NtXzNR87EzPoQhuNWYJ2str3hooXuXG+3/2gvegfjge1P1kPFqaqWuZTSAnADgLdKOMxJAHcC+ByASwHcAeD3AfxKCNFazAHVDPDL2+dh/Y7dWN4+z+3AAc5Egut37MaKjvluw1Mt37BzjzvTvd8s9uu2d0MTGnRNdztwANxzZZ5jefu8tGOvveTMnOfv6R9xO3Bq2eoHu3B0MJ41q/3bfcNuB04tW7u9G0cH48X82WiC9cWcxuXy9nlZcXHLD1/G8aEkVnTMz4ozFZs3P/wy+oeSgXH7iWVt7s/ebXLtp2u624FT69alYsxbzvzKkYp1daxcsamOlRn3fUOJCv21qVwy37vl7fPcmFl7yZlZdamqF/3WeeNm3fZuXHhWS85tbt21B5cufm9WTK/fsRsNdaGs5aaNrGXrtnfDsoBLF7/XXbb6wS70DiV8ty22fi0mxlkuiEjpiyXcDhyQfq9e81B3WntTLf/i938DKUXOOlltu2HnHhwfSuL4UNKtY4Pan6yHijMunTghxFwhxP8VQnQJId4UQnwgtfxLQogPebeVUj4gpewv9lxSypeklH8mpfxXKeUvpJT/AOAKAHPgJDvxu77Pp66tq7e3N2t9wrTSZp0PmpU+10z2Y81ir4n0WeaDzjEzGvJ9HXSe+rDuexzTsgvalqrPWHFr2jJnvNaH9TFjU32K5rdOSpnz2H77Zca5WmfaMquc+e2feQ1BsamOlXmMhGn5bk/jJ9/6VvHGQ67YyCdurFSZyLWNLf23MX32DTqeLZ1pK8baXy0vRjExznJRvLHilqga5YrboDrJ29b1W+69j+eqd1VboD6sp9Wxfu1P1kPFqXgnTgjxOwD+E8B1AH4L4HQAata90wF8sdLXkJpz7nUA5wesv09K2SGl7Ghpaclar2aAV7POB81Kn2sm+7Fmsbcl0rYJOseJWNL3ddB5hhOW73EMXStoW6o+Y8WtoYmc8TqcsMaMzeGEFbhOCJHz2H77Zca5WmdoIquc+e2feQ1BsamOlXmMsFH8+CMqj3zrW8UbD7liI5+40VNlItc2mvDfxvDZN+h4mvCOAQ3eXy0vRjExznJRvLHilqga5YrboDrJ29b1W+69j+eqd1VbYDhhpdWxfu1P1kPFGY/W+Z0AXgWwEMAfIX0yu+cBXDAO14DUeYv6yFPNAL+r+xA2r1yGXd2HcOc1S9NmhN+8chl2dh3ExuXpM9FvWrHEnenebxb7LZ3tsKUNy7Zw17Wjx1TnyjzHru5Dacfe+swbOc/f1hTB3Z86N2v2+tbGuqxZ7U9vrsfWzva0ZVs7290Jl6m2NEfD2NLZjl3dh7Li4s5rlmJWQwg7uw5mxZmKzbuuXYqmhlBg3D66u8f92btNrv0s28KWjPNtScWYt5z5lSMV6+pYuWJTHSsz7tVk3lS9Mt+7Xd2H3JjZ+swbWXWpqhf91nnjZktnO57f35tzm43Ll+Cpve9kxfTmlcswFE9mLTc0ZC3b0tkOXQee2vuOu2zbqg60NIR9ty22fi0mxlkupp4FX3l8zH80NTVHw7j3unbfe/W917WntTfV8rs/dS6EkDnrZLXtphVLMKshhFkNIbeODWp/sh4qjvA+8lGREwgxCOCPpZT/KoTQASQBdEgpdwshLgbwhJSyotNbCyE6APwHgK9KKf8q17Z+M9oD2dkpZSr7nmnLtAxpftkp46btZlRL2nZalr9o2NlP92SntFLHdLNTpjID+mWnjJu2//l9slNKKcfMDqSyU5qWDaN6s1Ny5GuGoLjNyk6ZiqWc2Sk1QNrp2SlVlklNAHZAdkpbShgVzE4ZS327pwlA1l52Ssasj1z1rZOd0oYm4GanNG2JcB7ZKZOWjVAB2SnjqZ9VvaiyU6p6cApnp2Tc+giKW6A2s0pOwmkIGLc+/OI2MzulJgDhk53Sue/mzk5p2RKC2SlLUfAfYDyyU+YaUDUbQCzH+oIJIXbASYyyG05myvMA/DmAwwC+VexxJ+MM8H6/k6YJnDYzGrAH1ZpIxMDcSGWKeVND+Y+Zs5wVeL7JWGanCk0TaJ0WKftxZxRQtWVuOzP1UaPfZfkti/iUu1BIx9ym8n1mWUyMs1wQkRKJGHhvhdoIaTLu36yHymM8vmJ5AU7GST/XAniuzOd7BcDHAdwPZ2qBLwH4ZwAfklIeK/O5iIiIiIiIxtV4fBN3B4CfCSGeBPA9ON/I/r4Q4osArgZwcTlPJqX8ewB/X85jEhERERERVYuKd+KklL8QQvwhgH8A8J3U4q8BOADgD6WU/1HpayiXkRETcWliKC5RZwhI6YylUGOJ1Dg3NUbIHYeUWt9Yp2EwbrvPEKuxG5omENadsUQjpjMGRI2rM3QNdYbASNKGJSUihg7TcsZ/GJpAxNAwmLAQDemAlEikxtRFQxpMe3TcSFjXIKXESOr55oihIW7a7hiT2fVh9MeSiCUt6Kntm6IhnBgx+cxyDVNj4tQYHBWb6nWdoWEokT62zUqtM3QNsaSzzi92E5Yz1kjFuYr/ltTg5N6hhLvMlhK6Njq2KGzomBkx0DuUQNKyEdY1hEMi67n5IHyefvIzTRvHhxNIWE5Mhj1j3CKGBksClm2nMkE6cdZQpyFhOuMwk5ZEwnLqu4awhpGkRNJyxiebdmp8na5BF4AFCSmFu16NudNTY9xiidEyIwQQDTnnSdrOVALea2is02BowIlYfrGcj3zjvZBywTJESj7j+CbhuDmC00aIWSaGPXVcSBdIWM69Pmk59aauCYQNDdKW0HWBeKodoaVyP2ipcXSmJRE3bTeHg2lLhPT0e7+3rrFtiROxBJKm0661pUQkpGN2Qx3rozyMxzdxkFI+DuBxIcT7ALQC6JNS7huPc5fLyIiJU8kkjp5KoOutY/i9xe/ByeGkO1Gyysjz7L4juOrcNnzr56/j+gsXupMWq0xk3/r56+gdSGTNcH/XtUsxoz6Eb/xkX0H7bV65DL947SguXtSC4YSFDTv3oKWxLmu7TSuWoD6s4/Yf7UXLtDBuvPSstGtX53hy71F3+9nT6rDzxYO495cH3OxBi+ZMY8GqESMjJvb3DbkTDK/53QW46tw297WKn+2/ehsnYgncdNnZaes2rViCrz+xDx9cMNPdzy+2Ni5fggeefwvXX7gQDzz/Fm667GzMiBr46mN7s2I5n/P1DsZzxpptS+w7MuBOFsrYnHxM08aB40PoHYhn1ZM/7OrB1cvm+sbgTZedjTnTnQ+kbrj/RfT0x3D54la3vvOL37uuXYpIyJmEPnP95Ytbs+L0/s90YCghcCzj2jKvYWZUw8fufq7k2Mw33gspFyxDROQkNUng2GAyq13w+MuHceXSubjnqf1uu3DLymVO0qdBG2sz7t31YR0NdQaklNjZdQhXLp2b1sZU9/7n3+xz6xoAONA3hIGRpNt+ZX1UmIpnp6w1QVmnDvcPQwL41H2/xvdWX4CEaeMz97+QNmGhc4M/Hzd890XcdtVi3PHY3qz1t121GAB8193xiQ8gYdkF7/e91RfgjaODuO3RV9DTH8O917XnPH7QcW67ajHWPNSdtn1bUxQfvetZd9kj6y+qhsGoLNUZ/OL2cP8wPnnfr933+ac3X4wbvvuib8y+eWwoMCbOmN3g7hcUWyre1f/fveGDeKN30Hfbsc635qHunLHWOxDH1Zufy9q3SmIzCGPWR1B9+9sTMex7d8Ct0xRvHRsUg9///AXYf2QQN3z3RQBIi9lcdeMN330xa73f9j/704/g0PFh32vzXoMAcNHGp0uOzXzjvZByUcC2jFsfky07ZblU0bd1jFsfmXF7uH8YtgT+eNuvc7Zlve3Cf1p9ge/2d3ziA2mvg9oaH73rWbeuAYBXDp8EAN/6tMrv6ZVQHdkphRCrCtleSvlgJa6jnEzb6ez29MdSadORFnBqna6JtNnqM9d7Z63PXFcf1lEPveD9pJSoD4/uF3Rudfyg46hzeLfXPZ+C9PTHkDDTJ3Gm6mXaMu19VrHppWI2V7x698u1nfd/TQRvO9b51M9BsZYwLd99GZuTR9Ky0+o0xVvHZi5XMWXZTn2oeGMtV93ot95ve00g8Nq81+BdXkps5hvvhZQLliEiMm0JibHbst7ltpQ561Agd1tD/azqGrUf66PiVCo75XcL+Hd/ha6hrIzUnEBtTVFoQqTNWK+0NUVh2RJtTdGcs9gHrVOz2he6nxACwwnLXTfW8XOdI3N7b2OkrSmKsFH8nEY0voxUvCoqNr1UzOaKCe9+ubbz/m/L4G3HOp/6OSjWwobuuy9jc/II6VpanaZ469jM5SqmdM2pDxVvrOWqG/3W+21vSwRem/cajFSDpdTYzDfeCykXLENEpOYbHqst612uBWw/nLDcf7naGurnsKEjbOjuPqyPilOpTtzCAv6dUaFrKKvmaBhhQ2BrZzue2vsOQobA5tSM9cDoLPU7uw5iS6cz0/3G5emz2KvlfjPc33XtUsxtihS83+aVy/Do7h7Magi56/y227RiCWY1hLD1mTewq/tQ1rWrc3i3b5sVxc6ug+6ybas60JxKWkHVrzkaxpbOdvd9VrGZGT/bnn0Tu7oPZa3btGIJtj7zRtp+frG1cfkSN27VccKG8I3lfM43Vqw1N4SxbVVH2r6MzcmltbEO82ZFfevJbc++GRiDKvbaZkXd9d76Lqjund0Y9l3vF6dS2mjzubbMa5gZ1coSm/nGeyHlgmWIiJqjYRg6fNsFO7sOYvPKZWntwi0rl8G0LWz1uXfPaghh3qwo2poi7r5+935vXdPcEMbpzfVp7Ve1Peuj/HBMXIZcz7oHZqeUErooX3ZKXTif9mZmp7SlRF0e2SltWyISkJ0ybjqvVXZKy5YIebJTjiQtd/sqzk5ZFRdRTYLiVmWnVDFYzuyUKmNVqdkpTctGaPJnp6zqi5souerbcmSnVPWrX3ZKK5U1zclO6exvWjbCKjulbUMXo9kpVRnShFPHT5HslIxbHxwT549j4qqbX9wyO2VVqY4xcZNVJGIgAgMzomNvG2RGffmuR2ku03FaQ9lfXbf4LKPaEYkYmBvJXcxn53msQmN3btj/vDM9xzltZkZhasjv2JomptqA5ynHMDS0To9M9GU48ozLtF3KeOn5xnsh5YJliIhUu7apiDquUDN92hCaJjCrgfVQsSqV2ORNAFdLKV8WQrwFZ4LvIFJKeWYlroOIiIiIiGiyqdQ3cb8AcMrzM5/ZJCIiIpqkyvEoaRU9kklU9SrSiZNS3uD5+TOVOAcREREREdFUxDFxBbBtiVgijqG4hGEIJFMD2+3UgHc1sD2UWudNeKIJZ/DnSCqZSDSkARJIpAbYh3UNugaY9mjCiIihIWE5g0f1rMH7GB3wr2vQNYG4aaUNsFeJKtT/dsYA/PqwhuFEesIUK5WAQheApmm1kDCCclCJTdSAZfWeq9fTohqSJtyEJt44joSc5DdSjibnUftHwxpMSyJpOXMmOokgUueoS8WbJ9mJBBA2NBiaQCwRnEhBJVuwbOd61DFbG+sQ4vjMKSMz6YZKgpO0bCeJE0brv1AqKVTctBHWNYjUzxFDhxBOghIpAStV/4V1DbPqw7BtG8eGEkimYjvqSfoT0p1YhZBImk4cNoZ1jJjO+qihwYZzntkNOk7ERsvUzKgGHQYGklZZE+8EJSLxLhdCZNXdNZgEiIjGyciIiaGk6dZtKsnZUMJCSNNg6M6ULUYqgUnctGHoGkK6k9xPE077IOlpy4YNp34xrdHl0ZCOWakEeoXURd76KxrWYdoSSXM0aRSAKV2/VbwTJ4S4C8BsKeV1PuseAnBESvlnlb6OUqkO3Dunkk5nKw4cG4hjw8496OmPuSmmH3j+Ldx46Vm456n9eHLv0bTlN116Fr711H70DiRw+8cXYzhhpe2/tbMd3/z563hy71FcvrgVN156Ftbv2J12/Gf3HcGK8+dnnfuua5fi7378GnoH46Pnu+xsHOg9hQUt07Fue3fWdd502dn41s9fR+9AAl++YpHv73LzRxdh0ZxpU6pQTBYjIyb29w2lvfdbOtvx2G96cO8vD7gxFwlp+Mz9L2bHx6VnIWQIfO6B9P0P9J7CkvmzcGo4iW89tR/XX7gQt+7ak7ZNY52Gv//xa1nrNq1Ygq8/sQ+9g3FsW9WRFlu2LbHvyADu+uk+32Oe09rIjtwUoOJg9YNd6OmPYc3vLsBV57Zh3fZutDTW+dad3vrP+/M/fvo8jCRt3PLDl9NiUH9PI945EcfaVNnwq2/vW9UOywbWbe/GhWc0o/PDp2P9jt1oaazDX/zBObj54Zfx+BcuxJt98awytrAZqA8B5//t02667FLq0cy/iTrmWS2N2N87mLbcW3f7rS/1WohochgZMXF0OI4Tw8m0um/zymXY/qu38fybfWn3bO/PWzvb0VCnwbKB3oz26I7VH0LCtLPaqfd2tuPuVBs3n7rIW++1NNZltVMf/OwHETftKV2/VWqeOK+PA3gyYN1PAPzhOFxDyfqGEjgRs3HoeAyGpqPneMwNJsCZXf7WXXuwvH0e1u/YjeXt87KWr0stX3vJmTg+lMzaf+32bnc/dZzM46/omO977psffhlrLzkz/Xzbu3He6c1uAyPrelLnW3vJmYG/y+oHu9A3lBi/PzSVTV8skfXer9vejRUd893Xa7d349DxmH987NjtxHrG/ued3oykKd14Vp0t7zaGpvuu27BzjxunmbHVN5TA6ge7Ao95dDA+Pn84mlAqDtT7v6JjvhvHQXWnt/7z/nx8KOl24NS2G3buwUjCdjtwgH99e+TkaOds9cVnuOvXXnImbn7YOebgiO1bxk7GbJyI2e6yUuvRzL+JOubRwXjWcm/d7beedToRAU4bIWHKrLpv/Y7dWH3xGVn3bO/Pa1P3+UM+7dGkKX3bqWs8bdx86iJvvefXTn27b3jK12/j8TjlXACHAtb1pNZXvYRpwbQl6sM6LOn8rwJH6emPYWY05P6fz3K//QG422eudx5pCz535vksW+Z1nbm2SZhW3n8nqh5mwHuvez6h6umPoT6sZ22j3vvMD7NUTGkifbvMbWwpA9d5Y84bWwnTynlM02Z+pKlAxYGia8J9PVZ9lflzUF2ZWTb8Ys67b+Y1qJ+DylhmrJZaj2b+TdQxk5ads+4OWs86nYhMz73cy9tOCKpb1X3er47VRHDdm9kGzlUXeeu9serofI852YzHN3H9AN4XsO59AAbG4RpKFjZ0GJrzbLAunP/V7PJKW1MUJ2JJ9/+g5SdiyZz7A3C3z1xv2XLMfb3n0zUx5nUGnUstDxt8hK0WGQHvveVpYLY1RTGcsLK2Ue99Zr9JxZQt07fL3EYTImdcqZ+9sRU29JzHNKbI4xFTnYoDxbKl+zqfutP7c9C2mWXDL+a8+2Zeg/o5qIyp8SPeZaXUo5l/E3XMkK7lrLuD1rNOJyLDcy/38rYTgupWdZ/3q2NtGVz3ZraNc9VF3npvrDo632NONuPRifsZgL8UQszxLky9/gsAPx2HayhZc0MYM6Ma5s2KwrQttM1yxlaoAFJjEXZ1H8Lmlcuwq/tQ1vItqeVbn3kDsxpCWftv7Wx391PHyTz+zq6Dvue+69ql2PrMG+nn62zHS2/3YUtnu+91bkmdb+szbwT+LttWdbiDR6m2NEfDWe/9ls527Ow66L7e2tmOebOi/vGxcpkT6xn7v/R2H0KGcON54/IlWduYtuW7btOKJW6cZsZWc0MY21Z1BB6ztZETgk4FKg7U+7+z66Abx0F1p7f+8/48qyGEO69ZmhWDkbCGrZ6y4VffzplR555327Nvuuu3PvMG7rrWOWZjRPMtYzOiGmZGNXdZqfVo5t9EHbO1sS5rubfu9lvPOp2IAKeNEDZEVt23eeUybHv2zax7tvfnran7/Dyf9mjIEL7t1Hs9bdx86iJvvefXTj29uX7K129Cyso+oiSEWADgRQB1AB7D6COUVwGIA/iQlPKtil5EATo6OmRXV5fvusDslKkslCpLn8pOqbJBagIQnuyUti0R8WSntFPZ0FR2StOyoY1rdkqgztBS2SmdR4eqPDtl1V3QRAuK26DslCrbZFp2SpkeZ5GQhoTpxEc+2SndbTzZKdW6QrNT2rYT9+qYkyA7JWPWR1DcBmWnNC0bEU92yc4UEwAAIABJREFUStuWMFLZKROmjVAqO2XCtFGXyk5p2U4M21JCFJydEkimsrb5ZadMmDaaJ3d2Ssatj1zthHLMlTaVlWmeOMatD7+4zSc7ZSxhQc8jO6Vqy2Zmp3TavMxOmYeCL7ziY+KklAeEEOcD+BsAHwXQDOAYgEcA/JWU8u1KX0O5aJpAQySChshEX0n5NDVM9BVQJUUiBuZG0ot5qe95yTGTY39NE2iZxm/cpjq/ODhtZjRg66LPgtPC5bkF+t0TIpHy3l6DysZYZYZlimpJPp1gTghePpGI4VtXNVfofIXWRfnUX1O5fhuXeeKklAcArBqPcxEREREREU1m4zEmDgAghNCEEB8QQnxECMHvf4iIiIiIiIowLt/ECSH+BMBfAZgNZ3jM+QB2CyH+BcBTUspvjsd1lIMaYxTShfu8rxprZNpAPDXmTdeEOzbO8jwnHEs645FU9jJLOgOGQobmZL5MLQ/pzrPHSUuiztAgJZDMGCuXtKS7vZYaCxEzbYR1gaQl3bFN7tg8DbBtQAjAEAISAiOm5V5L0lZj5iSkDRi6hqRlj9tzxiWO3SAfmWPiQqnYCOkCYUNgKG5njZerD2tImNId8xM0lu5UzIYA3DhrbayDEAJHB+MQkLDl6DopgbpQ+nvqfb9DuvN5Uizp/NzSEMaJEZOxMEWZpo3jwwl33G80rMO2JeKmDV0TCDsDjWHaNsK6hrhnPIemAZACdSGBWMLOiv2ZUR0nYpYbyyI1bjNsaLBtCSmlU9+m1keM0ePPiOpumYkYGkxbYmZUSxsTFw1rqNfHHhNXbH2XuV9TNFTQOBPTtHF0MI6k5YwhbG2sg2GM2+e5RFRFRkZMnIwnkUyNQQ/rGqZHBE7EUmPfTScHhNO2RVo9lzSl04aFQEtDGAMJE0nTThsfp3nG6AJwx7xbEpBSFlWHVUottkEr3okTQqwGcDeA78CZ9Pthz+pfAlgOoCY6cSMjJvb3DeGx3/TgqqVzsS41QeLli1vx5SvOSZu1/vLFrbjx0rPcSRQzX6ssaiFDw+an/xs3XLQQX39iH3oH49i0YgmaG8P4xk/2oXcgkTZL/eWLW3HTpWe551YZg+rDOqJhHc/v78WyBc2456n9uP7Che6kySpr2QPPv4X1v/c+JE3bnbBWHUOdX23nvaZtqzqwaM60igW0bUvsOzLgTtyosgxV8pyTnYpXNRmxyjr1+MuHsaJjHuKmdCc8Vln1DvSewoKW6ZgR1fHpbS+MxtxlZ6cdZ0tnO949MYy/fuxVN17+8qrFGIiZ+ObPX8f1Fy7EA8+/lRWD6j0FkPV+qxhsmRbOOh9jYeowTRsHjg+59WlLYx3+4g/OyaqvpkcM1NfpODaYSKtXNy5fgv3vnkT7wtnpMbtyGQ4cG8CClulpy731Xcu0ME4Mm/jSD36TVmbueWo//sdpM3DJ++dg3fZutDTW4ctXLMLZ72nAm33xrDLWVB/CrHod5//t077xW2x957ff1s52fPPnr+PJvUfHPI5p2njtyEBaud/a2Y5z5kxjR45oihkZMXHoVAzHPG1XVSe0Tg/h8Ik4Gus0WHHnC4o1D6W3ARrrNJyMmdj89H9jwxXnIGnZODmcTDvWndcsxbf//U3c/NFFqDM0fO3fXs1qFxRSh1VKrbZBx6PW/lMAd0opPw8nmYnXawAWjcM1lEVfLIF127uxomO+24kCgOXt87JmrV/ePs9tWPi97umP4eaHX0b/UBLL2+dhw849WHvJmejpd45zuH8Ey9vnZc1Sv7x9Xtq51fbHh5I43D+CSxe/F+t37Mby9nluIVHb3bprD5a3z0P/UNJtEHmPoc6vtvMuW/1gF/qGEpX72w4l3MKjrqnS55zsVLx6/6brd+zGio756OkfcRtyat267d047/RmrNveDQEtPeZ8tl182oy0eEmmOoUq9vxiUL2nfu+3ije/8zEWpo6jg/G0+nTtJWf61ldHBxIwND2rXr111x5cuvi92TG7Y7cb33714oadewBobgdOrVf16SeWtbn7qnp5JCF9y1jclDgRs91lmfFbbH3nt58qc/kc5+hgPKvcr93ejaOD8QLeISKaDPpiCfRktF1VnZA0gXXbu2FoOg4dj+HoqXhWG8DQdLcN23M8hiMn41nHuuWHL2N5+zysfrALb/cN+7YLCqnDKva3qNE26Hg8TrkQwE8C1g0BmDkO11AWpi3dmey9s8R7Z7D3Lsv1Wm1fH9ZRDz1tJnvv8rGO690ecB6dVMfy287vWv3Wqf29yxJm+qTQ5ZQwLd9rquQ5JzsVr14qfuvDuu86K7WP7Zl6JCiWzNRkoCpONJEdO7ne00Ljk7EwNSQtOy0+c9V5lvSPcTtguRVQJtQ5VAz7rfceU20fVMY0Abd8qGXe+C22vgvaT5WbsY6TtGz/smzZOc9LRJOPacvAtoDpaQuo9mXmNmqdaquq5ZnbqfrS297128b7erzv97XaBh2Pb+KOAVgQsG4RgMPjcA1lYWjCncneO0v8iVgya+b4zNnl/Wabb2uKYjhhuevUTPbe5YUcZzjhzBOnjuW3nd+1etd5f85cFjYqN09X2NB9r6mS55zsVLx6qfgNigE9tY8mRh8fCIolI/WIgYoTW2bHTtB7GvR++8W8dz+a/EK6lhafueo8XfjHuBawXA8oE+ocKob91nuPqbYPKmO2hFs+1DJv/BZb3+UqN/kcJ6Rr/mVZ56OURFONoYnAtoDhaQuo9mXmNmqdalfmaltmtnf9tvG+Hu/7fa22Qcej5v5XAP9HCHGGZ5kUQswGcDOAfxmHayiL5mgYWzrbsbPrILZ4Zrjf1X0oa9b6Xd2HsDljG+9rNSauqSGEXd2HsGnFEmx95g13vMfcpgh2dR/KmqV+V/ehtHOr7Wc1hDC3KYKn9r6DzSuXYVf3IWxcnj67/cblS7Cr+xCaGkK469qlWcdQ51fbeZdtW9XhDkytyN+2IYxtqzrSrqnS55zsVLx6/6abVy7Dzq6DaGuKYGvGui2d7Xjp7T5s6WyHhJ0ecz7b7v3tybR4CRkCWzvb3djzi0H1nvq93yre/M7HWJg6Whvr0urTrc+84VtftU4Lw7StrHp14/IleGrvO9kxu3KZG99+9eKmFUsA2PiHT56bVWZ2dR/Co7t73H1VvRwJC98yVmc4k36rZZnxW2x957efKnP5HKe1sS6r3G/tbEdr49SdZ4loqmqOhtGW0XZVdULIALZ0tsO0LcybFUXr9LqsNoBpW24btm1WFHNm1GUd685rlmJX9yFsW9WB05vrfdsFhdRhFftb1GgbVEjPY1MVOYEQzQCeBzAPwH8AuDj1+hwARwFcKKU8WdGLKIDfjPZe5chOqdZnZqeMJSxoGdkpTUsiHJCd0rSku32x2Snj5mhGzNHslIC0ZTVnp6zeUaYTJChuy56dUkroIv/slFoq858EUGeMnZ1yJGnBmJzZKWv64islKG6DslMmTKcO9MtOqbJNapoAJMbOTpmK5aDslKYt3fpWHX96KjulZTtZg2s5O6Vp2TDGzk7JuPWRq52Qz2TVVJo8Jvtm3Prwi1tvdkqVUTJXdkpVz6rslLaUkH7ZKaVESGN2ygIVfLKKj4mTUvYJIToAfAnAxwC8kTrvPQDuklKeqvQ1lFMkYmCuz+z2VDpNE2iZxk+Ey2mseJ1Zn/66KccMjpnrpkf8tzttZtR/RYax3u+WUHU/xkCVYxgaWoMCrBAB8dxQwqEzy0zQ8SJj3CeKre/89ivkOIah5V1GiWhyi0QM37qqmDpyVh737KC6qhrafrXYBh2X3oiUcgDAHal/REREREREVKSKdOKEEE8VsLmUUl5WiesgIiIiIiKabCr1TZwGZxiMsgjAewAcAHAEwBw4GSvfAbCvQtdAREREREQ06VSkEyelvET9LIT4QwB3A7hASvmCZ/mHAPwgta6mJZMWjg8nEAkJSADDcScBRFh3Br6rAfQhXTgJSmzpJh9Rg+EjIc0ZtC+cBCN6arC9Wl9naAjpwGDcdhNJqEQlbmKKOs05t2fQvWk71xEJCcRN57VlS4S00QH9IUPAsoCEZTsJWTSBSEhgMG7DlhKGpqExInDKM3g/YmgYTFhOUhUNkNJJrCJSSVJiiUmTkKKmZSY2yUxQEgk5SRukhLtMJSOpD2sYSdqw5ehgZhXD3qQnLQ1hnIw7SRwa6nQMJ2wkLdtNJmHaTtITTRMTPWiYakQyaTkJclJ1nWVLNNTpGEmO1kHTIk4yKTuVUMq7ra4JNNRpOBmzEDE0WLaThGp6xIlPM1UHGqmEUhFDg5VKHqVrAg2pcuI9ppY65+CIDU046frjpo1IWMNIIr1ujISQSgygY2bUP5mPtwyUY0C9bUuciCUQSzhJWyIhHbMb6ljGiCjQyIiJYctMSwI1M6rh+LDltmHNjDajN0melMC0qI6RpETStCGEgBASAk69po5p6Bps6bQ/rVTiPLYLSjceY+LuAHCbtwMHAFLK/xBC3A7gqwAeHYfrqIhk0sKB/mE01umIW8DRUwms3d6NlsY6fPmKRe7s9SrFtS0lfrznt7hy6Vys37E7bV1jRMe7J+O4/7m3cP2FC91Z7VXa6hn1IXzvVwdw8aI5eOD59G0uX9yKmy47G+sCzr21sx2WbeNPvveSu+zOa5biqVffxVVL52Kd51o2rViC2dPqsPPFg7j3lwew5ncX4Kpz27Bue3fa9Wz/1dt4/s0+bFy+xL2eB55/CzdctBBff2Ifegfj2LaqA4vmTGOhnAAjIyb29w2lvW9bOtvx2G96cO8vD7hxByDt/Vfv5xcuOxtCAGse6k6LU00Aa7bvzjpm/7CJzg+fnhbXm1cuw+MvH8Y1589H0pRY/VCXu46xQX6SSQuvHR3Et37+ulvHXXhGc1Zsbelsx/xZdejpH8E9T+3PqjO3dLbj1cMnMK+5ARt27sEn29twyfvnpJWHTSuW4JHdh3H1srnpdXUqpi9eNCfrmI/9pgdXt7chYUrsPtCH9oWzs+rGmfUhNNfr2HN4AHOmR7Cg2cmwsu/IAFY/mF4GzmppxP7ewazlhZQN25Y40DeEI6dG0n4PljEiCjIyYuJ4PIG+wWRWO2Fhcx3+8/BgWn1y5zVL8e1/f9Nt611/4UI8u+9IVvvwzmuWIhLS0tqbd127FCFDw42eZd+94Xy2C0o0HvPEnQWgN2DdUQDvG4drqJijg3H0HI/BsoGEKbE2FchrLznTDX7Amfl93Y7dOD6UxIqO+W5jxLvO0HRs2LkHy9vnuQ0HtX79jt1ImhIrOubj1l3Z2yxvn+cWIr9zr93ejeNDybRlt/zwZazomO824NXyDTv3oOd4DCs65gOAs03q2N7rWX3xGejpj6Vdz/L2ediwcw/WXnImevpjWP1gF/qGEuP0bpBXXyyR9b6t297tvq8q7o4NJtK2Ue/j2u3dOHoqnhWnRwcSvsdcffEZWXG9fsdurOiYj0PHY25FrdYxNsjP0cE41m3vTqvj/GJr3fZuDI7YWL9jt2+duW57Ny48q8WtCz+xrC2rPGzYuQerLz4ju65OxbTfMVd0zIeh6Vi/YzcuXfxe37oxYUqciNnYsHMP3u4bRt9QAn1DCbejprZd/WAXjg7GfZcXUjb6hhJ4u2846/dgGSOiIH2xBCwLvu2Ek6n6K7PN6G3r3bprj2/78JYfvpzV3rz54ZfRn7GM7YLSjUcn7i0AawLWrYEzTq5mmbZEfVh3vyJWwTgzGnJ/Vnr6Y6gP69A14bvOls7+QftqAu6+mdt4X+c6d+ayoGtR1wkgcBu13ns93v/VuoRpjfFXpEowPfGoeN839dovLtT76LcuKI5yxUl9WPddx9igTCpuvfVYUGz5betd762TVf2auU2uuA1aron0OjtzG02M/h71YR0J03nc2Pd3sOySy0bCtFjGiKggarhPrro1c3lmWy9XG3KsZayzSjcenbi/BvC/hBCvCCFuF0KsS/3/CoArAdw+DtdQMYYmMJyw3Easmu39RCzp/qy0NUUxnLBg2dJ3nSac/YP2VWOT/Lbxvs517sxlQdeirhNA4DZqvfd6vP+rdWGD831NBMMTj4r3fVOv/eJCvY9+64LiKFecDCcs33WMDcqk4tZbjwXFlt+23vXeOlnVr5nb5IrboOW2TK+zM7ex5ejvMZwaHxw2dP/fQddKLhthQ2cZI6KCGJqAHlCHBbUfMtt6udqQYy1jnVW6infipJTfhzPJ90kAfw7gH1P/nwDwMSnlDyp9DZXU2liHtllR6BoQNgS2drY7Y9CeeQObVixxA1SNJ5rVEMLOroPYvHJZ1jrTtrBpxRLs6j6EjcvT9928chlChsDOroPYuDx7m13dh7Alx7m3drZjVkMobdmd1yzFzq6D2JJxLZtWLEHbrCh2dh0EAGeb1LG917Pt2TfdMVTqenZ1H8KmFUuw9Zk33OebmxvC4/RukFdzNJz1vm3pbHffVxV3sxvDaduo93FrZztap9dlxWnrtLDvMbc9+2ZWXG9euQw7uw5i3qwotl3XkbaOsUF+WhvrsKWzPa2O84utLZ3taIxo2LxymW+duaWzHc/v73Xrwkd392SVh00rlmDbs29m19WpmPY75s6ugzBtC5tXLsNTe9/xrRvDhpMcYNOKJTi9uR7NDWE0N4SxbVV2GWhtrPNdXkjZaG4I4/Tm+qzfg2WMiII0R8PQdfi2E2ak6q/MNqO3rbdx+RLf9uGd1yzNam/ede1SNGUsY7ugdEJKOfZW5TqZEBqA2QCOSSntcTtxATo6OmRXV1dB+1QyO6XKtJYrO6Xaxs1OKSXq1LltiVBGdko7dc7M7JRJy4ZWQHbKodQ3kJrm/F7jmJ2SI14zBMVtUHZKW0poYjTubG92ylRclZKd0lSZTpmdUpkyv2ghguI2KzullGgI585OqWXUmQ11Gk7FLNSNkZ1yJJHaxic7pfeYadkpNSCkVXN2SiAS0sqRnZJx6yNXO2HBVx4f56uZeg587cqxNmHc+vCL23yyU1qqzZhqGxSUnVJKGMLJTinlaIZ0Zqf0VfAvPh7ZKV2pjtvR8TzneAiFdMyZMfqV8Mz6yp1rxhjHbirjuTPPNS2S/ro5184N5bsOKl4kYmBuJL2YN1XgvWkJj54jV/y3TKsr/8lp0gmFdMwtQ2XmF4v5xn/QdjOiPgt9tm2IZC/TNOFbBoKWF0LTBGY11LHuJaK8RSIGIjCy6g2/+iuX6QVu78V2QfHGY0wcERERERERlQk7cURERERERDWEnTgiIiIiIqIaMq6JTWqBEKIXwNsTfR0FmA3g2ERfRIVl/o7HpJRXTNTFVKM84nYqxAlQvb8nY9ZHifVtNb7Xk+2aGLc+arCd4FWNMVqosX4Hxq2PMeJ2MsRFLrXw+xUct+zE1TghRJeUsmOir6OSpsLvWGlT5W84VX5Pqs73mtdE1W4yxMNk+B2qzWT/m07W34+PUxIREREREdUQduKIiIiIiIhqCDtxte++ib6AcTAVfsdKmyp/w6nye1J1vte8Jqp2kyEeJsPvUG0m+990Uv5+HBNHRERERERUQ/hNHBERERERUQ1hJ46IiIiIiKiGsBNHRERERERUQ9iJIyIiIiIiqiHsxBEREREREdUQduKIiIiIiIhqCDtxRERERERENYSdOCIiIiIiohrCThwREREREVENYSeOiIiIiIiohrATR0REREREVEPYiSMiIiIiIqoh7MQRERERERHVEHbiiIiIiIiIagg7cURERERERDWEnTgiIiIiIqIawk4cERERERFRDWEnjoiIiIiIqIawE0dERERERFRD2IkjIiIiIiKqIezEERERERER1RB24oiIiIiIiGoIO3FEREREREQ1hJ04IiIiIiKiGsJOHBERERERUQ1hJ46IiIiIiKiGsBNHRERERERUQ9iJIyIiIiIiqiHsxBEREREREdUQduKIiIiIiIhqCDtxRERERERENYSdOCIiIiIiohrCThwREREREVENYSeOiIiIiIiohrATR0REREREVEPYiSMiIiIiIqoh7MQRERERERHVEHbiiIiIiIiIagg7cURERERERDWEnTgiIiIiIqIawk4cERERERFRDWEnjoiIiIiIqIawE0dERERERFRD2IkjIiIiIiKqIezEERERERER1RB24oiIiIiIiGoIO3FEREREREQ1hJ04IiIiIiKiGsJOHBERERERUQ1hJ46IiIiIiKiGsBNHRERERERUQ9iJIyIiIiIiqiHsxBEREREREdUQduKIiIiIiIhqCDtxRERERERENYSdOCIiIiIiohrCThwREREREVENYSeOiIiIiIiohrATR0REREREVEPYiSMiIiIiIqoh7MQRERERERHVkKI7cUKIs8t5IdXiiiuukAD4r7r/UQbGbdX/Ix+M26r/Rz4Yt1X/j3wwbqv+X8FK+SbuNSHEz4UQ1wghjBKOU1WOHTs20ZdAVDDGLdUixi3VIsYt1SLG7eRTSifuswCiAH4AoEcI8XdCiIXluSwiIiIiIiLyU3QnTkr5XSnlhQDOBbALwHoA+4UQTwghPiGE4Hg7IiIiIiKiMiu5oyWl3COl/BMApwFYA2AOgH8GcFAIcbsQYk6p5yAiIiIiIiJHOceyLQCwJPV/AsArAP4UwC1CiFVSykfKeK6qZdsSfUMJJEwLYUNHc0MYmiYm+rKIyoYxTjSK5YFofLCsVc6Crzw+5jYHvnblOFwJFaKkTpwQIgzgGjjfwF0E4G0AXwPwbSnlMSFEE4D7APxfAJO+E2fbEvuODGD1g13o6Y+hrSmKbas6sGjONFY0NCkwxolGsTwQjQ+WNaJspUwxcCeA3wJ4AMAAgI8DOFNKuVFKeQwApJT9AO4GcHoZrrXq9Q0l3AoGAHr6Y1j9YBf6hhITfGVE5cEYJxrF8kA0PljWiLKV8k3cdQC+DWCLlPJAju1eA3BDCeepGQnTcisYpac/hoRpTdAVEZUXY5xoFMsD0fhgWSPKVkpikzYp5a1jdOAgpTwmpXyghPPUjLCho60pmrasrSmKsKFP0BURlRdjnGgUywPR+GBZI8pWyhQD/A47Q3NDGNtWdbgVjXpmu7khPMFXRlQejHGiUSwPROODZY0oW6mJTT4GYC2ARQAiGaullPLMUo5fazRNYNGcaXhk/UUFZU9ixiWqlHLHVrExTjQZaZrAWS2NeHjNh5G0bIR0Da2NdSwPRCnlugfx3kOUrehOnBDiDwD8K4CfATgHwBMA6jGapfKX5bjAWqNpAi3T6vLenhmXqFIqFVuFxjjRZGXbEvt7B1l/E/ko9z2I9x6idKWMibsNwD8C+IPU6/8tpbwEwO8A0AH8W2mXNjUw4xJVCmOLqLJYxoiCsXwQVVYpnbhz4HwTZwOQSH2rJ6V8HcDtcDp5NAZmXKJKYWwRVRbLGFEwlg+iyiqlE2cDMKWUEkAvgPmedb8FMKXGwxWLGZeoUhhbRJXFMkYUjOWDqLJK6cTtA7Ag9XMXgC8JId4rhGgBcAuAA6Vd2tTAjEtUKYwtospiGSMKxvJBVFmlZKfcAeD9qZ//Ck6Ck57UawvAp0s49pTBjEtUKYwtospiGSMKxvJBVFlFd+KklP/o+blbCPE/AFwBJ0Plz6SUe8twfVMCMy5RpTC2iCqLZYwoGMsHUeWUMsXAfADvSCmTACCl7AHw/1LrDCHEfCnlwfJcJhEREREREQGljYl7C8B5AeuWptYTERERERFRGZXSicv1UHMITvZKIiIiIiIiKqOCHqcUQswEMMuzaK4Q4oyMzaIArgfwbonXRkRERERERBkKHRP3RTiZKGXq386A7URqOyIiIiIiIiqjQjtx/wJn/jcB4DsAvgrgjYxt4gD2Sin3lHx1RERERERElKagTpyU8mUALwOAEEICeFxKeawSF0ZERERERETZSpns+yFkJEYRQnwMwAcAPCWlfKmUCyMiIiIiIqJspXTi/gnOo5OrAEAIsRbA5tS6pBDiSinlz0q8PiIiIiIiIvIoZYqBCwD82PN6A5zJvmcA+GcAf1nCsScF25boHYjjcP8wegfisG050ZdEVBDGMFFuLCNE44NljShdKd/EtQI4DABCiPcBWAjgHinlgBDifgDfK8P11Szblth3ZACrH+xCT38MbU1RbFvVgUVzpkHTck2xV55z9w0lkDAthA0dzQ3hip+Tal9m3DRFQ9jfOzghMUxUC2xb4kDfEN7uG0Z9WMdwwsLpzfVY0NzAMkK+eH8uzkS2qYiqVSnfxJ0C0Jz6+RIAxzwZKS0AkRKOXfP6hhJuZQMAPf0xrH6wC31DiYqeV1V0V29+DhdtfBpXb34O+44M8BMrysk3bo4O4K6f7hv3GCaqFSdiCRw5NYLbHn0Fn7zv17jt0Vdw5NQITsRYRigb78/Fm6g2FVE1K6UT9zyArwghrgLwJaQ/Wvk+AD2lXFitS5iWW9koPf0xJEyroudlRUfF8IubNQ91Y3n7vLTtxiOGiWpFLGFhw849aeVmw849iCVYRigb78/Fm6g2FVE1K6UT92UAswD8CM63brd71n0SwK9KOHbNCxs62pqiacvamqIIG3pFz8uKjooRFDfNDeG0ZeMRw0S1wpLSt9xY/GKFfPD+XLyJalMRVbOiO3FSyv1SyrMBtEgp3yelPOBZ/UU4nTyXEGK+EKKUMXg1pbkhjG2rOtxKRz2/ndkoLjdWdFSMoLhpnVY37jFMVCsiIf9yEwmV8vkoTVa8PxdvotpURNWs5E6VlLLPZ9l/el8LIXQAbwE4H8DuUs9ZCzRNYNGcaXhk/UXjOoBZVXSZg39Z0VEuQXFz2ozouMcwUa2Y3VDnW25mN9RN9KVRFeL9uXgT1aYiqmbj+c3YlCtpmibQMm18b+as6KgYueJmvGOYqFawvqVCMF5Kw/sRUbop83hjpfmlDQYwIamEWdFRPvxitpi4YcpsmuokJCwpkTAtnIglMDPKMkD+JstrIea8AAAgAElEQVT9Od96n/cHosqZ8E6cEOISAE/7rDoppZzp2a4JwCYAfwggCidxys0+j25GANwBoBPATAC/AXCrlPLZivwCCJ6/pM7QsOo7L3BOE6o65Zpzh3P30FTmF/+bVizBnOkRzhVHk1a+9T7vD0SVVU2jr78A4MOef7+vVgghBJwsmFcAuAnAcgAhAE8LIdoyjvNtAKsB/B8AVwF4B8BPhBDnVurCg9IGv903zFTCVJXKleqaKbNpKvOL/w079+DtvmGWAZq08q33eX8gqqwJ/ybO41Up5a8D1n0cwP8EcKmU8mkAEEL8Ck6ylC/D6QBCCLEUwKcBfFZKeX9q2S8A/BeAv0kdp+yC0gbXh/WsZUwlTBNJPdoynDBx21WLsfWZN/DSoRMAiotPpsymqSxX3c8yQJNVvvV+ue8PfDSTKF01fROXy8cB/FZ14ABASnkSwL8C+ETGdkkAP/BsZwL4PoCPCSHK/iC6bUtYtvRNGzycMeErUwnTRFKPtly9+Tl8ZNMzuOOxvfizjy3CefOcp5aLiU+mzKapLGRovvEvAZYBmrTyrffLeX/w3r8u2vg0rt78HPYdGYBtc1JGmroK6sQJIZakxpxVwg4hhCWE6BNCfE8IMd+z7ncAvOKzz38BmC+EaPRs95aUcthnuzCA95X7ovuGEvjq43uxcfmStPlL7r2uHac313NOE6oafo+23LprD9ZecmbR8cm5e2gqMzSBTSvS6/5NK5bg9FlRlgGatPKt98t5f+CjmUTZCn2c8iU449VeEEK8CeBqKeXLY+0kpbSEEL8HYJ/P6pMA7gTwCwCnAJwH4C8A/EoIcZ6U8iiAWQAO+Ox7PPV/E4DB1Hb9ObabNda1FiphWnhy71H0DiRw21WLMTMawolYErMbwmiZFsEj6y+CbduwJCCl8ygAHwGgiRD0aMv73+OkvC4mLotJmc1HYmiyiCUsfP2JfWl1/9ef2Id7Pn0eY5omrXzr/XJOqcBH94myFdqJi8HJDAkACwDk/XiilPIXActfgtM5VH4hhHgWwAtwxrr9bzhzzPl9Z55ZE+S7XfpKIT4P4PMAMH/+/FybZlGPC7x06ATWPNQNwPm06ZH1FwFwUk/HTBumJRFLmDg2mMDpzfXMXEYlKzRuVax6b4RtTVFEw0ZBKa/zmZrANG0cHYwjadkI6RpaG+tgGBqzlVFJ9W21CRs6PrhgJs6Y3QBdE5jVEMYHF8yEZUvYtmRMTyKTKW5zyfdDtnynSijXlAohQ8Pli1uxvH2e+4HJru5DCBm1MipoYkyVuJ2qCu3EvQLgG0KIx1OvPyeEuCJgWymlvKOYi5JS7hZCvA7g/NSi4/D/Fq0p9X+/Zzu/KG3yrPc7330A7gOAjo6Ogh6wVo8LZDZKm6KhrMbqxuVL8E8vvI0bLlqIGfUGmhsq9WQqTQWFxm1QrBbyaEs+nTDTtPHakQGs3d7tbrO1sx3nzJmG/ljS95GYR9ZfNCnmTqKxlVLfVpsZdTquOrcNN3z3RTfWt3S246lX38WVS9sY05PIZIrbINX8IVtIB2689Cys37HbvbbNK5chxKGnOU2FuJ3KCu3EfQnAd+B8OyYBfC7HthLOfG3F8n6r9l8ALvfZZjGAg1LKQc92Vwsh6jPGxS0GkADw3yVcjyvzk6qzWhqzHhcIGn9021WLsWHnHnz/8xcADeW4GqKxqZidHjHw8JoPQxeApmkFP9oSNC7B2wk7Ohh3O3Bqm7Xbu/Hwmg9DSslHYmjS6B1KYF1GrK/b3o37P3M+Y5pqTj71e6WM9Q3gUNx2O3Dq2tbv2I3vf/4CNNVX9NKIqlZBnbjUFACLhRAaABPARXAeeywrIUQHgLMBPJxa9CMANwghPqIeyxRCTAfwvwB8z7PrjwD8NYBrADyQ2s4A8EkAT0op46VeW76fVAU9vz0zGkJPf6wsGZU4tojyUc5PV/MZl5C0bN9tTMtGNGz4PtKZK1sZ45yqlWn7fyhh6ILZKanmVGLcWT71dz73KCugrDE7JU1lRT1MLKW0AdwA4HUppRX0L59jCSF2CCG+KoT4IyHEpUKIWwA8AeAwgG+lNvsRgF8B2C6E+JQQ4mOpZQLA1z3X9Rs40wv8gxDic0KIy+BML7AQwF8V87tmyjdDUlBq3ROxJNqaogjppT3HzXS7lK9yZvXKJ2V0SPdPu27oWsHZyhjnVM1CmvCN9bpUrBPVknJPGZNv/Z3PPSoccF8ptS1FVMuKjn4p5QNSyj4hxCwhxJVCiOuEEH8ghCg0A+QrcOZ3ux/AT+A8svnPAD4kpTyWOpcN4CoAPwWwGcAjACwAvyelPJRxvBtSx/oqgMcBzANwhZRyd1G/qIdtS8SSZl6fVPk1VjcuX4Jd3YewtbMdLY2lPZrAdLuUr3J+uppPJ6y1sQ5bO9vTttna2Y7Wxrq0bGXP3fp7eGT9RTm/EWScUzULmmLA0AS/LaaaU+4pY/Ktv/O5R7UE3FdKbUsR1bJCx8SlEUJ8FcAtcOZgU3esuBDiG1LK2/I5hpTy7wH8fR7bHQfw2dS/XNvFAPxp6l/ZqE+U3j05ktfjYJmpdQFAE8DtH/+Am6mvFEy3S/lSExJnxmwxWb00TeCslkY8vObDMC0bRirzpLfBahgazpkzLWsbFfOFZCtjnFM1G7Fs3ykG7v7jcyf60qjGTcRj5OWcEgDIv/7O5x411n2FaCoquhMnhPgSnPncvg1gO4B3AbwHQCeAvxBC9Eopv1mWq6wC6hOllsY6bFy+BLfu2jNmhj9vY9VbIffHkiVVjLYtYdmy4LFFNDWpbws27ByNWfVtAVBYY8G2Jfb3Do45vs4wNJw2M+p7jEIETYvAOKdqENI19A7G3ellAD7iRaWr5iyRhci3/h7rHuVuV6b7CtFkUco3cWsB3C2lvNmzbB+ced4GAawHMGk6ceoTpZ7+GL7xk9FPXtuaonjvjOiYkxuXo0L2Nra/+vjerM7kvde1cxwGZck1IbEdLSw2xzt7Wa4pPHoH4kx2QhNKA3DXtUtx88Mvu/F517VLix+nQISJyxJZSFslnw//8p3WJtc9ilm8iYKV0olbAGfMmZ/HAawr4dhVx/uJkprYW03qPVbjsRwVsrdyvfOapXhy71H0DiTSKr3ZbMiSj7Ch+35bEDb0gmNzvB9v9Hu8pykayuvbQKJKi5s2/u7Hr6XVw3/349dw96f4OCUVb6IeI8/3fpBvZy/fxzNz3aOIKFgpHxj2AfhAwLrfSa2fNEoZ8FuOCtlbuaoMl6oz+cn7fo07HtsLTePnv5QtV+wWGpvlzl6WD/VY8tymerRMqwucMJzJTmi8GZpwG5+fvO/XWPNQN3oH41mPgREVYiLqWSD/tkohCacy62+/D9rKnVCFaKoo5Zu4RwDcIYToA/B9KWUyNSfbNQD+Bql52iYL9YnSj268CLGEBUtKREL5VajlGNfjrVy3PvNG3uPyFM61NXXl+jQ0bOi4fHErlrfPc79J2NV9KDA28308ppKY7ISqhdCAO69Zilt+OPo45Z3XLIUow+dprLOnromqZ/Ntq5S7Ds73GzuWCaJ0pXTi/hzAUjidte8IIY4DmAVAB/DvcJKeTDpHTsULfoyrHBVy5uOc3/jJPtzxiQ/gzNZGRENjJ6OYDIOkqXhBGSGboiF84bKzsXZ7txsbWzvb0RQNBR6rztBwxyc+gPqwjuGEhbpxzg7GZCdULXRNw7f//c20xymd178D25YlJa9inT21TUQ9m29bpRJ18FhZi1kmiLIJKYufNFcIIQBcCeB34XTgjgP4BYB/k6UceAJ1dHTIrq4u33W9A3Fcvfm5rIorn7FtpX6C5FeB3XtdOxa1ThszxW4p112lWGNnyBW3uRQaG9UQSzV6M6/aC5tIxcZttbBtiVffPYU1D3WnfRP37X9/E3979ZKiy0Q1lLMUxq2PSsftRL7/+bRVyp0AJR8F/k0Ytz5yxe2CrwSluBh14GtXlvuSKF3BcVvSPHGpjtpjqX+BUp29bwO4XUp5sJRzTqRCHiHwq7j8phvIp1JT27c0hrHjcx9C70AcfUMJ3P2z13HzRxeN2Xjl42ekZMZeobFRLbE00d8GEimzG8J44LMfhK4J6ELg+FAcvQOJkspEtZQzmhgT+f7nM49nIY8/Hugbwtt9w25dfXpzPRY0NxTckWOZIMpWUieuABqA6wHcA6BmO3H5PkKQq+ICkNcnWN7GtmVLfPXxvVjePg93PLY37fx73xkY89M5Pn42dXnjKGRoGBwxseo7L7ix9/3PX+A/yWrGPFfqOJaUuP8z5+ObP9+Plw6dcLcfz1jqG0q4v4P3mmv4m2WqUf2xOEZMC1I6ZcSUEpGQhts/vhjRcPFlgnX21JbP5NeVUs5xZydiCRw5NYLbHn0lbf63mfUhzGoorK6eyL8JUbUaz+iv+a+3882g5K24Pnnfr3Hbo6/gyKkRnIgl8srqpB5VuHrzc7ho49P49P/7D1x/4UKcNiNS1CdRzPw0NWXG0R9tfh5HTo2gpdG5efb0x2BLiU0rlqTFxqYVSwKPc/HXn8Ftj76CL1+xCOfNm8nEJjSl2TZwMmbiM/e/gEvv/AU+c/8LiJsShqbBtIsfUcA6e2pTk19n1suVznqaec+4evNz2HdkAHZGLOe7XSxhuRN4A049vWHnHsQShdfVE/U3Iapm4/VN3KSQ7yMEQRXXDz5/gfvaK7MB6tfRu3XXHvzT6guK+hYk3+umycUvjjbs3IPbrlrszsdjWtJ3klXvPFdBx/nB5y8oKJbK9Qkvv6WgapEwbazfsTutbKzfsRv/tPoCJE276OOyzp7aJmry63znict3O0tK3/aO5fP5xlj3B04ITpSNnbgCqefFTdPG0cE4DvUPI6RraG105j/pG0rAtCVuu2oxtj7zhtvZcr71ACKhsRugQd80/PZEDLc9+go2rViCrz+xD72D8bw/nc3nOXeaXILiaGYq8+S17W0IGxr+8sr3o28ogV3dPbhs8Rz85ZXvh64JN7te0HEAFDVZfT7JSHLd0KthmgMiADBtiZbGurSG5dZn3oAtJSKh0m6vrLOnroma/DrfpxwSpuUb95nbBbV3IqHsx/X3vTuA1Q956vTrOrDoPaP3h7Cho2Vaeh3fMi3MD+9oSmMnrgimaeO1IwNpadm/e8P5iJt2WpayjcuX4Bs/2YeXDp1AW1MUUko0RUNjNkCDvmk4EUsW/S0ITT1BcTScsHBtexs6P3w6PnXfr9HTH8Pli1tx46Vnud8qqOynsxvCEEKU/M1Xvp/cAmN3+PgtBVWLiKHhy1cscp+8UI941RkaZtSxcUnFqfZ54qJh3TfuM8eBzm6o8/09ZmeMhzs2GHc7cEDq/vBQF/553YVonR4BUNx0OESTHTtxRTj6/7P35nFSVOf6+HOqunqfnRlQZgREFkcchBYcMFdRclG/olwzgAqDgsoaJYui5t5wNZn4i4h8TVxYNBERUEHQXxK8LomKyRXcBoWYEUREnEFwhtl7r+V8/6iumq6lh54FGJjzfD79Yag+fepU91unznve932eYEyfSAB1wqluiOjFu9qx+7aqqWsV26qwrKwEv3mtCg/dUNLuAlRRKCgoNtx+CQ4eC+Hxt/ejLhjTHUKtb8A+CsLEMBk0pFoI9M10YWhfP25MOHAAUBYosqSFzV9fiaWTi7G1shqrywOGh2dHFxTH2+FNtltCCB776752HT4WpWDoCVAosGTLHkNUIhyX0RSOQ1Yo+jvZI5ah4+jIRlV3PvNzPAKemzMG1Q0RnZStKNdjcZQkhdqWjLyyaHynriMi2j8fomJbZK8xIlrWXQs2VDJCK4ZeDfaE6QREWbFMOF4nbzsJDSnwY+nkYj0i98B1csoFqF0EYnV5AAqleOBP/2q3Dk5RKJoicRxpimK+abHdw/WzGE4QzA9QQgh4AsgKICvGWoWCDJet/RZkuPBWVS0AYPP8caCUdmqh0N4Or53dLysrQV1r3JCOzIhLGHoaRFlBvt+Fe64ahvu2tkUlVs0cDcKmXAYbpOt0pbNR1d2amS0xEY2huIFN8rHpI5HrcyLX0TYWUbKugWoaI7Z1oOlcB8/ZZ3skXwMjtGJgsOJksVNSqCLgrSfpfCcUAs/pDEkawnHZcqwwx2Og5DxeCppdytmCDZXI8gh6LrhdFESbyHdXN+sOnPZ5M/MlQ+8CxxHk+ZxoiUqYvmYnLvntO7hh5fuggMFeMz2Crf1mJnZg36qqBaUU/XO8yM9wdXiB0B7bXioinwUTBhvGwmofGHoaeI5g8cQhugMHqPa7cOMu0M6TUzKcoUiX1TFdHAvFbNPUj4VineovEpfxs827Df39bPNuC5ukRvefjK7Q/Tt5Dk/NGIW1s8dg0zyVwO2pGaPgTJK60TYCzedkzwWG3oyT4sRRShVK6RWU0v0n43wnGgV+F1aXBwwL0hyfgBXTRlrob3++eTcqtlXh3quH4fnbxurOl6JQ1LXGcLgxjLrWGBSFptxpagqLePD6EXj/vivw6qJLLbts2iI4VTSQ7VT1XkiSgiPNEbgcHNbOHoPpgULUNEbw0GtVWJNkwwJPsKzMSN+8rKwEAk/0/3flYZkcFTTbcSq71+4VRlzC0FNBCDCwjzcFAx/z4noT7J7pZqQjMdQRRFOmIXaOGTVdNsnupvvP8QjgOc4gy8RznCGNk8luMDBY0el0SkKIE8AvANwM4BwA5ng5pZSekemaDgeH4X0zsHn+OEiyAgrgodeqUNcaR8WUERjYx4vvW2JY9vpePR1MyxfnEqx/dikQeX6nbUpBtlcAT4D8TI9tBERbBDdFREa9zqDDjoBn5czRAIDNlTVYOrkYG++4BA2hOAgI1u04aGAbU/9/AdbOHoMBed4uPyxTpdWkSrU8O9uD9++7gtV2MvRYcKRtk8OSCsbyKXsN0k1r7O6UQD4F6RRvY3rppHGmyybZ3XT/jVH7erdXFo1HgaCuXxihFQODFV2JxC0HsBTAfgC/A/Br06uiy6PrweA4AoHnwHMEHoHHzWMH4P5rhiMuKwhGJSx7fS8WTBiMTfNKsWZWAPl+F2hip+5Ic8R2N87BEUN0RIuGPPRaFaqOtKZMu9AWwau3H7BEU9aUB8Bz6HS6BsPpCzsCnkUbd+HOiUOwZcE4AAQPvVYFSaZQKMWSq4bp6StOnsOPrzgPNY1hLP3T54h1QfPqeEi1w9ov093p9E0GhpMBRaH47f98YTvvJqeCMZzZSDfC5nTwmFRcgDWzAvraYFJxQac3WgWes42ICbwNhX8aaZy5Hqcly2h1eQC5HuMGXnfT/acbUdQ2AtlzgYFBRVciZVMBPEApfai7BnO6QJsQH/vrPpQFijCsbwbisoLHX1dFuF+ce4mFfvex6SPRGpMwe+3HWDFtpO2EFYnL6ON3Gna3NEKU239wroGhL3lXzePkdRbCR9/ch0enjUTfTDe+ORbCL///z3U9OUZw0rtgR8BT0xiBJCuYunonCnM8eHLGKAgch4deq8LtPzjXUNC+auZovLrrMGoaI3jsr/vw4PUjOk1s0h44jmBIvl+PbDuSdBcZGHoyJErxVlUt6lrjhnk71y8g18vSvHoL0o2wdTdNPqUUHiePiikjdDZJj5MHNaXypivx0hITISuKoT9ZUdASEw3EJh25jnQigDwhmFRcgLJAkX4Pba2sto0oMjAwtKErTpwfwM7uGsjphJZoHD4Xb5nEVpUHoCgKsr1OzPzDh5bi4IopIwxpj/l+FxZMGKzTUjs4AoUCFduqUmrERUUZtS1RiLKCiKigNSqiKSzivAIfXlk0XmeHSqaPB4C5z3+CzfNKAUJOyEKcoedBI+Ax25Kc2H3N97vgdwmYvfYjLJ1cjLtfNha0L9y4C0snF2N/bRDzLhuM6Wt26rb+9KwACjJciMvUYE+AumCISTIIkGDoIyjwu+BIFL2bH+o5HgH764LdxrDWEdgtMLRrYCk7DMeDls72aXWTLsxcmOPBS/NKT5jNdBelPJOj6T6kq6/WEZr8dH4fjuOw8t2vUBYoghc84rKCle9+hYduKDG0S9fJjMRl/PiFTy3XsWleqSFNMt3rUBSKb+pDOFQf1p3CAXleDMzzGa7FLXD46Q+HYl6Szu7TswK2ouDMZhkY2tAVJ+4vAC4D8E43jeW0gCQpqG6M4lhrzKILt3BDJTbecQnqWmO2E6Y3IYS5evsBPDljFCJx2RCt+92NFyHDzWP51BLD8RXTRuLh1/eiMMcDUVZQse1fuHX8IAOl9fKpJTj/rAzUByWEYpJ9pE9ScOuzHzH5gV6CfJ8Tq8oDWGiqiXvm719jVFE27rlqGJrCcdQ0RpDtEWxtJtsj4L+uPR8/3fSZwdbnra/EC3MvwYxnPtT7fv62sYhJikUqYN2Og1g8cSiGJ2zNXDuyZlYAv//bl8fdJe5upKpjcTk43MLuE4Y0QAiwrKzEMBcvKytBc0SEwHPomxAq7i50F6V8d1PT93Zkux2WuXZVeQDZbuMSK11nKt3fJ8/nxM/+fdhxRcHTdTKVFMQm5mqMWIrriJmuoykSx/ctUUOGx/KpJcj2CshNEvwWZao7cFpf89ZXJtL+276TdBxCBobehA4l7RNCztVeAJ4AcDMh5L8JIRcnv5fU5oxDbTCGhRsqUzJB1rXGUB+K21LhhhM0vZ9WNyEYlSximT/d9Bm+a4rpBcOb5pWiYsoIKJSiLhjDUzNG4+HXv0BZoMhCab1kyx6EYrKeh293/m/rw0x+oBehKSph22c1WHfbWLy3ZAJenFuK13YfxubKGiyYMBj3bd2j24oWHU6GtmnQx2+vIadQGOzpUH3YViqgLFCEBRsqURuM2ab1zF9fibJAkaX/E82qmirF6BC7TxjSBKXQCYE2zSvF0snFWLfjoLqgFbvffruL3bC7WRJ7O+pCcTzx9pcGO3ji7S9RZ/o+SSJym4zCHA+IiQQn3d8nORX970smYPP8cRiS77c4NekyO/KcvXQAb44AArbtzAvK5I1q7TqWbNljkSyIpdCdiyfVYic7hBqD5fctUTRFmM0y9F50tPL6K6hEJvsB7AAwBMCDAD5MOp78OuOg1RmlWvRmeQS8XfW9pdB9VXkAOb42LS63YO8E9vE79RTLpoiIx9/ej76ZbiydXAyOqHpdqaImUkLA2Y7gZNXM0Xj87f2Wz0REmZGenKGISzI++qYJikIx8w8fYvGLn+KyYX1VxtOEDWm2srWy2mIzK2eOxrB+GXDw9gsP3rTwSLWxoZ1LkpXjygmMKsrGmlkBbFkwDoSQE2qbqcaiRcyTjzGZDgY7CDzB4olDUbGtCjc+/QEqtlXh1vGDsLWy+oREB1LZbCQupaS170g/zM47B1FW8FZVLeavr8SNT3+A+esr8VZVLSTZSMzBJyK3ZvIyc+1XRyJ2++uCmL5mJy5bvh3T1+zE/rqgxQ6Snb332nH2CKG24zMTrXIpJAbM/cmUIt/vMhC55PtdFskCTew7GYU5RjbudB1CBobehI6mU845IaM4jaDVGWmL3+Q0mlXlATSFRcyfMBgtkTh+d+NFyPE64XRwUCjFCx98g5fmleJocxRZCXHl5Il6UnEBKNpq4rSJsT4Yx/z1lXhx7iVYO3sMCjJctqkR2tg+rW7Co2+q0bzz8v2oC8aQ7RVQFzQKgBbmeHCgNohQTGJpNGcgPE4e9149DM2JesqaxohuFwUZLr2Q3OvkseSq4SAEeGFuKeqDMdS2xvDkO/vxi2vOh4MnlhTf5VNLABifxJrgvV09Z2GOBw6eS1mnp43n9h+cq9fmnegUr1QpRmHTooDJdDCkgihTNAQjehp9fSiOdTsO4q6JQ+HupPBxe0hls18cbUXFtqq075d00+sY0oMzxbxmZonkOM5WysVcw5bu75MuYYnm7B0vPVNRYDu+B667wHDemKTg1V2HsXb2GPAcgaxQPPP3r3HnlecZ2vkSzyDzs8PnNH4vAmf/jBGSxiYrKTTs2CY0Qy8GMbMY9XZcfPHF9JNPPrF9T5IUfNcSQUtEwoINlcj3u7B44hAMyPOC5wicDoJQTAFH1FoJSVYw57lPDE5eQYYTX9WGsG7HQb2uTevn3Hwfvq4L4fG39+v6coU5Hrw4txQbdh7EtSP7Y9HGXcj3uywT4/O3jYXf5UA4LuPgMbWPumAMz982Fh6Bw6GGCPIzXPi2Pqy/99SM0Xjwz/9CXTB2wuuPuhnM2zTBzm5rW6L40aodWDq5WN8YGFWUjQUTBmNAnhcegcdDr1Xhrapa3T63f/E9VvytLWK78xdXQN1MpoiKVCfS6Z/jRobbgWmrP9Bt8MkZoxATFYMTlk5N3DO3XIzz+vhQG4zphDzaOPN8Tpyd7UG/THe3O3InuSaO2awN2ptvTwc0haNoicqQZAqOEPAcgcADa//3IOb84Fz0y/Icv5MOwM5ml5WV6CzGhTmetObyDtTEMbu1gdlu60NR1DRE0BAS9XqtXJ+AwlwP8nxtdZHpfu+KQnG4KYyYRMERQKGAy0HQP9traHe4MYxLl71rGd/7912B/jle/f91rTHcsPJ9i1NotpWGUAzVDWHLdRTleg01bPXBKL78Pmhxuob29SPP33a93zVFdEKs5PNunj8OZ2e33RuhaBTftYioaYjo5y3M9eDsTAE+t9rfd41hPPiXf1kYLB+87gKcnXStCTC7tUF78+3A+1877ue/efja7h4SgxEdttuuiH2/A2ARpXSvzXtDAaymlF7Z2f57ImqDMcx45kNsvGMsNt5xCZrCIvxuBwSegBDgWGscCzfuanPaZo7G8qkleOQN9QG7cEOlmru+73ssuWo4HDzBywvGoT4YN7BcJj+U1dojilnjB+HXf/mXHlF55I19qJgyAoMLfPC7HPi+JYZbnt3RRhaRSN90cBzqWmO4J2lhvWrmaARjEjgC3VlkaTRnHiKijHy/Cz4nj3W3jUVDMJoCp1YAACAASURBVA6FUouTle1xYmJxX8REGT8KFKK6MYL9tUH8asoFqA+KFhrp8wp8eOLtr/DjK8/DC3NL0RwR4Xfy+Pnm3QCAiikjcF6BHxwBOAI8eP0IAzulltYjygoEnkO+z4mvjoV0Qh6NdCU5yn0iInKpxGMBMEFZhrQgK0BzWLTM+zeMLjxuhKAzTHvJNhuJS/jiaKv+rADST4lkwsndC0miiIqKgcBjxbSRkKTUaY3a/Gcnp6IoFM2JzeLkufesTGpo63TwmP9vAzH14nP0iNiWT761ROzSTc/MdKlrhuTrWFMeQKbLKB0gU9imNr6yaLyhXSqZG9GUZtoUUfD+l7W4svgsUEpBCME7VUfw7xecBc0HdvAEd105xHKvOZgOAUMvRlfyPSYAyEzxXgaAy7vQd4+EJCvI97sQlyhCMQnZXgHLXv8C//bIduz/PqRPLkAbRXtUVHDPVcMwqig7Ubem4PqLCjHnuY9xxaPvYe+RVgtV731b92DBhMEA1F2rvUdbcdPTH+DW8YMwqigbgOp8zXnuY/CEQFKolSxiQyUO1YfREpUw39S/Nq7vmqP6OVgazZkHl4PDvVcPw/2v/BMTV7yHYEyyyAhotlaxrQpTV+/EjU9/gFvGD8QjUy80bC5o7RdsqERUVLDj63p81xTBgdogvE4egoPDr6dcoEem+2d70D/Hi76ZakrR961R1LXGIEmKXsNxeaKG48vaIB776z6dZEUjXekI6YKiUNS1xnC4Mdyh2iA78VgmKMuQLqKiYjvvZ7gFSypdMtIVX7aDZp8epwMV26p0Bw7o2FzO7Lz7ICnUMrfe/fJuSKbfM7mG7fJ2athqgzHbubfWVBKR7XZgcmI9ceWK9zDnuY8x+aJCCytmuiLjDZG4Zb0wf0MlGkzkIWIKIhJRMjpnQopaN8Fka4QAQ/plYcYzH+Dy5dsx45kPMKRfFpKbSTLFE+/sN5LHvLMfkrnAjoGhF6ErEgOAuSimDYMBBLvYd4+Dtiie89zHqGmMYFJxAe6/5nwsnHAeMj0C8v0uQy756u0H4HXyuPvl3XpKGwExTJLtkUEkR+W0BffDP7oQobisa8sJHLGVFFA1wBw6hby5/wF5Xty9eXdKliqG0x+UAmvfP6jXwPXxu5BvYprU/l46uRirtx/Ap9VNWLChEhVTRiDba0+gIzg4bLj9EiiU4lB9GPds3o26YAwrpo3Eix8dws/+fRgAe0ro8wr8thsO2vmXlZXA5eA6RLrA6NIZThU0Mikt/Veb+wmBRXAZUFPya4MxiLKCo81R/X5MVcvUHjTGwePRyzOceKRbr1UfiuOxv+4zrBMe++s+PHRDieF3TzeCVReK67IGWhst4yc5XTHb7cCSq4ejpkFt5+Q5LLl6uMXZiyWyN8zrmJhoImhJOGfmNEkzi6WD52xr3RymDY5kltdUtXiEg0VaaVlZCUj3l54yMJw26JATRwiZgzZyEwrgaUJIq6mZB8AIAG93fXg9C7JC9cloVFE2bh0/CLc8+xHy/S48fvNFtgW8/bLcyPe7kOdzYllZCVqioj5Jnp3lRpbXiS0LxqE+FNcX0YU5HvTP8WDp5GJLqkxRrhdf14Xw8Ot7UReMYdXM0QjZEEosnjhEF2u2m2w9Ao8nZ4xiaTRnMgi11RPU0nsB1Rb21wZRsa1K1yP8tLoJRbmelCQkDo7gpqc/MDxIH31zH+5+eTce/tGF+mKU52DRCFp321jbxUmez6kT8jySYD1Ll3TBXNyf73fhaHMUPhcPj+Bg9s1wwuDgCCYVF1juszXlAeRnGJ0pSVKw9/vWdlPnO5LWzlIiew4cKZwahyVNUrGQN62YNhKKonSqv3SdvcaIiHBMMhwLxyQ0RkQUCG3zqpPnbNcxTlPKIgFsnTOz5cVlewKUn/xwiKEdR4B5lw3W9UgLc1TdXC7JQVMUWDI07tu6RxUiZ2DopejoHoYCQE68iOn/2qsewCoAt3ffMHsGpCQhzOSUrwUTBuNwU9Q2R/xIcxT3Xj0MhTkePPrmPrRERNx79TBsraxGS1TCjGc+wNTVO1GxrQr3XDUMk4oLsKysBEeaorapMnuPtmLpnz7HPVcNQ77fhYUbd8HlIBZa4HPyvKhptJcbeOaWi9E3083SaM5w2D30lmzZg8UT1Qeotohcvf2Anv5z96ShKMzxoLohgpaohCdnjLLQTWtsl1qfWkpmTWME/bLc+mLUjhL62/qwbXpNrs+pM6suf3MvVs4cbbHZVBGG5HoPrZ5u6Z8+x2WPbO9QmhoDQ0fhdXL4r2uLLffZ/A2VEOXjp8iZU+c7mtbOUiJ7BlwCZ5mzVs4cDZdgXGIpFLZpl+bpyeVI0Z+J8dSRIl3R4jxSinBcNmisheMyFFO0WEraqNbGt2TLHktaaFRSDHq2SycX45E39iFmSqd08hxuGN3fkO55w+j+NqnGRHfgtPP+dNNnAG27DinNaCcDQ29ChyJxlNJ1ANYBACHkXQAL7YhNzlS4hTba32SttmyPWvRrN8EQAPds2YMNt48FABRkutEYimPJVcP1tEyt7X1b92Dt7DG4d8seAMDq8oDtrq3WdvnUErREJeT4XMjyULw09xLEZQoHT9AUFrF29hh4nTzcAod1t40Fgfpw6JvR/Ux/DD0PqVJ8inI9eOfuy7G/NmiI9Ob7XRiQ58O628biSFMET7z9JX55bTFemFsKArVwf8sn32L0wDxLn1r6r9vBYe3sMZCpytZnTt98/O39FrteNXM0Vm8/gPW3jUVtawxNEREbdh7SU2sKczw4K8uT0maT6bhT1dOdZuyrDKcJInEFlNrP/eaFb6qoiWbjnU2F7AxBCkP3IioqeG23MeK05ZNvccv4QYZ2KSNnpkicQimyvQKemzNWZ6d0OojF6RJ4gpUzR2NREtnHypmjIZgiZ5JC9dR6LV1x7fsH8d8m6YB000IdHEFdMIb56yv1Y3bplKKs2DqF5uhZPI2IYqropPmcDAy9CZ2uiaOUXtGdAzkd0MenilbOX1+pa1/VNKrC36l0YpoSUQtKgXuuGoZbE9TlWxaMs520GkJxPaVSoRTP3zYWPEew18RCptW8JaczaHTu9149HDxHsPRPn+tyBD9+4VNWL9TL4EhhkwfqQnDynC47AKgRrHuvHoabnzGmSTZHVOY9zbbumjgU2z6rMZxH01ZbO/tiNEclQ/qkxoSqpXDWBWM4K9uFVxaORygu45tjIQRjEnZ8XY+JxX0NY9pcWaPTYLdnq8m1QcmbKxo6mqbGwJAuRIVCodReI8xks6nSk/snbLwzzpddPeiaWQH08TnBcRxz6E4SOA64bFhffWNWmz/N3z2XwhHhiNlWCJqjioVyP8NljNTKFLbO4+wfnGtoxxP7ejIzsWOqZ4a5ho0jwIppIy1poWZTS9cpTFVjl/z95fuclg3A1eUB5LMaUIZejE6XhBJCbmnnVU4IuY4QUtidgz3V4DiCPJ9Tr2d7aoaa7rB6+wHk+gQsn2pMW1wxbSRWbz+gRiicPOKSghXTRmLNrABEWbFNg9Ccw1XlAcREBS1REQJPLKmVWs2bOZJXFihCdUME89dX6pEJ807Y8Zj+GM4MOBICquZ0yK2V1Tg7243Hpo/U31s8cYjFTu7bugfZXtXe1+04iLJAERZuqMSNYwcY+lxdHtBlLjS70/pIZmidVFyA1eUBUAUghKB/pgtD+/oxIM+LNeUBbK2stk39PV50Irk2qDDHY3tfMfZVhhMBB0cQl2TLfWZH3lDgd2F1ecBy7/TLcHc6FdJO7Hn++kp8VtPMUolPIigltvVaZm4bwWZONotaA0BUpJiz9mPMee5j3Pj0B5jz3MeYs/ZjREWTZAGA/xhdiJrGCOpaY6hpjOA/RhdaFnYUKcZnqmLjCGzHZzZNnufgFjhUTBmBTfNKUTFlBNwCB95k83yiZtTMimmOnqXzvRBCbM9JCNukYOi96Ao75XNoY6dMvouSjymEkE0A5lBKT3uvQVEoYpJiEE5eO3uMLojpckh4aW4pZEpxpDmqk4+snX0xGkJxi4bMkzNG4c6kCNma8gAyPQ48f9tYhGISGsJxbK2sxm9/dCGev20sGkJx1IfUY1rNWzK01BztbwAsMtGLERFlPPLGPrw4txQNoTgyPaqm4c1jB+CP/ziImaXnoGLKCHidPPJMaY+AaiffNUVQsa0Ky8pKkOl2oKYxAgdHsOH2S9AYjiPP78Jvtv0Lb1XVYtO8Uts+NIbWF+aWYuPOgxg9MA9bK6uxeOJQfVd1UnEBfnltMVwODpvnjwOltEOpYVptkKLQk87Yx9LZei/cAoe4zOORNz43pKo98sY+/P6mi1DXGtPtweHgMLxvBjbPHwdJVuBIaIQ5HJ2n10ul/6XN+yyV+OSAUvuIk5mhVHBwyHQ79Hk3HJeR6XZAMNlAOumFgFqLJ8nUkv1grcVLb3zxpFo3sy0nQ5apnt2joTDHg5fnjzNebwptN8FhcuIcBH0yXIbvpU+Gy9CuLhTH7LUfW85pZuJkYOhN6IoTdymAjQD+AmALgO8B9AUwHcBkAIugslT+CsAhAP/ZpZH2ANSH4njotSo9pP9pdROaIyKmrt6JUUXZeGRqCW565gOMPzcPcy87Fyumj4RCAZ+Lx94jrXji5lHwuxyIijKOBePwOHm8NK8UskLRGIqjJSpCUihyfU4IPEGWR8Bv/mMEjjTHLCkETr4t/UCjty7M8SDLI0BWKF6aV4plr+81pH1qKMzxgBACRaFssXkGQ0jULUiKgo0fHMKdE4dAlBUMLvBj7mWDEI7LGJDnVReYKdJZtHTg+7buwQtzSzGpuABxWYFH4JHtFcARYMnVw7HkquFwOjhMKi5AWaBIXwBsrazW+6htieKyYX2R6XagLFCEx9/+0rBY+M1rVfjNDRdCADFsMqTjJCW3OSvLpYvpOjgCj5PD0ZZohx3DdMDkDXo3YqKC5rBoWx9ECMENK9832IPDwXVqwZnqHkiuB00+d1NEBGC/Ycc2HbofgsM+DdHsnGW6BDQIxv1sl8BbxLTTSS8EgHBcwTabdMpbLz0X2d62dunWk/EcwdiB2Ti3jw88R5Drc2LswGxLu1ROZtzkZEoKbHUUN5ucvUhcQWs4hqF9/ZAUCgdH8H1zGF6BB3xqG1FW9LVVMtOl2bFlYOhN6IoTdw+Alyilyc7ZlwD+kZAdmEcpvYEQkglgJs4AJy4uyahrjcPv4rHxjkt0J0hj1WsIqZpsmytrsLlSrRsaVZSNiv8YYdgpW1ZWghc/OoTFE4fi1coa/PO7Ztx55RDb+rb7rznfVvTz0Wkjsao8gCfe/hK3jh+EdTsO4tbxg/R0tsIcDx6bPhIvf1JjoQJeVlaCB//8OX7yw6GsduIMhoPnsKZ8NPwuB2aNG4AZSfVuq8sD4AjBLYkazUnFBZYCeY1IB4DuhC2eOBQAcKNJYmDdjoP4+aShll3XlTNHY8POQyjM8SAqyhBligy3gAu8ArJNVNvLykogSoqh7+dvG4uYpLTrJCU7UloNaLK9r5w5Gk++sx9vVdV2u5Nll87Goh+9Bw6ewOfiLXPsmlkBuATSLfbQ3kaBnVZc8n1rTiU23yuLJw7BoD4+eF08+vgYu2Vn4XIQ23otlyni1BCJ45E39qIsUAQveMRlBY+8sRe/ueFCFGS49XbOFPpqTt7MTglcO7K/oRZv5czRMAd3HZw9AYqZxdLt5HTxcD1yVh6A22nPimmpnTP1J6Vw9iST4+V1chAEwTD3ryoPwJt0Xq+Txx2XDdKzO8JxGXdcNgheJ0uVZ+i96IoT9+8AVqZ47x0Adyb+/juAe7twnh4Dj5PHvVcPQ/kfP7JNi7SLei2eOMSWVnrp5GIs2FCJtbPHYOhZmfrkqrVZt+Mgllw1HIBRiFl7nwB44u0v8d/XXYCbnv4ASydbaa5/tnk31t82FseCcWyaV4ojzVHUh+I6QUrVkVZdhJxFD85EUCgUCMdly26othGgHXurqhYAsHb2GHCE4NuGsIFIpzDHg/pQHBXbqlAxZYStPR9piumbFdp7izbuQsWUESgLFMIlcHr68NrZYyxt79u6B8/NGWs4dqg+bGk39/lP8Oc7L4WsqBsrhBA89leVtXXp5GJLbd+ihF7iW1W13e5kpUpnY+nKvQOSTDHnuU8MAsnhuIw+PidESU1V66o9HG+jIFkrTlYofvNalU6OZU4l1vrK97twz1XDDEQX7BnQeUTjCmRFMaQDyoqCaFxBVlLgVVYULLriPDSG1Eipk+ew6IrzrDT5hKKP32lML/Q7YRZiE2VqWTss2rjLwv4YlRQ8+c5+Q+bDk+/st7BTRuOKrXj4pnmlekQMUFmun519MQ43RvXx9c9xWyQQUhK52EQUU503J3FeRaGoDxrLUpZPLUGW2xjFZGDoTeiK1n0cQCDFe4HE+9o5Ql04T4+BnYbK3S/vhiRTrL99LIb3y8CLc0vx8vxxGFWUjcIcDwb2SV27VtMYgdPB4eyEtpYGTUhc01bRNORGFWUDaNPVuj3BQJXvd6WsfattjWHamp040hzF1NU7MX99pcEZTK6dYGQnZxZEmWLb7sMA7CnQ+/iNdWJvVdWiIRTHPS/vRn6GC3XBGACrnpx551OzI6+Ttz3PoD4+OB0Enx1qwAtzS7F9yQQM6evHjYFCS9vWqGg4Ztdnvt+FI01R3LDyfVy67F1MX7MTt44fhFFF2SnvA61WVPt/dzlZWjpbMhiRSu+BmGDf+7S6CfPXV+okFFFJ0SUGumoPx9soSNaKK8zx4qEbSvD+fVfg1UWXWpwyra9UUhzsGdA5iArFU+9+pacTxmUFT737FUTFTERC4BE4FOWqun5FuV54BM6yEBMlii2fVKMwx4P8DBcKczzY8kk1RMmcrmhf62aWt5AVireqanUbnb++Em9V1VpIb9LtT6EUokQNunOiRC0SCHbkWsunllgjdmmcV0yhYWf+jhkYehO6Eol7GcCvCCEy1Jq4WgAFAKYBeBDAs4l2FwHY14Xz9BiIkn1qQH6GC1/XhfD42/tRF4xh+dQSLJ9WAp/LAUWxp5/WonZf14WQ63NiUnGBHg2xe8Bq0Y6KbVVYO3sM6kNxEAD7vw/iweuLVQpim/Pk+pzYNK8UWQktoo7UTjCc3hB4ghmlA8AT+91Ql4PHqKJsQ7StKaLW9xRkurBpXilikqLXHmi7++G40U60z6WS2Th4LIQL+mfAwfOGlM5V5eoe0Iq/7W/rJ2x04uzsevHEIZhvE91eO3sMvm+Jtmvn2v+7y8myS2c70UQqDD0HqWqXeI7o911X7SFV3ZudDWsO3fH6YoRX3QuOwDbCZqbw5ziCuESxcGNbNs+qmaOtUgQEuOmSc+DgeMiUQuA53HTJOeBs0iTTSWtMJW9hZlBNtz9ZgWUOnr+h0kJsQgDk+ox6d7IimwOKaZ03lVwBY19l6M3oihP3cwAZAB5JvJLxAoC7E39/DmBnF87TY5DqYbr3aKvO4Pfom/uwZMsePDb9IrgcPBw8bGvS1u04qLevC8aw8Y5LAABlgSIMKfBbInMLJgzGsH4ZeGleKZrDarQkOaXA53Lo/ZYFipDncyI/w4VgVMLDr+9FfoYTq8oDesrC8WonGE5/8ISgOSIhJiq29RVHm6O49+pheOQN1QZXTBuJP/7v11g7ZwxqW2IWofmmSByLJw5FpsehC8mrTpYbTWEJGR4ev7/pIvzkpc8M6cYPv74XT8wYZZsu8+LcUqz42359setKIggozPHgvAKfXmsy/tw8zLt8MASe2D7MmyMi/G4HHps+Ej/b3HZ/aDVxQPqyBekiWd6AEUX0PmjU6NbaJQKOEPz5zkuR7emaPXTnRoHW19Fm+80O9gzoHByEIBKXLal+vIn+Pi4ptkQf5vRHl8ChPiRiwYa22rTV5QFkeYypg14nZ3mum2vJAMDBA2tnX4yapPTHwhw3zD+3W+Bsa+fcgpU9MzmFuCkiYvX2AzbsmQRiELh93UeG63AJplo8gcOqmaMtLJbJ5021SSjwXUkoY2A4vUHMFLMd7oCQoQAuAXAWgCMAPqSUftkNYzsluPjii+knn3xi+56iUHxxpEXfgUp2hLQoxdrZYxCMScjzOxGXVBY/MUEnTaCKc8qyDMHBo641htrWGFZvP4BHp5UgKiqYv6FSj7jVNKrMk1rtglaIfk6eF0eaIljx1pf6eV+cWwoHDzQERcP4lk8tQY5PwLHWODLcAnJ9Tn1HrCJBDX8a1kOcFoM8mbCz28ONYfzqL//CnEsHoY/fiZrGKPr4nfA4HeA54HBjBM/v/Ab/PbkYUUllBBN4Ao4jmLZ6p+FhOam4AA9cdwEUqtbZPfRalW47mvO3eOJQ+Fw8vjkWRrZXQIZbgFsgiCX6vnz5dsu431syAU1hEWdluQGiRrsJIeAJwHEcKCh++eo/8eD1F6A+JGKh6f7QUJjj0Y+/snC8oS8nTxCX6Qlhp+wAmM3aoL35tqdDUSjqWqMIxiUIPAcClTGP51SdHSfPoTEsWkh4OsoMqSgU39SHcKg+rC/AB+R5MTDP1yk7VhSKpkgcR5qihmdFimcAs1sbmO22pjGMmxKkHBoKczx4aV4pCnPaaCIP1Yds58G/L5mAc/Lais6+awzjwb/8y8L0++B1F+DspP4ON4bx/I6DmHrxOQZ2ylvGD0L/pHbfN0dQ0xQxbLD9/qaLUJjtQd+kor1D9SE89c5XFgbIO688zzC+o80RfN8SRUNI1G0y1yegb6Yb/ZL6O9wY1slKkr+XTfNKDeM70hTGsWAcx4JxQw1gH78TZyVoNo82R3DY5hr6Z3sM50yA2a0N2ptvB97/2nE//83D13b3kBiM6LDddiUSBwBIOGynrdPWEaipKk6snT0GTgeHuKToaWZAWzRg6uqdmP9vA3HdRYW6rklhjqoD53XxCMUVLHjWyCaV6RFw69odqGmMYPX2A1hWVoL7tu7RUyvtCtGTHciYJKPqSMiwuNVyxp+/bSzuf+WfBsduaD8/HrqhBA9cJ8Pj5CEpFEeaIyyScAaBEODW8YOwZItqP0/OGIWGUByz17btiv7uxosQEWXMee4Tww5ofkI3blRRNu69ehj8biNz2MqZo3HXlUPwXXMUf/zfr7HkquFojohwCzxe/+cRTBnVH3e9+Km+8TCkrz9luoxLIDpFe/KCsm+mC61RGWWBIlAKfbc5+f4w3ws1jaqWUvICgYGhu6GxPOb5BAg8h6awaIheaAvkZAISO5bJNeUBnJXtRqZLQGNEtHXu6kNxPPz6FwZGw4df/wIP3VDSKXIejiPI9bmQ7XGyCHI3QVaobWTKnOqXLtW/Nneb5zizrjVHgBsChbrgNSEENwQKLeLckkKx5r0DhvGtee8AHjARmzh5Dju+rtfZtbXx/ezfh5rOSxC2iTxypgGm0qcz184p1ChFoJ03OULJESDLIxjIXrI8guVaGRh6E7rsxBFC+gE4B4Db/B6l9O9d7b8nQZIU1LbGLWlm+2uDekQsKsrYcPtYDOrjM+xAaTnj624ba2GrXLRxFzbecYm+cAbUCWv9bWPBcWrqmB37ZHKd3NHmKPJ8TtsJU5M+0P6/ZMsefSeM6VyduaAUus3UNEZAYdXs+emmzyxskwsTjJKPv70f91w1DIQQSyqkxviopRErlGLq6p2G9MXkjYd8vwtPzRhl2LktzPXA4+RAIRh2sTWShYopIzDnuY9RmOPB+tvbWCs/rW7Co2+qgrTD+2Vg79FWQzScpYQxnGhoLI9bF4yDZMMQ+JOXPsNL80oNdWZ2LJPzEyyxfpfD8FxZMyugy78oimK7oFeU9PWxkiOAgoODg1PT/5wOHmdlWdkCGToGj4PDf/6f4YY07semj4TbzNZIgBXTRhqkVVZMG2mpdVOS5m6g7XlvTrsUeA6iRHGHqcbOnGKYyik0/+x8ivRgO504O5KRl0zj41LUY5udPZmmcIKTnD1JQUqxbwaG3opOO3GEkP4ANgC4TDsENYsk+e8zajVVG4zZygWsnT0Gy9/ci9t/cC5yfQJqGqMp2ZZ4Yl/PU9caw33XDMey1/finqvadK7Wzh7TbiF6ns+JNeUB/PmzGtw4doDthBkVZcvn5MTkmEw5rU2gR5uj6JvpQq6P6VydzlBMD8ZUheF2bJMD+/iweOKQBO3/GNvPaTZ531b14b1pXqlOXf2La86HpFBdbyjf74IoU8PO7ZpZAQRjMmS5/XHl+12QFRhs+9PqJlRsq8KLc0v16LMW2eA5MCF7hi5BkhTUBmMQZQUCz6HA74IjaUGusTyKCjVskmmoaYxATpBaaZsKqVgm+/idhsVpvt+F2pYYXA4O1Q0RnH9Whu2C/qV5pahrjR03giZJCvbVthoi3cunlui1sGzTruuQKNUdOKBN4mfzfKNTwxOCs7JdeHFuKWRKwRMCCgW8KYsq1VxtliKIpVljR9N0CqOijE8OqizCCqXgCMGfdtWgv0mgPt3xEQLbrAlzRNHNc/jl5PMtqZKuJGdUTFNzjoGhN6ErkbhVAEZA1YD7J4BYt4yoB0MTrtSIRgoyXPC7HPC7efz3dRdAVihkheLFjw5h6eQLUqZN2B2vD8UxvF8GHr95FFoiIh7+0YXqLpus6Hosdp87K8sNB0cw+wfnglKKNeUBS02c12llIeQJQX0wirik2KZqrikPdLkgn+HUwuXgdeHr8efmYfEPh+C1xT+Ak+cQjEmobY1ha2W1zjap2XWezwlKKYb1Uwl2HJx9QXkys6msUDz8+l7UBWNYVlaCYExCTGorfj8v349vG8LI97uQ73dhwYTBiMRlZLoFCDxs+8/1OfHqovHI8gjY9NEhy2JAc9hemlcKjqhaQ0+/dwA7vq43LEw7U4eUClpfiqJApjjVdXZnLERRRm0wBkLUBSjHAYqiLh45jsAtcMj1nhhxaklSsPf7VkNkbFV5AEU5bmS6xflXjQAAIABJREFU1d9ZY/uTFYqoaM8M7OQ5nYBESTwbbJ8JSRt7yTXQbZGVgCFLA1DvucONqsRNe06YolB81xzRHTjtHhd4Do9OH4nalihCMQlHW6Lol+lmNtxJSCk2oszU/DxPEIsqqGkMGwhGfE7jUoznCCYVF1hq4swRsXSdqXTbeZ08rht1NuKSAo6ozul1o862bPQJKdYxgml8lAL7jzYbnMJ3qo5gYJ4x3V0G8NbnR7B29hhDbd/shIyS9p3YRvWYzTL0YnTFifs3AIsppeu7azDdBUJIEYDHoAqSEwB/A/BTSum3ne1TFGV9Yk1OS5hUXIA7rxxiqIf43Y0XQbBhpVw+tQQNoZjOtpe8M6WxSm6trMadVw4x1LD97saLMLIo08JCtXLmaBwLxvDjFz7FY9NH4v/7H5WF8vnbxiIuyeA5dbHeEBLxn//nfExbs1M/n0Z4UZjj1SMu5jSfl+ePg8IWqactJFnB2vcP4qkZo8BxHH79l39ZUmpWzRyNLK9gsWvNviYVF4AQe4bVZGbTr+tCuOeqYXj0zX16dDoYk/Dg9cVoCIk4ltCc+23ZCDSHJUs60VMzRuHHCSFw7dzL39yrk6csKyvBnz49jKWTi5Hnc6JvpttAzLNy5mi8tvswpozqj/21Qb0WKc/n7LZ0YS31+LG/7rN8Vyya0X0QRRl7a4N44u0vcdeVQxCKy8jPcOHb+rAu47KmPABCgWBchsAROHgOEVGG28FBphSiTBPHCUSZwuvkkOFyGqJpqWCXcbFwQyWemzMWR5pjOK+PDy1RUd8gy3A7LPfHmlkB8AS6TdS1xvCb16psU+mOJsli2MnLLNxYicemX4SGcNywoNciE4/9dR8evH4EKKUGUqA8nxPHQip5lubAmR3E5VNL8NBrX7CIXBeRUmrCFHKSZIqWqGTISPj9TRch0yRYneHhcNfEoRbWyQyP0X4FnrN19szplOlKBwBAa0SysET6TU6mS+As6xiVddJ4Xq+TQ2BQH4u0jJk9kyfAtSP765kb2pyeLNGQignW7DgyMPQmdJqdkhDyHYA5lNI3u3dIXQMhxAtgN9TI4C+hpnX+BoAXQAmltF3hcTv2Ho2V0i1wIITglmc/0ifDNbMCtkx5L8wthZNXd5REWRXBdAkEB+vCGFzgQ0ykOBaMoT4Ux9bKatw6fhAefXMfFkwYbNvfpnml+FWCrersLDfcgrqL53c7cM/m3agLxrB0cjHmr6/USVUMNRblAWR5HYiKCo42R7HirS9RF4zhsekXoU+GE1c8+p7lu/j7kgn44mgrtlZW4/5rzoff7YAoKT3BqWOztgl2dvtdUxgHj4XRN9ONR974AkuuGq4/JDUU5njw7K0Xw+ty2LKrbbzjEgRjEkRZQWNIRLZXQJZHwMOvf2FwsDSpjId/dCFCcRlDC/xwJKQAkh+6Wr2cpomonefhH10IUVajf7ICNITiONoSxertB/RaN82+AWDLgnGYurpNuWRScQF+OfkC1LZEkeURcO+WPVhdPjpB2BNFfShu6Esjm7BDqshdXWsMN6x8PyU7Znt9gtmsLVKxqt749Ad44uaLEBUV282DumAMz982Fncn5r7lU0vw6q7DuGF0f739pOIC/Ne1xSBETWPTonp+N4dgVBXjdjlU1uCopMDBEbidHCJxBTxHsOXjal3DEADeuftyPPz6F3jgugtw49MfYPy5efjpD4dgeuJvjdFPoUCOx4FgXMaABKPf4cYw7nzhUzxwfTEak+pCc3wCPjvUgMCgPli4oRIrpo3EjU9/YPg+RhVlo+I/Rhjm85UzR4NSigf/XGXLXtwaEdE304W4THUJnPZYXeevr0xlw8xubWC22++awqhtjVl+24IMF87O9hraTV9jnWc3zy81tmsMY7rNfLx5XqmBnfJYaxTfNUctkgBnZ7nRJ6ONpqAlEkVtq4jqhog+vqJcDwoyBGR62tppjMZmp/CB6y4wkEUdbgxj16F6jBqQl2BkJfj0UD1GD8iztEu3v+OxWNa2RBARZYgydIZtgQc8Ao+CTMZOmQ4YO2WPR4fttitO3K8ADKaUlneqgxMEQshPAPxfAMMopV8ljg0CsB/AvZTS/9ve5+2MXFu8bbzjEgg8QTxBXZ7h5hGJq4sBniMQZRm1LXEolGJAnheirDpw2oTjc3FQFJUpSuBVMViZUlAKdfEQkyFT4PLl23VWwH6ZbsiUwiPweH9/HYryfMhwC5YdsHy/E26BIBRT07wqttlPnNpYBY5AoRQyBVwOThd15jmiFr2LMpw8B44DeMLhWDCGeeuNhfe5XgFKItWJUnUHmGifj3c8da0DaW9sgjYh1WI428OhKaLaqIMjhv/zHFG11VwOODiC1qiIQw0R3dkBgH/cOwEcIfrn3U4OMVFBXKIgRN3hdTkIopKaLuZ2qNHflqiE/AyXrWOY7Ixp+NvPL8PT732NW8YPtNh2nt+JlkgcWR4BUiKljucIXq2swYq/7ceoomzcf81wQ4RjdXkAmR4HDtSGDLIK9cE4RFnB2dnqLrTHySEqtskP5HgE7K8L2kbujjRHcOmyd7FpXqlloQ0A7993RXusmMxmbWBntxrN+U9+OMR2Ybf+trH4sjaIC87OQCSuMgRfc+FZGJzvA8+pzppCAUlRUP6HNsKHFdNGorYljPPPzrYsZv0uHhxRo3aarQs8ASEEoqRATNicR1DtX5uTo5ICmkgT05xEhcpoiSjom+lCKC6r943AoSUqYdYfP2p3g25ogR+zkjYJRxVlY/m0kQjFJHidPKKijOaICJeDR0GmCySRnjbsrCz43YIeuZlUXID7rzkfFEAoJiHTLUChFFeusG7WJduzjQ0zu7WB2W5rW6LwOWGZa0NxoCCzzUk6VB9CHx9vaVcfkg0U/t82hPDk2zZU/xPPwzm5be1qGsNYbyMxMGu8mmWjoSEURWNIxLdJdn9Orgc5PgG5vrbxfdcUhqRQVS5Dr9lTo9pnJTmZta0R8BxBNK4Yng2yQlGQ4TH0F5cVSEmOl4MHXDxn6O9QfQgfHjiG8UPy9fl9x/46XDK4j74RUtsStbAYr5mlphonf8cJMLu1AXPiejxOqsTAYQCzCCHvAPgfAA3mBpTSZ7vQf2dxPYAPNAcuMY6DhJD3AUyB6uB1CHFJxvhz85Dr43GoXk21uTFQiAnn97WkO7RE4vC7BTSGRMi0jbVsUnGBJT3i9zddBJ4jWPnuV7jzyiHI9jqgKGpUYdEV5yESl/UHurYwdQucoQi+pjGCBRsqUTFlBPpkuPQUJDsmqmBMwjf1YeT5nMj1ORGXZbxaWYPJI/sb0ieWTy2Bx8lj5btf4dbxg+Bz8rjzxU+N6ZbrK/Hwjy7E/a/8U08HvXX8IKzbcRBzLh3U4aJ5xpLZ/cj2cPi6PmZY2JltcPnUEvzy1c/1WratldV6WmRdMAaFAjebUmFyvA6U//FDS5/5fpdeg1fTGMGWBeNs6zD6mR64hTkeVDeoqWTJUW6zbcckBTMTi3ItyvLD4n5wChxutfnc87eNxYsfHcKt4wdZ7PunL32G/Ayn5ftYMyuA3//tS0NfWmqm08HrtYC2NVCMFbNb4HPyKB83AMeC9oQhta0xVGyrwlMzRuN/9hxG+bgBhkiENh/deeUQjD83D5sra1DTGMEf//drPHj9BThUH7ZQo2d5fDjSHDP0s6o8AE/SfJscSc72OFOe966JQ5Ht4SApFJcv366npLmdvO31xGUFb1XV4q2qWowqytZrP7X7KVkSZFlZCZ7f+Q3mXDrIYMPHWmNYsqWt7u3W8YNwy7Mf6X1obLLt1bYyG+48fE4Y5lrNfs7NM0bm+/j4tNr5Xeo9YE4v9LuMv4+TJ7ZpiE7e+MyUFaC2NWax+0yTeLhH4FDTZB1fYbZxfG4HwaEGa7sBuaZ2AodjwbglUliYY3wGZHt4nN8/W9/00/rL9rRdr0Kp7sABbeuQLQsYOyVD70VXpO5XAxgIYAKARwD8wfR6potj6ywuAPC5zfF/ASjuTIdOB495lw9GS0TRowRTRhdaaNcXbqhE8dlZWLJlDzI9goF2uixQZGn/k5c+Q2NIRFmgCIs27gIBh9++/gXuv+Z8NIZEC4Xvgg2VcAv2CwGvk8fCDZUoCxTBLfC2TFRepwMV26owdfVO3PLsR4hLFDNKB1rYrZZs2aOP676te5DttZcu6Jfl1vvW2pYFirBki6pvpy2A60Px437HdvTb6X6WwR5NEcVgc3Y2mPxbJf+OiycOwZryAB56rcpi4wCx7XPBhMEGm60PxVGYY0xz0QhLtOPawvTxt/enZPnTbFuUYVikzvzDh7jm8X+gMcXnGkJx/Xrsrtnu+5i/Xr2HzH3FJRl5PieeueVibK2sxrKyEsM1aAQWDF1HTFKwaOMueJ28rf00RUTUNEbw4xd2YerF51jo/TU7XrRxF+Ze1kaMUBYoQkyyp0aPSVaZgIUbKlHdEDEcW7RxF8oCRZh72bkpz7twQyU4wiMuKW19bdwFj2B/PXyCOAhok8+omDICT8wYZRmreY7VzudNchCT6+qS70lNXzHZbpdPLcHq7QeYDXcR5rlWs5+miNKpdpG4YrGvRRt3IRI3thNt5C0WbdwFUbayWNrbvbG/cNx+fGHTeVuj9u1ao527jmDMvr9grK1dTLJnp4xLjJ2SofeiK5G4Qd02iu5FLoBGm+MNAHLsPkAImQdgHgCcc845lvfzfE6E45JBNiCViKXWxswGlUoiwOvk4YX6AFYoxVtVtVg44TzDQzm5fSqGM21hk+0REIxJKRe15sn0xbmlxx3X8Yq2tfOa/9Xe03SS2kMq+u10PttbcTy7NctcpLLB5N9Ka3Nuvg8CTwy1a1qbZEaz5D7N/acS5Q7FRF32oCDDhZ9v3o1Pq5t0py+VbWsBWTP5Q6rP1YfiaV2z+T3zQlaLUHAcwbC+GXjoBlWja/P8cYydshNI126jopxS1B2APjel+n219zXk+ZztsvSlmgft+j7eebV0ZfM57K4nHBMNx+uCMfTLckNJMSa7OTY5OpzqnkzWVxxS4AcAfN8Sxf3XDEdhjofpxR0H7dltKkkhMzvlqWqXLjtlT74Oxk7ZORxvvmU4vdHpSByl9NDxXt050I4Oz+ZYyjudUvo0pfRiSunF+fn5lve5RJ2XxvAEtIlYJkNjeyrM8UCm1PC+9pA1tw/HZf09rc/a1liCeti+/+VTjbupy8radlObIiJqW2O2nz3aEjUcS3YK2xuXQqnlnMunluj9JaeYJf+rvZdOio6WqmYeB0vvSY3j2W2yvQKpbTD5t9LaODgCSpEickBs+zT3/2l1E9btOIi1s8dg07xSLJ1cjHU7DuJQgxrxuvvl3fiyNqjX322trMbq8kBK29ae56mcxeTPrZw5Glsrq9u95lTvFWS4UkbZOI4gP8OFvlkenJ3tQf8cL/IzTgzV/ZmKdO32u+Yo1u04iKWTi/G3n1+GiikjdFF3QP1tUs1fbXNX27FcnxPHgvbR4VTzuSa/Ye77eOd1JOqOk98jhOjXY74ftOjb3++9Aq8uuhTD+mbAwXPtniPZhpPvgePdkxXbqgAAd2/ejZuf+RB3v7xb36RgSI327NY81wL27I+nqp2Qop2Z2bEnX4dgs/Zh7JTHx/HmW4bTG11JpwQAEEJKCCF3EkIeIIT0Sxw7jxCS0fXhdQqNUKNxZuTAPkKXFrI9HLI8nL7I/NOuGjVXPGlCWVUeQNV3zVg+tQSbPzqEVTNH6+9vray2tP/9TRchxydga2W1yjYGBSumjcTWymrk+ATLhLW6PIBdh+rRx+/Eo9NG4u27L9cXNnXBGFaVB9SF8PYDtp/dWlltuKbCHA+awnHDOLWJURvXsrISvPDBN+iT4ULFlBHYNK8UFVNGwOvk8cgb+/SFttZ2a2V1p1J0tFS1VItnho4j28MZbM7OBpN/K+3306isD9S22No4QG37tLO7uyYOxfI39+LGpz9AxbYqzLl0kH6+VUk2WZjjwV1XDsG7X3yPNeUBvHvPBItta3pyqZzF5+aM1e3T5+KxeOJQ29RH7Zrtvo9nbrkYZ2epLH3v39e2oGYL3JOHPI9Tt41bxw9CxbYqLHl5D9wCh7qEVIVmr1s++TZRY2N0/LXfNibKuk04eII+ficemz7S0P53N14ECsXSz6ryAIpyPbabA8/8/et2z5vt4SA4SFtfM0fjT7tqMOdS9XrM94MWfSvM9uibAgV+l+2mRvIcq52vLhjTHcHiszL0OT3Vs2DTR4d0plY2z3YdfjdnO1f63cYllnlO1tple05su+S1i9ZudXkAWSf4vN3Zzu8ilnVInwwX/C42NzP0XnSFndIFYAOAH0GNclEAYyiluwghrwD4klJ6f7eNNP1xvQPASSn9gen4dqjXe3l7n0/F3qMoFJF4DBIFQjEKUVYM7JQOjsDn4tASkeEWeEiyomsUae97nRxCcQWKQuEws1MKHERZZcnTGB85EIgJkVgh0X9rTIGTJ5BkqrOwyYqqD+R1cYjGVeY0J89BSegl8RyBz8nhu6aYQQh8TYL5j+dU9kFRUlQhW45AAYWiqOM43BjFkaYQRg/IU5k1OaIyU0kKOEL0sXKMnfKUwc5uo1EJMqSU7JSavlY0oYGosYdleTiEYhQUgEsgCMfaPu91cRAlVYNOSxlzOzmdpczt4EABxGUFPFFZ+aJi2/lIwjYUCmR6eIhS2/3hFjiEYupYvC7OcF6Pk4OUYHslBGgMiRZb7uN3IiYrcHAcHBwBzwFRUUliDaRwcJxqu4lrNrNTnsC0SGazNkg130ajEuojccNv53LwBrvTiJp2fVOPGaUDwREYWCK9CbkAKcGyR0hinpUoRIVCSfRDEp8TeKLbarrslLEEOyVJOm+2h0NDWNZlYBxJ9ux2coiL6vND4Dk4BZXhL5XtSZKC2mAMkqzozMEkkcYeFWU4eA75PieaohJikgwC6JIKSuJ78zh5fX4XHBycPEEowYTsFjj08bUbSWZ2awOz3UajElpEEfEESy/PETgdBJmCALfbYWhnNyfzcLB2x2kXi0mQqLWNgzjgclkqg5jd2oCxU/Z4nFR2yocA/BDALAB/BfB90nuvA1gE4KQ7cQD+DOBRQsi5lNKvAYAQMhDApV0ZD8cReJwuBGPqTjCBWgSsUqyrTGQtUfWBrVB1oSnJ6oTucnBQFIrWqLpw9Lt5hOOK/mDWFg8apa/gIKAK9IWGk1cfvK0x1QHkHBxocsYoUel7ZVk96hV41XmU2yQDRBkYku/D5vnjIMrquP1uHjGJIp7QfuubYIyqD8UhSbL+8D8ry40BeT59kaE5Ww6unYWvDx2GlqrG0D1wux2IRgEgDoFX0yMbwjI4oi5GRZmCJ225xz4Xh9aogvqQasd+F4eYZNzkESXVkSrwu8DzHOpDcUTjMgghcDsIHA6CuEjVTgn0mgufk4eDJ/o94+A55HpcFvHlPH/b36nZ+oF+mRSvLro0HYef4TSE2+1A/6RFnjbnUAAepwM5HgFOhyqPcWH/LHVDS1EdlmxPmy3kpDkPKQrFsVAMMkVi08DYjwZRlFErxuBI0LlTqBsfed62zZGmiIJ8r8uwSAXased2xuhwcDg726KBZUG+0LG08+x27i2GjkP7resllYiLABYHTmunzckazA4Na2ffzuVyqOq/SW1SOHAMDL0GXbH+mwH8klL6AiHE/AQ5CJW58lTgGQB3AvgTIUQT+64AUA1gTVc65jiCTI8bVl3JjiO3E05Od+Bs5/F/8uM5UszZOn1gXgzbIS/p747adkftoLsWj8wGexfsfu/u/P05jqAgw6I1ZYEg8Cm1AH3H/zjDGYx05lrWrmvtXC4H+jOnjYFBR1fuhjwAX6R4jwNwSlZYlNIQIeRKAI8BWA91U+xtAD+llAZPxZgYGBgYGBgYGBgYTlewlMueh644cQcBjAPwjs17YwHs60LfXQKl9FsAZafq/AwMDAwMDAwMDAy9CczRO7noihP3PID/JIR8A+CVxDFKCLkCwM8APNi1oTEwMDAwMDAwMDAwnClIx9HrLpzpDmNX2Cl5ABsBTIdabuoCEAHgBvASpXRmdw3yZIIQUgfgVGrcdRR9ABw71YM4wTBf4zFK6dWnajA9EWnYbW+wE6DnXiezWRt0cb7tib/1mTYmZrc2OA3XCcnoiTbaURzvGpjd2uA4dnsm2EV7OB2ur8N222knTu+AkH8DcDWAfAD1AN6glL7XpU4Z0gYh5BNK6cWnehwnEr3hGk80est32Fuuk6Fn/tZsTAw9HWeCPZwJ19DTcKZ/p2fq9XWZ5odS+g8A/+iGsTAwMDAwMDAwMDAwMDAcBx1y4gghCoB0Q3eUUsq4YBkYGBgYGBgYGBgYGLoRHXWyfo30nTiGk4OnT/UATgJ6wzWeaPSW77C3XCdDz/yt2ZgYejrOBHs4E66hp+FM/07PyOvrck0cAwMDAwMDAwMDAwMDw8kDd6oHwMDAwMDAwMDAwMDAwJA+mBPXg0EImUAIoTavJlO7HELIHwghxwghIULI3wghF56qcbcHQkghIeQJQshOQkg4cT0Dbdq5CSHLCSFHCCGRRPvLbNpxhJBfEEK+IYRECSG7CSFM6B0AIaSIELKFENJMCGkhhLxCCDnnVI+rsyCETCWEbCWEHErYxD5CyG8JIRlJbQamuGcoIST7VI6foesghAwlhPyeELKHEBJMzA9/JoSMtGm7PYUd/LSbx3TK7jN2TzAcD+muI3oSunudwGCP03GN0N3r4tPdhhjxyOmBxQA+Tvq/pP1BCCEA/gxgEIC7ADQC+AWAdwkhF1FKa07mQNPAeVC1BSuhsppOStHujwCuBbAEwNcAfgzgTULIOErpZ0ntKgDcA+C/En3eBOBlQshkSun/nJhL6PkghHgBvANVw/FWqLWsv4FqFyWU0tCpHF8ncQ+AbwH8J4AaAKMAPAjgCkLIeEqpktT2t1Dvi2S0noxBMpxQTAJwBYB1AHYByAZwL4APCSGXUkorTe33AJhvOvZNdw2mB9xn7J5gSBcp1xE9EN29TmAwoQfMXV1Fd62LT28bopSyVw99AZgA9cb6YTttpiTaXJF0LAtAA4DHT/U12IyXS/r7jsTYB5rajEwcn5N0zAFgH4A/Jx0rgDoB/cr0+bf/X3tnHi1XUe3h7wcBFCWQEBkMaEAkiMigzDJERWUBMjjymIyIBtGnOD1BBAOIiE8GJxAEiRAEBQRikPmRCDIIIlOAMCWGKZAZQgYh7PfHrs49Ofd09+l7+97uTva3Vq97u7pOnao6u3btqtpVB3iw1WVtcT1/A1gCbJIJ2whXdN9qdf56WKa3FYQdlmTlw+n7sPT9iFbnNz59IgNDSHu5M2Fr4p30RbnwCcDtfZyflrazaBPxqfcpY0e026eZdkJ8qtZxR9oIzbSLlwcZCnfKzmdf4Hkzu7USYGbzgL/ggtxW2LIzw9XYF3gN+GPmuteBy4CPS1otBX8cWBUYm7t+LPA+SRv1Pscdy77AXWb2ZCXAzKYAf6cN5aIMZjajILgyEze0P/MStAYzm2mpp82EzQMepzUy0NJ2Fm0iWB5psp0QFLPc2QgZytrFHS9DMYjrDC6RtETSLEl/yPksvxd4uOCaScA7JL21f7LYVN4LTDGzBbnwSfigbZNMvMXAkwXxADbvsxy2P7XkYnmql93T30dz4adKej35+o8r8oUPlg8kDQa2oLsMAGyTZOC1tI/ui02+fTu2s2gTQRG17IhOpKydEBTTjrqrEZphF3e8DMWeuPZmHnA6MBF4Gd/v8H3gTknbmNlLwGCK93jMTn8HAfP7PqtNZTDuHpVndub3yt+5+Zn5gngrIrXqcFA/56VPkDQUf3flzWZ2bwpeDJwL3AjMADbD28wdkrY3syJDP+hsfgkIOCsX/jfgEnyVbi3czfB8Seub2Y+adO+2amfRJoICytgRnUhZOyEopq10VwM00y7ueBmKQVwbY2b/Av6VCZoo6W/AP/BNnT/AjZeil/2p73PYZ5Qt0/JY9may3NZNmkm7Bvff/0Il3MxeAI7MRL1N0vX4zNpxwCH9mc+gNpL2AG4qEXWimY0ouP5Y4CDgi1m3IAAzOyEX/RpJVwHHSTrLzJo1udUW7SzaRFBESTuiE4n+v/d0XP012S7ueBmKQVyHYWb3SXoc2C4FzaZ4tqAyk1I0y9DuzAaKXD0GZX6v/B0kSbnVuHy8FZE5VJeLTpSJpUh6E37y1MbA7lbnBFYze0bS7XS1maB9uAN4T4l4eXcXJB0J/Bj4gZn9ruT9LgX2B94H3Fk2kzVoi3YWbSJohAI7ohMpaycExbSF7moGvbCLO16GYhDXmWRnDyZRfPzu5sC0Js429yeTgAMkrZ7zVd4c+A9de+AmAasB72LZfXEVf+5H+jqjbcwk3N87z+Z0cL1IWgW4EtgeP53qobKXUjzjFrSQ1L4fa/Q6SYcCZwOnm9kpjVxauXWj96xCy9tZtImgh3T68y9rJwTFtFx3NZme2MUdL0NxsEmHIWlbYFPg7hQ0DhgqafdMnIHAJ+j+TqBOYRywCvCZSoCkAcDngBvNbHEKvh5vaAfnrj8EeDidtLSiMg7YUdLGlYD0stQP0qFyIWklfI/TR4D9zOyukte9Ay/33fXiBu2PpAOAC4Hzzew7DV5+ELAQKDvQqUdL21m0iaAnFNgRnUhZOyEoZrmxEXphF3e8DKn7mRBBuyDpEmAK/lLbufgGzmNx16L3m9nM1InfDmyIv6yw8lLDLYGtzOyZVuS9FpI+nf79CL5X4yh8w/0MM5uY4lyGv0Lgu3gdfAXYB9jZzO7LpPUT4Gh8Y+t9eOMbhRs0f+mXArUhkt4CPIAbrD/AZ6hOBtYAtuzEFVpJ5+DycgowPvfzs2b2rKTT8cmpO3GZGo63hzWBHcxscj9mOWgyknbDD+h4BPgakD2KfHHaL4GkXYFCVlljAAAV/ElEQVRjgD/jG9zXxF9ouy9wjJmd1qT8tLSdRZsI6lHGjmhh9qrSTDsh6E6rdVdPabZd3PEy1OoX1cWn+gcXugfx03heA54BzgPWz8UbDPwO999dgL/seqtW579GuazKZ0ImzpuBM4DpwCJ8hmVEQVor4wro3/gpbA8Cn251Gdvhg/t6X4mf4PQKcDW5F6Z20gc3xqvJzugU53D8PVlz8AMepgN/AIa3Ov/xaYoMjK4hA1Mz8TYBrgOeS3phPr7/7r/6IE8ta2fRJuJTQkZK2RHt9mmmnRCfqnXccTZCWXmmpF3c6TIUK3FBEARBEARBEAQdROyJC4IgCIIgCIIg6CBiEBcEQRAEQRAEQdBBxCAuCIIgCIIgCIKgg4hBXBAEQRAEQRAEQQcRg7ggCIIgCIIgCIIOIgZxQRAEQRAEQRAEHUQM4oIg6Iak0ZLi/SMlkTQs1dnGrc5LFkkTJE1odT6KqMiYpAGtzks7Imlkqp9hrc5LuyPpaEmfbHU+8vRV+5M0RtLUzPdhSVZGNvteQWfSLn140mOHVwk3SZv0Y152knS3pFfTvbduQppbp7oe3Iw8NkoM4oIgKOJ8YKdWZ6KDGAb8EGirQVzQ0VyLt8EXWp2RDuBooO0GcX3IycABrc5EEJRgJNBtENciLgAGAJ/AdevjTUhza7zvb8kgLmZAgyBYiqTVzGyxmT0LPNvMNJuRVhD0F5JWAV43s5bMZpvZDGBGK+7dDGq1+9AJvcPMnmp1HvoDSQJWMbP/tDovQWcjaSVgOHCKmf1fq/PTLGIlrkOQtImkiyVNkbRQ0tOSzpE0qCDuNyRNlbRI0j8k7Zy+j8nF20jSJZJmSFos6X5JMbu3HJBxVXufpFslLZD0gqSTkjJD0ogU55OSfitpBvBi9vpcmgMl/UrS80leJkv6ZupoKZHmppKukvRSks1pki5vxJ1O0lsk/UTSUykP0yVdKWndTJztJd0saX5ym7hF0va5dArdnPLtJOPysWNqKy+n8v9C0psqZQZuTZfclOJbCu83JB0o6bFUL5PybVnSmySdKenhVDfTJf1F0maZOB9Ied+vIP0xkp6VtHL6fpCkf6W05kl6SNKoHmT9PdVktGS5B0g6OcnEIkkzJd0uaZdcvC9Jui/pzzmSJkraOf1WcUc7StJPJT0PLAbWSr+X0pWStpI0LqW/UNLfJe2ai1Opx20k3ZbK/YSkI3PxurlTJvkcm571o0m+782XNcUt1Q/Uqdu6MpPL626pTc8F7s6VdydJd0haCPw091weyDy7C5RzTUpleTTz7O6t1L/crfCdwMGZtle6jM2iXvtLcYbI++3nUrzHJH25IN5G8v5+eor3tKSfZ35fxp2yRp52l+u/V5Ks3CBpiwbLtZ2kmyTNSrL6tKSzG8lvinNI7jlfLGn9XJyKfB8u6THgP8De6bfVJZ0mt4H+k/4epwZ0xYqMXE8em5HR5yWdrtSPpTgVPThKrodfkDQ3tfkNcumtnmR5VpKvq5KOWerWK+9jdwc+mGmbE3JZG6IqfWsDZatpm6T8LMHHPMenfEwtmXZVuyWle2GK+kSmjMPStV+TdKek2ake75K0dybt1eR9ypkF963o1M3yv2WJlbjO4e34ysjRwBzcbev7wF/JuL1JOgI4C182vhx4F/AHkjGSibch3sm+BHwTn/H9HHClpP3NbFwflyfoH64GfgecCnwcOB54AxidifNL4DrgUKBQeaaO8lrg/cAJwEN453oG8DZcFrMUpTkemAt8BZgJDAX2ouRkkqRVgZtw94VTgbuANVO5BgEvStoSmAg8grtxGHAMMFHSjmb2QJl7FXAxcCnusrUTXn9zcDeK+4CvAr8Gvg7ck655pIf3ahhJe+Dt/Frg2/gz+TmwCjA5RVsNWAP4Ee6iNxg4CrhL0mZmNt3M/inpHmAUcE0m/bWAzwI/NbMl8kHDWOAXwHfxZ7gZOT1TkjIyWovv4TrsOOB+YCCwLRn3Fkk/w+vlAvyZvQHsCLwDuCOT1nH48/sysDKwqKyulPR+4DbgX8CXgAXAkcDNknY2s39m7jMQf15nAScBXwDOkTTZzG6lNrviM8rHA4tw17rxkoaZ2dyUl1L9QAnqykwu/iV4O/k0y9oXawKXAT/DdcXClM+f4M+lIkdD0722SHW2RNLBwOl4Pd0GvBnYkq7newDeDz5Al8z06wpmmfYnaSDwdzz/o4EpuLyfI1+Z/GWKtxHwD1x+fgg8AWwIfKzBPO2Nt+FrgUNS8PeA2yRtaWbPlEjjrcANKT8jgVdw1/GdM3Hq5lc+UD0X+CNwLG7P/BjYQdL7zWx+5rYfwnX8iXibmyqf6LsB2ByX94fw9ns8LgffLlcrKzRjcTfC03Cd9x68LocBn8rFPTbFORxYB29/l+ADsgrnAZ/BZfle4CMpTpaj0n1XxvsUgJdzcWr1rXUpaZtcC+wC3I7rxPPxSboy1LJbrsX11Q/wuqh4L1Vc4Iele02ly41zvKS9zOw6M1ss6ULgCEnHmtmizH1HARPN7LGauTOz+HTgJwnELriRuk0KWwl4BvhrLu4nU7wxmbAL8I5u7Vzcm4D7W12++PRaPkanZ35MLvy3eEe8FjAixbmq2vWZ7/ukuCNz8SrKcEj6XpgmMCSF79uLMh1eLw3giqRw18qEDQRmA3/OhE0AJhRcPzXXTkame56YizceeDzzvVLuPVr0vP+ODxpXyoTtkPLUrZzp95WB1ZM8fDNX5iXAOzNhXwdeBzZI378DzO5rGS2Zzvjssy34fZNUnjNqxBmW8nIfoNxvpXQlcAvwKLBqro4fBa7OhI1J9/pQJmw13EA4r0D2huXkcw4wKBO2bYp3UPpeuh/owTOrJTMGnFlwTaW8+xXU+RLghFz4B1P8/dP3XwH31cnXVGBss9pTD+qlbvuja9D97gJ5nwkMSN8vAuYDb69xvzHA1AL5HZkJexK4JXfdwHSvs0qWqyJbW9aIUzO/SWZeBG7NhVfsl6/nnuMCYL1c3ENT3N1y4cfhq3XrtOrZt+uHTB+OT/wYcFguzsEpfOucHE3MxftOCn97+j4cnwj7n1y8XxTI4QTg9oL8VXRGzb61RDnL2iYDUrzRDaRd127JlGOTOmmtlPJwI3BNJnwjXA8emgnbMqV5YL08xjJ0hyBpVUnfT0vhC4HX8FlJ8AYFsEH6XJ67/BrcAMuyJz57OS8tCw/IzHZtlWYNg87nT7nvlwFvBbIuNVeVSGc3XGlfmgsfC6xK90NQ8mnOAp4GfiJ3n3p3iXvm+Rgw3WqvEu8GjLe0IgFgZi8D41h2FrFRrs19fwhfxWk5cvfG7YArzOyNSriZ3Y0bRdm4n5WfzjUX1wmv4vIwPBPtMnwg/KVM2CjgWvO9kuCrVYPkrk/7pJW6nlJGRmtxD7CXpFMk7ZJWbLPsgXeg55VI6+qK1ZOhrq6U9GZcvi4H3sjEEXAzLpdZFlhmxc18f9gTlJOpO81sTub7Q+lv5dpG+oG6lJSZCtV0yeu4cZblo/hzuSRXr3fjs/WVOrsH2FrSLyXtIWn1RsvQlzTQ/vbEyzalQI7WxleZwPXceDN7vhd5eje++pqv2wXAnXSXx2o8geuCc+XukBsWxKmX3+H4as4yqzRmdjvwb7rr5bus+wrvninuHbny3Iivdu5YsjwrKnvig90rC+oPustDUX8HXTpmB1y35XXMFT3IW2/71kZtk0bold0i354wXtKLuA58Ddd7S3WnmU3BdUB2K8IofOLwz/XuEYO4zuFUfGZlLL5UvD1dp3FV3NUq/uUvZS80syX47FuWdYDDcKHKfv43/b5287IetJAXq3wfmgkrc/rdYHzlJe+CMD3ze5Zl0kyG8Udxt4tTgcfl+ya+UuLeFdYGniuRz6LyTMddLnvK7Nz3xfjqSTswBDdk8s+abJikT+DuTI8CB+Ed8XZ4Z7HUjdbcpeNC4Iups98VNzB/k4kzEXcf2RA33GfI9yFu2YP8l5HRWvwYd73ZF5/YmiXpQklD0u8VXVbmoJ4i2SmjKwfjKw7HF8T7Gj7gzfa32UFYhcVUcWfOsYwsZtpkT/qBmpSVmQzVdMlL6f5Z1kl/n6R7nQ2k67ldhLsy7YAbO7Ml/Vnt8+qFUu0PL+9udC9rxRBeO/O3t4dKVer2goL77UPJ/t3M5uHujc8DZwPT5Psjs+539fJb6Ruq6eWafUdiHXzfY74s/8jkIajOOviAZj7L1l9FR+Trr6i/gzo6huI2UI/e9q2N2ial6Y3dkiY8bkn3/2/cBXk74Hq6686z8X2DW0h6C+7+fKGVONAn9sR1DgcCF5nZjyoByV89S0X5rZMNTDOFQ3JxZ+EGz2lV7tfjWcCgrVgXn0nKfgcfDFXaf37loYjZwGBJq+YUy3rp76xc/G5pmtnTwGGSBGyFG7dnS5pqZteVyMNM6q/OzM7kKct6LNtZLMINxTwtOSa4l8zEO+R1C35bF5/BBtchT5rZyMqP8hMYi8p8DvAtYD98z9FU3IBeipldAVyR9NAIXJdcL2mD7IpECWrJaF3M7LV079MkrYcbqWfgbn+fo2vgMpSu/YFVkysIK6MrB+Czwb/GBx1F+WykTnpDI/1APRqRGaiuS6rVK/hKTtGgdhYsNaTOxVeDBqX4p+ODyx3q5L8/KNv+ZuFG7zeqpFORzcq+m95Qqdtj8ZXgPKVPezSz+4FPpZWbbVOaf5K0lZk9XCK/Fb1bTS/fm79lQbxZ+B7Cz1a5x9Qa9w+8/hbhbpVFNGrvZXXMlEx4URvoaxq1TRqiF3bLnvhe4M9mPFio4knwV1yGR+F7e9egnOdIrMR1EKvjHUWWL+S+V46F/0wufH+6D9ivx/1uJ5nZvQWfOP55+SDf6R2Iz8Y93GA6E3F9kZetg3GD4K6yCZlzPz5IgPJuczcC66XVgVr53FvSGpWA9P8n0m8V/g1smnW9k7Qbrjx7QqW9vLmH1/eYtMJxD/Dp7GqPpB3wPQ4VVqe7O92h+ApSPs2n8Pr+Ln5IxW+rDULMbL6ZjccN7fVpfFa8WTKK+eEs5+OGa0WubsYHWN1OASxJXV1pZq/iA72t8P1b3eL18N49oZF+oB6lZaYH3IQ/l3dUqdcp+QvMbI6Z/RF3wc3qjcW0oO2lPJVtf9fjh/9Mq1LeV1K8G4F9lDu5sUEm40bhe6vc68FGEzSz183sLny1eSX8YIwy+Z2Mr9AcmA2Unwz7TpbVy9W4Hl/1n1+lPA2tMK+AVFZ/1qxSf40O4u7GB9t5HZP/Dn3fNptmm9Siht1Sre+vDNaW2u2SNsX3/ObTfgPvPw/FB4k3W8nXiMRKXOdwPfB5SQ/h7iefJHNCFLggSDoR+K2k83E3jY3x0/nm4R1mhRNwV4S/SfoVrvAH4YK5sZm1y8sZg97xpWRY3IOfhHYEvrF3rrreDFCG6/CTnX4j6W3AJPyEpiOAU+t1osnN7uf47PmTuBE4EjcQy76zZSy+T+tSSafiHckaqVxnmZ/idDK+EnOLpNPwjuZ7uEI9KZPWZbhR/zv5ceQb4cp5Xsm85Hk8leVwSbNxxT45Y5j1NT/EjamrJZ2Ln8p1Il0uJeA6ZH/5ccbjgQ/gB5bMpZiz8X1Ur+GnRy5F0kn4rOut+CzuBimt+83fb9YIVWW0zMWSrsFnL+/DV3S2wWdBzwUfkKYyfysN6MfhG8m3Bx5Lg4JalNWV3wL+Btwg6QJ8tnoIfmraymZ2TJny9JYG+4F6NCozjeTzqdRGfyVpOG6MLcKN9Y8C55vZrZLOww9SuRNfydoUN3ZuzCT3CLCrpH1wmZ9pZlN7m8cGKNP+zsRXhm9L9TkZeAs+sNvVzPbLpLU3vv/rx7i+HArsaWaHUAIzM0lfBa5JE1V/wlfM1sXthmlmdka9dFJ9fhk/QXZKyu/X6XoedfNrfsLoCfhK6lhcjw8FTsH33F1YokiX4JPWt0g6HW/vq+L7/vbFD8FZUCKdFRIzmyDpUtxz4gxcn72BTzLsBXzPzEq/+NrMJkv6A3By0t3/BD6MT5bCsjrmEeAoSZ8DngJeMbN6HhGN0CvbpBYl7ZbKKdRflfR7vL98EJ88fB24KMns+rhOmEbxAtoF+Japreh+Wmh1rA1O0YlP6VNyLsONlDm4UtuO4lN5jsZXGhbhrgq7pGvOzMXbAD/B5zl8xuIFfHb0kFaXNz69lpfRSTa2wA3thbhBcTLpBDVqnKhI7nTKFDYQPynuhSQvj+NHrisTpzBN3O3i9+maBbgLxETg4w2W6634XqR/Z2T2CjKnk+EuVjfjqzmv4n7p2xekNQo3Ihbixyl/gOqnU25Son5G4W6Br6drRvTzM/8v3DBcjHdkB5A5hRPvOH6ED7oWpPrfJl/mTHorp/q7vOC3vXH3yhfS/Z7BO6GqJ+r1REZLpvNtfLZ1Vkpjckp7lVy8I/HOdXGSvwnATum3YSkvR1S5Ryldia9OXIYPNhbjK2LjgL0yccYAzxbcY+mzysnesEzYVApOYaTg1DVK9gN16raUzFDjhLZq5c38fmh6fq/ibfZRXM9UTkL9fKqbSp1OwQdEAzNpbIavhC6glydw9lX7S3EGpbxPSXL0Usr30bm03oUf1DAzpfd09rlR4nTKFL4TPviek+RgapLPnUqWaThuwE5J18/AXb92aCS/Kc4h+OBrMd5WLwbWz8UplO/025vwdv0YXW34nhQ2oL+fd7t/6H7C9Eq4K+8D6VnOS///FF+hy8rREbm0RpDr0/CJ0XPSc5iP67m9yZ1Ei7s1/hUf+Btd/dFISvatJcpaxjbpyemUpewWfCLjOXxycKnOxr1MHkv1PQlfjR5Dpu3m0rmB5J5fNo9KFwbLMZK2w2deDjOzi1udn6DvkTQaVyyrmFnDJ9IFgaSP4qsLe5jZLa3OT9A7oh8IgqAvkfRdfO/wMDOb1ur8dBJpv+803Kvo+LLXhTvlcob8xZtfxWf3XsZnhr+Pz6Rd2cKsBUHQAUh6F+5+dya+vysGcB1G9ANBEPQlydV2C+B+3H1yV/x9cn+KAVx5kgvocHyVdCV8G0NpYhC3/LEQb1iH4a4bc3DXsmMsfMaDNiWdnFdrk94b1n+n+63oHE+X69NhjV6cTvGqefBF2dXhkIseU7cfSKcN1mKJhavOcksz22mwQvIKfljSMfheyefwl33/sFk36Ev930byvze+L3Qa8HkzK/PKp6WEO2UQBC1H0lT8pLJqnGhmo/snN0FvkDQC3+NWi42sxMETIRd9h6R6nf8XzGxMf+Ql6H8kjaTOoSJm1tDpV0HQTCRNoPvL4LP83jKvP2kw7RE0qZ9qJTGIC4Kg5Uh6H7Vf8Pm8NX4MctAC0gmQw+tEe9BKvMg05KLvkLRtnShTzKxX71gK2hdJa+On8lbF+vfVGEGwDOnU2lqv/enxKbTN7KdaSQzigiAIgiAIgiAIOoh42XcQBEEQBEEQBEEHEYO4IAiCIAiCIAiCDiIGcUEQBEEQBEEQBB1EDOKCIAiCIAiCIAg6iBjEBUEQBEEQBEEQdBD/D4/sd1bJreEdAAAAAElFTkSuQmCC\n",
- "text/plain": [
- "<Figure size 900x900 with 30 Axes>"
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- },
- {
- "data": {
- "text/html": [
- "<div>\n",
- "<style scoped>\n",
- " .dataframe tbody tr th:only-of-type {\n",
- " vertical-align: middle;\n",
- " }\n",
- "\n",
- " .dataframe tbody tr th {\n",
- " vertical-align: top;\n",
- " }\n",
- "\n",
- " .dataframe thead th {\n",
- " text-align: right;\n",
- " }\n",
- "</style>\n",
- "<table border=\"1\" class=\"dataframe\">\n",
- " <thead>\n",
- " <tr style=\"text-align: right;\">\n",
- " <th></th>\n",
- " <th>age</th>\n",
- " <th>priors_count</th>\n",
- " <th>days_b_screening_arrest</th>\n",
- " <th>decile_score</th>\n",
- " <th>length_of_stay</th>\n",
- " </tr>\n",
- " </thead>\n",
- " <tbody>\n",
- " <tr>\n",
- " <th>age</th>\n",
- " <td>1.00</td>\n",
- " <td>0.12</td>\n",
- " <td>-0.07</td>\n",
- " <td>-0.40</td>\n",
- " <td>0.01</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>priors_count</th>\n",
- " <td>0.12</td>\n",
- " <td>1.00</td>\n",
- " <td>0.02</td>\n",
- " <td>0.45</td>\n",
- " <td>0.19</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>days_b_screening_arrest</th>\n",
- " <td>-0.07</td>\n",
- " <td>0.02</td>\n",
- " <td>1.00</td>\n",
- " <td>0.09</td>\n",
- " <td>0.06</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>decile_score</th>\n",
- " <td>-0.40</td>\n",
- " <td>0.45</td>\n",
- " <td>0.09</td>\n",
- " <td>1.00</td>\n",
- " <td>0.21</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>length_of_stay</th>\n",
- " <td>0.01</td>\n",
- " <td>0.19</td>\n",
- " <td>0.06</td>\n",
- " <td>0.21</td>\n",
- " <td>1.00</td>\n",
- " </tr>\n",
- " </tbody>\n",
- "</table>\n",
- "</div>"
- ],
- "text/plain": [
- " age priors_count days_b_screening_arrest \\\n",
- "age 1.00 0.12 -0.07 \n",
- "priors_count 0.12 1.00 0.02 \n",
- "days_b_screening_arrest -0.07 0.02 1.00 \n",
- "decile_score -0.40 0.45 0.09 \n",
- "length_of_stay 0.01 0.19 0.06 \n",
- "\n",
- " decile_score length_of_stay \n",
- "age -0.40 0.01 \n",
- "priors_count 0.45 0.19 \n",
- "days_b_screening_arrest 0.09 0.06 \n",
- "decile_score 1.00 0.21 \n",
- "length_of_stay 0.21 1.00 "
- ]
- },
- "metadata": {},
- "output_type": "display_data"
- }
- ],
- "source": [
- "sns.pairplot(compas[[\n",
- " 'age', 'priors_count', 'days_b_screening_arrest', 'decile_score',\n",
- " 'length_of_stay'\n",
- "]])\n",
- "plt.show()\n",
- "\n",
- "display(compas[[\n",
- " 'age', 'priors_count', 'days_b_screening_arrest', 'decile_score',\n",
- " 'length_of_stay'\n",
- "]].corr().round(2))"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "**Notes:**\n",
- "\n",
- "* Some notable correlations: `age` and `decile_score` ($\\rho\\approx-0.40$, Spearman -0.44) and `decile_score` and `priors_count` ($\\rho\\approx0.45$, Spearman 0.44)\n",
- "* Spearman correlation was for `length_of_stay` and `priors_count` 0.27 and for `length_of_stay` and `decile_score` 0.27"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 59,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "1 1286\n",
- "2 822\n",
- "4 666\n",
- "3 647\n",
- "5 582\n",
- "6 529\n",
- "7 496\n",
- "9 420\n",
- "8 420\n",
- "10 304\n",
- "Name: decile_score, dtype: int64\n"
- ]
- }
- ],
- "source": [
- "# Decile scores should be evenly distributed but are not.\n",
- "print(compas.decile_score.value_counts())"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 60,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "25 - 45 3532\n",
- "Less than 25 1347\n",
- "Greater than 45 1293\n",
- "Name: age_cat, dtype: int64\n"
- ]
- }
- ],
- "source": [
- "print(compas.age_cat.value_counts())"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 61,
- "metadata": {
- "scrolled": false
- },
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "African-American 3175\n",
- "Caucasian 2103\n",
- "Hispanic 509\n",
- "Other 343\n",
- "Asian 31\n",
- "Native American 11\n",
- "Name: race, dtype: int64\n"
- ]
- }
- ],
- "source": [
- "print(compas.race.value_counts())"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "A very small number of Asian and Native American defendants."
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 62,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "Black defendants: 51.44%\n",
- "White defendants: 34.07%\n",
- "Hispanic defendants: 8.25%\n",
- "Asian defendants: 0.50%\n",
- "Native American defendants: 0.18%\n",
- "---\n",
- "Defendants of other race: 5.56%\n"
- ]
- }
- ],
- "source": [
- "print(\"Black defendants: %.2f%%\" % (3175 / 6172 * 100))\n",
- "print(\"White defendants: %.2f%%\" % (2103 / 6172 * 100))\n",
- "print(\"Hispanic defendants: %.2f%%\" % (509 / 6172 * 100))\n",
- "print(\"Asian defendants: %.2f%%\" % (31 / 6172 * 100))\n",
- "print(\"Native American defendants: %.2f%%\" % (11 / 6172 * 100))\n",
- "print(\"---\")\n",
- "print(\"Defendants of other race: %.2f%%\" % (343 / 6172 * 100))"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 63,
- "metadata": {},
- "outputs": [
- {
- "data": {
- "text/plain": [
- "Low 3421\n",
- "Medium 1607\n",
- "High 1144\n",
- "Name: score_text, dtype: int64"
- ]
- },
- "execution_count": 63,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
- "source": [
- "compas.score_text.value_counts()"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 64,
- "metadata": {},
- "outputs": [
- {
- "data": {
- "text/plain": [
- "F 3970\n",
- "M 2202\n",
- "Name: c_charge_degree, dtype: int64"
- ]
- },
- "execution_count": 64,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
- "source": [
- "compas.c_charge_degree.value_counts()"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 65,
- "metadata": {},
- "outputs": [
- {
- "data": {
- "text/plain": [
- "(-0.001, 0.5] 2085\n",
- "(0.5, 5.5] 2866\n",
- "(5.5, 10.5] 729\n",
- "(10.5, 20.5] 402\n",
- "(20.5, 40.5] 90\n",
- "Name: priors_count, dtype: int64"
- ]
- },
- "execution_count": 65,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
- "source": [
- "compas.priors_count.value_counts(\n",
- " sort=False, bins=[0, 0.5, 5.5, 10.5, 20.5, 40.5])"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 66,
- "metadata": {
- "scrolled": true
- },
- "outputs": [
- {
- "data": {
- "text/html": [
- "<div>\n",
- "<style scoped>\n",
- " .dataframe tbody tr th:only-of-type {\n",
- " vertical-align: middle;\n",
- " }\n",
- "\n",
- " .dataframe tbody tr th {\n",
- " vertical-align: top;\n",
- " }\n",
- "\n",
- " .dataframe thead th {\n",
- " text-align: right;\n",
- " }\n",
- "</style>\n",
- "<table border=\"1\" class=\"dataframe\">\n",
- " <thead>\n",
- " <tr style=\"text-align: right;\">\n",
- " <th>race</th>\n",
- " <th>African-American</th>\n",
- " <th>Asian</th>\n",
- " <th>Caucasian</th>\n",
- " <th>Hispanic</th>\n",
- " <th>Native American</th>\n",
- " <th>Other</th>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>sex</th>\n",
- " <th></th>\n",
- " <th></th>\n",
- " <th></th>\n",
- " <th></th>\n",
- " <th></th>\n",
- " <th></th>\n",
- " </tr>\n",
- " </thead>\n",
- " <tbody>\n",
- " <tr>\n",
- " <th>Female</th>\n",
- " <td>549</td>\n",
- " <td>2</td>\n",
- " <td>482</td>\n",
- " <td>82</td>\n",
- " <td>2</td>\n",
- " <td>58</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>Male</th>\n",
- " <td>2626</td>\n",
- " <td>29</td>\n",
- " <td>1621</td>\n",
- " <td>427</td>\n",
- " <td>9</td>\n",
- " <td>285</td>\n",
- " </tr>\n",
- " </tbody>\n",
- "</table>\n",
- "</div>"
- ],
- "text/plain": [
- "race African-American Asian Caucasian Hispanic Native American Other\n",
- "sex \n",
- "Female 549 2 482 82 2 58\n",
- "Male 2626 29 1621 427 9 285"
- ]
- },
- "execution_count": 66,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
- "source": [
- "tab = compas.groupby(['sex', 'race']).size()\n",
- "tab.unstack()"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 67,
- "metadata": {
- "scrolled": false
- },
- "outputs": [
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- "<Figure size 1008x504 with 1 Axes>"
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- },
- {
- "data": {
- "image/png": "iVBORw0KGgoAAAANSUhEUgAAA+gAAAHmCAYAAADgNqS9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzs3XmYJWV5N/7vrSCIygtEXh0ZG1CIica4xphIVEgiGEeMQdxXLpeYRBONiiTG17gG3KNxQf2hCSpEo2bERHFBzYL7kteJG4rMsMyrOIiyiCDP74+qjsfmzEyf6dNzqqc/n+uq6/SpeqrqPtUDT3/PU0u11gIAAADM1vVmXQAAAAAgoAMAAMAgCOgAAAAwAAI6AAAADICADgAAAAMgoAMAAMAACOgwI9V5RlV9rap+XFWtqn5/Eevdu2/7vJ1QJlvh9wAA41XVx6vKs5xhBwjosAyq6gN9ePvKNpo9IslLk1yW5BVJ/jrJ13ZGfUNQVb/WH6NWVQ+adT0AsFRV9ZtV9fdV9e2qurKqLq+qr1bV66rqzrOuDxi+as2XWzBNVbUmyaZ0X4BVkru11j47pt3bkzw8yZrW2uYJtr9XkrkkF7fWLp5O1TtfVb0uyZOTtCT/2lq734xLmsiu8nsAYOmq6vpJXpOuX7sqyUeT/He6Pu6Xkvx2khsm+f3W2vpZ1bmzVNVckr1aa6tm4AGmZbdZFwC7oMckuX6Slyf58yTHJblOQE+yJkkmCed9+yuywkfaq2rPJA9L98fLeUmOrKpbtNYunG1li7cr/B4AmJq/SRfOP5Pk2NbaxtGFVbVfkv+TZN8Z1LbTLfz8wOI5xR2m77FJLk3ynCTfSPLQPpAmSarqsf11WYf37+dP8/5O//5/rm2uqntW1ceq6odVde7C5Qt3XFV3rqrTq+qiqrqqqs6vqvdV1W+NtLlFVT2/qj5TVd/r251TVS+rqpuM2ebH+/3tXlV/XVXn9etsqKqH7eAxemCSfZKc2k/XT/LocQ2r6q39/m9VVcdX1bf60wa/VFVH9W32rqq/6z/3lf0xu81WtnenqnpXVf2//nN8q6peUlU3XtBuNfweAFiiqvrFJE9P8r0k9xsXTltrW1prf5rktH6dI6rqlKr6Rn8a/I+q6j+r6iFjtr+t/mZn90UH9P3tOX1/e3FVfaGqXrqg3XWuQdfvweIYQYcpqqp7JLlNkre01n5cVacmeX6SP0jyjr7Zl9Jdb/7YJAf2PyfJDxZs7h5J/iLJmUlen+Q6ndeCfT8kyT8kuTbJ+5Kcm+TmSX4ryTFJ/q1ves8kT0t3+t1/pjv97tfTjfbfs6ru0Vq7eswuTkty5yQfSPf/jocleUdV/aC19q/bqm2M4/r9vj3JxUl+lORx6UYgtuaVSe6S5P39/h+RZH1/zF/XzzstyUFJfj/JGVX1S621n85voKoe2Lf5SbpjtLn/TM9OcnhV3bO19pMF+92Vfw8ALN1j0g16vXF7lzy11q7qf3xWklsl+XSSC5Lsl+ToJKdV1ZrW2quWUtBy9EVVdaMk/5HkFknOSPLuJDdO93fPnyR55nbK0u/BYrTWTCbTlKYkb07X4dyrf39wus7xw2Pafrz7T/A68+/db6Mlefg2lj9vZN7Nk1ye5JIkv7ygfSW5xcj7/53kRmO2+5x+u48cV2e6zvTGI/Pv1c//0ITHaC7JT5OcNTLvrf227jGm/fyyryb5hZH5x/TzL0nyziTXH1n2mn7ZH4zMu2mSHyb59ujx6Jc9s2//jNXyezCZTCbTdKYkZ/X/Hz5ignUOHjPvRkm+nO4svL1G5l+nv9nWsuXqi9J9gdCSPHVM+19Y8P7jWfA3jn7PZFrc5BR3mJL+m+UHJ9mY5JNJ0lo7N13HckR1N0yZxOdaa+/YfrMk3bf3eyU5sbX21dEFrXPhyPvvttYuH7ON1/Wvv7OVffxFa+2yke18Isl3ktx1kTXOe2y6kYZTR+bN//y4baz34tba90fevzfdSPg+SZ7VRkbKk5zev95hZN6j041+P6td91r3l6c7NfGhY/a7q/4eAJiOm/evFyx2hf7vg4XzLk/ytiR7J7nbEupZ7r7oyoUzFvTPY+n3YHGc4g7T86B0AfC1rbXR667+Id1p0o9Nd7r7Yn1ugra/1r+euZjGVXVskicluWO6G9aMflm3ZiurfXHMvAvSjYjPb/fe6b7NH/Xx1trH++WV7g+HH6c7NW7ex/ptPbiq/nQrHfiXR9+01q6tqu+lG2XYtKDt/I33bjEy79f718Oq6lfGbP/qdHfaXWjF/R4AGLaq2jvdae4PSHeq+14LmmytD1iM5eqLPpGuf/27qvrdJB9M8u+ttW8stjD9HmyfgA7TMz/6e+qC+f+Y5G+TPLaqXrAgvG/LdyfY9//qX7d7F/SqemaSk/rt/2u6Tu7H/eL/k2SPceu11i4dM/ua/Hzneu9+Gwt9vH+9V7o/RN41ur0+bL8j3anmD0o3grDQD7ey/63NT5LdR+bt17/+6Zj227ISfw8A7Dyb033Be0CSr2+vcVXdIF3YvWOSz6e7lGtLusu/7pgutI/tAxZpWfqi1tqlVfWbSV6QZF2SY/ttfC3dKPd7p7WvUfo9VhsBHaagqm6d7uYnSbKhGyi+joPTBdizFrnZxQb55Gc3mLtFfjZ6fB1VtVu6a70uTHKHNnIzm6q6WcaH60VrrT0vyfO20eS4/vXYhXd3XdBmXEBfqvkgf2hr7ZwJ1ltxvwcAdqr/TNe/H57ujLDteUC6IH5ya+1Jowuq6vh++ahr+9frj9nW3mPmLVtf1J+a/8h+3TsluW+6L77fVVV3b62NPetMvweLJ6DDdDw23Y1Xzkp3E7KFbpquw31cFh/QJ/HZdCPP90nyhW20u2m6zvzD7bp3mr3HMtT1P/pHqByT7uY3795Ks/sk+a2qunVr7VtTLuEz6e6mf/ckkwT0SQz+9wDA1L0t3dNAnlhVr9rW9dhVtUeSW/dv3z+mybg+YD5wHzBm2Z3GzFv2vqi1dk2/n89W1bfTXc53v2z9sjD9HiySU0Ngiarqeumuq/5pkke01h6/cEp387jvJzmmv+5s2v4+yRVJjq+qX15QX1XV/HVd3013c5c7V9UNR9qsSfLiZahr1EPTXWP39nHHqD9Or0r3Rce2bha3o05JclmSE6vq0IULq2qfqhr3h84kVsLvAYAp6q/BfkW6u5S/v6rWLmzT9zEvT/KQdDeTTRYE06r6gyT3H7OLr6e7K/v9q2qfkfa3SvLUMe2XpS+qql+pqluO2d/N+tfr3DxuhH4PFklAh6X7nSS3TPLB1tpF4xq07tnab08XUB8y7QJaa5vTnRp+4yRfrKp3VtWLq+rN6Tr24/t21yZ5Q7rT7b9YVS+vqlOS/FeSDdOua4H50H3KNtqcmu66ssf0X3xMTWvtu+menb5fussQ3ldVL62q11XVB5NclO7GNUvZx0r4PQAwfSckeX2S30hyTlW9v6pO6qf3JtmU7hngl6QbOd+YLkCv79t8IMm70j2h5Oe07tnpr083Cv2FqnplVb0tyZfys3u8jLZfrr7od5J8p6rOqqo3VNVLqup9SU5M9ySUf9jawdHvweI5xR2Wbj54vnU77U5J9033cUneNO0iWmun96eZPTtdJ7p3um+sP5Ou05/37HSnyz06yR+nu0nL36X7BvuqadeVJFV1m3R/tGzY2vVp/Wf4bv9HygOS/G6SD02zjtba+qq6S7o75/52kt9Ld236xiSvzvZ/h4vZx2B/DwAsj/6U7z+qqren+7L3t/Kzx4adl+QdSd7QWvtiklTVEUleluSwdNeufyldn7QmyQPH7OKEdI8WfUySP0ryjf71/HSXby2sZzn6og+lC9j3THdm4J59+9cnOWlrgxQ7uC9YtWrxN5QGAAAAlotT3AEAAGAABHQAAAAYAAEdAAAABkBABwAAgAFwF/dFuOlNb9oOOuigWZcBAFPx+c9//uLW2v6zrmNb9L0A7EoW2/cK6Itw0EEH5XOf2+qToQBgRamq82Zdw/boewHYlSy273WKOwAAAAyAgA4AAAADIKADAADAAAjoAAAAMAACOgAAAAyAgA4AAAADIKADAADAAAjoAAAAMAACOgAAAAyAgA4AAAADIKADAADAAAjoAAAAMAACOgAAAAyAgA4AAAADIKADAADAAAjoAAAAMACDDehV9XtV9cmquqyqflhVn6uqI0aW71tVb66qi6vq8qr6SFXdfsx29qyql1bVRVV1ZVWdXVX33LmfBgAAALZtkAG9qp6U5J+TfD7JA5Mcm+RdSfbql1eS9UmOSvKUJMck2T3JWVW1dsHm3pLkCUmem2RdkouSfKiq7rj8nwQAAAAWZ7dZF7BQVR2U5FVJntlae9XIog+N/Hx0ksOSHNFaO6tf7+wk5yZ5VpKn9vPukOThSY5rrZ3Sz/tEkg1Jnt9vBwDYxa1ZO5fNF2ya2f5vfsAtc9H5G2e2fwBWhsEF9CTHJbk2yRu20eboJBfOh/Mkaa1dWlXvT/KA9AG9b3d1ktNH2l1TVacleXZV7dFau2raHwAAGJbNF2zKgcefMbP9n3fiupntG4CVY4inuB+W5GtJHlpV36qqa6rqnKr645E2t0vylTHrbkgyV1U3Hml3bmvtijHtbpDkkCnXDgAAADtkiAH9FkkOTfLSJH+T5D5JPpzktVX1p32b/ZJcMmbdLf3rvotst9/WiqiqJ1bVN6vqexs3OiUNAJabvheA1W6IAf16SW6S5EmttTe11j7WWntykg8mOaG/QVwlaWPWrTHvF9PuOlprJ7fWDm2t7T83NzfZJwAAJqbvBWC1G2JA/37/+uEF889McrMka9KNgI8b/Z4fOZ8fNd9euy1jlgEAAMBON8SAvmEr8+dHva/t29xuTJvbJtnYWrtsZFsHV9VeY9r9JMk5S6wVAAAApmKIAf29/euRC+YfmeT81trmdM9AP6Cq7jW/sKr2TnL/ftm89emej37sSLvdkjwkyZnu4A4AAMBQDPExa/+S5Kwkb6yqmyb5dpIHpbtZ3OP6NuuTnJ3k1Kp6ZrpT2k9IN8p+0vyGWmtfqqrTk7yqqnZP95z0Jyc5OMkjds7HAQAAgO0bXEBvrbWq+v0kL0ny1+muF/9akke01t7Rt7m2qtYleVmS1yXZM11gP7y1tmnBJh+X5EVJXphknyRfTnJUa+0LO+PzAAAAwGIMLqAnSWvth0n+uJ+21mZLkuP6aVvbujLJ0/sJAAAABmmI16ADAADAqiOgAwAAwAAI6AAAADAAAjoAAAAMgIAOAAAAAyCgAwAAwAAI6AAAADAAAjoAAAAMgIAOAAAAAyCgAwAAwAAI6AAAADAAAjoAAAAMgIAOAAAAAyCgAwAAwAAI6AAAADAAAjoAAAAMgIAOAAAAAyCgAwAAwAAI6AAAADAAAjoAAAAMgIAOAAAAAyCgAwAAwAAI6AAAADAAAjoAAAAMgIAOAAAAAyCgAwAAwAAI6AAAADAAAjoAAAAMgIAOAAAAAyCgAwAAwAAI6AAAADAAAjoAAAAMgIAOAAAAAyCgAwAAwAAI6AAAADAAAjoAAAAMgIAOAAAAAyCgAwAAwAAI6AAAADAAAjoAAAAMgIAOAAAAAyCgAwAAwAAI6AAAADAAAjoAAAAMgIAOAAAAAyCgAwAAwAAI6AAAADAAAjoAAAAMgIAOAAAAAyCgAwAAwAAI6AAAADAAgwzoVXXvqmpjph8saLdvVb25qi6uqsur6iNVdfsx29uzql5aVRdV1ZVVdXZV3XPnfSIAAADYtt1mXcB2PDXJZ0feXzP/Q1VVkvVJDk7ylCSXJDkhyVlVdcfW2vkj670lyf2SPDPJt5P8cZIPVdVvtNa+tLwfAQAAALZv6AH9q621T21l2dFJDktyRGvtrCSpqrOTnJvkWenCfarqDkkenuS41top/bxPJNmQ5Pn9dgAAAGCmBnmK+yIdneTC+XCeJK21S5O8P8kDFrS7OsnpI+2uSXJakiOrao+dUy4AAABs3dAD+tur6qdV9f2qekdVzY0su12Sr4xZZ0OSuaq68Ui7c1trV4xpd4Mkh0y9agAAAJjQUE9xvzTJy5N8IskPk9wpyV8kObuq7tRa+26S/ZJ8Z8y6W/rXfZNc1re7ZBvt9pte2QAAALBjBhnQW2tfTPLFkVmfqKpPJvlMumvLn5OkkrQxq9eY94tp9/MLq56Y7qZy++y///6LrBwA2FH6XgBWu6Gf4v4/WmtfSPKNJL/Wz9qS8aPf+/avlyyy3ZYxy9JaO7m1dmhrbf+5ublxTQCAKdL3ArDarZiA3hsdDd+Q7vryhW6bZGNr7bKRdgdX1V5j2v0kyTnLUSgAAABMYsUE9Kq6a5JfTPLpftb6JAdU1b1G2uyd5P79soy02z3JsSPtdkvykCRnttauWubSAQAAYLsGeQ16Vb093fPMv5DkB+luEndCkguSvKZvtj7J2UlOrapnpjul/YR0o+wnzW+rtfalqjo9yauqavd+u09OcnCSR+yUDwQAAADbMciAnu7xaQ9L8pQkeyXZnOQ9Sf5Pa+3iJGmtXVtV65K8LMnrkuyZLrAf3lrbtGB7j0vyoiQvTLJPki8nOaq/rh0AAABmbpABvbX2kiQvWUS7LUmO66dttbsyydP7CQAAAAZnxVyDDgAAALsyAR0AAAAGQEAHAACAARDQAQAAYAAEdAAAABgAAR0AAAAGQEAHAACAARDQAQAAYAAEdAAAABgAAR0AAAAGQEAHAACAARDQAQAAYAAEdAAAABgAAR0AAAAGQEAHAACAARDQAQAAYAAEdAAAABgAAR0AAAAGQEAHAACAARDQAQAAYAAEdAAAABgAAR0AAAAGQEAHAACAARDQAQAAYAAEdAAAABgAAR0AAAAGQEAHAACAARDQAQAAYAAEdAAAABgAAR0AAAAGQEAHAACAARDQAQAAYAAEdAAAABgAAR0AAAAGQEAHAACAARDQAQAAYAAEdAAAABgAAR0AAAAGQEAHAACAARDQAQAAYAAEdAAAABgAAR0AAAAGQEAHAACAARDQAQAAYAAEdAAAABgAAR0AAAAGQEAHAACAARDQAQAAYAAEdAAAABgAAR0AAAAGQEAHAACAARDQAQAAYABWRECvqg9WVauqFy6Yv29VvbmqLq6qy6vqI1V1+zHr71lVL62qi6rqyqo6u6ruufM+AQAAAGzb4AN6VT0syR3GzK8k65McleQpSY5JsnuSs6pq7YLmb0nyhCTPTbIuyUVJPlRVd1zG0gEAAGDRBh3Qq2qfJK9M8vQxi49OcliSR7XW3tla+2A/73pJnjWyjTskeXiSp7XW3tRa+2iSByfZmOT5y/wRAAAAYFEGHdCTnJRkQ2vtnWOWHZ3kwtbaWfMzWmuXJnl/kgcsaHd1ktNH2l2T5LQkR1bVHstROAAAAExisAG9qg5L8ugkf7SVJrdL8pUx8zckmauqG4+0O7e1dsWYdjdIcsgUygUAAIAlGWRAr6rdk7wxyctaa1/fSrP9klwyZv6W/nXfRbbbbys1PLGqvllV39u4cePiCgcAdpi+F4DVbpABPcnxSW6Y5EXbaFNJ2lbm70i7n9NaO7m1dmhrbf+5ubltNQUApkDfC8Bqt9usC1ioquaS/GWSxyfZY8E14nv0N477UboR8HGj3/Mj5/Oj5luSjOvl9x1ZDgAAADM1xBH0WyXZM8mp6UL2/JQkz+h/vn26a8hvN2b92ybZ2Fq7rH+/IcnBVbXXmHY/SXLOVKsHAACAHTDEgP6lJIePmZIutB+eLlSvT3JAVd1rfsWq2jvJ/ftl89anez76sSPtdkvykCRnttauWrZPAgAAAIs0uFPcW2s/SPLxhfOrKknOa619vH+/PsnZSU6tqmemG1k/Id215SeNbO9LVXV6klf1N587N8mTkxyc5BHL+VkAAABgsYY4gr4orbVrk6xL8uEkr0vy3iQ/TXJ4a23TguaPS3JKkhcm+UCSWyY5qrX2hZ1XMQAAAGzd4EbQt6a1dp27rrfWtiQ5rp+2te6VSZ7eTwAAADA4K3YEHQAAAHYlAjoAAAAMgIAOAAAAAyCgAwAAwAAI6AAAADAAAjoAAAAMgIAOAAAAAyCgAwAAwAAI6AAAADAAAjoAAAAMgIAOAAAAAyCgAwAAwAAI6AAAADAAAjoAAAAMgIAOAAAAAyCgz8CatXOpqplNa9bOzfoQAAAAsMBukzSuqhu31i5brmJWi80XbMqBx58xs/2fd+K6me0bAACA8SYdQd9cVW+sqjstSzUAAACwSk0a0F+VZF2Sz1XVZ6vquKraaxnqAgAAgFVlooDeWntOkrkkxybZkuRNSS6sqtdU1a8sQ30AAACwKkx8k7jW2k9ba+9prR2Z5NAkr08X2L9cVf9eVY+sqt2nXSgAAADsypZ6F/fvJ7kw3Wh6Jbl5krclOaeqfmOJ2wYAAIBVY4cCelX9elX9f0kuSPLiJP+R5M6ttUOS3D7JxiQnT61KAAAA2MVN+pi1Jyd5UroQ/o0kf5nkra21S+fbtNb+u6r+KslHplkoAAAA7MomCuhJXp3k/Un+vLX20W20+2aSl+xwVQAAALDKTBrQD2qtXbi9Rq21C5L81Y6VBAAAAKvPpNeg71VVh41bUFX3qKpbT6EmAAAAWHUmDeivTvIHW1n2wCSvXFo5DMGatXOpqplOa9bOzfowAAAA7FSTnuL+a0neuJVlH0/yqCVVwyBsvmBTDjz+jJnWcN6J62a6fwAAgJ1t0hH0vZP8eCvLfpLkfy2tHICVZdZnnDjbBABg1zHpCPq3kxye5Mwxyw5Pct6SKwJYQWZ9xomzTQAAdh2TjqCfmuTpVfWkqto9Sapq96p6UpKnJfn7aRcIAAAAq8GkI+gnJrlbktcneW1VfS/JTfvtvC/J30y3PAAAAFgdJgrorbWfJvn9qrpPkt9N8gtJLk5yZmvtI8tQH6xYa9bOZfMFm2a2/5sfcMtcdP7Gme0fAACYzKQj6EmS1tqZGX8dOtBzbTIAADCJHQroSVJV+yXZc+H81tqFS6oIAAAAVqGJAnpV3STJK5I8NMleW2l2/aUWBQAAAKvNpCPor03y4CRvTfJ/k1w17YIAAABgNZo0oN83ybNaa69ZjmIAAABgtZr0OejXS/LV5SgEAAAAVrNJA/o/JrnfchQCAAAAq9mkp7ifkeRvq+pGSf4lyZaFDVprn5xGYQAAALCa7EhAT5JbJXl8kjayrPr37uIOK8CatXPZfMGmmdZw8wNumYvO3zjTGgAAYCgmDei/uyxVADvd5gs25cDjz9h+w2V03onrZrp/gKHwpSkAyYQBvbX20eUqBABgtfKlKQDJ5DeJS5JU1b5VdVRVPaKq9u3n7T7d0oDVbs3auVTVTKc1a+dmfRgAAFglJj3FPVX1kiR/lmSPdNec/1qSS5J8oKo+2Vp74XRLBFYrI0qrg1N7AQA6EwX0qjo+ydOSvDjJh5P8x8ji9yd5eBIBHYBF80UMAEBn0hH0JyZ5QWvtRVW18G7t30xyyHTKAgAAgNVl0mvQ1yb5z60s+0mSGy+tHAAAAFidJg3oFya53VaW3T7Jd5ZUDQAAAKxSkwb0dyd5blX9+si8VlW3TvKMJKdPo6iqOrKqPlZVm6vqqqo6v6r+sapuu6DdLavq3VV1aVX9sKreU1XXueVyf9f5N1fVxVV1eVV9pKpuP41aAQAAYBomDejPS3JOutPcv9rPOy3JV5Kcm+QlU6prvySfT/InSe6T5IR0I/efqqoDk6Sq9krysSS/lOQxSR6V5NAkZ1XVjeY3VFWVZH2So5I8JckxSXbv262dUr0AAACwJBPdJK61dnlV3TNdGD4yyflJvp/kpCR/31q7ehpFtdbemeSdo/Oq6jNJvpbkQUlenuQJSW6V5DattXP6Nv+V7mZ1T0ryin7Vo5McluSI1tpZfbuz032h8KwkT51GzQAAALAUEz8HvbV2TZJT+mln+n7/Ov8lwNFJPjUfzvvazq2q/0jygPx8QL9wPpz37S6tqvf37QR0AAAAZm7SU9x3qqq6flXdoKoOTfLGJJvTnVKfdKe8f2XMahuSjF6rvq12c1XlzvMAAADM3EQj6FX1zSRtG01aa+02Syvp53w6yV36n89Jd5r6d/v3+yW5ZMw6W5LsO/J+v4y/u/yW/nXfJJctuVKAAVqzdi6bL9g00xpufsAtc9H5G2daAwDASjDpKe6fznUD+i8kuXuSHyb55DSKGvGoJHunu9b8GUk+XFWHtda+0y8f92VBjXm/mHY/v7DqiUmemWSf/ffff5KaAQZj8wWbcuDxZ8y0hvNOXDfT/bNy6HsBWO0mvUncI8fNr6r9knwwyQemUdTI/ubvFP/pqvrXdCPhz07yh+lGz/cbs9q++fmR9S3baJeMH4VPa+3kJCcnyV3vetdtnTXADBgVBNj16HsBWO0mvkncOK21LVV1UpLnZ0rPQh+zjx9U1TlJDulnbUh3fflCt03y3yPvN6R7VNu4dhtba05vX4GMCgIAALuaad4k7ookc1Pc3s+pqpule+b5t/pZ65PcvapuNdLmoCT36JdlpN0BVXWvkXZ7J7n/gnYAAAAwM0seQa+q66UbjX5ukq9up/lit/neJF9I8l/prm3/xSRPS3JNumegJ8mbkvxJkn+uqueku878BUk2pbvj+7z1Sc5OcmpVPTPdKe0npLsG/aRp1AsAAABLNeld3K/OdW+4dr10YfeyJPebUl2fSvLgJH+e5AbpQvfHk7xk/gZxrbXLq+qIJK9M8g99DR9N8mejp6231q6tqnVJXpbkdUn2TBfYD2+tzfYiZgAAAOhNOoJ+Yq4b0H+c5LwkH2itjb3h2qRaayf2+9peu41JjllEuy1JjusnAAAAGJxJ7+L+nOUqBAAAAFazad4kDgAAANhBk16DfvIEzVtr7UkT1gMAAACr0qTXoN83yU2S7J3k2nR3RN833Uj8D5P8aKTtwmvVAQAAgK2Y9BT3B6cL4o9McsPW2v5JbpjkUf38Y1trt+ynZXsmOgAAAOxqJh1Bf2WSk1pr75if0Vq7Osnbq2q/JK9O8utTrA8AAABWhUlH0O+Q5OtbWfb1JLdfWjkAAACwOk0a0P9fkgdtZdn6m9rSAAAZXElEQVSxSb67tHIAAABgdZr0FPdXJ3l5Vd08ybvSBfabpbs2/X5J/ny65QEAAMDqMFFAb629sqquSPJXSe4/sujCJE9urU3yGDYAAACgN+kIelprb6yqNyU5MMmaJBclOa+1du20iwMAAIDVYuKAniR9GD+3nwAAAIAlmvQmcamqX62qf6yqzVX1k6q6cz//hVV1n+mXCAAAALu+iQJ6Vf1mkk+ne9zae5Jcf8G2/nB6pQEAAMDqMekI+olJPprkl5M8NUmNLPtckrtMqS4AAABYVSa9Bv0uSY5prV1bVbVg2cXpHrkGAAAATGjSEfSrktxwK8tunuTSpZUDAAAAq9OkAf3fkzy1qkbXa/3rcUnOmkpVAAAAsMpMeor7c9OF9C8meVe6cP7Iqjopyd2T3G265QEAAMDqMNEIemvti0nuneQHSZ6X7iZxf5ZkzySHt9a+OuX6AAAAYFWYdAQ9rbXPJrlXVe2V5KZJLmmt/WjqlQEAAMAqsugR9Kq6QVV9t6runySttStaaxuFcwAAAFi6RQf01tpP0p3S/uPlKwcAAABWp0nv4r4+yTHLUQgAAACsZpNeg74+yWur6rQk70tyUX72mLUkSWvtk1OqDQAAAFaNSQP6e/vXB/fTaDiv/v31p1AXAAAArCqTBvTfXZYqAAAAYJXbbkCvqiOSfKa1dllr7aM7oSYAAABYdRZzk7gPJ7nt/Juqul5VfbKqDl2+sgAAAGB1WUxArzHvD0tyk+mXAwAAO2bN2rlU1cymNWvnZn0IgBVu0mvQAQBgkDZfsCkHHn/GzPZ/3onrZrZvYNcw6XPQAQAAgGWw2BH0A6rqVv3P1x+Z94OFDVtr355KZQAAALCKLDagv3vMvPdtpa3noAMA7GLWrJ3L5gs2zbSGmx9wy1x0/saZ1gCwnBYT0B+37FUAADBos76+O3GNN7Dr225Ab629bWcUAgAAAKuZm8QBAADAAAjoAAAAMAACOgAAAAyAgA4AAAADIKADwHasWTuXqprZtGbt3KwPAQCwEyz2OegAsGrN+vFSHi0FAKuDEXQAAAAYAAEdAAAABkBABwAAgAEQ0AEAAGAABHQAAAAYAAEdAAAABkBABwAAgAEQ0AEAAGAABHQAAAAYAAEdAAAABmBwAb2qHlRV/1RV51XVlVX19ap6SVXdZEG7favqzVV1cVVdXlUfqarbj9nenlX10qq6qN/e2VV1z533iQAAAGD7BhfQkzwjyU+T/EWSo5K8PsmTk3y4qq6XJFVVSdb3y5+S5Jgkuyc5q6rWLtjeW5I8Iclzk6xLclGSD1XVHZf/owAAAMDi7DbrAsa4f2vteyPvP1FVW5K8Lcm9k3wsydFJDktyRGvtrCSpqrOTnJvkWUme2s+7Q5KHJzmutXZKP+8TSTYkeX6/HQAAAJi5wY2gLwjn8z7bvx7Qvx6d5ML5cN6vd2mS9yd5wMh6Rye5OsnpI+2uSXJakiOrao8plg4AAAA7bHABfSvu1b9+tX+9XZKvjGm3IclcVd14pN25rbUrxrS7QZJDpl0oAAAA7IjBB/SqOiDd6egfaa19rp+9X5JLxjTf0r/uu8h2+21jv0+sqm9W1fc2btw4eeEAwET0vezq1qydS1XNdFqzdm7WhwHYhiFeg/4/+pHwf05yTZLHjS5K0satMub9YtpdR2vt5CQnJ8ld73rXcdsAAKZI38uubvMFm3Lg8WfMtIbzTlw30/0D2zbYEfSq2jPdndpvleTI1tr5I4u3ZPzo9/zI+SWLbLdlzDIAAADY6QYZ0Ktq9yT/lORuSX6vtfZ/FzTZkO768oVum2Rja+2ykXYHV9VeY9r9JMk506saAAAAdtzgAnr/rPO3J/ntJA9orX1qTLP1SQ6oqnuNrLd3kvv3y0bb7Z7k2JF2uyV5SJIzW2tXTf8TAAAAwOSGeA3636UL1C9KcnlV3X1k2fn9qe7rk5yd5NSqema6U9pPSHdt+UnzjVtrX6qq05O8qh+VPzfJk5McnOQRO+PDAAAAwGIMbgQ9yX37179MF8JHp8cnSWvt2iTrknw4yeuSvDfJT5Mc3lrbtGB7j0tySpIXJvlAklsmOaq19oXl/RgAAACweIMbQW+tHbTIdluSHNdP22p3ZZKn9xMAAAAM0hBH0AEAAGDVEdABAABgAAR0AAAAGAABHQAAAAZAQAcAAIABENABAABgAAR0AAAAGAABHQAAAAZAQAcAAIABENABAABgAAR0AAAAGAABHQAAAAZAQAcAAIABENABAABgAAR0AAAAGAABHQAAAAZAQAcAAIABENABAABgAAR0AAAgSbJm7VyqaqbTmrVzsz4MMDO7zboAAABgGDZfsCkHHn/GTGs478R1M90/zJIRdAAAABgAAR0AAAAGQEAHAACAARDQAQAAYAAEdAAAABgAAR0AAAAGQEAHAACAARDQAQAAYAAEdAAAABgAAR0AAAAGQEAHAACAARDQAQAAYAAEdAAAABgAAR0AAAAGQEAHAACAARDQAQAAYAAEdAAAABgAAR0AAAAGQEAHAACAARDQAQAAYAAEdAAAABgAAR0AAAAGQEAHAACAARDQAQAAYAAEdAAAABgAAR0AAAAGQEAHAACAARDQAQAAYAAEdAAAYMVYs3YuVTWzac3auVkfAnZhu826AAAAgMXafMGmHHj8GTPb/3knrpvZvtn1GUEHAACAARhkQK+qtVX1mqo6u6quqKpWVQeNabdnVb20qi6qqiv79vcc0+56VXVCVX2nqn5cVV+uqmN2xmcBAACAxRhkQE9ySJIHJ7kkyb9to91bkjwhyXOTrEtyUZIPVdUdF7R7QZLnJXltkvsm+VSSd1XV7023bAAAANgxQ70G/ZOttZslSVU9Psl9FjaoqjskeXiS41prp/TzPpFkQ5LnJzm6n/e/kzwjyd+01l7Wr35WVR2S5G+S/MsyfxYAAADYrkGOoLfWrl1Es6OTXJ3k9JH1rklyWpIjq2qPfvaRSW6Q5NQF65+a5PZVdfDSKwYAAIClGWRAX6TbJTm3tXbFgvkb0gXyQ0baXZXknDHtkuS2y1YhAAAALNJKDuj7pbtGfaEtI8vnX3/QWmvbaQcAAAAzs5IDeiVZGLrn5+9Iu59fWPXEqvpmVX1v48aNO1giALBY+l4AVruVHNC3ZPzo974jy+df962qhYF8Ybuf01o7ubV2aGtt/7m5uSUXCwBsm74XgNVuJQf0DUkOrqq9Fsy/bZKf5GfXnG9IskeSW49plyT/vWwVAgAAq8qatXOpqplOa9b6knOlGupj1hZjfZK/TnJskrclSVXtluQhSc5srV3Vt/tgusD+iL79vEcm+Upr7dydVjEAALBL23zBphx4/BkzreG8E9fNdP/suMEG9Kp6UP/jXfrX+1bV95J8r7X2idbal6rq9CSvqqrdk5yb5MlJDk4XxpMkrbXvVtUrk5xQVT9K8oV0If6IJA/YSR8HAAAAtmmwAT3Juxa8f13/+okk9+5/flySFyV5YZJ9knw5yVGttS8sWPcvk1yW5E+T3DzJ15M8uLX2/umXDQAAAJMbbEBvrW3zLut9myuTPL2fttXup+lC/AunUx0AAABM10q+SRwAAADsMgR0AAAAGAABHQAAAAZAQAcAAIABENABAABgAAR0AAAAGAABHQAAAAZAQAcAAFhF1qydS1XNdFqzdm7Wh2GQdpt1AQAAAOw8my/YlAOPP2OmNZx34rqZ7n+ojKADAADAAAjoAAAAMAACOgAAAAyAgA4AAAADIKADAADAAAjoAAAAMAACOgAAAAyAgA4AAAADIKADAADAAAjoAAAAMAACOgAAAAyAgA4AAAADIKADAADAAAjoAAAAMAACOgAAAAyAgA4AAAADIKADAAAwKGvWzqWqZjatWTs3k8+920z2CgAAAFux+YJNOfD4M2a2//NOXDeT/RpBBwAAgAEQ0AEAAGAABHQAAAAYAAEdAAAABkBABwAAgAEQ0AEAAGAABHQAAAAYAAEdAAAABkBABwAAgAEQ0AEAAGAABHQAAAAYAAEdAAAABkBABwAAgAEQ0AEAAGAABHQAAAAYAAEdAAAABkBABwAAgAEQ0AEAAGAABHQAAAAYAAEdAAAABkBABwAAgAEQ0AEAAGAABHQAAAAYAAEdAAAABkBABwAAgAFYNQG9qm5ZVe+uqkur6odV9Z6qmpt1XQAAAJCskoBeVXsl+ViSX0rymCSPSnJokrOq6kazrA0AAACSZLdZF7CTPCHJrZLcprV2TpJU1X8l+WaSJyV5xQxrAwAAgNUxgp7k6CSfmg/nSdJaOzfJfyR5wMyqAgAAgN5qCei3S/KVMfM3JLntTq4FAAAArqNaa7OuYdlV1U+SvKK19uwF81+Y5Nmtteuc6l9VT0zyzCT7JLlxujDP4t00ycWzLmKFcwyXzjGcDsdx6YZ2DA9sre0/6yIW0vcu2dD+na1EjuHSOYbT4Tgu3dCO4aL63tVyDXqSjPsmorbauLWTk5y8fOXs2qrqc621u866jpXMMVw6x3A6HMelcwwXR9+7NP6dLZ1juHSO4XQ4jku3Uo/hajnF/ZIk+42Zv2+/DAAAAGZqtQT0DemuQ1/otkn+eyfXAgAAANexWgL6+iR3r6pbzc+oqoOS3KNfxvQ5RXHpHMOlcwynw3FcOseQncG/s6VzDJfOMZwOx3HpVuQxXC03ibtRki8nuTLJc9Jdj/6CJDdJ8quttctmWB4AAACsjhH01trlSY5I8o0k/5Dk7UnOTXKEcA4AAMAQrIoRdAAAABi6VTGCzvKrqgdV1T9V1XlVdWVVfb2qXlJVN5l1bStZVX2wqlpVvXDWtawkVfV7VfXJqrqsqn5YVZ+rqiNmXddKUlX3qKozq+q7/TH8QlUdN+u6hqqq1lbVa6rq7Kq6ov/v9qAx7fasqpdW1UX9/yvPrqp77vyK2RXoe5eHvnfH6HuXTt87mV217xXQmZZnJPlpkr9IclSS1yd5cpIPV5V/Zzugqh6W5A6zrmOlqaonJfnnJJ9P8sAkxyZ5V5K9ZlnXSlJVv5rkI0l2T/KEJMck+WySt1TVk2dZ24AdkuTB6R7d+W/baPeWdMf0uUnWJbkoyYeq6o7LXiG7In3vlOl7d4y+d+n0vTtkl+x7neLOVFTV/q217y2Y9+gkb0vy2621j82mspWpqvZJ8rUkT0vyjiQvaq09Z7ZVDV//relXk5zQWnvVbKtZuarqxen+8N9v9D4dVfWpJK219hszK26gqup6rbVr+58fn+RNSQ5urX1npM0dknwpyXGttVP6ebulexTo11trR+/0wlnR9L3Tpe/dMfre6dD3Tm5X7Xt9u8pULPwDoffZ/vWAnVnLLuKkJBtaa++cdSErzHFJrk3yhlkXssLdIMnV6Z58MeoH0W+MNf8HwnYcne64nj6y3jVJTktyZFXtsUzlsYvS906dvnfH6HunQ987oV217/XLZjndq3/96kyrWGGq6rAkj07yR7OuZQU6LN3ox0Or6ltVdU1VnVNVfzzrwlaYt/avf1tVt6iqfarqCUl+O8krZ1fWine7JOe21q5YMH9Duj/MDtn5JbEL0vfuAH3vkuh7p+Ot/au+d7pWXN+726wLYNdUVQckeX6Sj7TWPjfrelaKqto9yRuTvKy19vVZ17MC3aKfXprumsxvpbsO7rVVtVtr7dWzLG6laK19paruneS9+dkfq1cn+cPW2mkzK2zl2y/ddXILbRlZDjtM37tj9L1Lpu+dAn3vsllxfa+AztRV1Y3T3SjkmiSPm3E5K83xSW6Y5EWzLmSFul6SmyR5bGvtPf28j/XXx51QVX/b3Hhju6rq0CT/lO7b5T9Md7rdA5K8oap+3Fp7+yzrW8Eqybh/f7WzC2HXo+9dEn3v0uh7p0Dfu2xWXN8roDNVVbVnkvVJbpXkXq2182dc0opRVXNJ/jLJ45PsseCamD36m9f8qLX205kUuDJ8P8mhST68YP6Z6e5wvCbJhTu7qBXoxem+tV/XWru6n/fRqvqFJK+uqncu8rovft6WJHNj5u87shwmpu/dcfreqdD3Toe+d3msuL7XNehMTX+K2D8luVuS/7+9e42VqyrjMP78pQKiEQu1wXhBMGBEsQRREUXERFruRCkIGkUTDBj5ICYCQQIoBNAgaqQBol+4BgtBEREqakEuBrkUkcQQaAtWKPcSKbYH0tcPe59kPMw5ndI558w0zy+ZrNO11571np22a97Za619QFU9OM0hDZsdgS2By2mm4oy+oNnV8wVg1+kJbWg8NE796LekDmy92RV4oOMDwqi7gW2B2VMf0ibhIWCHJGMfO7QLMAI8MvUhadg59m40x96N59jbH469k2Poxl4TdPVF+7zVK2g2sji0qv46zSENoyXAvl1e0Hxw2JcB/E9kwFzXlnPH1M8FVlTVyimOZ1itBHZLsvmY+o8DaxjAb5uHxPU0z7edP1rRPurlSGBRVa2drsA0nBx7+8Kxd+M59vaHY+/kGLqx1ynu6pcLaf7inw2sTrJnx7EVTrdbv6paBSweW58E4LGqes0xvcaNwJ+Bi5PMApYChwP74ZrMDfFzYCHw2yQLaNbBHQIcBVxQVSPTGdygSnJ4++NH2nL/JM8Az1TVrVW1JMnVwE/au57LgOOBHYAvTX3E2gQ49m4kx96+cOztD8fe12FTHHvjng3qhyTLge3HOXxmVZ0xddFsWpIUcHZVfW+6YxkGSd4KnEPz4WAmzaNfzq2qK6c1sCGTZH+ajZM+SDP981HgEuBi12J21/5b7ebWqvpM22Z0I6qjgbcBDwAnmQTo9XDsnTyOvRvGsbc/HHs33KY49pqgS5IkSZI0AFyDLkmSJEnSADBBlyRJkiRpAJigS5IkSZI0AEzQJUmSJEkaACbokiRJkiQNABN0SZIkSZIGgAm6JEmSJEkDwARdkqQBlOSYJNXxWp1keZLrkhyRZNLG8CRnJKkxdZXkjMnqU5IkmaBLkjTo5gOfAA4ATgPWAlcBi5K8aZL6/EXbpyRJmkIzpjsASZI0oSVV9UjHny9LshBYCPwQOKHfHVbVCmBFv993UCTZoqrWTncckiSN5R10SZKGTFVdC/wGODbJVqP1SbZKcl6SZUlG2vLUsdPhk7w9yYIk/0qyti0vS7JFe/w1U9y7STInyfVJXkjy3yR3JNm7h/N2bqfqP51kTZLHkyxMMqOjzYQxtm3mJbmr7fvFJL9O8v4xfS1OcnuSg5Pcn2Qt8M322IwkpyT5Z9vHE0nOT7Ll+n4HSZImg3fQJUkaTjcChwF7ALe1ye3NwC7AD4AHgT1ppsVvA3wHIMlM4M627izg78Bs4FBgc5op9OuVZHfgL8D9wLHAy8BxwC1J9qqqeyc4/QZgFXA88CzwTpop/G/oNcYk84DfAX8CjgTeAnwfuD3JblX1747+dgZ+1l6XpcDzbf3lwMHAeW1/H2jbvBf4Qi/XQZKkfjJBlyRpOD3elu9oy6OATwH7VNVtbd0fkwCcnuS8qnoa+DawI7BHVd3f8X5XbWD/P2pj+GxVjQAkuRn4B82XAod1OynJLGAn4NCqur7j0JUdP/cS41k0yfb+VfVq+953AQ/TfBlxYkfbWcB+VbWkI469aRL7r1bVpW31LUmeBy5vk/wlSJI0hZziLknScEpbjk5Fnwc8BtzZTt2e0d5VXwS8keZuOsB+wN/GJL4b1nGzOd0+NOvg13X0FeAW4NMTnP4cTWJ9bpJjk+zUpc2EMSZ5M7A7cPVocg5QVcuAO9rYOi3vkmzPA0aAa7tcL9bzO0iSNClM0CVJGk7vbssn23I2sD3wypjX3e3xbTvKjd0AbhtgM5o75WP7+xYwc7zHwFVVAZ8D7gHOAR5OsjTJ8R3N1hfjTJovA57scmxlG1+nbu1m00yXf2lM/E93xCBJ0pRyirskScPpQGANMLrW+zlgGXDEOO2Xt+Xomu+NsQpYB1wIXNqtQVWtG+/kqloKfCXN/Ps5NEn9giTLq+r3PcT4As3Mge26HNuO5lr8X5dd2j1Hc/3G29TuiQn6lyRpUngHXZKkIZPk88AhwEVV9XJbfRPNXfWXquqeLq9n23aLgI8lmfN6+6+q1TQbxM0B7uvWX4/vU+3U89H14h/qJca2/3uB+Uk2G61Psj2wF3BrD93fBGwJbD3O9TJBlyRNOe+gS5I02HZrN1bbHHgPcBAwH/gDcEpHuyuAr9FsDHc+8EB7zvtokvnD2mT+AuBomg3RzqLZ7X0WzQ7px1XVf3qM60TgNuDmJL+kmUY+i2Zt+GZVdXK3k5J8GPgpcDXwCM1U+WOAV2l2ZKfHGE+j2cX9hiQLaHZxPxN4ETh/fcFX1eIkVwHXJPkxzVKAdTQ7uB8AnFRVD/d4LSRJ6gsTdEmSBtvCtlxDsz76PuCLwDXtem4AquqVJHOBk4FvADsAq4FHaRLZkbbdqiSfpNkF/WSatdZP0STHI70GVVX3JfkocDrNI8y2Bp5p47toglNX0uz+fiLwrvb3ehA4aPTRbL3EWFU3JTmw7f9Xbf1i4LsbcPf7y8AJwNeBU2keMbec5nF1T/X4HpIk9U06xnZJkiRJkjRNXIMuSZIkSdIAMEGXJEmSJGkAmKBLkiRJkjQATNAlSZIkSRoAJuiSJEmSJA0AE3RJkiRJkgaACbokSZIkSQPABF2SJEmSpAHwPx2baMAZy05jAAAAAElFTkSuQmCC\n",
- "text/plain": [
- "<Figure size 1008x504 with 2 Axes>"
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- }
- ],
- "source": [
- "plt.bar(range(1, 11), compas.decile_score.value_counts(), ec='black')\n",
- "plt.title(\"Decile scores of all defendants\")\n",
- "plt.ylabel(\"Frequency\")\n",
- "plt.xlabel(\"Decile score\")\n",
- "plt.xticks(range(1, 11))\n",
- "plt.show()\n",
- "\n",
- "fig, ax = compas.query(\"race in ['Caucasian', 'African-American']\").hist(\n",
- " \"decile_score\",\n",
- " by=\"race\",\n",
- " figsize=(14, 7),\n",
- " sharey=True,\n",
- " xrot='horizontal',\n",
- " ec='black',\n",
- " bins=np.arange(0.5, 11.5, 1.0),\n",
- " rwidth=0.8)\n",
- "\n",
- "fig.text(-1.5, 350, \"Frequency\", rotation='vertical')\n",
- "fig.text(11.5, -60, \"Decile score\", horizontalalignment='center')\n",
- "plt.tight_layout(w_pad=-2)\n",
- "plt.show()"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 68,
- "metadata": {
- "scrolled": false
- },
- "outputs": [
- {
- "name": "stderr",
- "output_type": "stream",
- "text": [
- "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\scipy\\stats\\stats.py:1713: FutureWarning: Using a non-tuple sequence for multidimensional indexing is deprecated; use `arr[tuple(seq)]` instead of `arr[seq]`. In the future this will be interpreted as an array index, `arr[np.array(seq)]`, which will result either in an error or a different result.\n",
- " return np.add.reduce(sorted[indexer] * weights, axis=axis) / sumval\n"
- ]
- },
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- "<Figure size 1008x504 with 1 Axes>"
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- }
- ],
- "source": [
- "sns.distplot(compas.age)\n",
- "plt.title(\"Histogram of defendants' ages\")\n",
- "plt.xlabel(\"Age of defendant\")\n",
- "plt.show()"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 69,
- "metadata": {
- "scrolled": false
- },
- "outputs": [
- {
- "data": {
- "text/html": [
- "<div>\n",
- "<style scoped>\n",
- " .dataframe tbody tr th:only-of-type {\n",
- " vertical-align: middle;\n",
- " }\n",
- "\n",
- " .dataframe tbody tr th {\n",
- " vertical-align: top;\n",
- " }\n",
- "\n",
- " .dataframe thead th {\n",
- " text-align: right;\n",
- " }\n",
- "</style>\n",
- "<table border=\"1\" class=\"dataframe\">\n",
- " <thead>\n",
- " <tr style=\"text-align: right;\">\n",
- " <th>is_recid</th>\n",
- " <th>0</th>\n",
- " <th>1</th>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>age_cat</th>\n",
- " <th></th>\n",
- " <th></th>\n",
- " </tr>\n",
- " </thead>\n",
- " <tbody>\n",
- " <tr>\n",
- " <th>25 - 45</th>\n",
- " <td>1784</td>\n",
- " <td>1748</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>Greater than 45</th>\n",
- " <td>847</td>\n",
- " <td>446</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>Less than 25</th>\n",
- " <td>551</td>\n",
- " <td>796</td>\n",
- " </tr>\n",
- " </tbody>\n",
- "</table>\n",
- "</div>"
- ],
- "text/plain": [
- "is_recid 0 1\n",
- "age_cat \n",
- "25 - 45 1784 1748\n",
- "Greater than 45 847 446\n",
- "Less than 25 551 796"
- ]
- },
- "metadata": {},
- "output_type": "display_data"
- },
- {
- "data": {
- "text/html": [
- "<div>\n",
- "<style scoped>\n",
- " .dataframe tbody tr th:only-of-type {\n",
- " vertical-align: middle;\n",
- " }\n",
- "\n",
- " .dataframe tbody tr th {\n",
- " vertical-align: top;\n",
- " }\n",
- "\n",
- " .dataframe thead th {\n",
- " text-align: right;\n",
- " }\n",
- "</style>\n",
- "<table border=\"1\" class=\"dataframe\">\n",
- " <thead>\n",
- " <tr style=\"text-align: right;\">\n",
- " <th>is_recid</th>\n",
- " <th>0</th>\n",
- " <th>1</th>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>sex</th>\n",
- " <th></th>\n",
- " <th></th>\n",
- " </tr>\n",
- " </thead>\n",
- " <tbody>\n",
- " <tr>\n",
- " <th>Female</th>\n",
- " <td>740</td>\n",
- " <td>435</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>Male</th>\n",
- " <td>2442</td>\n",
- " <td>2555</td>\n",
- " </tr>\n",
- " </tbody>\n",
- "</table>\n",
- "</div>"
- ],
- "text/plain": [
- "is_recid 0 1\n",
- "sex \n",
- "Female 740 435\n",
- "Male 2442 2555"
- ]
- },
- "metadata": {},
- "output_type": "display_data"
- },
- {
- "data": {
- "text/html": [
- "<div>\n",
- "<style scoped>\n",
- " .dataframe tbody tr th:only-of-type {\n",
- " vertical-align: middle;\n",
- " }\n",
- "\n",
- " .dataframe tbody tr th {\n",
- " vertical-align: top;\n",
- " }\n",
- "\n",
- " .dataframe thead th {\n",
- " text-align: right;\n",
- " }\n",
- "</style>\n",
- "<table border=\"1\" class=\"dataframe\">\n",
- " <thead>\n",
- " <tr style=\"text-align: right;\">\n",
- " <th></th>\n",
- " <th>is_recid</th>\n",
- " <th>0</th>\n",
- " <th>1</th>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>race</th>\n",
- " <th>age_cat</th>\n",
- " <th></th>\n",
- " <th></th>\n",
- " </tr>\n",
- " </thead>\n",
- " <tbody>\n",
- " <tr>\n",
- " <th rowspan=\"3\" valign=\"top\">African-American</th>\n",
- " <th>25 - 45</th>\n",
- " <td>847.0</td>\n",
- " <td>1051.0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>Greater than 45</th>\n",
- " <td>261.0</td>\n",
- " <td>207.0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>Less than 25</th>\n",
- " <td>294.0</td>\n",
- " <td>515.0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th rowspan=\"3\" valign=\"top\">Asian</th>\n",
- " <th>25 - 45</th>\n",
- " <td>10.0</td>\n",
- " <td>4.0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>Greater than 45</th>\n",
- " <td>7.0</td>\n",
- " <td>4.0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>Less than 25</th>\n",
- " <td>4.0</td>\n",
- " <td>2.0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th rowspan=\"3\" valign=\"top\">Caucasian</th>\n",
- " <th>25 - 45</th>\n",
- " <td>620.0</td>\n",
- " <td>508.0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>Greater than 45</th>\n",
- " <td>442.0</td>\n",
- " <td>186.0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>Less than 25</th>\n",
- " <td>167.0</td>\n",
- " <td>180.0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th rowspan=\"3\" valign=\"top\">Hispanic</th>\n",
- " <th>25 - 45</th>\n",
- " <td>180.0</td>\n",
- " <td>111.0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>Greater than 45</th>\n",
- " <td>81.0</td>\n",
- " <td>28.0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>Less than 25</th>\n",
- " <td>51.0</td>\n",
- " <td>58.0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th rowspan=\"3\" valign=\"top\">Native American</th>\n",
- " <th>25 - 45</th>\n",
- " <td>5.0</td>\n",
- " <td>2.0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>Greater than 45</th>\n",
- " <td>NaN</td>\n",
- " <td>2.0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>Less than 25</th>\n",
- " <td>NaN</td>\n",
- " <td>2.0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th rowspan=\"3\" valign=\"top\">Other</th>\n",
- " <th>25 - 45</th>\n",
- " <td>122.0</td>\n",
- " <td>72.0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>Greater than 45</th>\n",
- " <td>56.0</td>\n",
- " <td>19.0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>Less than 25</th>\n",
- " <td>35.0</td>\n",
- " <td>39.0</td>\n",
- " </tr>\n",
- " </tbody>\n",
- "</table>\n",
- "</div>"
- ],
- "text/plain": [
- "is_recid 0 1\n",
- "race age_cat \n",
- "African-American 25 - 45 847.0 1051.0\n",
- " Greater than 45 261.0 207.0\n",
- " Less than 25 294.0 515.0\n",
- "Asian 25 - 45 10.0 4.0\n",
- " Greater than 45 7.0 4.0\n",
- " Less than 25 4.0 2.0\n",
- "Caucasian 25 - 45 620.0 508.0\n",
- " Greater than 45 442.0 186.0\n",
- " Less than 25 167.0 180.0\n",
- "Hispanic 25 - 45 180.0 111.0\n",
- " Greater than 45 81.0 28.0\n",
- " Less than 25 51.0 58.0\n",
- "Native American 25 - 45 5.0 2.0\n",
- " Greater than 45 NaN 2.0\n",
- " Less than 25 NaN 2.0\n",
- "Other 25 - 45 122.0 72.0\n",
- " Greater than 45 56.0 19.0\n",
- " Less than 25 35.0 39.0"
- ]
- },
- "metadata": {},
- "output_type": "display_data"
- }
- ],
- "source": [
- "tab = compas.groupby(['age_cat', 'is_recid']).size()\n",
- "display(tab.unstack())\n",
- "\n",
- "tab = compas.groupby(['sex', 'is_recid']).size()\n",
- "display(tab.unstack())\n",
- "\n",
- "tab = compas.groupby(['race', 'age_cat', 'is_recid']).size()\n",
- "display(tab.unstack())"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "From above it is clear that there are no Native American recidivists of age over 45 or under 25. There are some other value combinations that might be problematic. Therefore the procedure of estimating $P(X=x)$ has to be considered carefully."
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "### Synthetic data\n",
- "\n",
- "In the chunk below, we generate the synthetic data as described by Lakkaraju et al. The default values and definitions of $Y$ and $T$ values follow their description.\n",
- "\n",
- "**Parameters**\n",
- "\n",
- "* M = `nJudges_M`, number of judges\n",
- "* N = `nSubjects_N`, number of subjects assigned to each judge\n",
- "* betas $\\beta_i$ = `beta_i`, where $i \\in \\{X, Z, W\\}$ are coefficients for the respected variables\n",
- "* R = `acceptanceRate_R`, acceptance rates\n",
- "* X = `X`, invidual's features observable to all (models and judges)\n",
- "* Z = `Z`, information observable for judges only\n",
- "* W = `W`, unobservable / inaccessible information\n",
- "* T = `decision_T`, bail-or-jail decisions where $T=0$ represents jail decision and $T=1$ bail decision.\n",
- "* Y = `result_Y`, result variable, if $Y=0$ person will or would recidivate and if $Y=1$ person will or would not commit a crime."
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 104,
- "metadata": {
- "scrolled": false
- },
- "outputs": [
- {
- "data": {
- "text/html": [
- "<div>\n",
- "<style scoped>\n",
- " .dataframe tbody tr th:only-of-type {\n",
- " vertical-align: middle;\n",
- " }\n",
- "\n",
- " .dataframe tbody tr th {\n",
- " vertical-align: top;\n",
- " }\n",
- "\n",
- " .dataframe thead th {\n",
- " text-align: right;\n",
- " }\n",
- "</style>\n",
- "<table border=\"1\" class=\"dataframe\">\n",
- " <thead>\n",
- " <tr style=\"text-align: right;\">\n",
- " <th></th>\n",
- " <th>count</th>\n",
- " <th>mean</th>\n",
- " <th>std</th>\n",
- " <th>min</th>\n",
- " <th>25%</th>\n",
- " <th>50%</th>\n",
- " <th>75%</th>\n",
- " <th>max</th>\n",
- " </tr>\n",
- " </thead>\n",
- " <tbody>\n",
- " <tr>\n",
- " <th>judgeID_J</th>\n",
- " <td>50000.0</td>\n",
- " <td>49.500000</td>\n",
- " <td>28.866359</td>\n",
- " <td>0.000000</td>\n",
- " <td>24.750000</td>\n",
- " <td>49.500000</td>\n",
- " <td>74.250000</td>\n",
- " <td>99.000000</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>acceptanceRate_R</th>\n",
- " <td>50000.0</td>\n",
- " <td>0.489100</td>\n",
- " <td>0.241555</td>\n",
- " <td>0.113154</td>\n",
- " <td>0.258107</td>\n",
- " <td>0.474092</td>\n",
- " <td>0.714801</td>\n",
- " <td>0.898779</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>X</th>\n",
- " <td>50000.0</td>\n",
- " <td>-0.008054</td>\n",
- " <td>0.998408</td>\n",
- " <td>-4.050908</td>\n",
- " <td>-0.680597</td>\n",
- " <td>-0.008397</td>\n",
- " <td>0.660901</td>\n",
- " <td>4.099418</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>Z</th>\n",
- " <td>50000.0</td>\n",
- " <td>-0.004696</td>\n",
- " <td>0.993683</td>\n",
- " <td>-4.182233</td>\n",
- " <td>-0.680335</td>\n",
- " <td>-0.004356</td>\n",
- " <td>0.666608</td>\n",
- " <td>3.966532</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>W</th>\n",
- " <td>50000.0</td>\n",
- " <td>-0.000542</td>\n",
- " <td>0.995303</td>\n",
- " <td>-4.189579</td>\n",
- " <td>-0.671069</td>\n",
- " <td>0.002007</td>\n",
- " <td>0.671735</td>\n",
- " <td>4.276601</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>result_Y</th>\n",
- " <td>50000.0</td>\n",
- " <td>0.503380</td>\n",
- " <td>0.499994</td>\n",
- " <td>0.000000</td>\n",
- " <td>0.000000</td>\n",
- " <td>1.000000</td>\n",
- " <td>1.000000</td>\n",
- " <td>1.000000</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>probabilities_T</th>\n",
- " <td>50000.0</td>\n",
- " <td>0.498167</td>\n",
- " <td>0.278933</td>\n",
- " <td>-0.295551</td>\n",
- " <td>0.276483</td>\n",
- " <td>0.496720</td>\n",
- " <td>0.720596</td>\n",
- " <td>1.261540</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>decision_T</th>\n",
- " <td>50000.0</td>\n",
- " <td>0.488120</td>\n",
- " <td>0.499864</td>\n",
- " <td>0.000000</td>\n",
- " <td>0.000000</td>\n",
- " <td>0.000000</td>\n",
- " <td>1.000000</td>\n",
- " <td>1.000000</td>\n",
- " </tr>\n",
- " </tbody>\n",
- "</table>\n",
- "</div>"
- ],
- "text/plain": [
- " count mean std min 25% \\\n",
- "judgeID_J 50000.0 49.500000 28.866359 0.000000 24.750000 \n",
- "acceptanceRate_R 50000.0 0.489100 0.241555 0.113154 0.258107 \n",
- "X 50000.0 -0.008054 0.998408 -4.050908 -0.680597 \n",
- "Z 50000.0 -0.004696 0.993683 -4.182233 -0.680335 \n",
- "W 50000.0 -0.000542 0.995303 -4.189579 -0.671069 \n",
- "result_Y 50000.0 0.503380 0.499994 0.000000 0.000000 \n",
- "probabilities_T 50000.0 0.498167 0.278933 -0.295551 0.276483 \n",
- "decision_T 50000.0 0.488120 0.499864 0.000000 0.000000 \n",
- "\n",
- " 50% 75% max \n",
- "judgeID_J 49.500000 74.250000 99.000000 \n",
- "acceptanceRate_R 0.474092 0.714801 0.898779 \n",
- "X -0.008397 0.660901 4.099418 \n",
- "Z -0.004356 0.666608 3.966532 \n",
- "W 0.002007 0.671735 4.276601 \n",
- "result_Y 1.000000 1.000000 1.000000 \n",
- "probabilities_T 0.496720 0.720596 1.261540 \n",
- "decision_T 0.000000 1.000000 1.000000 "
- ]
- },
- "metadata": {},
- "output_type": "display_data"
- },
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "0 25594\n",
- "1 24406\n",
- "Name: decision_T, dtype: int64\n"
- ]
- },
- {
- "data": {
- "text/html": [
- "<div>\n",
- "<style scoped>\n",
- " .dataframe tbody tr th:only-of-type {\n",
- " vertical-align: middle;\n",
- " }\n",
- "\n",
- " .dataframe tbody tr th {\n",
- " vertical-align: top;\n",
- " }\n",
- "\n",
- " .dataframe thead th {\n",
- " text-align: right;\n",
- " }\n",
- "</style>\n",
- "<table border=\"1\" class=\"dataframe\">\n",
- " <thead>\n",
- " <tr style=\"text-align: right;\">\n",
- " <th>decision_T</th>\n",
- " <th>0</th>\n",
- " <th>1</th>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>result_Y</th>\n",
- " <th></th>\n",
- " <th></th>\n",
- " </tr>\n",
- " </thead>\n",
- " <tbody>\n",
- " <tr>\n",
- " <th>0.0</th>\n",
- " <td>19544</td>\n",
- " <td>5287</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>1.0</th>\n",
- " <td>6050</td>\n",
- " <td>19119</td>\n",
- " </tr>\n",
- " </tbody>\n",
- "</table>\n",
- "</div>"
- ],
- "text/plain": [
- "decision_T 0 1\n",
- "result_Y \n",
- "0.0 19544 5287\n",
- "1.0 6050 19119"
- ]
- },
- "metadata": {},
- "output_type": "display_data"
- }
- ],
- "source": [
- "# Set seed for reproducibility\n",
- "npr.seed(111)\n",
- "\n",
- "def generateData(nJudges_M=100,\n",
- " nSubjects_N=500,\n",
- " beta_X=1.0,\n",
- " beta_Z=1.0,\n",
- " beta_W=0.2):\n",
- "\n",
- " # Assign judge IDs as running numbering from 0 to nJudges_M - 1\n",
- " judgeID_J = np.repeat(np.arange(0, nJudges_M, dtype=np.int32), nSubjects_N)\n",
- "\n",
- " # Sample acceptance rates uniformly from a closed interval\n",
- " # from 0.1 to 0.9 and round to tenth decimal place.\n",
- " acceptance_rates = np.round(npr.uniform(.1, .9, nJudges_M), 10)\n",
- "\n",
- " # Replicate the rates so they can be attached to the corresponding judge ID.\n",
- " acceptanceRate_R = np.repeat(acceptance_rates, nSubjects_N)\n",
- "\n",
- " # Sample the variables from standard Gaussian distributions.\n",
- " X = npr.normal(size=nJudges_M * nSubjects_N)\n",
- " Z = npr.normal(size=nJudges_M * nSubjects_N)\n",
- " W = npr.normal(size=nJudges_M * nSubjects_N)\n",
- "\n",
- " probabilities_Y = 1 / (1 + np.exp(-(beta_X * X + beta_Z * Z + beta_W * W)))\n",
- "\n",
- " # 0 if P(Y = 0| X = x; Z = z; W = w) >= 0.5 , 1 otherwise\n",
- " result_Y = 1 - probabilities_Y.round()\n",
- "\n",
- " probabilities_T = 1 / (1 + np.exp(-(beta_X * X + beta_Z * Z)))\n",
- " probabilities_T += npr.normal(0, .1, nJudges_M * nSubjects_N)\n",
- "\n",
- " # Initialize decision values as 1\n",
- " decision_T = np.ones(nJudges_M * nSubjects_N)\n",
- "\n",
- " # Initialize the dataframe\n",
- " df_init = pd.DataFrame(\n",
- " np.column_stack((judgeID_J, acceptanceRate_R, X, Z, W, result_Y,\n",
- " probabilities_T, decision_T)),\n",
- " columns=[\n",
- " \"judgeID_J\", \"acceptanceRate_R\", \"X\", \"Z\", \"W\", \"result_Y\",\n",
- " \"probabilities_T\", \"decision_T\"\n",
- " ])\n",
- "\n",
- " # Sort by judges then probabilities\n",
- " data = df_init.sort_values(\n",
- " by=[\"judgeID_J\", \"probabilities_T\"], ascending=False)\n",
- "\n",
- " # Iterate over the data. Subject is in the top (1-r)*100% if\n",
- " # his within-judge-index is over acceptance threshold times\n",
- " # the number of subjects assigned to each judge. If subject\n",
- " # is over the limit they are assigned a zero, else one.\n",
- " data.reset_index(drop=True, inplace=True)\n",
- "\n",
- " data['decision_T'] = np.where(\n",
- " (data.index.values % nSubjects_N) <\n",
- " ((1 - data['acceptanceRate_R']) * nSubjects_N), 0, 1)\n",
- "\n",
- " return data\n",
- "\n",
- "\n",
- "df = []\n",
- "df = generateData()\n",
- "\n",
- "# Basic stats of the created data set.\n",
- "display(df.describe().T)\n",
- "\n",
- "print(df.decision_T.value_counts())\n",
- "\n",
- "tab = df.groupby(['result_Y', 'decision_T']).size()\n",
- "display(tab.unstack())"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 105,
- "metadata": {},
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "(25000, 8)\n",
- "(25000, 8)\n",
- "(12094, 8)\n",
- "(12312, 8)\n"
- ]
- },
- {
- "data": {
- "text/html": [
- "<div>\n",
- "<style scoped>\n",
- " .dataframe tbody tr th:only-of-type {\n",
- " vertical-align: middle;\n",
- " }\n",
- "\n",
- " .dataframe tbody tr th {\n",
- " vertical-align: top;\n",
- " }\n",
- "\n",
- " .dataframe thead th {\n",
- " text-align: right;\n",
- " }\n",
- "</style>\n",
- "<table border=\"1\" class=\"dataframe\">\n",
- " <thead>\n",
- " <tr style=\"text-align: right;\">\n",
- " <th>decision_T</th>\n",
- " <th>1</th>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>result_Y</th>\n",
- " <th></th>\n",
- " </tr>\n",
- " </thead>\n",
- " <tbody>\n",
- " <tr>\n",
- " <th>0.0</th>\n",
- " <td>2606</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <th>1.0</th>\n",
- " <td>9488</td>\n",
- " </tr>\n",
- " </tbody>\n",
- "</table>\n",
- "</div>"
- ],
- "text/plain": [
- "decision_T 1\n",
- "result_Y \n",
- "0.0 2606\n",
- "1.0 9488"
- ]
- },
- "execution_count": 105,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
- "source": [
- "# Split the data set to test and train\n",
- "from sklearn.model_selection import train_test_split\n",
- "train, test = train_test_split(df, test_size=0.5, random_state=0)\n",
- "\n",
- "print(train.shape)\n",
- "print(test.shape)\n",
- "\n",
- "train_labeled = train[train.decision_T == 1]\n",
- "test_labeled = test[test.decision_T == 1]\n",
- "\n",
- "print(train_labeled.shape)\n",
- "print(test_labeled.shape)\n",
- "\n",
- "tab = train_labeled.groupby(['result_Y', 'decision_T']).size()\n",
- "tab.unstack()"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "## Algorithms\n",
- "\n",
- "### Contraction algorithm\n",
- "\n",
- "Below is an implementation of Lakkaraju's team's algorithm presented in [their paper](https://helka.finna.fi/PrimoRecord/pci.acm3098066). Relevant parameters to be passed to the function are presented in the description."
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 106,
- "metadata": {},
- "outputs": [],
- "source": [
- "def contraction(df,\n",
- " judgeIDJ_col,\n",
- " decisionT_col,\n",
- " resultY_col,\n",
- " modelProbS_col,\n",
- " accRateR_col,\n",
- " r,\n",
- " binning=False):\n",
- " '''\n",
- " This is an implementation of the algorithm presented by Lakkaraju\n",
- " et al. in their paper \"The Selective Labels Problem: Evaluating \n",
- " Algorithmic Predictions in the Presence of Unobservables\" (2017).\n",
- " \n",
- " Parameters:\n",
- " df = The (Pandas) data frame containing the data, judge decisions,\n",
- " judge IDs, results and probability scores.\n",
- " judgeIDJ_col = String, the name of the column containing the judges' IDs\n",
- " in df.\n",
- " decisionT_col = String, the name of the column containing the judges' decisions\n",
- " resultY_col = String, the name of the column containing the realization\n",
- " modelProbS_col = String, the name of the column containing the probability\n",
- " scores from the black-box model B.\n",
- " accRateR_col = String, the name of the column containing the judges' \n",
- " acceptance rates\n",
- " r = Float between 0 and 1, the given acceptance rate.\n",
- " binning = Boolean, should judges with same acceptance rate be binned\n",
- " \n",
- " Returns:\n",
- " u = The estimated failure rate at acceptance rate r.\n",
- " '''\n",
- " # Sort first by acceptance rate and judge ID.\n",
- " sorted_df = df.sort_values(\n",
- " by=[accRateR_col, judgeIDJ_col], ascending=False)\n",
- "\n",
- " if binning:\n",
- " # Get maximum leniency\n",
- " max_leniency = sorted_df[accRateR_col].values[0].round(1)\n",
- "\n",
- " # Get list of judges that are the most lenient\n",
- " most_lenient_list = sorted_df.loc[sorted_df[accRateR_col].round(1) ==\n",
- " max_leniency, judgeIDJ_col]\n",
- "\n",
- " # Subset to obtain D_q\n",
- " D_q = sorted_df[sorted_df[judgeIDJ_col].isin(\n",
- " most_lenient_list.unique())]\n",
- " else:\n",
- " # Get most lenient judge\n",
- " most_lenient_ID = sorted_df[judgeIDJ_col].values[0]\n",
- "\n",
- " # Subset\n",
- " D_q = sorted_df[sorted_df[judgeIDJ_col] == most_lenient_ID]\n",
- "\n",
- " R_q = D_q[D_q[decisionT_col] == 1]\n",
- "\n",
- " R_sort_q = R_q.sort_values(by=modelProbS_col, ascending=False)\n",
- "\n",
- " number_to_remove = int(\n",
- " np.round((1 - r) * D_q.shape[0] - (D_q.shape[0] - R_q.shape[0])))\n",
- "\n",
- " R_B = R_sort_q[number_to_remove:R_sort_q.shape[0]]\n",
- "\n",
- " return np.sum(R_B[resultY_col] == 0) / D_q.shape[0]"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "### Causal model\n",
- "\n",
- "Our model is defined by the probabilistic expression \n",
- "\n",
- "\\begin{equation}\\label{model}\n",
- "P(Y=0 | \\text{do}(R=r)) = \\sum_x \\underbrace{P(Y=0|X=x, T=1)}_\\text{1} \n",
- "\\overbrace{P(T=1|R=r, X=x)}^\\text{2} \n",
- "\\underbrace{P(X=x)}_\\text{3}\n",
- "\\end{equation}\n",
- "\n",
- "As a picture (Z not in model):\n",
- "\n",
- "![Model as picture](../figures/intervention_model.png \"Intervention model\")\n",
- "\n",
- "<!---\n",
- "**Algorithm -- UPDATE!!**\n",
- "\n",
- "Our model will be constructed sequentially.\n",
- "\n",
- "Input: Training and test data sets $(\\mathbf{x}, t, y) \\in \\mathcal{D}$ and acceptance rate $r$. \n",
- "Returns: $P(Y=0 | \\text{do}(R=r))$\n",
- "\n",
- "Procedure:\n",
- "1. Model $P(X=x)$ in a suitable way and assign to $\\mathcal{M}_0$\n",
- "* Build model $\\mathcal{M}_1$ predicting response $Y$ with predictors $X$ from the labeled observations (where $T=1$) in training data.\n",
- "* Predict $P(Y=0|X=x)$ for every observation in the test data using model $\\mathcal{M}_1$.\n",
- "* Initialize `sum = 0`\n",
- "* For every point in the parameter space (for every $x$ in $X$)\n",
- " 1. $p_x \\leftarrow P(X=x)$ from $\\mathcal{M}_0$\n",
- " * $\\mathcal{D_x} \\leftarrow \\{\\mathcal{D} | X = x\\}$\n",
- " * Assign first $r\\cdot 100\\%$ observations from $\\mathcal{D_x}$ to $\\mathcal{D}_{rx}$\n",
- " * $p_t \\leftarrow \\dfrac{|\\{\\mathcal{D}_{rx}|T=1\\}|}{|\\mathcal{D}_{rx}|}$ (part 2 of eq. $\\ref{model}$) Pitääkö tähänkin treenaa joku oma luokittelija?\n",
- " * $p_y$ will be predicted from the model $\\mathcal{M}_1$\n",
- " * `sum +=` $p_y \\cdot p_t \\cdot p_x$\n",
- "* Return `sum`\n",
- "--->\n",
- "**Constructing $\\mathcal{M}_0$, preliminary ideas:**\n",
- "\n",
- "* Approximate $P(X=x)$ with frequencies (makes independence assumption, make variables factors first)\n",
- "* Construct Bayesian network using some well-known algorithm.\n",
- "\n",
- "Functions:\n",
- "\n",
- "* $f(x)$ gives probability of recidivism given personal properties and predictive model.\n",
- "* `ep` counts performance of the predictive model given a data, model and leniency rate like Michael's pdf."
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 107,
- "metadata": {
- "scrolled": false
- },
- "outputs": [],
- "source": [
- "def f(x, model, failure_value):\n",
- " '''\n",
- " Returns the probability of negative event (e.g. recidivism) given individual \n",
- " features (x), predictive model (model) and value for failure.\n",
- " '''\n",
- " if x.ndim == 1:\n",
- " # if x is vector, transform to column matrix.\n",
- " f_values = model.predict_proba(np.array(x).reshape(-1, 1))\n",
- " else:\n",
- " f_values = model.predict_proba(x)\n",
- "\n",
- " return f_values[:, model.classes_ == failure_value].flatten()\n",
- "\n",
- "\n",
- "def ep(r, df, result_col, feature_cols, model, failure_value):\n",
- " '''\n",
- " Returns:\n",
- " Empirical performance, i.e. percentage of recidivists. \n",
- " \n",
- " Parameters:\n",
- " r = leniency rate(s)\n",
- " df = test data, pandas DataFrame\n",
- " result_col = String (list), name of column containing the binarized results.\n",
- " feature_cols = String (list), name of columns containge individual features.\n",
- " model = trained sklearn classifier \n",
- " failure_value = value obtained from the model.classes_ representing the \n",
- " unwanted event label (usually 0 or 1).\n",
- " '''\n",
- " rates = np.zeros_like(r)\n",
- " for i in range(len(rates)):\n",
- " rates[i] = np.mean((df[result_col] == failure_value) &\n",
- " (f(df[feature_cols], model, failure_value) < r[i]))\n",
- " return rates\n",
- "\n",
- "def gp(r, x_values, y_model, x_model, failure_value):\n",
- " '''\n",
- " Returns:\n",
- " Generalized performance\n",
- " \n",
- " Parameters:\n",
- " r = leniency rate\n",
- " df = test data, pandas DataFrame\n",
- " feature_cols = String (list), name of columns containing individual features.\n",
- " y_model = trained sklearn classifier to predict response\n",
- " x_model = model of P(X=x)\n",
- " failure_value = value obtained from the model.classes_ representing the \n",
- " unwanted event label.\n",
- " '''\n",
- " preds = f(x_values, y_model, failure_value)\n",
- " \n",
- " return np.sum(preds * (preds < r) * x_model(x_values))"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "## Performance comparison\n",
- "\n",
- "Below we try to replicate the results obtained by Lakkaraju and compare their model's performance to the one of ours.\n",
- "\n",
- "### On synthetic data\n",
- "\n",
- "#### Predictive models\n",
- "\n",
- "Lakkaraju says that they used logistic regression to predict recidivism. We models using only *observed observations*, i.e. defendants that were granted bail and are in the train set. We then predict the probability of recidivism for all observations in the test data and attach it to our data set. I also applied random forest classifier."
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 108,
- "metadata": {},
- "outputs": [],
- "source": [
- "# instantiate the model (using the default parameters)\n",
- "logreg = LogisticRegression(solver='lbfgs')\n",
- "\n",
- "# fit, reshape X to be of shape (n_samples, n_features)\n",
- "logreg = logreg.fit(train_labeled.X.values.reshape(-1, 1), train_labeled.result_Y)\n",
- "\n",
- "# predict probabilities and attach to data\n",
- "label_probs_logreg = logreg.predict_proba(test.X.values.reshape(-1, 1))\n",
- "test = test.assign(B_prob_0_logreg=label_probs_logreg[:, 0])\n",
- "\n",
- "label_probs_logreg = logreg.predict_proba(test_labeled.X.values.reshape(-1, 1))\n",
- "test_labeled = test_labeled.assign(B_prob_0_logreg=label_probs_logreg[:, 0])\n",
- "test_labeled = test_labeled.assign(B_prob_1_logreg=label_probs_logreg[:, 1])\n",
- "\n",
- "########\n",
- "\n",
- "# instantiate the model (using the default parameters)\n",
- "forest = RandomForestClassifier(n_estimators=400, max_depth=8, random_state=0)\n",
- "\n",
- "# fit, reshape X to be of shape (n_samples, n_features)\n",
- "forest = forest.fit(train_labeled.X.values.reshape(-1, 1), train_labeled.result_Y)\n",
- "\n",
- "# predict probabilities and attach to data\n",
- "label_probs_forest = forest.predict_proba(test.X.values.reshape(-1, 1))\n",
- "test = test.assign(B_prob_0_forest=label_probs_forest[:, 0])\n",
- "\n",
- "label_probs_forest = forest.predict_proba(test_labeled.X.values.reshape(-1, 1))\n",
- "test_labeled = test_labeled.assign(B_prob_0_forest=label_probs_forest[:, 0])"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "#### Visual comparison\n",
- "\n",
- "Let's plot the failure rates against the acceptance rates using the difference."
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 109,
- "metadata": {
- "scrolled": false
- },
- "outputs": [
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- "<Figure size 1008x576 with 1 Axes>"
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- },
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "[[0. 0. 0. 0. 0.0165 0.0191]\n",
- " [0.0137 0.0137 0.0011 0.0123 0.0531 0.0568]\n",
- " [0.0457 0.0502 0.0076 0.0492 0.0905 0.0838]\n",
- " [0.0731 0.0731 0.049 0.0902 0.1227 0.1326]\n",
- " [0.1142 0.1187 0.0846 0.1311 0.1552 0.1607]\n",
- " [0.1644 0.1689 0.1742 0.1598 0.1786 0.1778]\n",
- " [0.2009 0.2146 0.277 0.2377 0.197 0.1974]\n",
- " [0.2785 0.2877 0.3625 0.3115 0.2095 0.2069]]\n"
- ]
- }
- ],
- "source": [
- "failure_rates = np.zeros((8, 6))\n",
- "\n",
- "# sort whole test data by \n",
- "#test_sorted = test.sort_values(by='B_prob_0_logreg', ascending=False)\n",
- "\n",
- "for r in np.arange(1, 9):\n",
- " ## Contraction, logistic regression\n",
- " failure_rates[r - 1, 0] = contraction(\n",
- " test_labeled, 'judgeID_J', 'decision_T', 'result_Y',\n",
- " 'B_prob_0_logreg', 'acceptanceRate_R', r / 10, False)\n",
- " \n",
- " ## Contraction, random forest\n",
- " failure_rates[r - 1, 1] = contraction(\n",
- " test_labeled, 'judgeID_J', 'decision_T', 'result_Y',\n",
- " 'B_prob_0_forest', 'acceptanceRate_R', r / 10, False)\n",
- "\n",
- " ## Human error rate - Correct?\n",
- " # Get judges with correct leniency as list\n",
- " correct_leniency_list = test_labeled.judgeID_J[test_labeled['acceptanceRate_R'].round(1) ==\n",
- " r / 10]\n",
- "\n",
- " # Released are the people they judged and released, T = 1\n",
- " released = test_labeled[test_labeled.judgeID_J.isin(correct_leniency_list)]\n",
- "\n",
- " # Get their failure rate, aka ratio of reoffenders to number of people judged in total\n",
- " failure_rates[r - 1, 2] = np.sum(\n",
- " released.result_Y == 0) / correct_leniency_list.shape[0]\n",
- " \n",
- " ## True evaluation -- didn't mention using contraction here???\n",
- " failure_rates[r - 1, 3] = contraction(test, 'judgeID_J', 'decision_T',\n",
- " 'result_Y', 'B_prob_0_logreg',\n",
- " 'acceptanceRate_R', r / 10, False)\n",
- "\n",
- " ## Causal model with logistic regression\n",
- " failure_rates[r - 1, 4] = ep([r / 10], test_labeled, 'result_Y', 'X', logreg, 0)\n",
- " \n",
- " ## Causal model with random forest classifier\n",
- " failure_rates[r - 1, 5] = ep([r / 10], test_labeled, 'result_Y', 'X', forest, 0)\n",
- " \n",
- "\n",
- "# klassifikaatioille scipy.stats semin kautta error barit xerr ja yerr argumenttien kautta\n",
- "\n",
- "plt.figure(figsize=(14, 8))\n",
- "plt.plot(np.arange(0.1, 0.9, .1), failure_rates[:, 0], label='Contraction, logistic')\n",
- "#plt.plot(np.arange(0.1, 0.9, .1), failure_rates[:, 1], label='Contraction, forest')\n",
- "plt.plot(np.arange(0.1, 0.9, .1), failure_rates[:, 2], label='\"Human judges\"')\n",
- "plt.plot(np.arange(0.1, 0.9, .1), failure_rates[:, 3], label='True Evaluation')\n",
- "\n",
- "plt.plot(np.arange(0.1, 0.9, .1), failure_rates[:, 4], label='Causal model, log.')\n",
- "plt.plot(np.arange(0.1, 0.9, .1), failure_rates[:, 5], label='Causal model, r.f.')\n",
- "\n",
- "plt.title('Failure rate vs. Acceptance rate')\n",
- "plt.xlabel('Acceptance rate')\n",
- "plt.ylabel('Failure rate')\n",
- "plt.legend()\n",
- "plt.show()\n",
- "\n",
- "with np.printoptions(precision=4, suppress=True):\n",
- " print(failure_rates)"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "**Thoughts:**\n",
- "\n",
- "Failure rates still too high for about 10 percentage points compared to Lakkaraju paper. Failure rates will change if seed is changed (e.g. with seed 0 contraction's failure rates are approximately 0.31, causal doesn't change that much). It seems like the contraction or our model is some how predicting the wrong thing. Behavior after 0.5 is not consistent? (Curves curve down in Lakkaraju's paper. + Human evaluation curve jumps to the wrong side.) Have to check some rounding rules."
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 110,
- "metadata": {
- "scrolled": false
- },
- "outputs": [
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- "<Figure size 1008x576 with 1 Axes>"
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- },
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "[0.48356 0.48564 0.48748 0.48964 0.49132 0.49248 0.4932 0.49368 0.49428\n",
- " 0.49432]\n",
- "[0.3268 0.34608 0.36604 0.3874 0.4092 0.43092 0.45352 0.47776 0.49836\n",
- " 0.50568]\n",
- "1.0 12642\n",
- "0.0 12358\n",
- "Name: result_Y, dtype: int64\n",
- "1.0 9631\n",
- "0.0 2681\n",
- "Name: result_Y, dtype: int64\n"
- ]
- }
- ],
- "source": [
- "x_vals = np.linspace(0, 1, 100)\n",
- "y_vals = ep(x_vals, test_labeled, 'result_Y', 'X', logreg, 0)\n",
- "y_vals2 = ep(x_vals, test_labeled, 'result_Y', 'X', logreg, 1)\n",
- "\n",
- "y_vals3 = ep(x_vals, test, 'result_Y', 'X', logreg, 0)\n",
- "y_vals4 = ep(x_vals, test, 'result_Y', 'X', logreg, 1)\n",
- "\n",
- "plt.figure(figsize=(14, 8))\n",
- "plt.plot(x_vals, y_vals)\n",
- "plt.plot(x_vals, y_vals2)\n",
- "plt.plot(x_vals, y_vals3)\n",
- "plt.plot(x_vals, y_vals4)\n",
- "plt.show()\n",
- "\n",
- "print(y_vals3[-10:])\n",
- "print(y_vals4[-10:])\n",
- "print(test.result_Y.value_counts())\n",
- "print(test_labeled.result_Y.value_counts())"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 111,
- "metadata": {},
- "outputs": [
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- "<Figure size 1008x504 with 1 Axes>"
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- }
- ],
- "source": [
- "# Mindless comparison as X is continuous (we should integrate).\n",
- "\n",
- "thresholds = np.linspace(.0, 1.0)\n",
- "\n",
- "x_values = np.linspace(-10, 10, 1000)\n",
- "\n",
- "rates_logistic = np.zeros(0)\n",
- "rates_forest = np.zeros(0)\n",
- "\n",
- "for leniency in thresholds:\n",
- " rates_logistic = np.append(rates_logistic, gp(leniency, x_values, logreg, lambda x: scs.norm.pdf(x), 0))\n",
- " rates_forest = np.append(rates_forest, gp(leniency, x_values, forest, lambda x: scs.norm.pdf(x), 0))\n",
- "\n",
- "plt.plot(thresholds, rates_logistic, label=\"Logistic model\")\n",
- "plt.plot(thresholds, rates_forest, label=\"Random forest\")\n",
- "plt.title(\"Generalized performance\")\n",
- "plt.legend()\n",
- "plt.show()"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "### On COMPAS data\n",
- "\n",
- "\n",
- "#### Predictive models\n",
- "\n",
- "Let's build the predictive models (first here random forest and logistic regression). Some of our variables are string so they will first have to be transformed to be dummy / indicator variables."
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 112,
- "metadata": {},
- "outputs": [],
- "source": [
- "# convert string values to dummies, drop first so full rank\n",
- "compas_dummy = pd.get_dummies(compas, columns=['c_charge_degree', 'race', 'age_cat', 'score_text', 'sex'], drop_first=True)\n",
- "\n",
- "########\n",
- "\n",
- "predict_columns = ['priors_count', 'days_b_screening_arrest', 'length_of_stay',\n",
- " 'c_charge_degree_M', 'race_Asian', 'race_Caucasian', 'race_Hispanic',\n",
- " 'race_Native American', 'race_Other', 'age_cat_Greater than 45',\n",
- " 'age_cat_Less than 25', 'score_text_Low', 'score_text_Medium', 'sex_Male']\n",
- "\n",
- "response_column = 'two_year_recid'\n",
- "\n",
- "# instantiate the model (using the default parameters)\n",
- "logreg_c = LogisticRegression(solver='lbfgs', max_iter=1000)\n",
- "\n",
- "# fit, reshape X to be of shape (n_samples, n_features)\n",
- "logreg_c = logreg_c.fit(compas_dummy[predict_columns], compas_dummy[response_column])\n",
- "\n",
- "# predict probabilities and attach to data\n",
- "#label_probs_logreg = logreg_c.predict_proba(test.X.values.reshape(-1, 1))\n",
- "#test = test.assign(B_prob_0_machine=label_probs_logreg[:, 0])\n",
- "\n",
- "########\n",
- "\n",
- "# instantiate the model\n",
- "forest_c = RandomForestClassifier(n_estimators=300, max_depth=5, random_state=0)\n",
- "\n",
- "# fit, reshape X to be of shape (n_samples, n_features)\n",
- "forest_c = forest_c.fit(compas_dummy[predict_columns], compas_dummy[response_column])\n",
- "\n",
- "# predict probabilities and attach to data\n",
- "#label_probs_forest = forest.predict_proba(test.X.values.reshape(-1, 1))\n",
- "#test = test.assign(B_prob_0_forest=label_probs_forest[:, 0])"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 113,
- "metadata": {},
- "outputs": [
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- "<Figure size 1008x576 with 1 Axes>"
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- }
- ],
- "source": [
- "failures_compas = np.zeros((11, 2))\n",
- "\n",
- "for r in np.arange(0, 11):\n",
- " ## Causal model with logistic regression\n",
- " failures_compas[r, 0] = ep([r / 10], compas_dummy, response_column, predict_columns, logreg_c, 1)\n",
- " \n",
- " ## Causal model with random forest classifier\n",
- " failures_compas[r, 1] = ep([r / 10], compas_dummy, response_column, predict_columns, forest_c, 1)\n",
- "\n",
- "# klassifikaatioille scipy.stats semin kautta error barit xerr ja yerr argumenttien kautta\n",
- "\n",
- "plt.figure(figsize=(14, 8))\n",
- "plt.plot(np.arange(0, 11) / 10, failures_compas[:, 0], label='Causal model, log.')\n",
- "plt.plot(np.arange(0, 11) / 10, failures_compas[:, 1], label='Causal model, for.')\n",
- "\n",
- "plt.title('Failure rate vs. Acceptance rate - COMPAS')\n",
- "plt.xlabel('Leniency')\n",
- "plt.ylabel('Empirical performance')\n",
- "plt.legend()\n",
- "plt.show()\n"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "Of course if leniency is one, then the empirical performance should always converge to the proportion of false positives in the data."
- ]
- }
- ],
- "metadata": {
- "kernelspec": {
- "display_name": "Python 3",
- "language": "python",
- "name": "python3"
- },
- "language_info": {
- "codemirror_mode": {
- "name": "ipython",
- "version": 3
- },
- "file_extension": ".py",
- "mimetype": "text/x-python",
- "name": "python",
- "nbconvert_exporter": "python",
- "pygments_lexer": "ipython3",
- "version": "3.6.7"
- },
- "toc": {
- "base_numbering": 1,
- "nav_menu": {
- "height": "168px",
- "width": "176px"
- },
- "number_sections": true,
- "sideBar": true,
- "skip_h1_title": true,
- "title_cell": "Table of Contents",
- "title_sidebar": "Contents",
- "toc_cell": true,
- "toc_position": {
- "height": "calc(100% - 180px)",
- "left": "10px",
- "top": "150px",
- "width": "251.4px"
- },
- "toc_section_display": true,
- "toc_window_display": true
- },
- "varInspector": {
- "cols": {
- "lenName": 16,
- "lenType": 16,
- "lenVar": 40
- },
- "kernels_config": {
- "python": {
- "delete_cmd_postfix": "",
- "delete_cmd_prefix": "del ",
- "library": "var_list.py",
- "varRefreshCmd": "print(var_dic_list())"
- },
- "r": {
- "delete_cmd_postfix": ") ",
- "delete_cmd_prefix": "rm(",
- "library": "var_list.r",
- "varRefreshCmd": "cat(var_dic_list()) "
- }
- },
- "position": {
- "height": "465.65px",
- "left": "843.6px",
- "right": "16.2px",
- "top": "159px",
- "width": "676.2px"
- },
- "types_to_exclude": [
- "module",
- "function",
- "builtin_function_or_method",
- "instance",
- "_Feature"
- ],
- "window_display": false
- }
- },
- "nbformat": 4,
- "nbformat_minor": 2
-}
diff --git a/analysis_and_scripts/Bachelors_thesis_analyses_OLD.ipynb b/analysis_and_scripts/Bachelors_thesis_analyses_OLD.ipynb
new file mode 100644
index 0000000000000000000000000000000000000000..7ba50d2cfff4e1bfbad0a9ba94eb3ef8e4501d2e
--- /dev/null
+++ b/analysis_and_scripts/Bachelors_thesis_analyses_OLD.ipynb
@@ -0,0 +1,3237 @@
+{
+ "cells": [
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "toc": true
+ },
+ "source": [
+ "<h1>Table of Contents<span class=\"tocSkip\"></span></h1>\n",
+ "<div class=\"toc\"><ul class=\"toc-item\"><li><span><a href=\"#Data-sets\" data-toc-modified-id=\"Data-sets-1\"><span class=\"toc-item-num\">1 </span>Data sets</a></span><ul class=\"toc-item\"><li><span><a href=\"#COMPAS-data\" data-toc-modified-id=\"COMPAS-data-1.1\"><span class=\"toc-item-num\">1.1 </span>COMPAS data</a></span></li><li><span><a href=\"#Synthetic-data\" data-toc-modified-id=\"Synthetic-data-1.2\"><span class=\"toc-item-num\">1.2 </span>Synthetic data</a></span></li></ul></li><li><span><a href=\"#Algorithms\" data-toc-modified-id=\"Algorithms-2\"><span class=\"toc-item-num\">2 </span>Algorithms</a></span><ul class=\"toc-item\"><li><span><a href=\"#Contraction-algorithm\" data-toc-modified-id=\"Contraction-algorithm-2.1\"><span class=\"toc-item-num\">2.1 </span>Contraction algorithm</a></span></li><li><span><a href=\"#Causal-model\" data-toc-modified-id=\"Causal-model-2.2\"><span class=\"toc-item-num\">2.2 </span>Causal model</a></span></li></ul></li><li><span><a href=\"#Performance-comparison\" data-toc-modified-id=\"Performance-comparison-3\"><span class=\"toc-item-num\">3 </span>Performance comparison</a></span><ul class=\"toc-item\"><li><span><a href=\"#On-synthetic-data\" data-toc-modified-id=\"On-synthetic-data-3.1\"><span class=\"toc-item-num\">3.1 </span>On synthetic data</a></span><ul class=\"toc-item\"><li><span><a href=\"#Predictive-models\" data-toc-modified-id=\"Predictive-models-3.1.1\"><span class=\"toc-item-num\">3.1.1 </span>Predictive models</a></span></li><li><span><a href=\"#Visual-comparison\" data-toc-modified-id=\"Visual-comparison-3.1.2\"><span class=\"toc-item-num\">3.1.2 </span>Visual comparison</a></span></li></ul></li><li><span><a href=\"#On-COMPAS-data\" data-toc-modified-id=\"On-COMPAS-data-3.2\"><span class=\"toc-item-num\">3.2 </span>On COMPAS data</a></span><ul class=\"toc-item\"><li><span><a href=\"#Predictive-models\" data-toc-modified-id=\"Predictive-models-3.2.1\"><span class=\"toc-item-num\">3.2.1 </span>Predictive models</a></span></li></ul></li></ul></li></ul></div>"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# Bachelors thesis' analyses\n",
+ "\n",
+ "*This Jupyter notebook is for the analyses and model building for Riku Laine's bachelors thesis*\n",
+ "\n",
+ "Table of contents is provided above. First I will briefly present the COMPAS data set and then create the synthetic data set as done by Lakkaraju *et al.* ([link](https://helka.finna.fi/PrimoRecord/pci.acm3098066)). Then I will proceed to implement algorithms. Finally I will do the side-by-side comparisons of the results on the synthetic data. Finally I run the causal model on the COMPAS data.\n",
+ "\n",
+ "## Data sets\n",
+ "\n",
+ "*Below I load the COMPAS data set and generate the synthetic one.*\n",
+ "\n",
+ "### COMPAS data\n",
+ "\n",
+ "The following data filtering procedure follows the one described in the [ProPublica methodology](https://www.propublica.org/article/how-we-analyzed-the-compas-recidivism-algorithm)."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 35,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "# Imports\n",
+ "\n",
+ "import numpy as np\n",
+ "import pandas as pd\n",
+ "from datetime import datetime\n",
+ "import matplotlib.pyplot as plt\n",
+ "import scipy.stats as scs\n",
+ "import scipy.integrate as si\n",
+ "import seaborn as sns\n",
+ "import numpy.random as npr\n",
+ "from sklearn.preprocessing import OneHotEncoder\n",
+ "from sklearn.linear_model import LogisticRegression\n",
+ "from sklearn.ensemble import RandomForestClassifier\n",
+ "\n",
+ "# Settings\n",
+ "\n",
+ "%matplotlib inline\n",
+ "\n",
+ "plt.rcParams.update({'font.size': 16})\n",
+ "plt.rcParams.update({'figure.figsize': (14, 7)})\n",
+ "\n",
+ "# Suppress deprecation warnings.\n",
+ "\n",
+ "import warnings\n",
+ "\n",
+ "def fxn():\n",
+ " warnings.warn(\"deprecated\", DeprecationWarning)\n",
+ "\n",
+ "with warnings.catch_warnings():\n",
+ " warnings.simplefilter(\"ignore\")\n",
+ " fxn()"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 36,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "(7214, 53)\n",
+ "['id' 'name' 'first' 'last' 'compas_screening_date' 'sex' 'dob' 'age'\n",
+ " 'age_cat' 'race' 'juv_fel_count' 'decile_score' 'juv_misd_count'\n",
+ " 'juv_other_count' 'priors_count' 'days_b_screening_arrest' 'c_jail_in'\n",
+ " 'c_jail_out' 'c_case_number' 'c_offense_date' 'c_arrest_date'\n",
+ " 'c_days_from_compas' 'c_charge_degree' 'c_charge_desc' 'is_recid'\n",
+ " 'r_case_number' 'r_charge_degree' 'r_days_from_arrest' 'r_offense_date'\n",
+ " 'r_charge_desc' 'r_jail_in' 'r_jail_out' 'violent_recid'\n",
+ " 'is_violent_recid' 'vr_case_number' 'vr_charge_degree' 'vr_offense_date'\n",
+ " 'vr_charge_desc' 'type_of_assessment' 'decile_score.1' 'score_text'\n",
+ " 'screening_date' 'v_type_of_assessment' 'v_decile_score' 'v_score_text'\n",
+ " 'v_screening_date' 'in_custody' 'out_custody' 'priors_count.1' 'start'\n",
+ " 'end' 'event' 'two_year_recid']\n"
+ ]
+ }
+ ],
+ "source": [
+ "# Read file\n",
+ "compas_raw = pd.read_csv(\"../data/compas-scores-two-years.csv\")\n",
+ "\n",
+ "# Check dimensions, number of rows should be 7214\n",
+ "print(compas_raw.shape)\n",
+ "print(compas_raw.columns.values)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 37,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "(6172, 13)"
+ ]
+ },
+ "execution_count": 37,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "# Select columns\n",
+ "compas = compas_raw[[\n",
+ " 'age', 'c_charge_degree', 'race', 'age_cat', 'score_text', 'sex',\n",
+ " 'priors_count', 'days_b_screening_arrest', 'decile_score', 'is_recid',\n",
+ " 'two_year_recid', 'c_jail_in', 'c_jail_out'\n",
+ "]]\n",
+ "\n",
+ "# Subset values, see reasons in ProPublica methodology.\n",
+ "compas = compas.query('days_b_screening_arrest <= 30 and \\\n",
+ " days_b_screening_arrest >= -30 and \\\n",
+ " is_recid != -1 and \\\n",
+ " c_charge_degree != \"O\"')\n",
+ "\n",
+ "# Drop row if score_text is na\n",
+ "compas = compas[compas.score_text.notnull()]\n",
+ "\n",
+ "compas.shape"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 38,
+ "metadata": {
+ "scrolled": false
+ },
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "<div>\n",
+ "<style scoped>\n",
+ " .dataframe tbody tr th:only-of-type {\n",
+ " vertical-align: middle;\n",
+ " }\n",
+ "\n",
+ " .dataframe tbody tr th {\n",
+ " vertical-align: top;\n",
+ " }\n",
+ "\n",
+ " .dataframe thead th {\n",
+ " text-align: right;\n",
+ " }\n",
+ "</style>\n",
+ "<table border=\"1\" class=\"dataframe\">\n",
+ " <thead>\n",
+ " <tr style=\"text-align: right;\">\n",
+ " <th></th>\n",
+ " <th>0</th>\n",
+ " <th>1</th>\n",
+ " <th>2</th>\n",
+ " <th>3</th>\n",
+ " </tr>\n",
+ " </thead>\n",
+ " <tbody>\n",
+ " <tr>\n",
+ " <th>id</th>\n",
+ " <td>1</td>\n",
+ " <td>3</td>\n",
+ " <td>4</td>\n",
+ " <td>5</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>name</th>\n",
+ " <td>miguel hernandez</td>\n",
+ " <td>kevon dixon</td>\n",
+ " <td>ed philo</td>\n",
+ " <td>marcu brown</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>first</th>\n",
+ " <td>miguel</td>\n",
+ " <td>kevon</td>\n",
+ " <td>ed</td>\n",
+ " <td>marcu</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>last</th>\n",
+ " <td>hernandez</td>\n",
+ " <td>dixon</td>\n",
+ " <td>philo</td>\n",
+ " <td>brown</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>compas_screening_date</th>\n",
+ " <td>2013-08-14</td>\n",
+ " <td>2013-01-27</td>\n",
+ " <td>2013-04-14</td>\n",
+ " <td>2013-01-13</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>sex</th>\n",
+ " <td>Male</td>\n",
+ " <td>Male</td>\n",
+ " <td>Male</td>\n",
+ " <td>Male</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>dob</th>\n",
+ " <td>1947-04-18</td>\n",
+ " <td>1982-01-22</td>\n",
+ " <td>1991-05-14</td>\n",
+ " <td>1993-01-21</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>age</th>\n",
+ " <td>69</td>\n",
+ " <td>34</td>\n",
+ " <td>24</td>\n",
+ " <td>23</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>age_cat</th>\n",
+ " <td>Greater than 45</td>\n",
+ " <td>25 - 45</td>\n",
+ " <td>Less than 25</td>\n",
+ " <td>Less than 25</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>race</th>\n",
+ " <td>Other</td>\n",
+ " <td>African-American</td>\n",
+ " <td>African-American</td>\n",
+ " <td>African-American</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>juv_fel_count</th>\n",
+ " <td>0</td>\n",
+ " <td>0</td>\n",
+ " <td>0</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>decile_score</th>\n",
+ " <td>1</td>\n",
+ " <td>3</td>\n",
+ " <td>4</td>\n",
+ " <td>8</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>juv_misd_count</th>\n",
+ " <td>0</td>\n",
+ " <td>0</td>\n",
+ " <td>0</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>juv_other_count</th>\n",
+ " <td>0</td>\n",
+ " <td>0</td>\n",
+ " <td>1</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>priors_count</th>\n",
+ " <td>0</td>\n",
+ " <td>0</td>\n",
+ " <td>4</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>days_b_screening_arrest</th>\n",
+ " <td>-1</td>\n",
+ " <td>-1</td>\n",
+ " <td>-1</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>c_jail_in</th>\n",
+ " <td>2013-08-13 06:03:42</td>\n",
+ " <td>2013-01-26 03:45:27</td>\n",
+ " <td>2013-04-13 04:58:34</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>c_jail_out</th>\n",
+ " <td>2013-08-14 05:41:20</td>\n",
+ " <td>2013-02-05 05:36:53</td>\n",
+ " <td>2013-04-14 07:02:04</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>c_case_number</th>\n",
+ " <td>13011352CF10A</td>\n",
+ " <td>13001275CF10A</td>\n",
+ " <td>13005330CF10A</td>\n",
+ " <td>13000570CF10A</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>c_offense_date</th>\n",
+ " <td>2013-08-13</td>\n",
+ " <td>2013-01-26</td>\n",
+ " <td>2013-04-13</td>\n",
+ " <td>2013-01-12</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>c_arrest_date</th>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>c_days_from_compas</th>\n",
+ " <td>1</td>\n",
+ " <td>1</td>\n",
+ " <td>1</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>c_charge_degree</th>\n",
+ " <td>F</td>\n",
+ " <td>F</td>\n",
+ " <td>F</td>\n",
+ " <td>F</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>c_charge_desc</th>\n",
+ " <td>Aggravated Assault w/Firearm</td>\n",
+ " <td>Felony Battery w/Prior Convict</td>\n",
+ " <td>Possession of Cocaine</td>\n",
+ " <td>Possession of Cannabis</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>is_recid</th>\n",
+ " <td>0</td>\n",
+ " <td>1</td>\n",
+ " <td>1</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>r_case_number</th>\n",
+ " <td>NaN</td>\n",
+ " <td>13009779CF10A</td>\n",
+ " <td>13011511MM10A</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>r_charge_degree</th>\n",
+ " <td>NaN</td>\n",
+ " <td>(F3)</td>\n",
+ " <td>(M1)</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>r_days_from_arrest</th>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>0</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>r_offense_date</th>\n",
+ " <td>NaN</td>\n",
+ " <td>2013-07-05</td>\n",
+ " <td>2013-06-16</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>r_charge_desc</th>\n",
+ " <td>NaN</td>\n",
+ " <td>Felony Battery (Dom Strang)</td>\n",
+ " <td>Driving Under The Influence</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>r_jail_in</th>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>2013-06-16</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>r_jail_out</th>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>2013-06-16</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>violent_recid</th>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>is_violent_recid</th>\n",
+ " <td>0</td>\n",
+ " <td>1</td>\n",
+ " <td>0</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>vr_case_number</th>\n",
+ " <td>NaN</td>\n",
+ " <td>13009779CF10A</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>vr_charge_degree</th>\n",
+ " <td>NaN</td>\n",
+ " <td>(F3)</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>vr_offense_date</th>\n",
+ " <td>NaN</td>\n",
+ " <td>2013-07-05</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>vr_charge_desc</th>\n",
+ " <td>NaN</td>\n",
+ " <td>Felony Battery (Dom Strang)</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>type_of_assessment</th>\n",
+ " <td>Risk of Recidivism</td>\n",
+ " <td>Risk of Recidivism</td>\n",
+ " <td>Risk of Recidivism</td>\n",
+ " <td>Risk of Recidivism</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>decile_score.1</th>\n",
+ " <td>1</td>\n",
+ " <td>3</td>\n",
+ " <td>4</td>\n",
+ " <td>8</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>score_text</th>\n",
+ " <td>Low</td>\n",
+ " <td>Low</td>\n",
+ " <td>Low</td>\n",
+ " <td>High</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>screening_date</th>\n",
+ " <td>2013-08-14</td>\n",
+ " <td>2013-01-27</td>\n",
+ " <td>2013-04-14</td>\n",
+ " <td>2013-01-13</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>v_type_of_assessment</th>\n",
+ " <td>Risk of Violence</td>\n",
+ " <td>Risk of Violence</td>\n",
+ " <td>Risk of Violence</td>\n",
+ " <td>Risk of Violence</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>v_decile_score</th>\n",
+ " <td>1</td>\n",
+ " <td>1</td>\n",
+ " <td>3</td>\n",
+ " <td>6</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>v_score_text</th>\n",
+ " <td>Low</td>\n",
+ " <td>Low</td>\n",
+ " <td>Low</td>\n",
+ " <td>Medium</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>v_screening_date</th>\n",
+ " <td>2013-08-14</td>\n",
+ " <td>2013-01-27</td>\n",
+ " <td>2013-04-14</td>\n",
+ " <td>2013-01-13</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>in_custody</th>\n",
+ " <td>2014-07-07</td>\n",
+ " <td>2013-01-26</td>\n",
+ " <td>2013-06-16</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>out_custody</th>\n",
+ " <td>2014-07-14</td>\n",
+ " <td>2013-02-05</td>\n",
+ " <td>2013-06-16</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>priors_count.1</th>\n",
+ " <td>0</td>\n",
+ " <td>0</td>\n",
+ " <td>4</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>start</th>\n",
+ " <td>0</td>\n",
+ " <td>9</td>\n",
+ " <td>0</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>end</th>\n",
+ " <td>327</td>\n",
+ " <td>159</td>\n",
+ " <td>63</td>\n",
+ " <td>1174</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>event</th>\n",
+ " <td>0</td>\n",
+ " <td>1</td>\n",
+ " <td>0</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>two_year_recid</th>\n",
+ " <td>0</td>\n",
+ " <td>1</td>\n",
+ " <td>1</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " </tbody>\n",
+ "</table>\n",
+ "</div>"
+ ],
+ "text/plain": [
+ " 0 \\\n",
+ "id 1 \n",
+ "name miguel hernandez \n",
+ "first miguel \n",
+ "last hernandez \n",
+ "compas_screening_date 2013-08-14 \n",
+ "sex Male \n",
+ "dob 1947-04-18 \n",
+ "age 69 \n",
+ "age_cat Greater than 45 \n",
+ "race Other \n",
+ "juv_fel_count 0 \n",
+ "decile_score 1 \n",
+ "juv_misd_count 0 \n",
+ "juv_other_count 0 \n",
+ "priors_count 0 \n",
+ "days_b_screening_arrest -1 \n",
+ "c_jail_in 2013-08-13 06:03:42 \n",
+ "c_jail_out 2013-08-14 05:41:20 \n",
+ "c_case_number 13011352CF10A \n",
+ "c_offense_date 2013-08-13 \n",
+ "c_arrest_date NaN \n",
+ "c_days_from_compas 1 \n",
+ "c_charge_degree F \n",
+ "c_charge_desc Aggravated Assault w/Firearm \n",
+ "is_recid 0 \n",
+ "r_case_number NaN \n",
+ "r_charge_degree NaN \n",
+ "r_days_from_arrest NaN \n",
+ "r_offense_date NaN \n",
+ "r_charge_desc NaN \n",
+ "r_jail_in NaN \n",
+ "r_jail_out NaN \n",
+ "violent_recid NaN \n",
+ "is_violent_recid 0 \n",
+ "vr_case_number NaN \n",
+ "vr_charge_degree NaN \n",
+ "vr_offense_date NaN \n",
+ "vr_charge_desc NaN \n",
+ "type_of_assessment Risk of Recidivism \n",
+ "decile_score.1 1 \n",
+ "score_text Low \n",
+ "screening_date 2013-08-14 \n",
+ "v_type_of_assessment Risk of Violence \n",
+ "v_decile_score 1 \n",
+ "v_score_text Low \n",
+ "v_screening_date 2013-08-14 \n",
+ "in_custody 2014-07-07 \n",
+ "out_custody 2014-07-14 \n",
+ "priors_count.1 0 \n",
+ "start 0 \n",
+ "end 327 \n",
+ "event 0 \n",
+ "two_year_recid 0 \n",
+ "\n",
+ " 1 \\\n",
+ "id 3 \n",
+ "name kevon dixon \n",
+ "first kevon \n",
+ "last dixon \n",
+ "compas_screening_date 2013-01-27 \n",
+ "sex Male \n",
+ "dob 1982-01-22 \n",
+ "age 34 \n",
+ "age_cat 25 - 45 \n",
+ "race African-American \n",
+ "juv_fel_count 0 \n",
+ "decile_score 3 \n",
+ "juv_misd_count 0 \n",
+ "juv_other_count 0 \n",
+ "priors_count 0 \n",
+ "days_b_screening_arrest -1 \n",
+ "c_jail_in 2013-01-26 03:45:27 \n",
+ "c_jail_out 2013-02-05 05:36:53 \n",
+ "c_case_number 13001275CF10A \n",
+ "c_offense_date 2013-01-26 \n",
+ "c_arrest_date NaN \n",
+ "c_days_from_compas 1 \n",
+ "c_charge_degree F \n",
+ "c_charge_desc Felony Battery w/Prior Convict \n",
+ "is_recid 1 \n",
+ "r_case_number 13009779CF10A \n",
+ "r_charge_degree (F3) \n",
+ "r_days_from_arrest NaN \n",
+ "r_offense_date 2013-07-05 \n",
+ "r_charge_desc Felony Battery (Dom Strang) \n",
+ "r_jail_in NaN \n",
+ "r_jail_out NaN \n",
+ "violent_recid NaN \n",
+ "is_violent_recid 1 \n",
+ "vr_case_number 13009779CF10A \n",
+ "vr_charge_degree (F3) \n",
+ "vr_offense_date 2013-07-05 \n",
+ "vr_charge_desc Felony Battery (Dom Strang) \n",
+ "type_of_assessment Risk of Recidivism \n",
+ "decile_score.1 3 \n",
+ "score_text Low \n",
+ "screening_date 2013-01-27 \n",
+ "v_type_of_assessment Risk of Violence \n",
+ "v_decile_score 1 \n",
+ "v_score_text Low \n",
+ "v_screening_date 2013-01-27 \n",
+ "in_custody 2013-01-26 \n",
+ "out_custody 2013-02-05 \n",
+ "priors_count.1 0 \n",
+ "start 9 \n",
+ "end 159 \n",
+ "event 1 \n",
+ "two_year_recid 1 \n",
+ "\n",
+ " 2 3 \n",
+ "id 4 5 \n",
+ "name ed philo marcu brown \n",
+ "first ed marcu \n",
+ "last philo brown \n",
+ "compas_screening_date 2013-04-14 2013-01-13 \n",
+ "sex Male Male \n",
+ "dob 1991-05-14 1993-01-21 \n",
+ "age 24 23 \n",
+ "age_cat Less than 25 Less than 25 \n",
+ "race African-American African-American \n",
+ "juv_fel_count 0 0 \n",
+ "decile_score 4 8 \n",
+ "juv_misd_count 0 1 \n",
+ "juv_other_count 1 0 \n",
+ "priors_count 4 1 \n",
+ "days_b_screening_arrest -1 NaN \n",
+ "c_jail_in 2013-04-13 04:58:34 NaN \n",
+ "c_jail_out 2013-04-14 07:02:04 NaN \n",
+ "c_case_number 13005330CF10A 13000570CF10A \n",
+ "c_offense_date 2013-04-13 2013-01-12 \n",
+ "c_arrest_date NaN NaN \n",
+ "c_days_from_compas 1 1 \n",
+ "c_charge_degree F F \n",
+ "c_charge_desc Possession of Cocaine Possession of Cannabis \n",
+ "is_recid 1 0 \n",
+ "r_case_number 13011511MM10A NaN \n",
+ "r_charge_degree (M1) NaN \n",
+ "r_days_from_arrest 0 NaN \n",
+ "r_offense_date 2013-06-16 NaN \n",
+ "r_charge_desc Driving Under The Influence NaN \n",
+ "r_jail_in 2013-06-16 NaN \n",
+ "r_jail_out 2013-06-16 NaN \n",
+ "violent_recid NaN NaN \n",
+ "is_violent_recid 0 0 \n",
+ "vr_case_number NaN NaN \n",
+ "vr_charge_degree NaN NaN \n",
+ "vr_offense_date NaN NaN \n",
+ "vr_charge_desc NaN NaN \n",
+ "type_of_assessment Risk of Recidivism Risk of Recidivism \n",
+ "decile_score.1 4 8 \n",
+ "score_text Low High \n",
+ "screening_date 2013-04-14 2013-01-13 \n",
+ "v_type_of_assessment Risk of Violence Risk of Violence \n",
+ "v_decile_score 3 6 \n",
+ "v_score_text Low Medium \n",
+ "v_screening_date 2013-04-14 2013-01-13 \n",
+ "in_custody 2013-06-16 NaN \n",
+ "out_custody 2013-06-16 NaN \n",
+ "priors_count.1 4 1 \n",
+ "start 0 0 \n",
+ "end 63 1174 \n",
+ "event 0 0 \n",
+ "two_year_recid 1 0 "
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "# Calculate length of stay\n",
+ "out = pd.to_datetime(compas.c_jail_out, format=\"%Y-%m-%d %H:%M:%S\")\n",
+ "in_ = pd.to_datetime(compas.c_jail_in, format=\"%Y-%m-%d %H:%M:%S\")\n",
+ "\n",
+ "compas['length_of_stay'] = (out - in_).astype('timedelta64[D]')\n",
+ "\n",
+ "# Structure of the data\n",
+ "display(compas_raw.head(4).T)\n",
+ "#print(np.sum(compas_raw.c_arrest_date.isnull()))"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "**Columns:**\n",
+ "\n",
+ "* id = identification number\n",
+ "* name \n",
+ "* first (name)\n",
+ "* last (name)\n",
+ "* compas_screening_date = date of COMPAS filling\n",
+ "* sex\n",
+ "* dob = date of birth\n",
+ "* age\n",
+ "* age_cat\n",
+ "* race\n",
+ "* juv_fel_count = No. of juvenile felonies\n",
+ "* decile_score = decile score of COMPAS\n",
+ "* juv_misd_count = No. of juvenile misdemeanors\n",
+ "* juv_other_count = No. of other crimes juvenile \n",
+ "* priors_count = No. of priors \n",
+ "* days_b_screening_arrest = date of a defendants Compas scored crime - date for person's arrest (c_offense_date - screening_date) \n",
+ "* c_jail_in = jailing date of COMPAS scored crime\n",
+ "* c_jail_out = jailing date of COMPAS scored crime\n",
+ "* c_case_number = case number of COMPAS scored crime\n",
+ "* c_offense_date = offense date of COMPAS scored crime\n",
+ "* c_arrest_date = arrest date of COMPAS scored crime\n",
+ "* c_days_from_compas = \n",
+ "* c_charge_degree\n",
+ "* c_charge_desc\n",
+ "* is_recid\n",
+ "* r_case_number\n",
+ "* r_charge_degree\n",
+ "* r_days_from_arrest\n",
+ "* r_offense_date\n",
+ "* r_charge_desc\n",
+ "* r_jail_in\n",
+ "* r_jail_out\n",
+ "* violent_recid\n",
+ "* is_violent_recid\n",
+ "* vr_case_number\n",
+ "* vr_charge_degree\n",
+ "* vr_offense_date\n",
+ "* vr_charge_desc\n",
+ "* type_of_assessment\n",
+ "* decile_score.1\n",
+ "* score_text\n",
+ "* screening_date\n",
+ "* v_type_of_assessment\n",
+ "* v_decile_score\n",
+ "* v_score_text\n",
+ "* v_screening_date\n",
+ "* in_custody\n",
+ "* out_custody\n",
+ "* priors_count.1\n",
+ "* start\n",
+ "* end\n",
+ "* event\n",
+ "* two_year_recid\n",
+ "\n",
+ "Let's obtain the basic statistics for each of the variables."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 39,
+ "metadata": {
+ "scrolled": false
+ },
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "<div>\n",
+ "<style scoped>\n",
+ " .dataframe tbody tr th:only-of-type {\n",
+ " vertical-align: middle;\n",
+ " }\n",
+ "\n",
+ " .dataframe tbody tr th {\n",
+ " vertical-align: top;\n",
+ " }\n",
+ "\n",
+ " .dataframe thead th {\n",
+ " text-align: right;\n",
+ " }\n",
+ "</style>\n",
+ "<table border=\"1\" class=\"dataframe\">\n",
+ " <thead>\n",
+ " <tr style=\"text-align: right;\">\n",
+ " <th></th>\n",
+ " <th>count</th>\n",
+ " <th>unique</th>\n",
+ " <th>top</th>\n",
+ " <th>freq</th>\n",
+ " <th>mean</th>\n",
+ " <th>std</th>\n",
+ " <th>min</th>\n",
+ " <th>25%</th>\n",
+ " <th>50%</th>\n",
+ " <th>75%</th>\n",
+ " <th>max</th>\n",
+ " </tr>\n",
+ " </thead>\n",
+ " <tbody>\n",
+ " <tr>\n",
+ " <th>age</th>\n",
+ " <td>6172</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>34.5345</td>\n",
+ " <td>11.7309</td>\n",
+ " <td>18</td>\n",
+ " <td>25</td>\n",
+ " <td>31</td>\n",
+ " <td>42</td>\n",
+ " <td>96</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>c_charge_degree</th>\n",
+ " <td>6172</td>\n",
+ " <td>2</td>\n",
+ " <td>F</td>\n",
+ " <td>3970</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>race</th>\n",
+ " <td>6172</td>\n",
+ " <td>6</td>\n",
+ " <td>African-American</td>\n",
+ " <td>3175</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>age_cat</th>\n",
+ " <td>6172</td>\n",
+ " <td>3</td>\n",
+ " <td>25 - 45</td>\n",
+ " <td>3532</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>score_text</th>\n",
+ " <td>6172</td>\n",
+ " <td>3</td>\n",
+ " <td>Low</td>\n",
+ " <td>3421</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>sex</th>\n",
+ " <td>6172</td>\n",
+ " <td>2</td>\n",
+ " <td>Male</td>\n",
+ " <td>4997</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>priors_count</th>\n",
+ " <td>6172</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>3.24644</td>\n",
+ " <td>4.74377</td>\n",
+ " <td>0</td>\n",
+ " <td>0</td>\n",
+ " <td>1</td>\n",
+ " <td>4</td>\n",
+ " <td>38</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>days_b_screening_arrest</th>\n",
+ " <td>6172</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>-1.74028</td>\n",
+ " <td>5.08471</td>\n",
+ " <td>-30</td>\n",
+ " <td>-1</td>\n",
+ " <td>-1</td>\n",
+ " <td>-1</td>\n",
+ " <td>30</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>decile_score</th>\n",
+ " <td>6172</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>4.4185</td>\n",
+ " <td>2.83946</td>\n",
+ " <td>1</td>\n",
+ " <td>2</td>\n",
+ " <td>4</td>\n",
+ " <td>7</td>\n",
+ " <td>10</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>is_recid</th>\n",
+ " <td>6172</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>0.484446</td>\n",
+ " <td>0.499799</td>\n",
+ " <td>0</td>\n",
+ " <td>0</td>\n",
+ " <td>0</td>\n",
+ " <td>1</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>two_year_recid</th>\n",
+ " <td>6172</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>0.45512</td>\n",
+ " <td>0.498022</td>\n",
+ " <td>0</td>\n",
+ " <td>0</td>\n",
+ " <td>0</td>\n",
+ " <td>1</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>c_jail_in</th>\n",
+ " <td>6172</td>\n",
+ " <td>6172</td>\n",
+ " <td>2014-01-05 10:19:57</td>\n",
+ " <td>1</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>c_jail_out</th>\n",
+ " <td>6172</td>\n",
+ " <td>6161</td>\n",
+ " <td>2013-09-14 05:58:00</td>\n",
+ " <td>3</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>length_of_stay</th>\n",
+ " <td>6172</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>NaN</td>\n",
+ " <td>14.6228</td>\n",
+ " <td>46.6935</td>\n",
+ " <td>-1</td>\n",
+ " <td>0</td>\n",
+ " <td>1</td>\n",
+ " <td>5</td>\n",
+ " <td>799</td>\n",
+ " </tr>\n",
+ " </tbody>\n",
+ "</table>\n",
+ "</div>"
+ ],
+ "text/plain": [
+ " count unique top freq mean \\\n",
+ "age 6172 NaN NaN NaN 34.5345 \n",
+ "c_charge_degree 6172 2 F 3970 NaN \n",
+ "race 6172 6 African-American 3175 NaN \n",
+ "age_cat 6172 3 25 - 45 3532 NaN \n",
+ "score_text 6172 3 Low 3421 NaN \n",
+ "sex 6172 2 Male 4997 NaN \n",
+ "priors_count 6172 NaN NaN NaN 3.24644 \n",
+ "days_b_screening_arrest 6172 NaN NaN NaN -1.74028 \n",
+ "decile_score 6172 NaN NaN NaN 4.4185 \n",
+ "is_recid 6172 NaN NaN NaN 0.484446 \n",
+ "two_year_recid 6172 NaN NaN NaN 0.45512 \n",
+ "c_jail_in 6172 6172 2014-01-05 10:19:57 1 NaN \n",
+ "c_jail_out 6172 6161 2013-09-14 05:58:00 3 NaN \n",
+ "length_of_stay 6172 NaN NaN NaN 14.6228 \n",
+ "\n",
+ " std min 25% 50% 75% max \n",
+ "age 11.7309 18 25 31 42 96 \n",
+ "c_charge_degree NaN NaN NaN NaN NaN NaN \n",
+ "race NaN NaN NaN NaN NaN NaN \n",
+ "age_cat NaN NaN NaN NaN NaN NaN \n",
+ "score_text NaN NaN NaN NaN NaN NaN \n",
+ "sex NaN NaN NaN NaN NaN NaN \n",
+ "priors_count 4.74377 0 0 1 4 38 \n",
+ "days_b_screening_arrest 5.08471 -30 -1 -1 -1 30 \n",
+ "decile_score 2.83946 1 2 4 7 10 \n",
+ "is_recid 0.499799 0 0 0 1 1 \n",
+ "two_year_recid 0.498022 0 0 0 1 1 \n",
+ "c_jail_in NaN NaN NaN NaN NaN NaN \n",
+ "c_jail_out NaN NaN NaN NaN NaN NaN \n",
+ "length_of_stay 46.6935 -1 0 1 5 799 "
+ ]
+ },
+ "execution_count": 39,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "compas.describe(include='all').T"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "**Notes:**\n",
+ "\n",
+ "* Mean age is roughly 34.5 years ranging from 18 to 96\n",
+ "* Defendants have an average of 3.2 priors (sd 4.7) and more than half have 1 or more prior.\n",
+ "* 48.4% have recidivated in general and 45.5% recidivated within a two-year period following their arrest."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 40,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 900x900 with 30 Axes>"
+ ]
+ },
+ "metadata": {
+ "needs_background": "light"
+ },
+ "output_type": "display_data"
+ },
+ {
+ "data": {
+ "text/html": [
+ "<div>\n",
+ "<style scoped>\n",
+ " .dataframe tbody tr th:only-of-type {\n",
+ " vertical-align: middle;\n",
+ " }\n",
+ "\n",
+ " .dataframe tbody tr th {\n",
+ " vertical-align: top;\n",
+ " }\n",
+ "\n",
+ " .dataframe thead th {\n",
+ " text-align: right;\n",
+ " }\n",
+ "</style>\n",
+ "<table border=\"1\" class=\"dataframe\">\n",
+ " <thead>\n",
+ " <tr style=\"text-align: right;\">\n",
+ " <th></th>\n",
+ " <th>age</th>\n",
+ " <th>priors_count</th>\n",
+ " <th>days_b_screening_arrest</th>\n",
+ " <th>decile_score</th>\n",
+ " <th>length_of_stay</th>\n",
+ " </tr>\n",
+ " </thead>\n",
+ " <tbody>\n",
+ " <tr>\n",
+ " <th>age</th>\n",
+ " <td>1.00</td>\n",
+ " <td>0.12</td>\n",
+ " <td>-0.07</td>\n",
+ " <td>-0.40</td>\n",
+ " <td>0.01</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>priors_count</th>\n",
+ " <td>0.12</td>\n",
+ " <td>1.00</td>\n",
+ " <td>0.02</td>\n",
+ " <td>0.45</td>\n",
+ " <td>0.19</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>days_b_screening_arrest</th>\n",
+ " <td>-0.07</td>\n",
+ " <td>0.02</td>\n",
+ " <td>1.00</td>\n",
+ " <td>0.09</td>\n",
+ " <td>0.06</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>decile_score</th>\n",
+ " <td>-0.40</td>\n",
+ " <td>0.45</td>\n",
+ " <td>0.09</td>\n",
+ " <td>1.00</td>\n",
+ " <td>0.21</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>length_of_stay</th>\n",
+ " <td>0.01</td>\n",
+ " <td>0.19</td>\n",
+ " <td>0.06</td>\n",
+ " <td>0.21</td>\n",
+ " <td>1.00</td>\n",
+ " </tr>\n",
+ " </tbody>\n",
+ "</table>\n",
+ "</div>"
+ ],
+ "text/plain": [
+ " age priors_count days_b_screening_arrest \\\n",
+ "age 1.00 0.12 -0.07 \n",
+ "priors_count 0.12 1.00 0.02 \n",
+ "days_b_screening_arrest -0.07 0.02 1.00 \n",
+ "decile_score -0.40 0.45 0.09 \n",
+ "length_of_stay 0.01 0.19 0.06 \n",
+ "\n",
+ " decile_score length_of_stay \n",
+ "age -0.40 0.01 \n",
+ "priors_count 0.45 0.19 \n",
+ "days_b_screening_arrest 0.09 0.06 \n",
+ "decile_score 1.00 0.21 \n",
+ "length_of_stay 0.21 1.00 "
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "# Distributions of the continuous variables \n",
+ "sns.pairplot(compas[[\n",
+ " 'age', 'priors_count', 'days_b_screening_arrest', 'decile_score',\n",
+ " 'length_of_stay'\n",
+ "]])\n",
+ "plt.show()\n",
+ "\n",
+ "# Correlations of the continuous variables\n",
+ "display(compas[[\n",
+ " 'age', 'priors_count', 'days_b_screening_arrest', 'decile_score',\n",
+ " 'length_of_stay'\n",
+ "]].corr().round(2))"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "**Notes:**\n",
+ "\n",
+ "* Some notable correlations: `age` and `decile_score` ($\\rho\\approx-0.40$, Spearman -0.44) and `decile_score` and `priors_count` ($\\rho\\approx0.45$, Spearman 0.44)\n",
+ "* Spearman correlation was for `length_of_stay` and `priors_count` 0.27 and for `length_of_stay` and `decile_score` 0.27"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 41,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "1 1286\n",
+ "2 822\n",
+ "4 666\n",
+ "3 647\n",
+ "5 582\n",
+ "6 529\n",
+ "7 496\n",
+ "9 420\n",
+ "8 420\n",
+ "10 304\n",
+ "Name: decile_score, dtype: int64\n"
+ ]
+ }
+ ],
+ "source": [
+ "# Decile scores should be evenly distributed but are not.\n",
+ "print(compas.decile_score.value_counts())"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 42,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "25 - 45 3532\n",
+ "Less than 25 1347\n",
+ "Greater than 45 1293\n",
+ "Name: age_cat, dtype: int64\n"
+ ]
+ }
+ ],
+ "source": [
+ "print(compas.age_cat.value_counts())"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 43,
+ "metadata": {
+ "scrolled": false
+ },
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "African-American 3175\n",
+ "Caucasian 2103\n",
+ "Hispanic 509\n",
+ "Other 343\n",
+ "Asian 31\n",
+ "Native American 11\n",
+ "Name: race, dtype: int64\n"
+ ]
+ }
+ ],
+ "source": [
+ "print(compas.race.value_counts())"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "A very small number of Asian and Native American defendants."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 44,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Black defendants: 51.44%\n",
+ "White defendants: 34.07%\n",
+ "Hispanic defendants: 8.25%\n",
+ "Asian defendants: 0.50%\n",
+ "Native American defendants: 0.18%\n",
+ "---\n",
+ "Defendants of other race: 5.56%\n"
+ ]
+ }
+ ],
+ "source": [
+ "print(\"Black defendants: %.2f%%\" % (3175 / 6172 * 100))\n",
+ "print(\"White defendants: %.2f%%\" % (2103 / 6172 * 100))\n",
+ "print(\"Hispanic defendants: %.2f%%\" % (509 / 6172 * 100))\n",
+ "print(\"Asian defendants: %.2f%%\" % (31 / 6172 * 100))\n",
+ "print(\"Native American defendants: %.2f%%\" % (11 / 6172 * 100))\n",
+ "print(\"---\")\n",
+ "print(\"Defendants of other race: %.2f%%\" % (343 / 6172 * 100))"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 45,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Low 3421\n",
+ "Medium 1607\n",
+ "High 1144\n",
+ "Name: score_text, dtype: int64\n"
+ ]
+ },
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 1008x504 with 1 Axes>"
+ ]
+ },
+ "metadata": {
+ "needs_background": "light"
+ },
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "print(compas.score_text.value_counts())\n",
+ "\n",
+ "# Recidivists spent longer in incarceration in some point\n",
+ "compas.boxplot(column=['length_of_stay'], by=['is_recid'])\n",
+ "plt.show()"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 46,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "F 3970\n",
+ "M 2202\n",
+ "Name: c_charge_degree, dtype: int64"
+ ]
+ },
+ "execution_count": 46,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "compas.c_charge_degree.value_counts()"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 47,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "(-0.001, 0.5] 2085\n",
+ "(0.5, 5.5] 2866\n",
+ "(5.5, 10.5] 729\n",
+ "(10.5, 20.5] 402\n",
+ "(20.5, 40.5] 90\n",
+ "Name: priors_count, dtype: int64"
+ ]
+ },
+ "execution_count": 47,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "compas.priors_count.value_counts(\n",
+ " sort=False, bins=[0, 0.5, 5.5, 10.5, 20.5, 40.5])"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 48,
+ "metadata": {
+ "scrolled": true
+ },
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "<div>\n",
+ "<style scoped>\n",
+ " .dataframe tbody tr th:only-of-type {\n",
+ " vertical-align: middle;\n",
+ " }\n",
+ "\n",
+ " .dataframe tbody tr th {\n",
+ " vertical-align: top;\n",
+ " }\n",
+ "\n",
+ " .dataframe thead th {\n",
+ " text-align: right;\n",
+ " }\n",
+ "</style>\n",
+ "<table border=\"1\" class=\"dataframe\">\n",
+ " <thead>\n",
+ " <tr style=\"text-align: right;\">\n",
+ " <th>race</th>\n",
+ " <th>African-American</th>\n",
+ " <th>Asian</th>\n",
+ " <th>Caucasian</th>\n",
+ " <th>Hispanic</th>\n",
+ " <th>Native American</th>\n",
+ " <th>Other</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>sex</th>\n",
+ " <th></th>\n",
+ " <th></th>\n",
+ " <th></th>\n",
+ " <th></th>\n",
+ " <th></th>\n",
+ " <th></th>\n",
+ " </tr>\n",
+ " </thead>\n",
+ " <tbody>\n",
+ " <tr>\n",
+ " <th>Female</th>\n",
+ " <td>549</td>\n",
+ " <td>2</td>\n",
+ " <td>482</td>\n",
+ " <td>82</td>\n",
+ " <td>2</td>\n",
+ " <td>58</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>Male</th>\n",
+ " <td>2626</td>\n",
+ " <td>29</td>\n",
+ " <td>1621</td>\n",
+ " <td>427</td>\n",
+ " <td>9</td>\n",
+ " <td>285</td>\n",
+ " </tr>\n",
+ " </tbody>\n",
+ "</table>\n",
+ "</div>"
+ ],
+ "text/plain": [
+ "race African-American Asian Caucasian Hispanic Native American Other\n",
+ "sex \n",
+ "Female 549 2 482 82 2 58\n",
+ "Male 2626 29 1621 427 9 285"
+ ]
+ },
+ "execution_count": 48,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "tab = compas.groupby(['sex', 'race']).size()\n",
+ "tab.unstack()"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 49,
+ "metadata": {
+ "scrolled": false
+ },
+ "outputs": [
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 1008x504 with 1 Axes>"
+ ]
+ },
+ "metadata": {
+ "needs_background": "light"
+ },
+ "output_type": "display_data"
+ },
+ {
+ "data": {
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAA/sAAAH5CAYAAADeLy4YAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAIABJREFUeJzs3Xm4JHV9L/73R0CRIBeIXBkZDuASE4xxTTRXgoJxiwgxiMYlLjwuP5JoohEJiTHGJQpqJFeDcbvEG1yIJuqAieICmgVXxFwnbkRkhmUiZBDCIot8f390ndgeemZOz+k+fU7N6/U89fTpqk9Vf7oOTJ13f6uqq7UWAAAAoD9uN+sGAAAAgMkS9gEAAKBnhH0AAADoGWEfAAAAekbYBwAAgJ4R9gEAAKBnhH1YpWrgJVX1jar6QVW1qvrVRaz38K72FcvQJlvg9wDAjqaqzq0q3/sNy0TYhxWoqj7aBcGvbaXsaUlen+TaJH+W5E+SfGM5+lsJqurnu33UquqJs+4HALakqv5XVf3fqvpOVd1QVddV1der6tSqesCs+wP6qVrz4RqsJFW1JsnGDD6MqyS/0Fr74oi69yR5apI1rbVNY2x/tyRzSa5srV05ma6XX1WdmuS4JC3JP7TWHjfjlsbSl98DAFtWVTsleXMGx6sbk3wqyb9lcOz66SSPSHLHJL/aWls3qz6XS1XNJdmttbbDDE7ALO086waA23hmkp2SvDHJ7yU5Nsltwn6SNUkyTtDv6q/PKj8DoKp2TfKUDP5gujjJo6vqrq21y2bb2eL14fcAwDa9LoOg/4Ukx7TWNgwvrKq9k/xxkr1m0NuyW/j+gelyGj+sPM9KcnWSlyX5VpJf78JtkqSqntVd73ZY93z+VPbvds//+1rwqjq0qj5dVddU1UULly984ap6QFWdUVWXV9WNVXVJVX24qn5pqOauVfXKqvpCVV3R1V1YVW+oqjuN2Oa53evtUlV/UlUXd+usr6qnbOc+ekKSPZOc3k07JXnGqMKq+qvu9e9WVSdU1b93p1BeUFWP6Wr2qKq/6N73Dd0+u9cWtnf/qvpAVf1H9z7+vapeW1W7L6jbEX4PAGxBVf1UkhcnuSLJ40YF3dba5tba7yR5f7fO4VV1WlV9qzvV/7+q6l+q6skjtr+148hyH2P2646jF3bH0Sur6vyqev2Cuttcs+94BtNjZB9WkKp6aJJ7JXlXa+0HVXV6klcm+bUk7+3KLsjg+vxnJTmg+zlJvr9gcw9N8gdJzk7y1iS3OWAueO0nJ/nrJLcm+XCSi5Lsm+SXkhyd5B+70kOTvCiDUxH/JYNTER+cwVkIh1bVQ1trN494ifcneUCSj2bwb89Tkry3qr7fWvuHrfU2wrHd674nyZVJ/ivJszMYQdmSNyV5YJIzu9d/WpJ13T4/tZv3/iQHJvnVJGdV1U+31n44v4GqekJXc1MG+2hT955+P8lhVXVoa+2mBa/b598DAFv2zAwG1t62rcu1Wms3dj++NMndknw+yaVJ9k5yZJL3V9Wa1topS2loGseYqvqJJP+c5K5JzkrywSS7Z/D3zG8nOX4bbTmewbS01kwm0wqZkrwzg4Pcw7rnB2VwQP7EiNpzB/8L32b+w7tttCRP3cryVwzN2zfJdUmuSvIzC+oryV2Hnv/PJD8xYrsv67b79FF9ZnAA331o/sO6+R8fcx/NJflhknOG5v1Vt62HjqifX/b1JD85NP/obv5VSd6XZKehZW/ulv3a0Lw7J7kmyXeG90e37Piu/iU7yu/BZDKZTFufkpzT/ft6+BjrHDRi3k8k+WoGZ/3tNjT/NseRrS2b1jEmgw8jWpIXjqj/yQXPz82Cv10cz0ym6U1O44cVovtk/ElJNiT5bJK01i7K4GB2eA1uajOOL7XW3rvtsiSD0YfdkpzUWvv68II2cNnQ8++11q4bsY1Tu8df3sJr/EFr7dqh7XwmyXeTPGiRPc57VgYjJacPzZv/+dlbWe9PW2v/OfT8QxmM0O+Z5KVtaAQ/yRnd432H5j0jg1H5l7bb3hvgjRmcpvnrI163r78HALZu3+7x0sWu0B33F867Lsm7k+yR5BeW0M+0jzE3LJyx4Lg7kuMZTI/T+GHleGIGYfItrbXh69n+OoNTwZ+VwSn9i/WlMWp/vns8ezHFVXVMkucnuV8GNxUa/uBwzRZW+8qIeZdmMFI/v92HZzAaMezc1tq53fLK4I+VH2RwmuC8T3fbelJV/c4W/mj46vCT1tqtVXVFBqMkGxfUzt/08K5D8x7cPR5SVT87Yvs3Z3Bn5YVW3e8BgNmoqj0yOJX/qAxO599tQcmW/m1fjGkdYz6TwXHzL6rqkUk+luSfWmvfWmxjjmcwHcI+rBzzo9KnL5j/N0n+d5JnVdWrFnwQsDXfG+O1/0f3uM272VfV8UlO7rb/DxkcWH/QLf7jJHcYtV5r7eoRs2/Jjx/QH95tY6Fzu8eHZfDHzweGt9cF9/dmcDr9EzMYAVnomi28/pbmJ8kuQ/P27h5/Z0T91qzG3wMAS7cpgw+B90vyzW0VV9XtMwjO90vy5QwuQ9ucwaVr98vgA4CR/7Yv0lSOMa21q6vqfyV5VZIjkhzTbeMbGYy+f2hSrzXM8Qy2TdiHFaCq7p7BDWqSZP1gAPs2DsogDJ+zyM0u9kOB5Ec397trfjSqfRtVtXMG19BdluS+beiGQ1V1l4wO6ovWWntFkldspeTY7vGYhXfzXVAzKuwv1fyHAvdsrV04xnqr7vcAwET8SwbH7cMyOANtW47KINS/vbX2/OEFVXVCt3zYrd3jTiO2tceIeVM7xnSXHzy9W/f+SR6bwYfjH6iqh7TWRp7l5ngG0yXsw8rwrAxujnNOBjeAW+jOGRzkn53Fh/1xfDGDEfFHJTl/K3V3zuAPiE+0295Z+KFT6Ou/dV+/c3QGNyj64BbKHpXkl6rq7q21f59wC1/I4FsRHpJknLA/jhX/ewBg0d6dwbe1PK+qTtna9etVdYckd++enjmiZNS/7fPhfb8Ry+4/Yt7UjzGttVu61/liVX0ng0sRH5ctX9LmeAZT5DQXmLGqul0G16H/MMnTWmvPWThlcOO+/0xydHc936T93yTXJzmhqn5mQX9VVfPXy30vgxvwPKCq7jhUsybJn06hr2G/nsG1i+8ZtY+6/XRKBh+abO1GfdvrtCTXJjmpqu65cGFV7VlVo/64Gsdq+D0AsAjdNet/lsHd5s+sqrULa7pjxxuTPDmDG/QmC0JuVf1aksePeIlvZnB3/cdX1Z5D9XdL8sIR9VM5xlTVz1bV/iNe7y7d421u3DfE8QymSNiH2fvlJPsn+Vhr7fJRBW3w3e3vySDsPnnSDbTWNmVw+vvuSb5SVe+rqj+tqndm8MfECV3drUn+MoNLCr5SVW+sqtOS/GuS9ZPua4H5AH/aVmpOz+B6vWd2H6JMTGvte0melsG1++ur6sNV9fqqOrWqPpbk8gxuLrSU11gNvwcAFu/EJG9N8otJLqyqM6vq5G76UJKNGXzH/FUZjOhvyCCMr+tqPprkAxl8g8yPaa3d2G37zknOr6o3VdW7k1yQH93rZrh+WseYX07y3ao6p6r+sqpeW1UfTnJSBt9U89db2jmOZzBdTuOH2ZsPsX+1jbrTMvik/tgk75h0E621M7pT7n4/gwP3Hhl84v6FDP7QmPf7GZw6+Iwkv5XBjXT+IoNP4G+cdF9JUlX3yuAPpfVbuu6vew/f6/4wOirJI5N8fJJ9tNbWVdUDM7hT8iOS/EoG1/JvSPLn2fbvcDGvsWJ/DwCMpzut/Ter6j0ZfCD8S/nRV8ldnOS9Sf6ytfaVJKmqw5O8IckhGVzrf0EGx5o1SZ4w4iVOzOBrZJ+Z5DeTfKt7vCSDS88W9jONY8zHMwjrh2ZwJuKuXf1bk5y8pYGM7XwtYAy1+Bt7AwAAAKuB0/gBAACgZ4R9AAAA6BlhHwAAAHpG2AcAAICeEfYBAACgZ3z13pjufOc7twMPPHDWbQDAqvHlL3/5ytbaPqOWOa4CwHi2dlwdJuyP6cADD8yXvrTFr/kGABaoqou3tMxxFQDGs7Xj6jCn8QMAAEDPCPsAAADQM8I+AAAA9IywDwAAAD0j7AMAAEDPCPsAAADQM8I+AAAA9IywDwAAAD0j7AMAAEDPCPsAAADQM8I+AAAA9IywDwAAAD0j7AMAAEDPCPsAAADQM8I+AAAA9IywDwAAAD2zasJ+Vf1KVX22qq6tqmuq6ktVdfjQ8r2q6p1VdWVVXVdVn6yq+4zYzq5V9fqquryqbqiq86rq0OV9NwAAADA9qyLsV9Xzk3wkyZeTPCHJMUk+kGS3bnklWZfkMUlekOToJLskOaeq1i7Y3LuSPDfJy5MckeTyJB+vqvtN/50AAADA9O086wa2paoOTHJKkuNba6cMLfr40M9HJjkkyeGttXO69c5LclGSlyZ5YTfvvkmemuTY1tpp3bzPJFmf5JXddgCAVWzN2rlsunTjrNvIvvvtn8sv2TDrNgDYQa34sJ/k2CS3JvnLrdQcmeSy+aCfJK21q6vqzCRHpQv7Xd3NSc4Yqrulqt6f5Per6g6ttRsn/QYAgOWz6dKNOeCEs2bdRi4+6YhZtwDADmw1nMZ/SJJvJPn1qvr3qrqlqi6sqt8aqrl3kq+NWHd9krmq2n2o7qLW2vUj6m6f5B4T7h0AAACW3WoI+3dNcs8kr0/yuiSPSvKJJG+pqt/pavZOctWIdTd3j3stsm7vSTQMAAAAs7Qawv7tktwpyfNba+9orX26tXZcko8lObG7OV8laSPWrRHPF1P34wurnldV366qKzZscO0dACyF4yoATN9qCPv/2T1+YsH8s5PcJcmaDEbmR43Kz4/oz4/mb6tu84hlaa29vbV2z9baPnNzc4vtGwAYwXEVAKZvNYT99VuYPz8af2tXc+8RNQcn2dBau3ZoWwdV1W4j6m5KcuESewUAAICZWw1h/0Pd46MXzH90kktaa5uSrEuyX1U9bH5hVe2R5PHdsnnrkuyS5Jihup2TPDnJ2e7EDwAAQB+shq/e+/sk5yR5W1XdOcl3kjwxgxv1PburWZfkvCSnV9XxGZy2f2IGo/8nz2+otXZBVZ2R5JSq2iXJRUmOS3JQkqctz9sBAACA6VrxYb+11qrqV5O8NsmfZHB9/TeSPK219t6u5taqOiLJG5KcmmTXDML/Ya21jQs2+ewkr0ny6iR7Jvlqkse01s5fjvcDAAAA07biw36StNauSfJb3bSlms1Jju2mrW3rhiQv7iYAAADondVwzT4AAAAwBmEfAAAAekbYBwAAgJ4R9gEAAKBnhH0AAADoGWEfAAAAekbYBwAAgJ4R9gEAAKBnhH0AAADoGWEfAAAAekbYBwAAgJ4R9gEAAKBnhH0AAADoGWEfAAAAekbYBwAAgJ4R9gEAAKBnhH0AAADoGWEfAAAAekbYBwAAgJ4R9gEAAKBnhH0AAADoGWEfAAAAekbYBwAAgJ4R9gEAAKBnhH0AAADoGWEfAAAAekbYBwAAgJ4R9gEAAKBnhH0AAADoGWEfAAAAekbYBwAAgJ4R9gEAAKBnhH0AAADoGWEfAAAAekbYBwAAgJ4R9gEAAKBnhH0AAADoGWEfAAAAekbYBwAAgJ4R9gEAAKBnhH0AAADoGWEfAAAAekbYBwAAgJ4R9gEAAKBnhH0AAADoGWEfAAAAekbYBwAAgJ4R9gEAAKBnhH0AAADoGWEfAAAAekbYBwAAgJ4R9gEAAKBnhH0AAADoGWEfAAAAembFh/2qenhVtRHT9xfU7VVV76yqK6vquqr6ZFXdZ8T2dq2q11fV5VV1Q1WdV1WHLt87AgAAgOnaedYNjOGFSb449PyW+R+qqpKsS3JQkhckuSrJiUnOqar7tdYuGVrvXUkel+T4JN9J8ltJPl5Vv9hau2C6bwEAAACmbzWF/a+31j63hWVHJjkkyeGttXOSpKrOS3JRkpdm8EFBquq+SZ6a5NjW2mndvM8kWZ/kld12AAAAYFVb8afxL9KRSS6bD/pJ0lq7OsmZSY5aUHdzkjOG6m5J8v4kj66qOyxPuwAAADA9qynsv6eqflhV/1lV762quaFl907ytRHrrE8yV1W7D9Vd1Fq7fkTd7ZPcY+JdAwAAwDJbDafxX53kjUk+k+SaJPdP8gdJzquq+7fWvpdk7yTfHbHu5u5xryTXdnVXbaVu78m1DQAAALOx4sN+a+0rSb4yNOszVfXZJF/I4Fr8lyWpJG3E6jXi+WLqfnxh1fMyuKHfnvvss88iOwcARnFcBYDpW02n8f+31tr5Sb6V5Oe7WZszelR+r+7xqkXWbR6xLK21t7fW7tla22dubm5UCQCwSI6rADB9qzLsd4ZH6ddncD3+Qgcn2dBau3ao7qCq2m1E3U1JLpxGowAAALCcVmXYr6oHJfmpJJ/vZq1Lsl9VPWyoZo8kj++WZahulyTHDNXtnOTJSc5urd045dYBAABg6lb8NftV9Z4kFyU5P8n3M7hB34lJLk3y5q5sXZLzkpxeVcdncNr+iRmM/p88v63W2gVVdUaSU6pql267xyU5KMnTluUNAQAAwJSt+LCfwVfqPSXJC5LslmRTkr9L8settSuTpLV2a1UdkeQNSU5NsmsG4f+w1trGBdt7dpLXJHl1kj2TfDXJY7r7AAAAAMCqt+LDfmvttUleu4i6zUmO7aat1d2Q5MXdBAAAAL2zKq/ZBwAAALZM2AcAAICeEfYBAACgZ4R9AAAA6BlhHwAAAHpG2AcAAICeEfYBAACgZ4R9AAAA6BlhHwAAAHpG2AcAAICeEfYBAACgZ4R9AAAA6BlhHwAAAHpG2AcAAICeEfYBAACgZ4R9AAAA6BlhHwAAAHpG2AcAAICeEfYBAACgZ4R9AAAA6BlhHwAAAHpG2AcAAICeEfYBAACgZ4R9AAAA6BlhHwAAAHpG2AcAAICeEfYBAACgZ4R9AAAA6BlhHwAAAHpG2AcAAICeEfYBAACgZ4R9AAAA6BlhHwAAAHpG2AcAAICeEfYBAACgZ4R9AAAA6BlhHwAAAHpG2AcAAICeEfYBAACgZ4R9AAAA6BlhHwAAAHpG2AcAAICeEfYBAACgZ4R9AAAA6BlhHwAAAHpG2AcAAICeEfYBAACgZ4R9AAAA6BlhHwAAAHpG2AcAAICeEfYBAACgZ4R9AAAA6BlhHwAAAHpG2AcAAICeWZVhv6o+VlWtql69YP5eVfXOqrqyqq6rqk9W1X1GrL9rVb2+qi6vqhuq6ryqOnT53gEAAABMz6oL+1X1lCT3HTG/kqxL8pgkL0hydJJdkpxTVWsXlL8ryXOTvDzJEUkuT/LxqrrfFFsHAACAZbGqwn5V7ZnkTUlePGLxkUkOSfIbrbX3tdY+1s27XZKXDm3jvkmemuRFrbV3tNY+leRJSTYkeeWU3wIAAABM3aoK+0lOTrK+tfa+EcuOTHJZa+2c+RmttauTnJnkqAV1Nyc5Y6juliTvT/LoqrrDNBoHAACA5bJqwn5VHZLkGUl+cwsl907ytRHz1yeZq6rdh+ouaq1dP6Lu9knuMYF2AQAAYGZWRdivql2SvC3JG1pr39xC2d5Jrhoxf3P3uNci6/be3j4BAABgJVgVYT/JCUnumOQ1W6mpJG0L87en7kcLqp5XVd+uqis2bNiwrV4BgK1wXAWA6VvxYb+q5pL8YZI/SnKHqtqzu1Ffhp7vlMHI/KhR+fkR/fnR/G3VbV64oLX29tbaPVtr+8zNzW3vWwEA4rgKAMthxYf9JHdLsmuS0zMI7PNTkryk+/k+GVxzf+8R6x+cZENr7dru+fokB1XVbiPqbkpy4US7BwAAgGW2GsL+BUkOGzElgw8ADssgoK9Lsl9VPWx+xaraI8nju2Xz1iXZJckxQ3U7J3lykrNbazdO7Z0AAADAMth51g1sS2vt+0nOXTi/qpLk4tbaud3zdUnOS3J6VR2fwYj/iRlci3/y0PYuqKozkpzS3fjvoiTHJTkoydOm+V4AAABgOayGkf1Faa3dmuSIJJ9IcmqSDyX5YZLDWmsbF5Q/O8lpSV6d5KNJ9k/ymNba+cvXMQAAAEzHih/Z35LW2m3unt9a25zk2G7a2ro3JHlxNwEAAECv9GZkHwAAABgQ9gEAAKBnhH0AAADoGWEfAAAAekbYBwAAgJ4R9gEAAKBnhH0AAADoGWEfAAAAekbYBwAAgJ4R9gEAAKBnhH0AAADoGWEfAAAAekbYBwAAgJ4R9gEAAKBnhH0AAADoGWEfAAAAekbYXwHWrJ1LVa2Iac3auVnvDgAAAJZo53GKq2r31tq102pmR7Xp0o054ISzZt1GkuTik46YdQsAAAAs0bgj+5uq6m1Vdf+pdAMAAAAs2bhh/5QkRyT5UlV9saqOrardptAXAAAAsJ3GCvuttZclmUtyTJLNSd6R5LKqenNV/ewU+gMAAADGNPYN+lprP2yt/V1r7dFJ7pnkrRmE/69W1T9V1dOrapdJNwoAAAAszlLvxv+fSS7LYJS/kuyb5N1JLqyqX1zitgEAAIDtsF1hv6oeXFX/J8mlSf40yT8neUBr7R5J7pNkQ5K3T6xLAAAAYNHG/eq945I8P4NA/60kf5jkr1prV8/XtNb+rar+KMknJ9koAAAAsDhjhf0kf57kzCS/11r71Fbqvp3ktdvdFQAAALDdxg37B7bWLttWUWvt0iR/tH0tAQAAAEsx7jX7u1XVIaMWVNVDq+ruE+gJAAAAWIJxw/6fJ/m1LSx7QpI3La0dAAAAYKnGDfs/n+TcLSw7N8mDl9IMK9+atXOpqhUxrVk7N+vdAQAAsCKNe83+Hkl+sIVlNyX5H0trh5Vu06Ubc8AJZ826jSTJxScdMesWAAAAVqRxR/a/k+SwLSw7LMnFS2sHYMe0Us6accYMAEA/jDuyf3qSP66q7yb5P621m6tqlyTHJnlRkldNuD+AHcJKOWvGGTMAAP0wbtg/KckvJHlrkrdU1RVJ7txt58NJXjfZ9gAAAIBxjRX2W2s/TPKrVfWoJI9M8pNJrkxydmvtk1PoDwAAABjTuCP7SZLW2tlJzp5wL7DDWrN2Lpsu3TjrNrLvfvvn8ks2zLoNAABgibYr7CdJVe2dZNeF81trly2pI9gBuV4bAACYpLHCflXdKcmfJfn1JLttoWynpTYFAAAAbL9xR/bfkuRJSf4qyf9LcuOkGwIAAACWZtyw/9gkL22tvXkazQAAAABLd7vtqP/6NBoBAAAAJmPcsP83SR43jUYAAACAyRj3NP6zkvzvqvqJJH+fZPPCgtbaZyfRGAAAALB9tifsJ8ndkjwnSRtaVt1zd+MHAACAGRo37D9yKl0AAAAAEzNW2G+tfWpajQCrw5q1c9l06cZZt5F999s/l1+yYdZtAADAijTuyH6SpKr2SvLgJD+Z5O9ba1dV1S6ttZsn2h2w4my6dGMOOOGsbRdO2cUnHTHrFgCWzAeoAEzL2GG/ql6b5HeT3CGDa/R/PslVST5aVZ9trb16si0CAPSTD1ABmJaxvnqvqk5I8qIkr03y0AxuyjfvzPhaPmAFWbN2LlU182nN2rlZ7woAAHYw447sPy/Jq1prr6mqhXfd/3aSe0ymLYClM2KGU6QBgB3VuGF/bZJ/2cKym5LsvrR2AGByfOADAOyoxjqNP8llSe69hWX3SfLdJXUDAAAALNm4Yf+DSV5eVQ8emteq6u5JXpLkjIl1BgAAAGyXccP+K5JcmMGp/F/v5r0/ydeSXJTBjfsAAACAGRor7LfWrktyaJLnJDk/yblJ/jXJbyd5RGvtxkk3WFWPrqpPV9Wmqrqxqi6pqr+pqoMX1O1fVR+sqqur6pqq+ruqus0tsKtqr6p6Z1VdWVXXVdUnq+o+k+4bAAAAZmXcG/SltXZLktO6aTnsneTLSU5NckWSuSS/n+RzVXWf1trFVbVbkk8nuTHJM5O0JK9Ock5V/Vz3IUWqqpKsS3JQkhckuSrJiV3d/VprlyzTewIAAICpGTvsL7fW2vuSvG94XlV9Ick3kjwxyRuTPDfJ3ZLcq7V2YVfzrxl8HeDzk/xZt+qRSQ5Jcnhr7Zyu7rwMLkF4aZIXTvv9AAAAwLSNFfar6tsZjJpvSWut3WtpLS3Kf3aPN3ePRyb53HzQ7xq5qKr+OclR+fGwf9l80O/qrq6qM7s6YR8AAIBVb9yR/c/ntmH/J5M8JMk1ST47iaZGqaqdkuyU5IAkr0uyKYObAyaDrwP8yIjV1ic5Zuj5vTO4meCoumdU1e6ttWsn1jQAAADMwFhhv7X29FHzq2rvJB9L8tFJNLUFn0/ywO7nCzM4Ff973fO9M7j+fqHNSfYaer53ku9uoS5drbAPsBVr1s5l06UbZ91G9t1v/1x+yYZZtwEAsCJN5Jr91trmqjo5ySuTnDGJbY7wG0n2yODa/Jck+URVHdJa++58GyPWqRHPF1P34wurnpfk+CR77rPPPuP0DNA7my7dmANOOGvWbeTik46YdQtsJ8dVAJi+sb56bxuuz+BO+VPRWvt6a+3z3Q37HpFk9wzuyp8MRvX3HrHaXvnxEf/NW6lLRp8dkNba21tr92yt7TM3N7W3yIStWTuXqloR05q1/rsBmOe4CgDTt+SR/aq6XZKDk7w8ydeX3NEitNa+X1UXJrlHN2t9BtfjL3Rwkn8ber4+yaO2ULfB9fr9slJGHxMjkAAAwPIaa2S/qm6uqpuGpyQ3Jflqkp9J8qJpNDmij7sk+ekk/97NWpfkIVV1t6GaA5M8tFuWobr9quphQ3V7JHn8gjoAAABYtcYd2T8pt73m/QdJLk7y0dbayNPgl6KqPpTk/CT/msEd/38qgw8Vbknyxq7sHUl+O8lHquplXY+vSrIxyduGNrcuyXlJTq+q4zM4bf/EDK7ZP3nSvQMAAMAsjHs3/pdNq5Gt+FySJyX5vSS3zyDAn5vktfM352utXVdVhyd5U5K/ziC8fyrJ7w6fmt9au7WqjkjyhiSnJtk1g/B/WGtt9reWBgAAgAmYyN34p6m1dlIGZxRsq25DkqMXUbc5ybHdBAAAAL0zVtivqrePUd5aa88fsx8AAABgicYd2X9skjtl8H33t2ZwzfteGdyIIX9cAAAaEklEQVTo75ok/zVUO+r77AEAAIApG+tu/BlcO39NkqcnuWNrbZ8kd0zyG938Y1pr+3eTL84FAACAGRh3ZP9NSU5urb13fkZr7eYk76mqvZP8eZIHT7A/AAAAYEzjjuzfN8k3t7Dsm0nus7R2AAAAgKUaN+z/R5InbmHZMUm+t7R2AAAAgKUa9zT+P0/yxqraN8kHMgj/d8ngWv7HJfm9ybYHAAAAjGussN9ae1NVXZ/kj5I8fmjRZUmOa62N89V8AAAAwBSMO7Kf1trbquodSQ5IsibJ5Ukubq3dOunmAAAAgPGNHfaTpAv2F3UTAAAAsIKMe4O+VNXPVdXfVNWmqrqpqh7QzX91VT1q8i0CAAAA4xgr7FfV/0ry+Qy+gu/vkuy0YFv/3+RaAwAAALbHuCP7JyX5VJKfSfLCJDW07EtJHjihvgAAAIDtNO41+w9McnRr7daqqgXLrszga/gAAACAGRp3ZP/GJHfcwrJ9k1y9tHYAAACApRo37P9TkhdW1fB6rXs8Nsk5E+kKAAAA2G7jnsb/8gwC/1eSfCCDoP/0qjo5yUOS/MJk2wMAAADGNdbIfmvtK0kenuT7SV6RwQ36fjfJrkkOa619fcL9AQAAAGMad2Q/rbUvJnlYVe2W5M5Jrmqt/dfEOwMAAAC2y6JH9qvq9lX1vap6fJK01q5vrW0Q9AEAAGBlWXTYb63dlMFp+z+YXjsAAADAUo17N/51SY6eRiMAAADAZIx7zf66JG+pqvcn+XCSy/Ojr95LkrTWPjuh3gAAAIDtMG7Y/1D3+KRuGg761T3faQJ9AQAAANtp3LD/yKl0AQAAAEzMNsN+VR2e5AuttWtba59ahp4AAACAJVjMDfo+keTg+SdVdbuq+mxV3XN6bQEAAADbazFhv0Y8PyTJnSbfDgAAALBU4371HgAAALDCCfsAAADQM4u9G/9+VXW37uedhuZ9f2Fha+07E+kMAAAA2C6LDfsfHDHvw1uo3WkL8wEAYOrWrJ3Lpks3zrqN7Lvf/rn8kg2zbgPYQS0m7D976l0AAMCEbLp0Yw444axZt5GLTzpi1i0AO7Bthv3W2ruXoxEAAABgMtygDwAAAHpG2AcAYJvWrJ1LVc18WrN2bta7AmBVWOwN+gAA2IG5Dh5gdTGyDwAAAD0j7AMAAEDPCPsAAADQM8I+AAAA9IywDwAAAD0j7APACuBrzQCASfLVewCwAvhaMwBgkozsAwAAQM8I+wAAANAzwj4AAAD0jLAPAAAAPSPsAwAAQM8I+wAAANAzwj4AAAD0jLAPAAAAPSPsAwAAQM8I+wAAANAzKz7sV9UTq+pvq+riqrqhqr5ZVa+tqjstqNurqt5ZVVdW1XVV9cmqus+I7e1aVa+vqsu77Z1XVYcu3zsCAACA6VrxYT/JS5L8MMkfJHlMkrcmOS7JJ6rqdklSVZVkXbf8BUmOTrJLknOqau2C7b0ryXOTvDzJEUkuT/Lxqrrf9N8KAAAATN/Os25gER7fWrti6PlnqmpzkncneXiSTyc5MskhSQ5vrZ2TJFV1XpKLkrw0yQu7efdN8tQkx7bWTuvmfSbJ+iSv7LYDAAAAq9qKH9lfEPTnfbF73K97PDLJZfNBv1vv6iRnJjlqaL0jk9yc5IyhuluSvD/Jo6vqDhNsHQAAAGZixYf9LXhY9/j17vHeSb42om59krmq2n2o7qLW2vUj6m6f5B6TbhQAAACW26oL+1W1Xwan3H+ytfalbvbeSa4aUb65e9xrkXV7T6pPAAAAmJVVFfa7EfqPJLklybOHFyVpo1YZ8XwxdQtf93lV9e2qumLDhg1jdAwALOS4Cj+yZu1cqmrm05q1c7PeFcCErYYb9CUZfGVeBnfcv1uSh7XWLhlavDmjR+XnR/SvGqob9S/ZXkPLb6O19vYkb0+SBz3oQaM+LAAAFslxFX5k06Ubc8AJZ826jVx80hGzbgGYsFUxsl9VuyT52yS/kORXWmv/b0HJ+gyux1/o4CQbWmvXDtUdVFW7jai7KcmFk+saAAAAZmPFh/2qul2S9yR5RJKjWmufG1G2Lsl+VfWwofX2SPL4btlw3S5Jjhmq2znJk5Oc3Vq7cfLvAAAAAJbXajiN/y8yCOevSXJdVT1kaNkl3en865Kcl+T0qjo+g9P2T8zgWvyT54tbaxdU1RlJTunOFrgoyXFJDkrytOV4MwAAADBtK35kP8lju8c/zCDQD0/PSZLW2q1JjkjyiSSnJvlQkh8mOay1tnHB9p6d5LQkr07y0ST7J3lMa+386b4NAAAAWB4rfmS/tXbgIus2Jzm2m7ZWd0OSF3cTAAAA9M5qGNkHAAAAxiDsAwAAQM8I+wAAANAzwj4AAAD0jLAPAAAAPSPsAwAAQM8I+wAAANAzwj4AAAD0jLAPAAAAPSPsAwAAQM8I+wAAANAzwj4AAAD0jLAPAAAAPSPsAwAAQM8I+wAAANAzwj4AAAD0jLAPAAAAPSPsAwAAQM8I+wAAANAzwj4AAAD0jLAPAABs05q1c6mqmU9r1s7NelfAqrDzrBsAAABWvk2XbswBJ5w16zZy8UlHzLoFWBWM7AMAAEDPCPsAAADQM8I+AAAA9IywDwAAAD0j7AMAAEDPCPsAAADQM8I+AAAA9IywDwAAAD0j7AMAAEDPCPsAAADQM8I+AAAA9IywDwAAAD0j7AMAAEDPCPsAAADQM8I+AAAA9IywDwAAAD0j7AMAAEDPCPsAAADQM8I+AAAA9IywDwAAAD0j7AMAAEDPCPsAAADQM8I+AAAA9IywDwAAAD0j7AMAAEDPCPsAAADQM8I+AAAA9IywDwAAAD0j7AMAAEDPCPsAAADQM8I+AADQK2vWzqWqZj6tWTs3613BDmznWTcAAAAwSZsu3ZgDTjhr1m3k4pOOmHUL7MCM7AMAAEDPrIqwX1Vrq+rNVXVeVV1fVa2qDhxRt2tVvb6qLq+qG7r6Q0fU3a6qTqyq71bVD6rqq1V19HK8FwAAAJi2VRH2k9wjyZOSXJXkH7dS964kz03y8iRHJLk8ycer6n4L6l6V5BVJ3pLksUk+l+QDVfUrk20bAAAAlt9quWb/s621uyRJVT0nyaMWFlTVfZM8NcmxrbXTunmfSbI+ySuTHNnN+59JXpLkda21N3Srn1NV90jyuiR/P+X3AgAAAFO1Kkb2W2u3LqLsyCQ3JzljaL1bkrw/yaOr6g7d7EcnuX2S0xesf3qS+1TVQUvvGAAAAGZnVYT9Rbp3kotaa9cvmL8+g3B/j6G6G5NcOKIuSQ6eWocAAACwDPoU9vfO4Jr+hTYPLZ9//H5rrW2j7r9V1fOq6ttVdcWGDRsm0iwA7KgcVwFg+voU9ivJwgA/P3976v5ba+3trbV7ttb2mZubW0KLAIDjKgBMX5/C/uaMGJVPstfQ8vnHvapqYbhfWAcAAACrUp/C/vokB1XVbgvmH5zkpvzoGv31Se6Q5O4j6pLk36bWIQAAwJA1a+dSVTOf1qx1plXfrJav3luMdUn+JMkxSd6dJFW1c5InJzm7tXZjV/exDML/07r6eU9P8rXW2kXL1jEAALBD23TpxhxwwlmzbiMXn3TErFtgwlZN2K+qJ3Y/PrB7fGxVXZHkitbaZ1prF1TVGUlOqapdklyU5LgkB2UQ7JMkrbXvVdWbkpxYVf+V5PwMPhA4PMlRy/R2AAAAYGpWTdhP8oEFz0/tHj+T5OHdz89O8pokr06yZ5KvJnlMa+38Bev+YZJrk/xOkn2TfDPJk1prZ06+bQAAAFheqybst9a2eLf8oZobkry4m7ZW98MMPhB49WS6AwAAgJWjTzfoAwAAACLsAwAAQO8I+wAAANAzwj4AAAD0jLAPAAAAPSPsAwAAQM8I+wAAANAzwj4AAAD0jLAPAADAVq1ZO5eqWhHTmrVzs94dq8LOs24AAACAlW3TpRtzwAlnzbqNJMnFJx0x6xZWBSP7AAAA0DPCPgAAAPSMsA8AAAA9I+wDAABAzwj7AAAA0DPCPgAAAPSMsA8AAAA9I+wDAABAzwj7AAAA0DPCPgAAAPSMsA8AAAA9I+wDAABAzwj7AAAA0DPCPgAAAPSMsA8AAAA9I+wDAABAzwj7AAAA9MaatXOpqhUxrVk7N7P9sPPMXhkAAAAmbNOlG3PACWfNuo0kycUnHTGz1zayDwAAAD0j7AMAAEDPCPsAAADQM8I+AAAA9IywDwAAAD0j7AMAAEDPCPsAAADQM8I+AAAA9IywDwAAAD0j7AMAAEDPCPsAAADQM8I+AAAA9IywDwAAAD0j7AMAAEDPCPsAAADQM8I+AAAA9IywDwAAAD0j7AMAAEDPCPsAAADQM8I+AAAA9IywDwAAAD0j7AMAAEDPCPsAAADQM8I+AAAA9IywDwAAAD0j7AMAAEDP7JBhv6r2r6oPVtXVVXVNVf1dVc3Nui8AAACYhB0u7FfVbkk+neSnkzwzyW8kuWeSc6rqJ2bZGwAAAEzCzrNuYAaem+RuSe7VWrswSarqX5N8O8nzk/zZDHsDAACAJdvhRvaTHJnkc/NBP0laaxcl+eckR82sKwAAAJiQHTHs3zvJ10bMX5/k4GXuBQAAACZuRwz7eye5asT8zUn2WuZeAAAAYOKqtTbrHpZVVd2U5I2ttRMXzH9NkhNaa7e5j0FVPS/J8Un2TLJ7BmcBsP3unOTKWTfRQ/brdNiv02G/TsdK3a8HtNb2mX/iuDpxK/X3vtrZr9Nhv06H/TodK3W//thxdUt2xLD/H0k+3Fp7/oL5pyY5ZjE7jaWpqi+11h406z76xn6dDvt1OuzX6bBfd0x+79Nhv06H/Tod9ut0rPb9uiOexr8+g+v2Fzo4yb8tcy8AAAAwcTti2F+X5CFVdbf5GVV1YJKHdssAAABgVdsRw/47knw3yUeq6qiqOjLJR5JsTPK2WTa2A3n7rBvoKft1OuzX6bBfp8N+3TH5vU+H/Tod9ut02K/Tsar36w53zX6SVNVckjcleWSSSvKpJL/bWvvuLPsCAACASdghwz4AAAD02Y54Gj/LrKqeWFV/W1UXV9UNVfXNqnptVd1p1r31TVV9rKpaVb161r2sdlX1K1X12aq6tqquqaovVdXhs+5rNauqh1bV2VX1vW6fnl9Vx866r9WiqtZW1Zur6ryqur77f/3AEXW7VtXrq+ry7t/c86rq0OXvmGlxXF0+jquT47g6eY6rS7MjHFeFfZbDS5L8MMkfJHlMkrcmOS7JJ6rKf4MTUlVPSXLfWffRB1X1/Azu5fHlJE9IckySDyTZbZZ9rWZV9XNJPplklyTPTXJ0ki8meVdVHTfL3laReyR5UpKrkvzjVurelcE+fnmSI5JcnuTjVXW/qXfIcnFcXQaOq5PjuDp5jqsT0fvjqtP4mbqq2qe1dsWCec9I8u4kj2itfXo2nfVHVe2Z5BtJXpTkvUle01p72Wy7Wp26T3S/nuTE1tops+2mP6rqTzMIKHu31q4dmv+5JK219osza26VqKrbtdZu7X5+TgY3nD1o+H4zVXXfJBckOba1dlo3b+cMvnb2m621I5e9cSbOcXX6HFcnx3F1OhxXl25HOK769JepW/gHSeeL3eN+y9lLj52cZH1r7X2zbqQHjv3/27v3WMvK8o7j35+MDFpvAyPx0oJQ0XiFWGutxaI1chOBqHiPYlMM2PpHaVIg1IgKKlWqECXQtIlREHFEFG8wogJeowhDkdQSYEYdEUFgjIJz6zz9410n7m73zNnDnL33nMX3k+ysmXettd/n7MyZZz9rve+7gC3AebMOpGd2BTYBvxtqX4e5aCxzX0jmcSTtc7544LzNwKeAQ5IsnVB4miLz6lSYVxeOeXUyzKs76MGQV/2HoFk5qNv+90yj6IEkBwJvAt4261h64kDa3ZzXJrk1yeYktyT5+1kHtsh9rNuek+QJSR6T5DjgJbSno2hhPANYXVX3D7XfRPti+OTph6QpMa8uEPPqgjOvTsbHuq15dbIWdV5dMusA9OCT5InAu4Erq+raWcezmCV5KHA+8MGq+p9Zx9MTT+heH6DNh72VNrfwI0mWVNXZswxusaqqHyV5EXApv/8CvQk4vqo+NbPA+md32tzDYfcM7FfPmFcXjnl1IsyrE2BenZpFnVct9jVVSR5BW6BlM/CWGYfTBycBDwPOmHUgPfIQ4JHAsVX12a7t692cw1OSnFMudrLdkuwHXEK7En48bdjhUcB5SdZX1YWzjK9HAoz695lpB6LpMK8uOPPqwjOvToB5dWoWdV612NfUJNkNuAzYFzioqtbOOKRFLclewKnA3wFLh+YMLe0WF/pNVf3vTAJcvO4G9gO+OtS+krbq9eOB26cdVA+8l3bH4Yiq2tS1fS3JHsDZSS4ac+6ctu0eYK8R7csG9qsnzKsLy7w6MebVyTCvTseizqvO2ddUdMPiLgGeBxxeVTfOOKQ+2BfYDbiANrxo7gVtddZ7gWfNJrRF7aattM9dwTVxPjDPAm4Y+EIy5/vAHsCe0w+pl24C9kky/DirpwMbgVumH5Imwbw6EebVyTCvToZ5dToWdV612NfEdc/8vZC2YMhRVfW9GYfUF6uAF494Qfui8mJ28v+AdlKXdttDhtoPAdZW1R1Tjqcv7gAOSLLrUPtfAOvZya+MLyKX0Z65fMxcQ/eIoNcAK6tqw6wC08Ixr06MeXUyzKuTYV6djkWdVx3Gr2n4KO0X5AzgviTPH9i31mGHD0xVrQOuGm5PAvCTqvqDfRrLl4FvAOcnWQ7cBrwKOBjnw+6IjwArgC8kOZc2t/BI4HXAh6pq4yyDWyySvKr7459128OS3AXcVVVXV9WqJBcDH+7u/K4GTgD2Ad4w/Yg1IebVCTCvTox5dTLMqwug73k1roehSUuyBth7K7vfVVWnTS+a/ktSwBlV9S+zjmWxSvIo4H20LyPLaI8Men9VfXKmgS1ySQ6jLX71DNpQ2VuBfwfOdw7seLrf71GurqoXdcfMLS72euAxwA3ASRYq/WFenS7z6o4zr06GeXXH9T2vWuxLkiRJktQzztmXJEmSJKlnLPYlSZIkSeoZi31JkiRJknrGYl+SJEmSpJ6x2JckSZIkqWcs9iVJkiRJ6hmLfUmSJEmSesZiX5IkSZKknrHYlyRJACQ5NkkNvO5LsibJpUlenWRi3xuSnJakhtoqyWmT6lOSpD6z2JckScOOAf4SOBx4B7ABuAhYmeRhE+rzP7o+JUnSAlgy6wAkSdJOZ1VV3TLw908kWQGsAP4VePtCd1hVa4G1C/2+O4skS6tqw6zjkCQ9eHhnX5IkzauqLgE+DxyX5OFz7UkenuTMJKuTbOy2pw4P+U/y2CTnJvlZkg3d9hNJlnb7/2AY/yhJ9k9yWZJ7k/wuybeTvHCM857STUe4M8n6JD9NsiLJkoFjthljd8yhSb7b9f3rJJ9L8tShvq5K8q0kL09yfZINwNu6fUuSnJLkx10ftyc5K8lu8/0MkiRtD+/sS5KkcX0ZOBp4LnBNVyhfATwdeA9wI/B82tD/3YF/AkiyDPhO13Y68F/AnsBRwK60aQLzSvIc4JvA9cBxwP3A8cCVSV5QVT/cxulfBNYBJwC/Ap5Im6bwkHFjTHIo8CXg68BrgEcA7wa+leSAqvr5QH9PAc7pPpfbgHu69guAlwNndv09rTvmScArx/kcJEkah8W+JEka10+77eO77euAA4GDquqaru1rSQDemeTMqroT+EdgX+C5VXX9wPtdtJ39f6CL4W+qaiNAkiuAH9EuMBw96qQky4H9gKOq6rKBXZ8c+PM4MZ5OK9wPq6rN3Xt/F7iZdmHjxIFjlwMHV9WqgTheSLtI8Oaq+njXfGWSe4ALugsGq5AkaQE4jF+SJI0r3XZuuP2hwE+A73TD05d0d/tXAg+l3eUHOBj4wVARvX0dt4UBD6KtG7BloK8AVwJ/vY3T76YV6e9PclyS/UYcs80Yk/wR8Bzg4rlCH6CqVgPf7mIbtGZE4X4osBG4ZMTnxTw/gyRJ28ViX5IkjetPuu0vuu2ewN7ApqHX97v9ewxsd3Txvd2BXWh38If7+wdg2dYeDVhVBbwUuBZ4H3BzktuSnDBw2HwxLqNdWPjFiH13dPENGnXcnrQpAb8div/OgRgkSVoQDuOXJEnjehmwHpibG383sBp49VaOX9Nt5+bI74h1wBbgo8DHRx1QVVu2dnJV3Qa8KW2Owf60CwTnJllTVV8ZI8Z7aSMaHjdi3+Non8X/63LEcXfTPr+tLSh4+zb6lyRpu3hnX5IkzSvJK4AjgfOq6v6u+XLa3f7fVtW1I16/6o5bCTwvyf4PtP+quo+2ON/+wHWj+hvzfaobXj83v/6Z48TY9f9D4Jgku8y1J9kbeAFw9RjdXw7sBjx6K5+Xxb4kacF4Z1+SJA07oFvUbldgL+AI4Bjgq8ApA8ddCLyFtijfWcAN3Tl/SrswcHR3YeBDwOtpi9GdTlu1fzltpfvjq+o3Y8Z1InANcEWS/6QNlV9Om0u/S1WdPOqkJM8GzgYuBm6hTQc4FthMW1mfMWN8B201/i8mOZe2Gv+7gF8DZ80XfFVdleQi4DNJ/o023WELbSX+w4GTqurmMT8LSZK2yWJfkiQNW9Ft19Pmk18HvBb4TDf/HYCq2pTkEOBk4K3APsB9wK20onhjd9y6JH9FW83+ZNrc9F/SCu2N4wZVVdcl+XPgnbTH2j0auKuL77xtnHoHbRX/E4E/7n6uG4Ej5h7XN06MVXV5kpd1/X+6a78K+OftuCv/RuDtwN8Cp9IeO7iG9gjDX475HpIkzSsDOVuSJEmSJPWAc/YlSZIkSeoZi31JkiRJknrGYl+SJEmSpJ6x2JckSZIkqWcs9iVJkiRJ6hmLfUmSJEmSesZiX5IkSZKknrHYlyRJkiSpZyz2JUmSJEnqmf8DXtbF+6mpYNkAAAAASUVORK5CYII=\n",
+ "text/plain": [
+ "<Figure size 1008x504 with 2 Axes>"
+ ]
+ },
+ "metadata": {
+ "needs_background": "light"
+ },
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "plt.bar(range(1, 11), compas.decile_score.value_counts(), ec='black')\n",
+ "plt.title(\"Decile scores of all defendants\")\n",
+ "plt.ylabel(\"Frequency\")\n",
+ "plt.xlabel(\"Decile score\")\n",
+ "plt.xticks(range(1, 11))\n",
+ "plt.show()\n",
+ "\n",
+ "fig, ax = compas.query(\"race in ['Caucasian', 'African-American']\").hist(\n",
+ " \"decile_score\",\n",
+ " by=\"race\",\n",
+ " sharey=True,\n",
+ " xrot='horizontal',\n",
+ " ec='black',\n",
+ " bins=np.arange(0.5, 11.5, 1.0),\n",
+ " rwidth=0.8)\n",
+ "\n",
+ "fig.text(-1.5, 350, \"Frequency\", rotation='vertical')\n",
+ "fig.text(11.5, -60, \"Decile score\", horizontalalignment='center')\n",
+ "plt.tight_layout(w_pad=0)\n",
+ "plt.show()"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 50,
+ "metadata": {
+ "scrolled": false
+ },
+ "outputs": [
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 1008x504 with 1 Axes>"
+ ]
+ },
+ "metadata": {
+ "needs_background": "light"
+ },
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "compas['age'].plot.kde()\n",
+ "plt.title(\"Histogram of defendants' ages\")\n",
+ "plt.xlabel(\"Age of defendant\")\n",
+ "plt.xlim(10,90)\n",
+ "plt.show()"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 51,
+ "metadata": {
+ "scrolled": false
+ },
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "<div>\n",
+ "<style scoped>\n",
+ " .dataframe tbody tr th:only-of-type {\n",
+ " vertical-align: middle;\n",
+ " }\n",
+ "\n",
+ " .dataframe tbody tr th {\n",
+ " vertical-align: top;\n",
+ " }\n",
+ "\n",
+ " .dataframe thead th {\n",
+ " text-align: right;\n",
+ " }\n",
+ "</style>\n",
+ "<table border=\"1\" class=\"dataframe\">\n",
+ " <thead>\n",
+ " <tr style=\"text-align: right;\">\n",
+ " <th>is_recid</th>\n",
+ " <th>0</th>\n",
+ " <th>1</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>age_cat</th>\n",
+ " <th></th>\n",
+ " <th></th>\n",
+ " </tr>\n",
+ " </thead>\n",
+ " <tbody>\n",
+ " <tr>\n",
+ " <th>25 - 45</th>\n",
+ " <td>1784</td>\n",
+ " <td>1748</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>Greater than 45</th>\n",
+ " <td>847</td>\n",
+ " <td>446</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>Less than 25</th>\n",
+ " <td>551</td>\n",
+ " <td>796</td>\n",
+ " </tr>\n",
+ " </tbody>\n",
+ "</table>\n",
+ "</div>"
+ ],
+ "text/plain": [
+ "is_recid 0 1\n",
+ "age_cat \n",
+ "25 - 45 1784 1748\n",
+ "Greater than 45 847 446\n",
+ "Less than 25 551 796"
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ },
+ {
+ "data": {
+ "text/html": [
+ "<div>\n",
+ "<style scoped>\n",
+ " .dataframe tbody tr th:only-of-type {\n",
+ " vertical-align: middle;\n",
+ " }\n",
+ "\n",
+ " .dataframe tbody tr th {\n",
+ " vertical-align: top;\n",
+ " }\n",
+ "\n",
+ " .dataframe thead th {\n",
+ " text-align: right;\n",
+ " }\n",
+ "</style>\n",
+ "<table border=\"1\" class=\"dataframe\">\n",
+ " <thead>\n",
+ " <tr style=\"text-align: right;\">\n",
+ " <th>is_recid</th>\n",
+ " <th>0</th>\n",
+ " <th>1</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>sex</th>\n",
+ " <th></th>\n",
+ " <th></th>\n",
+ " </tr>\n",
+ " </thead>\n",
+ " <tbody>\n",
+ " <tr>\n",
+ " <th>Female</th>\n",
+ " <td>740</td>\n",
+ " <td>435</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>Male</th>\n",
+ " <td>2442</td>\n",
+ " <td>2555</td>\n",
+ " </tr>\n",
+ " </tbody>\n",
+ "</table>\n",
+ "</div>"
+ ],
+ "text/plain": [
+ "is_recid 0 1\n",
+ "sex \n",
+ "Female 740 435\n",
+ "Male 2442 2555"
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ },
+ {
+ "data": {
+ "text/html": [
+ "<div>\n",
+ "<style scoped>\n",
+ " .dataframe tbody tr th:only-of-type {\n",
+ " vertical-align: middle;\n",
+ " }\n",
+ "\n",
+ " .dataframe tbody tr th {\n",
+ " vertical-align: top;\n",
+ " }\n",
+ "\n",
+ " .dataframe thead th {\n",
+ " text-align: right;\n",
+ " }\n",
+ "</style>\n",
+ "<table border=\"1\" class=\"dataframe\">\n",
+ " <thead>\n",
+ " <tr style=\"text-align: right;\">\n",
+ " <th></th>\n",
+ " <th>is_recid</th>\n",
+ " <th>0</th>\n",
+ " <th>1</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>race</th>\n",
+ " <th>age_cat</th>\n",
+ " <th></th>\n",
+ " <th></th>\n",
+ " </tr>\n",
+ " </thead>\n",
+ " <tbody>\n",
+ " <tr>\n",
+ " <th rowspan=\"3\" valign=\"top\">African-American</th>\n",
+ " <th>25 - 45</th>\n",
+ " <td>847.0</td>\n",
+ " <td>1051.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>Greater than 45</th>\n",
+ " <td>261.0</td>\n",
+ " <td>207.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>Less than 25</th>\n",
+ " <td>294.0</td>\n",
+ " <td>515.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th rowspan=\"3\" valign=\"top\">Asian</th>\n",
+ " <th>25 - 45</th>\n",
+ " <td>10.0</td>\n",
+ " <td>4.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>Greater than 45</th>\n",
+ " <td>7.0</td>\n",
+ " <td>4.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>Less than 25</th>\n",
+ " <td>4.0</td>\n",
+ " <td>2.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th rowspan=\"3\" valign=\"top\">Caucasian</th>\n",
+ " <th>25 - 45</th>\n",
+ " <td>620.0</td>\n",
+ " <td>508.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>Greater than 45</th>\n",
+ " <td>442.0</td>\n",
+ " <td>186.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>Less than 25</th>\n",
+ " <td>167.0</td>\n",
+ " <td>180.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th rowspan=\"3\" valign=\"top\">Hispanic</th>\n",
+ " <th>25 - 45</th>\n",
+ " <td>180.0</td>\n",
+ " <td>111.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>Greater than 45</th>\n",
+ " <td>81.0</td>\n",
+ " <td>28.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>Less than 25</th>\n",
+ " <td>51.0</td>\n",
+ " <td>58.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th rowspan=\"3\" valign=\"top\">Native American</th>\n",
+ " <th>25 - 45</th>\n",
+ " <td>5.0</td>\n",
+ " <td>2.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>Greater than 45</th>\n",
+ " <td>NaN</td>\n",
+ " <td>2.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>Less than 25</th>\n",
+ " <td>NaN</td>\n",
+ " <td>2.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th rowspan=\"3\" valign=\"top\">Other</th>\n",
+ " <th>25 - 45</th>\n",
+ " <td>122.0</td>\n",
+ " <td>72.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>Greater than 45</th>\n",
+ " <td>56.0</td>\n",
+ " <td>19.0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>Less than 25</th>\n",
+ " <td>35.0</td>\n",
+ " <td>39.0</td>\n",
+ " </tr>\n",
+ " </tbody>\n",
+ "</table>\n",
+ "</div>"
+ ],
+ "text/plain": [
+ "is_recid 0 1\n",
+ "race age_cat \n",
+ "African-American 25 - 45 847.0 1051.0\n",
+ " Greater than 45 261.0 207.0\n",
+ " Less than 25 294.0 515.0\n",
+ "Asian 25 - 45 10.0 4.0\n",
+ " Greater than 45 7.0 4.0\n",
+ " Less than 25 4.0 2.0\n",
+ "Caucasian 25 - 45 620.0 508.0\n",
+ " Greater than 45 442.0 186.0\n",
+ " Less than 25 167.0 180.0\n",
+ "Hispanic 25 - 45 180.0 111.0\n",
+ " Greater than 45 81.0 28.0\n",
+ " Less than 25 51.0 58.0\n",
+ "Native American 25 - 45 5.0 2.0\n",
+ " Greater than 45 NaN 2.0\n",
+ " Less than 25 NaN 2.0\n",
+ "Other 25 - 45 122.0 72.0\n",
+ " Greater than 45 56.0 19.0\n",
+ " Less than 25 35.0 39.0"
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "tab = compas.groupby(['age_cat', 'is_recid']).size()\n",
+ "display(tab.unstack())\n",
+ "\n",
+ "tab = compas.groupby(['sex', 'is_recid']).size()\n",
+ "display(tab.unstack())\n",
+ "\n",
+ "tab = compas.groupby(['race', 'age_cat', 'is_recid']).size()\n",
+ "display(tab.unstack())"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "From above it is clear that there are no Native American recidivists of age over 45 or under 25. There are some other value combinations that might be problematic. Therefore the procedure of estimating $P(X=x)$ has to be considered carefully."
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "### Synthetic data\n",
+ "\n",
+ "In the chunk below, we generate the synthetic data as described by Lakkaraju et al. The default values and definitions of $Y$ and $T$ values follow their description.\n",
+ "\n",
+ "**Parameters**\n",
+ "\n",
+ "* M = `nJudges_M`, number of judges\n",
+ "* N = `nSubjects_N`, number of subjects assigned to each judge\n",
+ "* betas $\\beta_i$ = `beta_i`, where $i \\in \\{X, Z, W\\}$ are coefficients for the respected variables\n",
+ "\n",
+ "**Columns of the data:**\n",
+ "\n",
+ "* `judgeID_J` = judge IDs as running numbering from 0 to `nJudges_M - 1`\n",
+ "* R = `acceptanceRate_R`, acceptance rates\n",
+ "* X = `X`, invidual's features observable to all (models and judges)\n",
+ "* Z = `Z`, information observable for judges only\n",
+ "* W = `W`, unobservable / inaccessible information\n",
+ "* T = `decision_T`, bail-or-jail decisions where $T=0$ represents jail decision and $T=1$ bail decision.\n",
+ "* Y = `result_Y`, result variable, if $Y=0$ person will or would recidivate and if $Y=1$ person will or would not commit a crime."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 52,
+ "metadata": {
+ "scrolled": false
+ },
+ "outputs": [
+ {
+ "data": {
+ "text/html": [
+ "<div>\n",
+ "<style scoped>\n",
+ " .dataframe tbody tr th:only-of-type {\n",
+ " vertical-align: middle;\n",
+ " }\n",
+ "\n",
+ " .dataframe tbody tr th {\n",
+ " vertical-align: top;\n",
+ " }\n",
+ "\n",
+ " .dataframe thead th {\n",
+ " text-align: right;\n",
+ " }\n",
+ "</style>\n",
+ "<table border=\"1\" class=\"dataframe\">\n",
+ " <thead>\n",
+ " <tr style=\"text-align: right;\">\n",
+ " <th></th>\n",
+ " <th>count</th>\n",
+ " <th>mean</th>\n",
+ " <th>std</th>\n",
+ " <th>min</th>\n",
+ " <th>25%</th>\n",
+ " <th>50%</th>\n",
+ " <th>75%</th>\n",
+ " <th>max</th>\n",
+ " </tr>\n",
+ " </thead>\n",
+ " <tbody>\n",
+ " <tr>\n",
+ " <th>judgeID_J</th>\n",
+ " <td>50000.0</td>\n",
+ " <td>49.500000</td>\n",
+ " <td>28.866359</td>\n",
+ " <td>0.000000</td>\n",
+ " <td>24.750000</td>\n",
+ " <td>49.500000</td>\n",
+ " <td>74.250000</td>\n",
+ " <td>99.000000</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>acceptanceRate_R</th>\n",
+ " <td>50000.0</td>\n",
+ " <td>0.478235</td>\n",
+ " <td>0.230644</td>\n",
+ " <td>0.103756</td>\n",
+ " <td>0.264643</td>\n",
+ " <td>0.473985</td>\n",
+ " <td>0.647587</td>\n",
+ " <td>0.890699</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>X</th>\n",
+ " <td>50000.0</td>\n",
+ " <td>-0.003875</td>\n",
+ " <td>0.996715</td>\n",
+ " <td>-4.659953</td>\n",
+ " <td>-0.671782</td>\n",
+ " <td>-0.001726</td>\n",
+ " <td>0.668077</td>\n",
+ " <td>3.831790</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>Z</th>\n",
+ " <td>50000.0</td>\n",
+ " <td>0.006964</td>\n",
+ " <td>0.998001</td>\n",
+ " <td>-4.852118</td>\n",
+ " <td>-0.666258</td>\n",
+ " <td>0.004730</td>\n",
+ " <td>0.679477</td>\n",
+ " <td>4.241772</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>W</th>\n",
+ " <td>50000.0</td>\n",
+ " <td>0.010863</td>\n",
+ " <td>0.996944</td>\n",
+ " <td>-4.029138</td>\n",
+ " <td>-0.666574</td>\n",
+ " <td>0.012306</td>\n",
+ " <td>0.679578</td>\n",
+ " <td>4.285856</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>result_Y</th>\n",
+ " <td>50000.0</td>\n",
+ " <td>0.496500</td>\n",
+ " <td>0.499993</td>\n",
+ " <td>0.000000</td>\n",
+ " <td>0.000000</td>\n",
+ " <td>0.000000</td>\n",
+ " <td>1.000000</td>\n",
+ " <td>1.000000</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>probabilities_T</th>\n",
+ " <td>50000.0</td>\n",
+ " <td>0.500627</td>\n",
+ " <td>0.410701</td>\n",
+ " <td>-1.350727</td>\n",
+ " <td>0.214651</td>\n",
+ " <td>0.500403</td>\n",
+ " <td>0.786029</td>\n",
+ " <td>2.005477</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>decision_T</th>\n",
+ " <td>50000.0</td>\n",
+ " <td>0.477260</td>\n",
+ " <td>0.499488</td>\n",
+ " <td>0.000000</td>\n",
+ " <td>0.000000</td>\n",
+ " <td>0.000000</td>\n",
+ " <td>1.000000</td>\n",
+ " <td>1.000000</td>\n",
+ " </tr>\n",
+ " </tbody>\n",
+ "</table>\n",
+ "</div>"
+ ],
+ "text/plain": [
+ " count mean std min 25% \\\n",
+ "judgeID_J 50000.0 49.500000 28.866359 0.000000 24.750000 \n",
+ "acceptanceRate_R 50000.0 0.478235 0.230644 0.103756 0.264643 \n",
+ "X 50000.0 -0.003875 0.996715 -4.659953 -0.671782 \n",
+ "Z 50000.0 0.006964 0.998001 -4.852118 -0.666258 \n",
+ "W 50000.0 0.010863 0.996944 -4.029138 -0.666574 \n",
+ "result_Y 50000.0 0.496500 0.499993 0.000000 0.000000 \n",
+ "probabilities_T 50000.0 0.500627 0.410701 -1.350727 0.214651 \n",
+ "decision_T 50000.0 0.477260 0.499488 0.000000 0.000000 \n",
+ "\n",
+ " 50% 75% max \n",
+ "judgeID_J 49.500000 74.250000 99.000000 \n",
+ "acceptanceRate_R 0.473985 0.647587 0.890699 \n",
+ "X -0.001726 0.668077 3.831790 \n",
+ "Z 0.004730 0.679477 4.241772 \n",
+ "W 0.012306 0.679578 4.285856 \n",
+ "result_Y 0.000000 1.000000 1.000000 \n",
+ "probabilities_T 0.500403 0.786029 2.005477 \n",
+ "decision_T 0.000000 1.000000 1.000000 "
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ },
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "0 26137\n",
+ "1 23863\n",
+ "Name: decision_T, dtype: int64\n"
+ ]
+ },
+ {
+ "data": {
+ "text/html": [
+ "<div>\n",
+ "<style scoped>\n",
+ " .dataframe tbody tr th:only-of-type {\n",
+ " vertical-align: middle;\n",
+ " }\n",
+ "\n",
+ " .dataframe tbody tr th {\n",
+ " vertical-align: top;\n",
+ " }\n",
+ "\n",
+ " .dataframe thead th {\n",
+ " text-align: right;\n",
+ " }\n",
+ "</style>\n",
+ "<table border=\"1\" class=\"dataframe\">\n",
+ " <thead>\n",
+ " <tr style=\"text-align: right;\">\n",
+ " <th>decision_T</th>\n",
+ " <th>0</th>\n",
+ " <th>1</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>result_Y</th>\n",
+ " <th></th>\n",
+ " <th></th>\n",
+ " </tr>\n",
+ " </thead>\n",
+ " <tbody>\n",
+ " <tr>\n",
+ " <th>0.0</th>\n",
+ " <td>17790</td>\n",
+ " <td>7385</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>1.0</th>\n",
+ " <td>8347</td>\n",
+ " <td>16478</td>\n",
+ " </tr>\n",
+ " </tbody>\n",
+ "</table>\n",
+ "</div>"
+ ],
+ "text/plain": [
+ "decision_T 0 1\n",
+ "result_Y \n",
+ "0.0 17790 7385\n",
+ "1.0 8347 16478"
+ ]
+ },
+ "metadata": {},
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "# Set seed for reproducibility\n",
+ "npr.seed(0)\n",
+ "\n",
+ "def generateData(nJudges_M=100,\n",
+ " nSubjects_N=500,\n",
+ " beta_X=1.0,\n",
+ " beta_Z=1.0,\n",
+ " beta_W=0.2):\n",
+ "\n",
+ " # Assign judge IDs as running numbering from 0 to nJudges_M - 1\n",
+ " judgeID_J = np.repeat(np.arange(0, nJudges_M, dtype=np.int32), nSubjects_N)\n",
+ "\n",
+ " # Sample acceptance rates uniformly from a closed interval\n",
+ " # from 0.1 to 0.9 and round to tenth decimal place.\n",
+ " acceptance_rates = np.round(npr.uniform(.1, .9, nJudges_M), 10)\n",
+ "\n",
+ " # Replicate the rates so they can be attached to the corresponding judge ID.\n",
+ " acceptanceRate_R = np.repeat(acceptance_rates, nSubjects_N)\n",
+ "\n",
+ " # Sample the variables from standard Gaussian distributions.\n",
+ " X = npr.normal(size=nJudges_M * nSubjects_N)\n",
+ " Z = npr.normal(size=nJudges_M * nSubjects_N)\n",
+ " W = npr.normal(size=nJudges_M * nSubjects_N)\n",
+ "\n",
+ " probabilities_Y = 1 / (1 + np.exp(-(beta_X * X + beta_Z * Z + beta_W * W)))\n",
+ "\n",
+ " # 0 if P(Y = 0| X = x; Z = z; W = w) >= 0.5 , 1 otherwise\n",
+ " result_Y = 1 - probabilities_Y.round()\n",
+ "\n",
+ " probabilities_T = 1 / (1 + np.exp(-(beta_X * X + beta_Z * Z)))\n",
+ " probabilities_T += npr.normal(0, np.sqrt(0.1), nJudges_M * nSubjects_N)\n",
+ "\n",
+ " # Initialize decision values as 1\n",
+ " decision_T = np.ones(nJudges_M * nSubjects_N)\n",
+ "\n",
+ " # Initialize the dataframe\n",
+ " df_init = pd.DataFrame(\n",
+ " np.column_stack((judgeID_J, acceptanceRate_R, X, Z, W, result_Y,\n",
+ " probabilities_T, decision_T)),\n",
+ " columns=[\n",
+ " \"judgeID_J\", \"acceptanceRate_R\", \"X\", \"Z\", \"W\", \"result_Y\",\n",
+ " \"probabilities_T\", \"decision_T\"\n",
+ " ])\n",
+ "\n",
+ " # Sort by judges then probabilities\n",
+ " data = df_init.sort_values(\n",
+ " by=[\"judgeID_J\", \"probabilities_T\"], ascending=False)\n",
+ "\n",
+ " # Iterate over the data. Subject is in the top (1-r)*100% if\n",
+ " # his within-judge-index is over acceptance threshold times\n",
+ " # the number of subjects assigned to each judge. If subject\n",
+ " # is over the limit they are assigned a zero, else one.\n",
+ " data.reset_index(drop=True, inplace=True)\n",
+ "\n",
+ " data['decision_T'] = np.where(\n",
+ " (data.index.values % nSubjects_N) <\n",
+ " ((1 - data['acceptanceRate_R']) * nSubjects_N), 0, 1)\n",
+ "\n",
+ " return data\n",
+ "\n",
+ "\n",
+ "df = []\n",
+ "df = generateData()\n",
+ "\n",
+ "# Basic stats of the created data set.\n",
+ "display(df.describe().T)\n",
+ "\n",
+ "print(df.decision_T.value_counts())\n",
+ "\n",
+ "tab = df.groupby(['result_Y', 'decision_T']).size()\n",
+ "display(tab.unstack())"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 53,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "(25000, 8)\n",
+ "(25000, 8)\n",
+ "(25000, 8)\n",
+ "(25000, 8)\n"
+ ]
+ },
+ {
+ "data": {
+ "text/html": [
+ "<div>\n",
+ "<style scoped>\n",
+ " .dataframe tbody tr th:only-of-type {\n",
+ " vertical-align: middle;\n",
+ " }\n",
+ "\n",
+ " .dataframe tbody tr th {\n",
+ " vertical-align: top;\n",
+ " }\n",
+ "\n",
+ " .dataframe thead th {\n",
+ " text-align: right;\n",
+ " }\n",
+ "</style>\n",
+ "<table border=\"1\" class=\"dataframe\">\n",
+ " <thead>\n",
+ " <tr style=\"text-align: right;\">\n",
+ " <th>decision_T</th>\n",
+ " <th>1</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>result_Y</th>\n",
+ " <th></th>\n",
+ " </tr>\n",
+ " </thead>\n",
+ " <tbody>\n",
+ " <tr>\n",
+ " <th>0.0</th>\n",
+ " <td>3650</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <th>1.0</th>\n",
+ " <td>8216</td>\n",
+ " </tr>\n",
+ " </tbody>\n",
+ "</table>\n",
+ "</div>"
+ ],
+ "text/plain": [
+ "decision_T 1\n",
+ "result_Y \n",
+ "0.0 3650\n",
+ "1.0 8216"
+ ]
+ },
+ "execution_count": 53,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "# Split the data set to test and train\n",
+ "from sklearn.model_selection import train_test_split\n",
+ "train, test = train_test_split(df, test_size=0.5, random_state=0)\n",
+ "\n",
+ "print(train.shape)\n",
+ "print(test.shape)\n",
+ "\n",
+ "train_labeled = train.copy()\n",
+ "test_labeled = test.copy()\n",
+ "\n",
+ "# Set results as NA if decision is negative.\n",
+ "train_labeled.result_Y = np.where(train.decision_T == 0, np.nan, train.result_Y)\n",
+ "test_labeled.result_Y = np.where(test.decision_T == 0, np.nan, test.result_Y)\n",
+ "\n",
+ "print(train_labeled.shape)\n",
+ "print(test_labeled.shape)\n",
+ "\n",
+ "tab = train_labeled.groupby(['result_Y', 'decision_T']).size()\n",
+ "tab.unstack()"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "## Algorithms\n",
+ "\n",
+ "### Contraction algorithm\n",
+ "\n",
+ "Below is an implementation of Lakkaraju's team's algorithm presented in [their paper](https://helka.finna.fi/PrimoRecord/pci.acm3098066). Relevant parameters to be passed to the function are presented in the description."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 54,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "def contraction(df,\n",
+ " judgeIDJ_col,\n",
+ " decisionT_col,\n",
+ " resultY_col,\n",
+ " modelProbS_col,\n",
+ " accRateR_col,\n",
+ " r,\n",
+ " binning=False):\n",
+ " '''\n",
+ " This is an implementation of the algorithm presented by Lakkaraju\n",
+ " et al. in their paper \"The Selective Labels Problem: Evaluating \n",
+ " Algorithmic Predictions in the Presence of Unobservables\" (2017).\n",
+ " \n",
+ " Parameters:\n",
+ " df = The (Pandas) data frame containing the data, judge decisions,\n",
+ " judge IDs, results and probability scores.\n",
+ " judgeIDJ_col = String, the name of the column containing the judges' IDs\n",
+ " in df.\n",
+ " decisionT_col = String, the name of the column containing the judges' decisions\n",
+ " resultY_col = String, the name of the column containing the realization\n",
+ " modelProbS_col = String, the name of the column containing the probability\n",
+ " scores from the black-box model B.\n",
+ " accRateR_col = String, the name of the column containing the judges' \n",
+ " acceptance rates\n",
+ " r = Float between 0 and 1, the given acceptance rate.\n",
+ " binning = Boolean, should judges with same acceptance rate be binned\n",
+ " \n",
+ " Returns:\n",
+ " u = The estimated failure rate at acceptance rate r.\n",
+ " '''\n",
+ " # Sort first by acceptance rate and judge ID.\n",
+ " sorted_df = df.sort_values(\n",
+ " by=[accRateR_col, judgeIDJ_col], ascending=False)\n",
+ "\n",
+ " if binning:\n",
+ " # Get maximum leniency\n",
+ " max_leniency = sorted_df[accRateR_col].values[0].round(1)\n",
+ "\n",
+ " # Get list of judges that are the most lenient\n",
+ " most_lenient_list = sorted_df.loc[sorted_df[accRateR_col].round(1) ==\n",
+ " max_leniency, judgeIDJ_col]\n",
+ "\n",
+ " # Subset to obtain D_q\n",
+ " D_q = sorted_df[sorted_df[judgeIDJ_col].isin(\n",
+ " most_lenient_list.unique())].copy()\n",
+ " else:\n",
+ " # Get most lenient judge\n",
+ " most_lenient_ID = sorted_df[judgeIDJ_col].values[0]\n",
+ "\n",
+ " # Subset\n",
+ " D_q = sorted_df[sorted_df[judgeIDJ_col] == most_lenient_ID].copy()\n",
+ "\n",
+ " # All observations of R_q have observed outcome labels\n",
+ " R_q = D_q[D_q[decisionT_col] == 1]\n",
+ "\n",
+ " # \"Observations deemed as high risk by B are at the top of this list\"\n",
+ " R_sort_q = R_q.sort_values(by=modelProbS_col, ascending=False)\n",
+ "\n",
+ " number_to_remove = int(\n",
+ " round((1.0 - r) * D_q.shape[0] - (D_q.shape[0] - R_q.shape[0])))\n",
+ "\n",
+ " # \"R_B is the list of observations assigned to t = 1 by B\"\n",
+ " R_B = R_sort_q[number_to_remove:R_sort_q.shape[0]]\n",
+ "\n",
+ " return np.sum(R_B[resultY_col] == 0) / D_q.shape[0]"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "### Causal model\n",
+ "\n",
+ "Our model is defined by the probabilistic expression \n",
+ "\n",
+ "\\begin{equation}\\label{model_disc}\n",
+ "P(Y=0 | \\text{do}(R=r)) = \\sum_x \\underbrace{P(Y=0|X=x, T=1)}_\\text{1} \n",
+ "\\overbrace{P(T=1|R=r, X=x)}^\\text{2} \n",
+ "\\underbrace{P(X=x)}_\\text{3}\n",
+ "\\end{equation}\n",
+ "\n",
+ "which is equal to \n",
+ "\n",
+ "\\begin{equation}\\label{model_cont}\n",
+ "P(Y=0 | \\text{do}(R=r)) = \\int_x P(Y=0|X=x, T=1)P(T=1|R=r, X=x)P(X=x)\n",
+ "\\end{equation}\n",
+ "\n",
+ "for continuous $x$. As a picture (Z is latent variable, and can be excluded from the expression as per do-calculus):\n",
+ "\n",
+ "![Model as picture](../figures/intervention_model.png \"Intervention model\")\n",
+ "\n",
+ "Does this hold for predicting probability for each individual? Of course we build models for each of the terms separately, that is we model the probability of negative outcome and the probability of positive decision when $R=r$ separately.\n",
+ "\n",
+ "\\begin{equation}\n",
+ "P(Y=0 | \\text{do}(R=r), X=x) = P(Y=0|X=x, T=1)P(T=1|R=r, X=x)\n",
+ "\\end{equation}\n",
+ "\n",
+ "<!---\n",
+ "**Algorithm -- UPDATE!!**\n",
+ "\n",
+ "Our model will be constructed sequentially.\n",
+ "\n",
+ "Input: Training and test data sets $(\\mathbf{x}, t, y) \\in \\mathcal{D}$ and acceptance rate $r$. \n",
+ "Returns: $P(Y=0 | \\text{do}(R=r))$\n",
+ "\n",
+ "Procedure:\n",
+ "1. Model $P(X=x)$ in a suitable way and assign to $\\mathcal{M}_0$\n",
+ "* Build model $\\mathcal{M}_1$ predicting response $Y$ with predictors $X$ from the labeled observations (where $T=1$) in training data.\n",
+ "* Predict $P(Y=0|X=x)$ for every observation in the test data using model $\\mathcal{M}_1$.\n",
+ "* Initialize `sum = 0`\n",
+ "* For every point in the parameter space (for every $x$ in $X$)\n",
+ " 1. $p_x \\leftarrow P(X=x)$ from $\\mathcal{M}_0$\n",
+ " * $\\mathcal{D_x} \\leftarrow \\{\\mathcal{D} | X = x\\}$\n",
+ " * Assign first $r\\cdot 100\\%$ observations from $\\mathcal{D_x}$ to $\\mathcal{D}_{rx}$\n",
+ " * $p_t \\leftarrow \\dfrac{|\\{\\mathcal{D}_{rx}|T=1\\}|}{|\\mathcal{D}_{rx}|}$ (part 2 of eq. $\\ref{model}$) Pitääkö tähänkin treenaa joku oma luokittelija?\n",
+ " * $p_y$ will be predicted from the model $\\mathcal{M}_1$\n",
+ " * `sum +=` $p_y \\cdot p_t \\cdot p_x$\n",
+ "* Return `sum`\n",
+ "--->\n",
+ "**Constructing $P(X=x)$, preliminary ideas:**\n",
+ "\n",
+ "* Approximate it with frequencies (makes independence assumption, make variables factors first)\n",
+ "* Construct Bayesian network using some well-known algorithm.\n",
+ "\n",
+ "<!---\n",
+ "Functions:\n",
+ "\n",
+ "* $f(x)$ gives probability of recidivism given personal properties and predictive model. Corresponds to parts 1 and 2 of eq $\\ref{model_disc}$ and $\\ref{model_cont}$.\n",
+ "* `ep` counts performance of the predictive model given a data, model and leniency rate like Michael's pdf. That is:\n",
+ "--->"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 55,
+ "metadata": {
+ "scrolled": false
+ },
+ "outputs": [],
+ "source": [
+ "def f(x, model, class_value):\n",
+ " '''\n",
+ " Returns the probabilities (as vector) of class value (class_value) given \n",
+ " individual features (x) and the trained, predictive model (model).\n",
+ " '''\n",
+ " if x.ndim == 1:\n",
+ " # if x is vector, transform to column matrix.\n",
+ " f_values = model.predict_proba(np.array(x).reshape(-1, 1))\n",
+ " else:\n",
+ " f_values = model.predict_proba(x)\n",
+ "\n",
+ " return f_values[:, model.classes_ == class_value].flatten()\n",
+ "\n",
+ "def ep(r, df, result_col, feature_cols, model, failure_value):\n",
+ " '''\n",
+ " Returns:\n",
+ " Empirical performance (float), i.e. percentage of recidivists. \n",
+ " \n",
+ " Parameters:\n",
+ " r = leniency rate(s)\n",
+ " df = test data, pandas DataFrame\n",
+ " result_col = String (or list of), column(s) containing the binarized results.\n",
+ " feature_cols = String (or list of), column(s) containing the individual features.\n",
+ " model = trained sklearn classifier. \n",
+ " failure_value = value obtained from the model.classes_ representing the \n",
+ " unwanted event label (usually 0 or 1).\n",
+ " '''\n",
+ " # Initialize DataFrame\n",
+ " ep_data = pd.DataFrame()\n",
+ "\n",
+ " # Attach booleans indicating failure\n",
+ " ep_data = ep_data.assign(failed=df[result_col] == failure_value)\n",
+ "\n",
+ " # Attach prediction values\n",
+ " ep_data = ep_data.assign(\n",
+ " failure_predictions=f(df[feature_cols], model, failure_value))\n",
+ "\n",
+ " # sort by predictions, most harmless at top\n",
+ " ep_data.sort_values(by='failure_predictions', inplace=True, ascending=True)\n",
+ "\n",
+ " # calculate number of subjects to which assign a positive decision\n",
+ " if isinstance(r, float):\n",
+ " to_release = int(round(ep_data.shape[0] * r))\n",
+ "\n",
+ " # subset data\n",
+ " released = ep_data[0:to_release]\n",
+ "\n",
+ " return np.sum(released.failed)/df.shape[0]\n",
+ " else:\n",
+ " results = np.zeros(0)\n",
+ " for r_value in r:\n",
+ " to_release = int(round(ep_data.shape[0] * r_value))\n",
+ "\n",
+ " # subset data\n",
+ " released = ep_data[0:to_release]\n",
+ "\n",
+ " results = np.append(results, np.sum(released.failed)/df.shape[0])\n",
+ " return results\n",
+ "\n",
+ "\n",
+ "def ep2_0(r, df, result_col, decision_col, feature_cols, r_col, y_model,\n",
+ " t_model, failure_value, pos_decision_value):\n",
+ " '''\n",
+ " Returns:\n",
+ " Empirical performance, i.e. percentage of recidivists. \n",
+ " \n",
+ " Parameters:\n",
+ " r = leniency rate(s)\n",
+ " df = test data, pandas DataFrame\n",
+ " result_col = String (list), name of column containing the binarized results.\n",
+ " feature_cols = String (list), name of columns containge individual features.\n",
+ " model = trained sklearn classifier \n",
+ " failure_value = value obtained from the model.classes_ representing the \n",
+ " unwanted event label (usually 0 or 1).\n",
+ " '''\n",
+ " # Initialize DataFrame\n",
+ " ep_data = pd.DataFrame()\n",
+ "\n",
+ " # Attach booleans indicating failure\n",
+ " ep_data = ep_data.assign(failed=df[result_col] == failure_value)\n",
+ "\n",
+ " # Attach prediction values\n",
+ " ep_data = ep_data.assign(\n",
+ " failure_predictions=f(df[feature_cols], y_model, failure_value))\n",
+ "\n",
+ " ep_data.failure_predictions = ep_data.failure_predictions * f(\n",
+ " df[[feature_cols, r_col]], t_model, pos_decision_value) * scs.norm.pdf(df[feature_cols])\n",
+ " # sort by predictions, most harmless at top\n",
+ " ep_data.sort_values(by='failure_predictions', inplace=True, ascending=True)\n",
+ "\n",
+ " # calculate number of subjects to which assign a positive decision\n",
+ " if isinstance(r, float):\n",
+ " to_release = int(round(ep_data.shape[0] * r))\n",
+ "\n",
+ " # subset data\n",
+ " released = ep_data[0:to_release]\n",
+ "\n",
+ " return np.sum(released.failed)/df.shape[0]\n",
+ " else:\n",
+ " results = np.zeros(0)\n",
+ " for r_value in r:\n",
+ " to_release = int(round(ep_data.shape[0] * r_value))\n",
+ "\n",
+ " # subset data\n",
+ " released = ep_data[0:to_release]\n",
+ "\n",
+ " results = np.append(results, np.sum(released.failed)/df.shape[0])\n",
+ " return results\n",
+ "\n",
+ "\n",
+ "def gp(r, x_values, y_model, x_model, failure_value):\n",
+ " '''\n",
+ " Returns:\n",
+ " Generalized performance\n",
+ " \n",
+ " Parameters:\n",
+ " r = leniency rate\n",
+ " df = test data, pandas DataFrame\n",
+ " feature_cols = String (list), name of columns containing individual features.\n",
+ " y_model = trained sklearn classifier to predict response\n",
+ " x_model = model of P(X=x)\n",
+ " failure_value = value obtained from the model.classes_ representing the \n",
+ " unwanted event label.\n",
+ " '''\n",
+ " preds = f(x_values, y_model, failure_value)\n",
+ "\n",
+ " return np.sum(preds * (preds < r) * x_model(x_values))"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "## Performance comparison\n",
+ "\n",
+ "Below we try to replicate the results obtained by Lakkaraju and compare their model's performance to the one of ours.\n",
+ "\n",
+ "### On synthetic data\n",
+ "\n",
+ "#### Predictive models\n",
+ "\n",
+ "Lakkaraju says that they used logistic regression to predict recidivism. We construct the models using only *observed observations*, i.e. defendants that were granted bail and are in the train set. We then predict the probability of recidivism for all observations in the test data and attach it to our data set. I also applied random forest classifier."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 56,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "# instantiate the model (using the default parameters)\n",
+ "logreg = LogisticRegression(solver='lbfgs')\n",
+ "\n",
+ "# fit, reshape X to be of shape (n_samples, n_features)\n",
+ "logreg = logreg.fit(\n",
+ " train_labeled.X[train_labeled.decision_T == 1].values.reshape(-1, 1),\n",
+ " train_labeled.result_Y[train_labeled.decision_T == 1])\n",
+ "\n",
+ "# predict probabilities and attach to data\n",
+ "label_probs_logreg = logreg.predict_proba(test.X.values.reshape(-1, 1))\n",
+ "test = test.assign(B_prob_0_logreg=label_probs_logreg[:, 0])\n",
+ "\n",
+ "test_labeled = test_labeled.assign(B_prob_0_logreg=label_probs_logreg[:, 0])\n",
+ "\n",
+ "########\n",
+ "\n",
+ "# instantiate the model (using the default parameters)\n",
+ "forest = RandomForestClassifier(n_estimators=400, max_depth=8, random_state=0)\n",
+ "\n",
+ "# fit, reshape X to be of shape (n_samples, n_features)\n",
+ "forest = forest.fit(\n",
+ " train_labeled.X[train_labeled.decision_T == 1].values.reshape(-1, 1),\n",
+ " train_labeled.result_Y[train_labeled.decision_T == 1])\n",
+ "\n",
+ "# predict probabilities and attach to data\n",
+ "label_probs_forest = forest.predict_proba(test.X.values.reshape(-1, 1))\n",
+ "test = test.assign(B_prob_0_forest=label_probs_forest[:, 0])\n",
+ "\n",
+ "test_labeled = test_labeled.assign(B_prob_0_forest=label_probs_forest[:, 0])"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 57,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "0.037128152319060165 [0 1] [0. 1.]\n"
+ ]
+ },
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 1008x504 with 1 Axes>"
+ ]
+ },
+ "metadata": {
+ "needs_background": "light"
+ },
+ "output_type": "display_data"
+ },
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 1008x504 with 1 Axes>"
+ ]
+ },
+ "metadata": {
+ "needs_background": "light"
+ },
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "# instantiate the model (using the default parameters)\n",
+ "t_malli = LogisticRegression(solver='lbfgs')\n",
+ "\n",
+ "# fit, reshape X to be of shape (n_samples, n_features)\n",
+ "t_malli = t_malli.fit(train[['X', 'acceptanceRate_R']], train.decision_T)\n",
+ "\n",
+ "#test_labeled['B_prob_0_causal'] =\n",
+ "\n",
+ "test_labeled['do_R'] = 0.1\n",
+ "\n",
+ "tmp = t_malli.predict_proba(test_labeled[['X', 'do_R']])[:, 1] * logreg.predict_proba(test_labeled.X.values.reshape(-1, 1))[:, 0]\n",
+ "\n",
+ "print(max(tmp), t_malli.classes_, logreg.classes_)\n",
+ "plt.hist(tmp, bins=30);plt.show()\n",
+ "\n",
+ "plt.scatter(test_labeled.X, t_malli.predict_proba(test_labeled[['X', 'do_R']])[:, 1], c=test_labeled.result_Y);plt.show()"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 58,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "#from sklearn.tree import export_graphviz\n",
+ "## Export as dot file\n",
+ "#export_graphviz(forest.estimators_[0], out_file='tree.dot', \n",
+ "# feature_names = 'X',\n",
+ "# class_names = ['0', '1'],\n",
+ "# rounded = True, proportion = False, \n",
+ "# precision = 2, filled = True)\n",
+ "#\n",
+ "#import pydot\n",
+ "#\n",
+ "#(graph,) = pydot.graph_from_dot_file('tree.dot')\n",
+ "#graph.write_png('tree.png')\n",
+ "#\n",
+ "## Display in jupyter notebook\n",
+ "#from IPython.display import Image\n",
+ "#Image(filename = 'tree.png')"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "#### Visual comparison\n",
+ "\n",
+ "Let's plot the failure rates against the acceptance rates using the difference."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 59,
+ "metadata": {
+ "scrolled": false
+ },
+ "outputs": [
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 1008x576 with 1 Axes>"
+ ]
+ },
+ "metadata": {
+ "needs_background": "light"
+ },
+ "output_type": "display_data"
+ },
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "[[0.0516 0.0196 0.0153 0.012 0.0287]\n",
+ " [0.1066 0.0436 0.0249 0.0359 0.0456]\n",
+ " [0.156 0.0585 0.0547 0.0677 0.0701]\n",
+ " [0.2053 0.073 0.0892 0.1116 0.103 ]\n",
+ " [0.2558 0.0917 0.1323 0.1793 0.1437]\n",
+ " [0.3033 0.1058 0.1852 0.2231 0.1895]\n",
+ " [0.3535 0.1194 0.2403 0.2749 0.2359]\n",
+ " [0.4022 0.1322 0.3315 0.3586 0.2783]]\n"
+ ]
+ }
+ ],
+ "source": [
+ "failure_rates = np.zeros((8, 5))\n",
+ "\n",
+ "for r in np.arange(1, 9):\n",
+ " \n",
+ " #### True evaluation\n",
+ " # Sort by failure probabilities, subjects with the smallest risk are first. \n",
+ " df_sorted = test.sort_values(\n",
+ " by='B_prob_0_logreg', inplace=False, ascending=True)\n",
+ "\n",
+ " to_release = int(round(df_sorted.shape[0] * r / 10))\n",
+ "\n",
+ " # Failure was coded as zero.\n",
+ " failure_rates[r - 1, 0] = np.mean(df_sorted.result_Y[0:to_release] == 0)\n",
+ " \n",
+ " #### Labeled outcomes only\n",
+ " # Sort by failure probabilities, subjects with the smallest risk are first. \n",
+ " df_sorted = test_labeled.sort_values(\n",
+ " by='B_prob_0_logreg', inplace=False, ascending=True)\n",
+ "\n",
+ " to_release = int(round(df_sorted.shape[0] * r / 10))\n",
+ "\n",
+ " failure_rates[r - 1, 1] = np.mean(df_sorted.result_Y[0:to_release] == 0) # keskiarvo resulteista, mutta kun siellä ne NAt\n",
+ " \n",
+ " #### Human error rate\n",
+ " # Get judges with correct leniency as list\n",
+ " correct_leniency_list = test_labeled.judgeID_J[\n",
+ " test_labeled['acceptanceRate_R'].round(1) == r / 10].values\n",
+ "\n",
+ " # Released are the people they judged and released, T = 1\n",
+ " released = test_labeled[test_labeled.judgeID_J.isin(correct_leniency_list)\n",
+ " & (test_labeled.decision_T == 1)]\n",
+ "\n",
+ " # Get their failure rate, aka ratio of reoffenders to number of people judged in total\n",
+ " failure_rates[r - 1, 2] = np.sum(\n",
+ " released.result_Y == 0) / correct_leniency_list.shape[0]\n",
+ " # onko jakaja oikein\n",
+ " \n",
+ " #### Contraction, logistic regression\n",
+ " failure_rates[r - 1, 3] = contraction(\n",
+ " test_labeled, 'judgeID_J', 'decision_T', 'result_Y', 'B_prob_0_logreg',\n",
+ " 'acceptanceRate_R', r / 10, False)\n",
+ "\n",
+ " #### P(Y=0 | T=1, X=x)*P(T=1 | R=r, X=x)*P(X=x)\n",
+ " failure_rates[r - 1, 4] = si.quad(lambda x: f(np.array([x]), logreg, 0)*f(np.array([[x, r/10]]), t_malli, 1)*scs.norm.pdf(x), -np.inf, np.inf)[0]\n",
+ "\n",
+ "# klassifikaatioille scipy.stats semin kautta error barit xerr ja yerr argumenttien kautta\n",
+ "\n",
+ "plt.figure(figsize=(14, 8))\n",
+ "plt.plot(np.arange(0.1, 0.9, .1), failure_rates[:, 0], label='True Evaluation', c='green')\n",
+ "plt.plot(np.arange(0.1, 0.9, .1), failure_rates[:, 1], label='Labeled outcomes', c='lime')\n",
+ "plt.plot(np.arange(0.1, 0.9, .1), failure_rates[:, 2], label='Human evaluation', c='red')\n",
+ "plt.plot(np.arange(0.1, 0.9, .1), failure_rates[:, 3], label='Contraction, log.', c='blue')\n",
+ "plt.plot(np.arange(0.1, 0.9, .1), failure_rates[:, 4], label='Integrand', c='magenta')\n",
+ "\n",
+ "plt.title('Failure rate vs. Acceptance rate')\n",
+ "plt.xlabel('Acceptance rate')\n",
+ "plt.ylabel('Failure rate')\n",
+ "plt.legend()\n",
+ "plt.grid()\n",
+ "plt.show()\n",
+ "\n",
+ "with np.printoptions(precision=4, suppress=True):\n",
+ " print(failure_rates)"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 60,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Mean absolute errors:\n",
+ "0.0\n",
+ "0.1488280357142857\n",
+ "0.09513124234987873\n",
+ "0.07143974791310947\n",
+ "integrand: 0.0924364389817069\n",
+ "[0.04311835 0.01373383 0.03947055 0.04332996 0.03225757]\n"
+ ]
+ }
+ ],
+ "source": [
+ "print(\"Mean absolute errors:\")\n",
+ "for i in range(failure_rates.shape[1]):\n",
+ " if i == 4:\n",
+ " print(\"integrand: \", end=\"\")\n",
+ " print(np.mean(np.abs(failure_rates[:, 0] - failure_rates[:, i])))\n",
+ "\n",
+ "print(scs.sem(failure_rates, axis=0))\n",
+ "# true, labeled, human, contraction, integrand"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 61,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "[0.0001962]\n",
+ "[0.00089983]\n",
+ "[0.00387493]\n",
+ "[0.01469389]\n",
+ "[0.04412076]\n",
+ "[0.08669258]\n",
+ "[0.08755081]\n",
+ "[0.05041144]\n",
+ "[0.02300916]\n",
+ "[0.00982404]\n"
+ ]
+ }
+ ],
+ "source": [
+ "for x in np.arange(-5, 5):\n",
+ " r = 0.3\n",
+ " print(f(np.array([[x, r]]), t_malli, 1)*f(np.array([x]), logreg, 0))"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 62,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "0.0 (0.017619869764739768, 7.585898195449956e-11)\n",
+ "0.02040816326530612 (0.019499240205571863, 8.655956024063706e-11)\n",
+ "0.04081632653061224 (0.021562205154989997, 9.880436235634218e-11)\n",
+ "0.061224489795918366 (0.02382327833090757, 1.1274381060923062e-10)\n",
+ "0.08163265306122448 (0.026297456959911286, 1.2849031088594973e-10)\n",
+ "0.1020408163265306 (0.02900010492672288, 1.4609844185002997e-10)\n",
+ "0.12244897959183673 (0.0319468078833058, 1.6554943762895352e-10)\n",
+ "0.14285714285714285 (0.03515319858110887, 1.8674473764395006e-10)\n",
+ "0.16326530612244897 (0.038634751352196184, 2.095119758676625e-10)\n",
+ "0.18367346938775508 (0.04240654552924697, 2.3362449214498625e-10)\n",
+ "0.2040816326530612 (0.04648299866372316, 2.5883188075795255e-10)\n",
+ "0.22448979591836732 (0.050877571666397205, 2.8489597344974724e-10)\n",
+ "0.24489795918367346 (0.05560244943011784, 3.1162530874696104e-10)\n",
+ "0.26530612244897955 (0.06066820205896193, 3.389008915464566e-10)\n",
+ "0.2857142857142857 (0.06608343346006608, 3.6668878631143796e-10)\n",
+ "0.3061224489795918 (0.07185442567550888, 3.950372592961083e-10)\n",
+ "0.32653061224489793 (0.07798478884648338, 4.240593484446999e-10)\n",
+ "0.3469387755102041 (0.08447512800332105, 4.539029413703657e-10)\n",
+ "0.36734693877551017 (0.09132273884923851, 4.847103787620876e-10)\n",
+ "0.3877551020408163 (0.09852134524130086, 5.165699403846364e-10)\n",
+ "0.4081632653061224 (0.10606089106859018, 5.494592801492646e-10)\n",
+ "0.42857142857142855 (0.11392739860585298, 5.831822383147574e-10)\n",
+ "0.44897959183673464 (0.1221029041332785, 6.172998775429425e-10)\n",
+ "0.4693877551020408 (0.13056547965190782, 6.50998996053682e-10)\n",
+ "0.4897959183673469 (0.1392893469282113, 6.831022211771777e-10)\n",
+ "0.5102040816326531 (0.14824508695845484, 7.121281083428359e-10)\n",
+ "0.5306122448979591 (0.15739994438907143, 7.360615587244717e-10)\n",
+ "0.5510204081632653 (0.16671822264028938, 7.524755669431017e-10)\n",
+ "0.5714285714285714 (0.17616176166439657, 7.586667272574147e-10)\n",
+ "0.5918367346938775 (0.18569048665110674, 7.519459735769229e-10)\n",
+ "0.6122448979591836 (0.19526301279342237, 7.301802499164036e-10)\n",
+ "0.6326530612244897 (0.20483728865197567, 6.927757357768648e-10)\n",
+ "0.6530612244897959 (0.2143712588715651, 6.426429112279573e-10)\n",
+ "0.673469387755102 (0.22382352612718304, 5.921440913935919e-10)\n",
+ "0.6938775510204082 (0.23315399226401498, 6.074283572483916e-10)\n",
+ "0.7142857142857142 (0.2423244596368379, 6.818423366197291e-10)\n",
+ "0.7346938775510203 (0.2512991755747974, 8.178925862612687e-10)\n",
+ "0.7551020408163265 (0.260045305566269, 1.0267506205637807e-09)\n",
+ "0.7755102040816326 (0.2685333239986697, 1.3174097463721993e-09)\n",
+ "0.7959183673469387 (0.27673731489308406, 1.693465127253933e-09)\n",
+ "0.8163265306122448 (0.28463517882427547, 2.150800685998324e-09)\n",
+ "0.836734693877551 (0.292208745898638, 2.6764016466967304e-09)\n",
+ "0.8571428571428571 (0.2994437980818355, 3.2487043441125506e-09)\n",
+ "0.8775510204081632 (0.30633000716306324, 3.8396644375453745e-09)\n",
+ "0.8979591836734693 (0.3128607970943802, 4.4183558134388365e-09)\n",
+ "0.9183673469387754 (0.3190331412775102, 4.955636569829531e-09)\n",
+ "0.9387755102040816 (0.3248473065588803, 5.429888447004714e-09)\n",
+ "0.9591836734693877 (0.3303065562506764, 5.846557896489933e-09)\n",
+ "0.9795918367346939 (0.33541682447132515, 6.200391636431277e-09)\n",
+ "1.0 (0.3401863735703823, 6.4855735760226974e-09)\n"
+ ]
+ }
+ ],
+ "source": [
+ "for r in np.linspace(0,1):\n",
+ " print(r, si.quad(lambda x: f(np.array([[x, r]]), t_malli, 1)*f(np.array([x]), logreg, 0)*scs.norm.pdf(x), -np.inf, np.inf))"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 63,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "0.0 12543\n",
+ "1.0 12457\n",
+ "Name: result_Y, dtype: int64"
+ ]
+ },
+ "execution_count": 63,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "test.result_Y.value_counts()"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "**Thoughts:**\n",
+ "\n",
+ "Failure rates still too high for about 10 percentage points compared to Lakkaraju paper. Failure rates will change if seed is changed (e.g. with seed 0 contraction's failure rates are approximately 0.31, causal doesn't change that much). It seems like the contraction or our model is some how predicting the wrong thing. Behavior after 0.5 is not consistent? (Curves curve down in Lakkaraju's paper. + Human evaluation curve jumps to the wrong side.) Have to check some rounding rules."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 64,
+ "metadata": {
+ "scrolled": false
+ },
+ "outputs": [
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 1008x504 with 1 Axes>"
+ ]
+ },
+ "metadata": {
+ "needs_background": "light"
+ },
+ "output_type": "display_data"
+ },
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "0.14484459336296954\n"
+ ]
+ }
+ ],
+ "source": [
+ "r_vals = np.linspace(0, 1, 200)\n",
+ "\n",
+ "probs = np.zeros(0)\n",
+ "\n",
+ "for r in r_vals:\n",
+ " probs = np.append(probs, si.quad(lambda x: f(np.array([[x, r]]), t_malli, 1)*f(np.array([x]), logreg, 0)*scs.norm.pdf(x), -np.inf, np.inf)[0])\n",
+ "\n",
+ "plt.plot(r_vals, probs)\n",
+ "plt.grid()\n",
+ "plt.show()\n",
+ "\n",
+ "\n",
+ "print(probs[np.argmin(np.abs(r_vals-0.5))])"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 65,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 1008x504 with 1 Axes>"
+ ]
+ },
+ "metadata": {
+ "needs_background": "light"
+ },
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "# Mindless comparison as X is continuous (we should integrate).\n",
+ "\n",
+ "thresholds = np.linspace(.0, 1.0)\n",
+ "\n",
+ "x_values = np.linspace(-10, 10, 1000)\n",
+ "\n",
+ "rates_logistic = np.zeros(0)\n",
+ "rates_forest = np.zeros(0)\n",
+ "\n",
+ "for leniency in thresholds:\n",
+ " rates_logistic = np.append(rates_logistic, gp(leniency, x_values, logreg, lambda x: scs.norm.pdf(x), 0))\n",
+ " rates_forest = np.append(rates_forest, gp(leniency, x_values, forest, lambda x: scs.norm.pdf(x), 0))\n",
+ "\n",
+ "plt.plot(thresholds, rates_logistic, label=\"Logistic model\")\n",
+ "plt.plot(thresholds, rates_forest, label=\"Random forest\")\n",
+ "plt.title(\"Generalized performance over synthetic data\")\n",
+ "plt.legend()\n",
+ "plt.show()"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "### On COMPAS data\n",
+ "\n",
+ "\n",
+ "#### Predictive models\n",
+ "\n",
+ "Let's build the predictive models (first here random forest and logistic regression). Some of our variables are string so they will first have to be transformed to be dummy / indicator variables."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 66,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "# convert string values to dummies, drop first so full rank\n",
+ "compas_dummy = pd.get_dummies(compas, columns=['c_charge_degree', 'race', 'age_cat', 'score_text', 'sex'], drop_first=True)\n",
+ "\n",
+ "########\n",
+ "\n",
+ "predict_columns = ['priors_count', 'days_b_screening_arrest', 'length_of_stay',\n",
+ " 'c_charge_degree_M', 'race_Asian', 'race_Caucasian', 'race_Hispanic',\n",
+ " 'race_Native American', 'race_Other', 'age_cat_Greater than 45',\n",
+ " 'age_cat_Less than 25', 'score_text_Low', 'score_text_Medium', 'sex_Male']\n",
+ "\n",
+ "response_column = 'two_year_recid'\n",
+ "\n",
+ "# instantiate the model (using the default parameters)\n",
+ "logreg_c = LogisticRegression(solver='lbfgs', max_iter=1000)\n",
+ "\n",
+ "# fit, reshape X to be of shape (n_samples, n_features)\n",
+ "logreg_c = logreg_c.fit(compas_dummy[predict_columns], compas_dummy[response_column])\n",
+ "\n",
+ "########\n",
+ "\n",
+ "# instantiate the model\n",
+ "forest_c = RandomForestClassifier(n_estimators=300, max_depth=5, random_state=0)\n",
+ "\n",
+ "# fit, reshape X to be of shape (n_samples, n_features)\n",
+ "forest_c = forest_c.fit(compas_dummy[predict_columns], compas_dummy[response_column])"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 67,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "<Figure size 1008x576 with 1 Axes>"
+ ]
+ },
+ "metadata": {
+ "needs_background": "light"
+ },
+ "output_type": "display_data"
+ }
+ ],
+ "source": [
+ "failures_compas = np.zeros((11, 2))\n",
+ "\n",
+ "for r in np.arange(0, 11):\n",
+ " ## Causal model with logistic regression\n",
+ " failures_compas[r, 0] = ep([r / 10], compas_dummy, response_column, predict_columns, logreg_c, 0)\n",
+ " \n",
+ " ## Causal model with random forest classifier\n",
+ " failures_compas[r, 1] = ep([r / 10], compas_dummy, response_column, predict_columns, forest_c, 0)\n",
+ "\n",
+ "# klassifikaatioille scipy.stats semin kautta error barit xerr ja yerr argumenttien kautta\n",
+ "\n",
+ "plt.figure(figsize=(14, 8))\n",
+ "plt.plot(np.arange(0, 11) / 10, failures_compas[:, 0], label='Causal model, log.')\n",
+ "plt.plot(np.arange(0, 11) / 10, failures_compas[:, 1], label='Causal model, for.')\n",
+ "\n",
+ "plt.title('Failure rate vs. Acceptance rate - COMPAS')\n",
+ "plt.xlabel('Leniency')\n",
+ "plt.ylabel('Failure rate')\n",
+ "plt.legend()\n",
+ "plt.show()\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "Of course if leniency is one, then the empirical performance should always converge to the proportion of false positives in the data."
+ ]
+ }
+ ],
+ "metadata": {
+ "kernelspec": {
+ "display_name": "Python 3",
+ "language": "python",
+ "name": "python3"
+ },
+ "language_info": {
+ "codemirror_mode": {
+ "name": "ipython",
+ "version": 3
+ },
+ "file_extension": ".py",
+ "mimetype": "text/x-python",
+ "name": "python",
+ "nbconvert_exporter": "python",
+ "pygments_lexer": "ipython3",
+ "version": "3.7.0"
+ },
+ "toc": {
+ "base_numbering": 1,
+ "nav_menu": {
+ "height": "265.233px",
+ "width": "283.233px"
+ },
+ "number_sections": true,
+ "sideBar": true,
+ "skip_h1_title": true,
+ "title_cell": "Table of Contents",
+ "title_sidebar": "Contents",
+ "toc_cell": true,
+ "toc_position": {
+ "height": "calc(100% - 180px)",
+ "left": "10px",
+ "top": "150px",
+ "width": "251.4px"
+ },
+ "toc_section_display": true,
+ "toc_window_display": true
+ },
+ "varInspector": {
+ "cols": {
+ "lenName": 16,
+ "lenType": 16,
+ "lenVar": 40
+ },
+ "kernels_config": {
+ "python": {
+ "delete_cmd_postfix": "",
+ "delete_cmd_prefix": "del ",
+ "library": "var_list.py",
+ "varRefreshCmd": "print(var_dic_list())"
+ },
+ "r": {
+ "delete_cmd_postfix": ") ",
+ "delete_cmd_prefix": "rm(",
+ "library": "var_list.r",
+ "varRefreshCmd": "cat(var_dic_list()) "
+ }
+ },
+ "position": {
+ "height": "465.65px",
+ "left": "843.6px",
+ "right": "16.2px",
+ "top": "159px",
+ "width": "676.2px"
+ },
+ "types_to_exclude": [
+ "module",
+ "function",
+ "builtin_function_or_method",
+ "instance",
+ "_Feature"
+ ],
+ "window_display": false
+ }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}
diff --git a/analysis_and_scripts/Compas Analysis.ipynb b/analysis_and_scripts/Compas Analysis.ipynb
index 43596e7b77ba390700a1d737e29eada1b3696cda..cce5687f22aaa7cd9aedb253b426425ab111eb45 100644
--- a/analysis_and_scripts/Compas Analysis.ipynb
+++ b/analysis_and_scripts/Compas Analysis.ipynb
@@ -1,5 +1,15 @@
{
"cells": [
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "toc": true
+ },
+ "source": [
+ "<h1>Table of Contents<span class=\"tocSkip\"></span></h1>\n",
+ "<div class=\"toc\"><ul class=\"toc-item\"><li><span><a href=\"#Loading-the-Data\" data-toc-modified-id=\"Loading-the-Data-1\"><span class=\"toc-item-num\">1 </span>Loading the Data</a></span></li><li><span><a href=\"#Racial-Bias-in-Compas\" data-toc-modified-id=\"Racial-Bias-in-Compas-2\"><span class=\"toc-item-num\">2 </span>Racial Bias in Compas</a></span><ul class=\"toc-item\"><li><span><a href=\"#Risk-of-Violent-Recidivism\" data-toc-modified-id=\"Risk-of-Violent-Recidivism-2.1\"><span class=\"toc-item-num\">2.1 </span>Risk of Violent Recidivism</a></span></li></ul></li><li><span><a href=\"#Predictive-Accuracy-of-COMPAS\" data-toc-modified-id=\"Predictive-Accuracy-of-COMPAS-3\"><span class=\"toc-item-num\">3 </span>Predictive Accuracy of COMPAS</a></span></li><li><span><a href=\"#Directions-of-the-Racial-Bias\" data-toc-modified-id=\"Directions-of-the-Racial-Bias-4\"><span class=\"toc-item-num\">4 </span>Directions of the Racial Bias</a></span></li><li><span><a href=\"#Risk-of-Violent-Recidivism\" data-toc-modified-id=\"Risk-of-Violent-Recidivism-5\"><span class=\"toc-item-num\">5 </span>Risk of Violent Recidivism</a></span></li><li><span><a href=\"#Gender-differences-in-Compas-scores\" data-toc-modified-id=\"Gender-differences-in-Compas-scores-6\"><span class=\"toc-item-num\">6 </span>Gender differences in Compas scores</a></span></li></ul></div>"
+ ]
+ },
{
"cell_type": "markdown",
"metadata": {},
@@ -2383,6 +2393,48 @@
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.7"
+ },
+ "toc": {
+ "base_numbering": 1,
+ "nav_menu": {},
+ "number_sections": true,
+ "sideBar": true,
+ "skip_h1_title": true,
+ "title_cell": "Table of Contents",
+ "title_sidebar": "Contents",
+ "toc_cell": true,
+ "toc_position": {},
+ "toc_section_display": true,
+ "toc_window_display": true
+ },
+ "varInspector": {
+ "cols": {
+ "lenName": 16,
+ "lenType": 16,
+ "lenVar": 40
+ },
+ "kernels_config": {
+ "python": {
+ "delete_cmd_postfix": "",
+ "delete_cmd_prefix": "del ",
+ "library": "var_list.py",
+ "varRefreshCmd": "print(var_dic_list())"
+ },
+ "r": {
+ "delete_cmd_postfix": ") ",
+ "delete_cmd_prefix": "rm(",
+ "library": "var_list.r",
+ "varRefreshCmd": "cat(var_dic_list()) "
+ }
+ },
+ "types_to_exclude": [
+ "module",
+ "function",
+ "builtin_function_or_method",
+ "instance",
+ "_Feature"
+ ],
+ "window_display": false
}
},
"nbformat": 4,
diff --git a/analysis_and_scripts/tree.dot b/analysis_and_scripts/tree.dot
new file mode 100644
index 0000000000000000000000000000000000000000..4b473230d4b6e4d43a9a24c1ce13de64589c289f
--- /dev/null
+++ b/analysis_and_scripts/tree.dot
@@ -0,0 +1,609 @@
+digraph Tree {
+node [shape=box, style="filled, rounded", color="black", fontname=helvetica] ;
+edge [fontname=helvetica] ;
+0 [label="X <= 0.06\ngini = 0.42\nsamples = 22638\nvalue = [10938, 24937]\nclass = 1", fillcolor="#399de58f"] ;
+1 [label="X <= -0.58\ngini = 0.26\nsamples = 14836\nvalue = [3590, 19957]\nclass = 1", fillcolor="#399de5d1"] ;
+0 -> 1 [labeldistance=2.5, labelangle=45, headlabel="True"] ;
+2 [label="X <= -1.07\ngini = 0.14\nsamples = 8795\nvalue = [1085, 12887]\nclass = 1", fillcolor="#399de5ea"] ;
+1 -> 2 ;
+3 [label="X <= -1.54\ngini = 0.08\nsamples = 4657\nvalue = [294, 7122]\nclass = 1", fillcolor="#399de5f4"] ;
+2 -> 3 ;
+4 [label="X <= -1.73\ngini = 0.03\nsamples = 2168\nvalue = [53, 3392]\nclass = 1", fillcolor="#399de5fb"] ;
+3 -> 4 ;
+5 [label="X <= -2.12\ngini = 0.02\nsamples = 1502\nvalue = [21, 2321]\nclass = 1", fillcolor="#399de5fd"] ;
+4 -> 5 ;
+6 [label="X <= -2.26\ngini = 0.0\nsamples = 647\nvalue = [2, 998]\nclass = 1", fillcolor="#399de5fe"] ;
+5 -> 6 ;
+7 [label="gini = 0.0\nsamples = 457\nvalue = [0, 696]\nclass = 1", fillcolor="#399de5ff"] ;
+6 -> 7 ;
+8 [label="X <= -2.26\ngini = 0.01\nsamples = 190\nvalue = [2, 302]\nclass = 1", fillcolor="#399de5fd"] ;
+6 -> 8 ;
+9 [label="gini = 0.0\nsamples = 1\nvalue = [2, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+8 -> 9 ;
+10 [label="gini = 0.0\nsamples = 189\nvalue = [0, 302]\nclass = 1", fillcolor="#399de5ff"] ;
+8 -> 10 ;
+11 [label="X <= -2.12\ngini = 0.03\nsamples = 855\nvalue = [19, 1323]\nclass = 1", fillcolor="#399de5fb"] ;
+5 -> 11 ;
+12 [label="gini = 0.0\nsamples = 1\nvalue = [1, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+11 -> 12 ;
+13 [label="X <= -2.08\ngini = 0.03\nsamples = 854\nvalue = [18, 1323]\nclass = 1", fillcolor="#399de5fc"] ;
+11 -> 13 ;
+14 [label="gini = 0.09\nsamples = 54\nvalue = [4, 83]\nclass = 1", fillcolor="#399de5f3"] ;
+13 -> 14 ;
+15 [label="gini = 0.02\nsamples = 800\nvalue = [14, 1240]\nclass = 1", fillcolor="#399de5fc"] ;
+13 -> 15 ;
+16 [label="X <= -1.73\ngini = 0.06\nsamples = 666\nvalue = [32, 1071]\nclass = 1", fillcolor="#399de5f7"] ;
+4 -> 16 ;
+17 [label="gini = 0.0\nsamples = 1\nvalue = [3, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+16 -> 17 ;
+18 [label="X <= -1.57\ngini = 0.05\nsamples = 665\nvalue = [29, 1071]\nclass = 1", fillcolor="#399de5f8"] ;
+16 -> 18 ;
+19 [label="X <= -1.57\ngini = 0.06\nsamples = 580\nvalue = [29, 934]\nclass = 1", fillcolor="#399de5f7"] ;
+18 -> 19 ;
+20 [label="gini = 0.06\nsamples = 579\nvalue = [28, 934]\nclass = 1", fillcolor="#399de5f7"] ;
+19 -> 20 ;
+21 [label="gini = 0.0\nsamples = 1\nvalue = [1, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+19 -> 21 ;
+22 [label="gini = 0.0\nsamples = 85\nvalue = [0, 137]\nclass = 1", fillcolor="#399de5ff"] ;
+18 -> 22 ;
+23 [label="X <= -1.54\ngini = 0.11\nsamples = 2489\nvalue = [241, 3730]\nclass = 1", fillcolor="#399de5ef"] ;
+3 -> 23 ;
+24 [label="gini = 0.0\nsamples = 1\nvalue = [3, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+23 -> 24 ;
+25 [label="X <= -1.27\ngini = 0.11\nsamples = 2488\nvalue = [238, 3730]\nclass = 1", fillcolor="#399de5ef"] ;
+23 -> 25 ;
+26 [label="X <= -1.29\ngini = 0.1\nsamples = 1321\nvalue = [107, 1999]\nclass = 1", fillcolor="#399de5f1"] ;
+25 -> 26 ;
+27 [label="X <= -1.29\ngini = 0.1\nsamples = 1196\nvalue = [106, 1805]\nclass = 1", fillcolor="#399de5f0"] ;
+26 -> 27 ;
+28 [label="gini = 0.1\nsamples = 1195\nvalue = [104, 1805]\nclass = 1", fillcolor="#399de5f0"] ;
+27 -> 28 ;
+29 [label="gini = 0.0\nsamples = 1\nvalue = [2, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+27 -> 29 ;
+30 [label="X <= -1.29\ngini = 0.01\nsamples = 125\nvalue = [1, 194]\nclass = 1", fillcolor="#399de5fe"] ;
+26 -> 30 ;
+31 [label="gini = 0.0\nsamples = 65\nvalue = [0, 104]\nclass = 1", fillcolor="#399de5ff"] ;
+30 -> 31 ;
+32 [label="gini = 0.02\nsamples = 60\nvalue = [1, 90]\nclass = 1", fillcolor="#399de5fc"] ;
+30 -> 32 ;
+33 [label="X <= -1.27\ngini = 0.13\nsamples = 1167\nvalue = [131, 1731]\nclass = 1", fillcolor="#399de5ec"] ;
+25 -> 33 ;
+34 [label="X <= -1.27\ngini = 0.46\nsamples = 10\nvalue = [5, 9]\nclass = 1", fillcolor="#399de571"] ;
+33 -> 34 ;
+35 [label="gini = 0.18\nsamples = 8\nvalue = [1, 9]\nclass = 1", fillcolor="#399de5e3"] ;
+34 -> 35 ;
+36 [label="gini = 0.0\nsamples = 2\nvalue = [4, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+34 -> 36 ;
+37 [label="X <= -1.1\ngini = 0.13\nsamples = 1157\nvalue = [126, 1722]\nclass = 1", fillcolor="#399de5ec"] ;
+33 -> 37 ;
+38 [label="gini = 0.14\nsamples = 956\nvalue = [117, 1408]\nclass = 1", fillcolor="#399de5ea"] ;
+37 -> 38 ;
+39 [label="gini = 0.05\nsamples = 201\nvalue = [9, 314]\nclass = 1", fillcolor="#399de5f8"] ;
+37 -> 39 ;
+40 [label="X <= -0.86\ngini = 0.21\nsamples = 4138\nvalue = [791, 5765]\nclass = 1", fillcolor="#399de5dc"] ;
+2 -> 40 ;
+41 [label="X <= -1.07\ngini = 0.17\nsamples = 1657\nvalue = [244, 2387]\nclass = 1", fillcolor="#399de5e5"] ;
+40 -> 41 ;
+42 [label="gini = 0.0\nsamples = 1\nvalue = [2, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+41 -> 42 ;
+43 [label="X <= -1.04\ngini = 0.17\nsamples = 1656\nvalue = [242, 2387]\nclass = 1", fillcolor="#399de5e5"] ;
+41 -> 43 ;
+44 [label="X <= -1.05\ngini = 0.23\nsamples = 204\nvalue = [43, 289]\nclass = 1", fillcolor="#399de5d9"] ;
+43 -> 44 ;
+45 [label="X <= -1.07\ngini = 0.21\nsamples = 200\nvalue = [40, 287]\nclass = 1", fillcolor="#399de5db"] ;
+44 -> 45 ;
+46 [label="gini = 0.12\nsamples = 48\nvalue = [5, 70]\nclass = 1", fillcolor="#399de5ed"] ;
+45 -> 46 ;
+47 [label="gini = 0.24\nsamples = 152\nvalue = [35, 217]\nclass = 1", fillcolor="#399de5d6"] ;
+45 -> 47 ;
+48 [label="X <= -1.04\ngini = 0.48\nsamples = 4\nvalue = [3, 2]\nclass = 0", fillcolor="#e5813955"] ;
+44 -> 48 ;
+49 [label="gini = 0.0\nsamples = 1\nvalue = [2, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+48 -> 49 ;
+50 [label="gini = 0.44\nsamples = 3\nvalue = [1, 2]\nclass = 1", fillcolor="#399de57f"] ;
+48 -> 50 ;
+51 [label="X <= -1.04\ngini = 0.16\nsamples = 1452\nvalue = [199, 2098]\nclass = 1", fillcolor="#399de5e7"] ;
+43 -> 51 ;
+52 [label="gini = 0.0\nsamples = 43\nvalue = [0, 69]\nclass = 1", fillcolor="#399de5ff"] ;
+51 -> 52 ;
+53 [label="X <= -1.04\ngini = 0.16\nsamples = 1409\nvalue = [199, 2029]\nclass = 1", fillcolor="#399de5e6"] ;
+51 -> 53 ;
+54 [label="gini = 0.0\nsamples = 1\nvalue = [2, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+53 -> 54 ;
+55 [label="gini = 0.16\nsamples = 1408\nvalue = [197, 2029]\nclass = 1", fillcolor="#399de5e6"] ;
+53 -> 55 ;
+56 [label="X <= -0.85\ngini = 0.24\nsamples = 2481\nvalue = [547, 3378]\nclass = 1", fillcolor="#399de5d6"] ;
+40 -> 56 ;
+57 [label="X <= -0.86\ngini = 0.43\nsamples = 52\nvalue = [26, 58]\nclass = 1", fillcolor="#399de58d"] ;
+56 -> 57 ;
+58 [label="X <= -0.86\ngini = 0.39\nsamples = 49\nvalue = [21, 58]\nclass = 1", fillcolor="#399de5a3"] ;
+57 -> 58 ;
+59 [label="X <= -0.86\ngini = 0.47\nsamples = 6\nvalue = [5, 3]\nclass = 0", fillcolor="#e5813966"] ;
+58 -> 59 ;
+60 [label="gini = 0.38\nsamples = 4\nvalue = [1, 3]\nclass = 1", fillcolor="#399de5aa"] ;
+59 -> 60 ;
+61 [label="gini = 0.0\nsamples = 2\nvalue = [4, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+59 -> 61 ;
+62 [label="X <= -0.86\ngini = 0.35\nsamples = 43\nvalue = [16, 55]\nclass = 1", fillcolor="#399de5b5"] ;
+58 -> 62 ;
+63 [label="gini = 0.09\nsamples = 13\nvalue = [1, 21]\nclass = 1", fillcolor="#399de5f3"] ;
+62 -> 63 ;
+64 [label="gini = 0.42\nsamples = 30\nvalue = [15, 34]\nclass = 1", fillcolor="#399de58e"] ;
+62 -> 64 ;
+65 [label="gini = 0.0\nsamples = 3\nvalue = [5, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+57 -> 65 ;
+66 [label="X <= -0.6\ngini = 0.23\nsamples = 2429\nvalue = [521, 3320]\nclass = 1", fillcolor="#399de5d7"] ;
+56 -> 66 ;
+67 [label="X <= -0.6\ngini = 0.24\nsamples = 2263\nvalue = [500, 3084]\nclass = 1", fillcolor="#399de5d6"] ;
+66 -> 67 ;
+68 [label="X <= -0.72\ngini = 0.23\nsamples = 2238\nvalue = [481, 3066]\nclass = 1", fillcolor="#399de5d7"] ;
+67 -> 68 ;
+69 [label="gini = 0.22\nsamples = 1216\nvalue = [238, 1669]\nclass = 1", fillcolor="#399de5db"] ;
+68 -> 69 ;
+70 [label="gini = 0.25\nsamples = 1022\nvalue = [243, 1397]\nclass = 1", fillcolor="#399de5d3"] ;
+68 -> 70 ;
+71 [label="X <= -0.6\ngini = 0.5\nsamples = 25\nvalue = [19, 18]\nclass = 0", fillcolor="#e581390d"] ;
+67 -> 71 ;
+72 [label="gini = 0.48\nsamples = 16\nvalue = [9, 14]\nclass = 1", fillcolor="#399de55b"] ;
+71 -> 72 ;
+73 [label="gini = 0.41\nsamples = 9\nvalue = [10, 4]\nclass = 0", fillcolor="#e5813999"] ;
+71 -> 73 ;
+74 [label="X <= -0.59\ngini = 0.15\nsamples = 166\nvalue = [21, 236]\nclass = 1", fillcolor="#399de5e8"] ;
+66 -> 74 ;
+75 [label="X <= -0.6\ngini = 0.07\nsamples = 58\nvalue = [3, 86]\nclass = 1", fillcolor="#399de5f6"] ;
+74 -> 75 ;
+76 [label="gini = 0.16\nsamples = 22\nvalue = [3, 32]\nclass = 1", fillcolor="#399de5e7"] ;
+75 -> 76 ;
+77 [label="gini = 0.0\nsamples = 36\nvalue = [0, 54]\nclass = 1", fillcolor="#399de5ff"] ;
+75 -> 77 ;
+78 [label="X <= -0.59\ngini = 0.19\nsamples = 108\nvalue = [18, 150]\nclass = 1", fillcolor="#399de5e0"] ;
+74 -> 78 ;
+79 [label="gini = 0.49\nsamples = 4\nvalue = [4, 3]\nclass = 0", fillcolor="#e5813940"] ;
+78 -> 79 ;
+80 [label="gini = 0.16\nsamples = 104\nvalue = [14, 147]\nclass = 1", fillcolor="#399de5e7"] ;
+78 -> 80 ;
+81 [label="X <= -0.31\ngini = 0.39\nsamples = 6041\nvalue = [2505, 7070]\nclass = 1", fillcolor="#399de5a5"] ;
+1 -> 81 ;
+82 [label="X <= -0.41\ngini = 0.32\nsamples = 2550\nvalue = [808, 3212]\nclass = 1", fillcolor="#399de5bf"] ;
+81 -> 82 ;
+83 [label="X <= -0.58\ngini = 0.3\nsamples = 1590\nvalue = [454, 2034]\nclass = 1", fillcolor="#399de5c6"] ;
+82 -> 83 ;
+84 [label="X <= -0.58\ngini = 0.49\nsamples = 9\nvalue = [8, 6]\nclass = 0", fillcolor="#e5813940"] ;
+83 -> 84 ;
+85 [label="gini = 0.0\nsamples = 1\nvalue = [2, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+84 -> 85 ;
+86 [label="X <= -0.58\ngini = 0.5\nsamples = 8\nvalue = [6, 6]\nclass = 0", fillcolor="#e5813900"] ;
+84 -> 86 ;
+87 [label="gini = 0.0\nsamples = 1\nvalue = [0, 2]\nclass = 1", fillcolor="#399de5ff"] ;
+86 -> 87 ;
+88 [label="X <= -0.58\ngini = 0.48\nsamples = 7\nvalue = [6, 4]\nclass = 0", fillcolor="#e5813955"] ;
+86 -> 88 ;
+89 [label="gini = 0.0\nsamples = 1\nvalue = [2, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+88 -> 89 ;
+90 [label="gini = 0.5\nsamples = 6\nvalue = [4, 4]\nclass = 0", fillcolor="#e5813900"] ;
+88 -> 90 ;
+91 [label="X <= -0.41\ngini = 0.3\nsamples = 1581\nvalue = [446, 2028]\nclass = 1", fillcolor="#399de5c7"] ;
+83 -> 91 ;
+92 [label="X <= -0.41\ngini = 0.3\nsamples = 1564\nvalue = [446, 2000]\nclass = 1", fillcolor="#399de5c6"] ;
+91 -> 92 ;
+93 [label="X <= -0.42\ngini = 0.3\nsamples = 1559\nvalue = [440, 1997]\nclass = 1", fillcolor="#399de5c7"] ;
+92 -> 93 ;
+94 [label="gini = 0.3\nsamples = 1529\nvalue = [435, 1944]\nclass = 1", fillcolor="#399de5c6"] ;
+93 -> 94 ;
+95 [label="gini = 0.16\nsamples = 30\nvalue = [5, 53]\nclass = 1", fillcolor="#399de5e7"] ;
+93 -> 95 ;
+96 [label="X <= -0.41\ngini = 0.44\nsamples = 5\nvalue = [6, 3]\nclass = 0", fillcolor="#e581397f"] ;
+92 -> 96 ;
+97 [label="gini = 0.0\nsamples = 2\nvalue = [4, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+96 -> 97 ;
+98 [label="gini = 0.48\nsamples = 3\nvalue = [2, 3]\nclass = 1", fillcolor="#399de555"] ;
+96 -> 98 ;
+99 [label="gini = 0.0\nsamples = 17\nvalue = [0, 28]\nclass = 1", fillcolor="#399de5ff"] ;
+91 -> 99 ;
+100 [label="X <= -0.4\ngini = 0.36\nsamples = 960\nvalue = [354, 1178]\nclass = 1", fillcolor="#399de5b2"] ;
+82 -> 100 ;
+101 [label="X <= -0.4\ngini = 0.5\nsamples = 31\nvalue = [18, 22]\nclass = 1", fillcolor="#399de52e"] ;
+100 -> 101 ;
+102 [label="X <= -0.41\ngini = 0.46\nsamples = 27\nvalue = [12, 21]\nclass = 1", fillcolor="#399de56d"] ;
+101 -> 102 ;
+103 [label="gini = 0.0\nsamples = 3\nvalue = [3, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+102 -> 103 ;
+104 [label="X <= -0.4\ngini = 0.42\nsamples = 24\nvalue = [9, 21]\nclass = 1", fillcolor="#399de592"] ;
+102 -> 104 ;
+105 [label="gini = 0.46\nsamples = 21\nvalue = [9, 16]\nclass = 1", fillcolor="#399de570"] ;
+104 -> 105 ;
+106 [label="gini = 0.0\nsamples = 3\nvalue = [0, 5]\nclass = 1", fillcolor="#399de5ff"] ;
+104 -> 106 ;
+107 [label="X <= -0.4\ngini = 0.24\nsamples = 4\nvalue = [6, 1]\nclass = 0", fillcolor="#e58139d4"] ;
+101 -> 107 ;
+108 [label="gini = 0.0\nsamples = 2\nvalue = [5, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+107 -> 108 ;
+109 [label="X <= -0.4\ngini = 0.5\nsamples = 2\nvalue = [1, 1]\nclass = 0", fillcolor="#e5813900"] ;
+107 -> 109 ;
+110 [label="gini = 0.0\nsamples = 1\nvalue = [0, 1]\nclass = 1", fillcolor="#399de5ff"] ;
+109 -> 110 ;
+111 [label="gini = 0.0\nsamples = 1\nvalue = [1, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+109 -> 111 ;
+112 [label="X <= -0.4\ngini = 0.35\nsamples = 929\nvalue = [336, 1156]\nclass = 1", fillcolor="#399de5b5"] ;
+100 -> 112 ;
+113 [label="X <= -0.4\ngini = 0.09\nsamples = 25\nvalue = [2, 42]\nclass = 1", fillcolor="#399de5f3"] ;
+112 -> 113 ;
+114 [label="X <= -0.4\ngini = 0.38\nsamples = 6\nvalue = [2, 6]\nclass = 1", fillcolor="#399de5aa"] ;
+113 -> 114 ;
+115 [label="gini = 0.0\nsamples = 4\nvalue = [0, 6]\nclass = 1", fillcolor="#399de5ff"] ;
+114 -> 115 ;
+116 [label="gini = 0.0\nsamples = 2\nvalue = [2, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+114 -> 116 ;
+117 [label="gini = 0.0\nsamples = 19\nvalue = [0, 36]\nclass = 1", fillcolor="#399de5ff"] ;
+113 -> 117 ;
+118 [label="X <= -0.35\ngini = 0.35\nsamples = 904\nvalue = [334, 1114]\nclass = 1", fillcolor="#399de5b3"] ;
+112 -> 118 ;
+119 [label="X <= -0.35\ngini = 0.39\nsamples = 512\nvalue = [212, 602]\nclass = 1", fillcolor="#399de5a5"] ;
+118 -> 119 ;
+120 [label="gini = 0.36\nsamples = 485\nvalue = [183, 588]\nclass = 1", fillcolor="#399de5b0"] ;
+119 -> 120 ;
+121 [label="gini = 0.44\nsamples = 27\nvalue = [29, 14]\nclass = 0", fillcolor="#e5813984"] ;
+119 -> 121 ;
+122 [label="X <= -0.33\ngini = 0.31\nsamples = 392\nvalue = [122, 512]\nclass = 1", fillcolor="#399de5c2"] ;
+118 -> 122 ;
+123 [label="gini = 0.24\nsamples = 174\nvalue = [40, 247]\nclass = 1", fillcolor="#399de5d6"] ;
+122 -> 123 ;
+124 [label="gini = 0.36\nsamples = 218\nvalue = [82, 265]\nclass = 1", fillcolor="#399de5b0"] ;
+122 -> 124 ;
+125 [label="X <= -0.31\ngini = 0.42\nsamples = 3491\nvalue = [1697, 3858]\nclass = 1", fillcolor="#399de58f"] ;
+81 -> 125 ;
+126 [label="gini = 0.0\nsamples = 1\nvalue = [4, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+125 -> 126 ;
+127 [label="X <= -0.04\ngini = 0.42\nsamples = 3490\nvalue = [1693, 3858]\nclass = 1", fillcolor="#399de58f"] ;
+125 -> 127 ;
+128 [label="X <= -0.05\ngini = 0.42\nsamples = 2579\nvalue = [1217, 2896]\nclass = 1", fillcolor="#399de594"] ;
+127 -> 128 ;
+129 [label="X <= -0.05\ngini = 0.42\nsamples = 2493\nvalue = [1187, 2780]\nclass = 1", fillcolor="#399de592"] ;
+128 -> 129 ;
+130 [label="X <= -0.06\ngini = 0.42\nsamples = 2491\nvalue = [1183, 2780]\nclass = 1", fillcolor="#399de592"] ;
+129 -> 130 ;
+131 [label="gini = 0.42\nsamples = 2432\nvalue = [1144, 2721]\nclass = 1", fillcolor="#399de594"] ;
+130 -> 131 ;
+132 [label="gini = 0.48\nsamples = 59\nvalue = [39, 59]\nclass = 1", fillcolor="#399de556"] ;
+130 -> 132 ;
+133 [label="gini = 0.0\nsamples = 2\nvalue = [4, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+129 -> 133 ;
+134 [label="X <= -0.04\ngini = 0.33\nsamples = 86\nvalue = [30, 116]\nclass = 1", fillcolor="#399de5bd"] ;
+128 -> 134 ;
+135 [label="X <= -0.04\ngini = 0.34\nsamples = 81\nvalue = [30, 106]\nclass = 1", fillcolor="#399de5b7"] ;
+134 -> 135 ;
+136 [label="gini = 0.32\nsamples = 80\nvalue = [26, 106]\nclass = 1", fillcolor="#399de5c0"] ;
+135 -> 136 ;
+137 [label="gini = 0.0\nsamples = 1\nvalue = [4, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+135 -> 137 ;
+138 [label="gini = 0.0\nsamples = 5\nvalue = [0, 10]\nclass = 1", fillcolor="#399de5ff"] ;
+134 -> 138 ;
+139 [label="X <= 0.04\ngini = 0.44\nsamples = 911\nvalue = [476, 962]\nclass = 1", fillcolor="#399de581"] ;
+127 -> 139 ;
+140 [label="X <= 0.03\ngini = 0.46\nsamples = 726\nvalue = [405, 749]\nclass = 1", fillcolor="#399de575"] ;
+139 -> 140 ;
+141 [label="X <= 0.01\ngini = 0.44\nsamples = 652\nvalue = [342, 698]\nclass = 1", fillcolor="#399de582"] ;
+140 -> 141 ;
+142 [label="gini = 0.46\nsamples = 511\nvalue = [292, 536]\nclass = 1", fillcolor="#399de574"] ;
+141 -> 142 ;
+143 [label="gini = 0.36\nsamples = 141\nvalue = [50, 162]\nclass = 1", fillcolor="#399de5b0"] ;
+141 -> 143 ;
+144 [label="X <= 0.03\ngini = 0.49\nsamples = 74\nvalue = [63, 51]\nclass = 0", fillcolor="#e5813931"] ;
+140 -> 144 ;
+145 [label="gini = 0.44\nsamples = 30\nvalue = [36, 17]\nclass = 0", fillcolor="#e5813987"] ;
+144 -> 145 ;
+146 [label="gini = 0.49\nsamples = 44\nvalue = [27, 34]\nclass = 1", fillcolor="#399de534"] ;
+144 -> 146 ;
+147 [label="X <= 0.05\ngini = 0.38\nsamples = 185\nvalue = [71, 213]\nclass = 1", fillcolor="#399de5aa"] ;
+139 -> 147 ;
+148 [label="X <= 0.05\ngini = 0.32\nsamples = 85\nvalue = [27, 108]\nclass = 1", fillcolor="#399de5bf"] ;
+147 -> 148 ;
+149 [label="gini = 0.39\nsamples = 58\nvalue = [24, 67]\nclass = 1", fillcolor="#399de5a4"] ;
+148 -> 149 ;
+150 [label="gini = 0.13\nsamples = 27\nvalue = [3, 41]\nclass = 1", fillcolor="#399de5ec"] ;
+148 -> 150 ;
+151 [label="X <= 0.05\ngini = 0.42\nsamples = 100\nvalue = [44, 105]\nclass = 1", fillcolor="#399de594"] ;
+147 -> 151 ;
+152 [label="gini = 0.0\nsamples = 3\nvalue = [5, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+151 -> 152 ;
+153 [label="gini = 0.39\nsamples = 97\nvalue = [39, 105]\nclass = 1", fillcolor="#399de5a0"] ;
+151 -> 153 ;
+154 [label="X <= 0.86\ngini = 0.48\nsamples = 7802\nvalue = [7348, 4980]\nclass = 0", fillcolor="#e5813952"] ;
+0 -> 154 [labeldistance=2.5, labelangle=-45, headlabel="False"] ;
+155 [label="X <= 0.39\ngini = 0.5\nsamples = 5371\nvalue = [4259, 4246]\nclass = 0", fillcolor="#e5813901"] ;
+154 -> 155 ;
+156 [label="X <= 0.39\ngini = 0.49\nsamples = 2632\nvalue = [1792, 2365]\nclass = 1", fillcolor="#399de53e"] ;
+155 -> 156 ;
+157 [label="X <= 0.28\ngini = 0.49\nsamples = 2621\nvalue = [1792, 2344]\nclass = 1", fillcolor="#399de53c"] ;
+156 -> 157 ;
+158 [label="X <= 0.26\ngini = 0.49\nsamples = 1790\nvalue = [1171, 1652]\nclass = 1", fillcolor="#399de54a"] ;
+157 -> 158 ;
+159 [label="X <= 0.26\ngini = 0.49\nsamples = 1668\nvalue = [1108, 1516]\nclass = 1", fillcolor="#399de545"] ;
+158 -> 159 ;
+160 [label="X <= 0.16\ngini = 0.49\nsamples = 1665\nvalue = [1102, 1516]\nclass = 1", fillcolor="#399de546"] ;
+159 -> 160 ;
+161 [label="gini = 0.48\nsamples = 878\nvalue = [547, 833]\nclass = 1", fillcolor="#399de558"] ;
+160 -> 161 ;
+162 [label="gini = 0.49\nsamples = 787\nvalue = [555, 683]\nclass = 1", fillcolor="#399de530"] ;
+160 -> 162 ;
+163 [label="gini = 0.0\nsamples = 3\nvalue = [6, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+159 -> 163 ;
+164 [label="X <= 0.27\ngini = 0.43\nsamples = 122\nvalue = [63, 136]\nclass = 1", fillcolor="#399de589"] ;
+158 -> 164 ;
+165 [label="X <= 0.27\ngini = 0.45\nsamples = 112\nvalue = [61, 120]\nclass = 1", fillcolor="#399de57d"] ;
+164 -> 165 ;
+166 [label="gini = 0.36\nsamples = 45\nvalue = [17, 56]\nclass = 1", fillcolor="#399de5b2"] ;
+165 -> 166 ;
+167 [label="gini = 0.48\nsamples = 67\nvalue = [44, 64]\nclass = 1", fillcolor="#399de550"] ;
+165 -> 167 ;
+168 [label="X <= 0.28\ngini = 0.2\nsamples = 10\nvalue = [2, 16]\nclass = 1", fillcolor="#399de5df"] ;
+164 -> 168 ;
+169 [label="gini = 0.0\nsamples = 4\nvalue = [0, 8]\nclass = 1", fillcolor="#399de5ff"] ;
+168 -> 169 ;
+170 [label="gini = 0.32\nsamples = 6\nvalue = [2, 8]\nclass = 1", fillcolor="#399de5bf"] ;
+168 -> 170 ;
+171 [label="X <= 0.32\ngini = 0.5\nsamples = 831\nvalue = [621, 692]\nclass = 1", fillcolor="#399de51a"] ;
+157 -> 171 ;
+172 [label="X <= 0.32\ngini = 0.5\nsamples = 340\nvalue = [293, 247]\nclass = 0", fillcolor="#e5813928"] ;
+171 -> 172 ;
+173 [label="X <= 0.29\ngini = 0.5\nsamples = 321\nvalue = [270, 239]\nclass = 0", fillcolor="#e581391d"] ;
+172 -> 173 ;
+174 [label="gini = 0.47\nsamples = 74\nvalue = [76, 46]\nclass = 0", fillcolor="#e5813965"] ;
+173 -> 174 ;
+175 [label="gini = 0.5\nsamples = 247\nvalue = [194, 193]\nclass = 0", fillcolor="#e5813901"] ;
+173 -> 175 ;
+176 [label="X <= 0.32\ngini = 0.38\nsamples = 19\nvalue = [23, 8]\nclass = 0", fillcolor="#e58139a6"] ;
+172 -> 176 ;
+177 [label="gini = 0.44\nsamples = 16\nvalue = [17, 8]\nclass = 0", fillcolor="#e5813987"] ;
+176 -> 177 ;
+178 [label="gini = 0.0\nsamples = 3\nvalue = [6, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+176 -> 178 ;
+179 [label="X <= 0.32\ngini = 0.49\nsamples = 491\nvalue = [328, 445]\nclass = 1", fillcolor="#399de543"] ;
+171 -> 179 ;
+180 [label="X <= 0.32\ngini = 0.32\nsamples = 30\nvalue = [10, 39]\nclass = 1", fillcolor="#399de5be"] ;
+179 -> 180 ;
+181 [label="gini = 0.44\nsamples = 17\nvalue = [8, 17]\nclass = 1", fillcolor="#399de587"] ;
+180 -> 181 ;
+182 [label="gini = 0.15\nsamples = 13\nvalue = [2, 22]\nclass = 1", fillcolor="#399de5e8"] ;
+180 -> 182 ;
+183 [label="X <= 0.33\ngini = 0.49\nsamples = 461\nvalue = [318, 406]\nclass = 1", fillcolor="#399de537"] ;
+179 -> 183 ;
+184 [label="gini = 0.47\nsamples = 22\nvalue = [23, 14]\nclass = 0", fillcolor="#e5813964"] ;
+183 -> 184 ;
+185 [label="gini = 0.49\nsamples = 439\nvalue = [295, 392]\nclass = 1", fillcolor="#399de53f"] ;
+183 -> 185 ;
+186 [label="gini = 0.0\nsamples = 11\nvalue = [0, 21]\nclass = 1", fillcolor="#399de5ff"] ;
+156 -> 186 ;
+187 [label="X <= 0.67\ngini = 0.49\nsamples = 2739\nvalue = [2467, 1881]\nclass = 0", fillcolor="#e581393d"] ;
+155 -> 187 ;
+188 [label="X <= 0.67\ngini = 0.5\nsamples = 1835\nvalue = [1593, 1352]\nclass = 0", fillcolor="#e5813927"] ;
+187 -> 188 ;
+189 [label="X <= 0.67\ngini = 0.5\nsamples = 1822\nvalue = [1589, 1330]\nclass = 0", fillcolor="#e581392a"] ;
+188 -> 189 ;
+190 [label="X <= 0.39\ngini = 0.5\nsamples = 1818\nvalue = [1581, 1330]\nclass = 0", fillcolor="#e5813928"] ;
+189 -> 190 ;
+191 [label="X <= 0.39\ngini = 0.28\nsamples = 11\nvalue = [15, 3]\nclass = 0", fillcolor="#e58139cc"] ;
+190 -> 191 ;
+192 [label="gini = 0.38\nsamples = 8\nvalue = [9, 3]\nclass = 0", fillcolor="#e58139aa"] ;
+191 -> 192 ;
+193 [label="gini = 0.0\nsamples = 3\nvalue = [6, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+191 -> 193 ;
+194 [label="X <= 0.39\ngini = 0.5\nsamples = 1807\nvalue = [1566, 1327]\nclass = 0", fillcolor="#e5813927"] ;
+190 -> 194 ;
+195 [label="gini = 0.0\nsamples = 4\nvalue = [0, 6]\nclass = 1", fillcolor="#399de5ff"] ;
+194 -> 195 ;
+196 [label="gini = 0.5\nsamples = 1803\nvalue = [1566, 1321]\nclass = 0", fillcolor="#e5813928"] ;
+194 -> 196 ;
+197 [label="gini = 0.0\nsamples = 4\nvalue = [8, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+189 -> 197 ;
+198 [label="X <= 0.67\ngini = 0.26\nsamples = 13\nvalue = [4, 22]\nclass = 1", fillcolor="#399de5d1"] ;
+188 -> 198 ;
+199 [label="X <= 0.67\ngini = 0.17\nsamples = 11\nvalue = [2, 20]\nclass = 1", fillcolor="#399de5e6"] ;
+198 -> 199 ;
+200 [label="X <= 0.67\ngini = 0.44\nsamples = 2\nvalue = [1, 2]\nclass = 1", fillcolor="#399de57f"] ;
+199 -> 200 ;
+201 [label="gini = 0.0\nsamples = 1\nvalue = [0, 2]\nclass = 1", fillcolor="#399de5ff"] ;
+200 -> 201 ;
+202 [label="gini = 0.0\nsamples = 1\nvalue = [1, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+200 -> 202 ;
+203 [label="X <= 0.67\ngini = 0.1\nsamples = 9\nvalue = [1, 18]\nclass = 1", fillcolor="#399de5f1"] ;
+199 -> 203 ;
+204 [label="gini = 0.2\nsamples = 5\nvalue = [1, 8]\nclass = 1", fillcolor="#399de5df"] ;
+203 -> 204 ;
+205 [label="gini = 0.0\nsamples = 4\nvalue = [0, 10]\nclass = 1", fillcolor="#399de5ff"] ;
+203 -> 205 ;
+206 [label="X <= 0.67\ngini = 0.5\nsamples = 2\nvalue = [2, 2]\nclass = 0", fillcolor="#e5813900"] ;
+198 -> 206 ;
+207 [label="gini = 0.0\nsamples = 1\nvalue = [2, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+206 -> 207 ;
+208 [label="gini = 0.0\nsamples = 1\nvalue = [0, 2]\nclass = 1", fillcolor="#399de5ff"] ;
+206 -> 208 ;
+209 [label="X <= 0.85\ngini = 0.47\nsamples = 904\nvalue = [874, 529]\nclass = 0", fillcolor="#e5813965"] ;
+187 -> 209 ;
+210 [label="X <= 0.67\ngini = 0.47\nsamples = 884\nvalue = [863, 508]\nclass = 0", fillcolor="#e5813969"] ;
+209 -> 210 ;
+211 [label="gini = 0.0\nsamples = 8\nvalue = [16, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+210 -> 211 ;
+212 [label="X <= 0.67\ngini = 0.47\nsamples = 876\nvalue = [847, 508]\nclass = 0", fillcolor="#e5813966"] ;
+210 -> 212 ;
+213 [label="gini = 0.0\nsamples = 3\nvalue = [0, 7]\nclass = 1", fillcolor="#399de5ff"] ;
+212 -> 213 ;
+214 [label="X <= 0.68\ngini = 0.47\nsamples = 873\nvalue = [847, 501]\nclass = 0", fillcolor="#e5813968"] ;
+212 -> 214 ;
+215 [label="gini = 0.33\nsamples = 40\nvalue = [52, 14]\nclass = 0", fillcolor="#e58139ba"] ;
+214 -> 215 ;
+216 [label="gini = 0.47\nsamples = 833\nvalue = [795, 487]\nclass = 0", fillcolor="#e5813963"] ;
+214 -> 216 ;
+217 [label="X <= 0.85\ngini = 0.45\nsamples = 20\nvalue = [11, 21]\nclass = 1", fillcolor="#399de579"] ;
+209 -> 217 ;
+218 [label="gini = 0.0\nsamples = 3\nvalue = [0, 7]\nclass = 1", fillcolor="#399de5ff"] ;
+217 -> 218 ;
+219 [label="X <= 0.86\ngini = 0.49\nsamples = 17\nvalue = [11, 14]\nclass = 1", fillcolor="#399de537"] ;
+217 -> 219 ;
+220 [label="X <= 0.86\ngini = 0.5\nsamples = 15\nvalue = [11, 11]\nclass = 0", fillcolor="#e5813900"] ;
+219 -> 220 ;
+221 [label="gini = 0.5\nsamples = 14\nvalue = [9, 11]\nclass = 1", fillcolor="#399de52e"] ;
+220 -> 221 ;
+222 [label="gini = 0.0\nsamples = 1\nvalue = [2, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+220 -> 222 ;
+223 [label="gini = 0.0\nsamples = 2\nvalue = [0, 3]\nclass = 1", fillcolor="#399de5ff"] ;
+219 -> 223 ;
+224 [label="X <= 1.56\ngini = 0.31\nsamples = 2431\nvalue = [3089, 734]\nclass = 0", fillcolor="#e58139c2"] ;
+154 -> 224 ;
+225 [label="X <= 1.08\ngini = 0.35\nsamples = 1849\nvalue = [2240, 663]\nclass = 0", fillcolor="#e58139b4"] ;
+224 -> 225 ;
+226 [label="X <= 1.06\ngini = 0.4\nsamples = 824\nvalue = [942, 359]\nclass = 0", fillcolor="#e581399e"] ;
+225 -> 226 ;
+227 [label="X <= 0.99\ngini = 0.39\nsamples = 756\nvalue = [877, 318]\nclass = 0", fillcolor="#e58139a3"] ;
+226 -> 227 ;
+228 [label="X <= 0.99\ngini = 0.42\nsamples = 510\nvalue = [573, 241]\nclass = 0", fillcolor="#e5813994"] ;
+227 -> 228 ;
+229 [label="X <= 0.86\ngini = 0.41\nsamples = 508\nvalue = [573, 236]\nclass = 0", fillcolor="#e5813996"] ;
+228 -> 229 ;
+230 [label="gini = 0.0\nsamples = 12\nvalue = [18, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+229 -> 230 ;
+231 [label="gini = 0.42\nsamples = 496\nvalue = [555, 236]\nclass = 0", fillcolor="#e5813993"] ;
+229 -> 231 ;
+232 [label="gini = 0.0\nsamples = 2\nvalue = [0, 5]\nclass = 1", fillcolor="#399de5ff"] ;
+228 -> 232 ;
+233 [label="X <= 1.0\ngini = 0.32\nsamples = 246\nvalue = [304, 77]\nclass = 0", fillcolor="#e58139be"] ;
+227 -> 233 ;
+234 [label="X <= 0.99\ngini = 0.19\nsamples = 56\nvalue = [76, 9]\nclass = 0", fillcolor="#e58139e1"] ;
+233 -> 234 ;
+235 [label="gini = 0.31\nsamples = 29\nvalue = [34, 8]\nclass = 0", fillcolor="#e58139c3"] ;
+234 -> 235 ;
+236 [label="gini = 0.05\nsamples = 27\nvalue = [42, 1]\nclass = 0", fillcolor="#e58139f9"] ;
+234 -> 236 ;
+237 [label="X <= 1.05\ngini = 0.35\nsamples = 190\nvalue = [228, 68]\nclass = 0", fillcolor="#e58139b3"] ;
+233 -> 237 ;
+238 [label="gini = 0.38\nsamples = 168\nvalue = [193, 66]\nclass = 0", fillcolor="#e58139a8"] ;
+237 -> 238 ;
+239 [label="gini = 0.1\nsamples = 22\nvalue = [35, 2]\nclass = 0", fillcolor="#e58139f0"] ;
+237 -> 239 ;
+240 [label="X <= 1.06\ngini = 0.47\nsamples = 68\nvalue = [65, 41]\nclass = 0", fillcolor="#e581395e"] ;
+226 -> 240 ;
+241 [label="gini = 0.0\nsamples = 1\nvalue = [0, 2]\nclass = 1", fillcolor="#399de5ff"] ;
+240 -> 241 ;
+242 [label="X <= 1.06\ngini = 0.47\nsamples = 67\nvalue = [65, 39]\nclass = 0", fillcolor="#e5813966"] ;
+240 -> 242 ;
+243 [label="gini = 0.0\nsamples = 1\nvalue = [3, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+242 -> 243 ;
+244 [label="X <= 1.06\ngini = 0.47\nsamples = 66\nvalue = [62, 39]\nclass = 0", fillcolor="#e581395f"] ;
+242 -> 244 ;
+245 [label="gini = 0.47\nsamples = 10\nvalue = [5, 8]\nclass = 1", fillcolor="#399de560"] ;
+244 -> 245 ;
+246 [label="gini = 0.46\nsamples = 56\nvalue = [57, 31]\nclass = 0", fillcolor="#e5813974"] ;
+244 -> 246 ;
+247 [label="X <= 1.51\ngini = 0.31\nsamples = 1025\nvalue = [1298, 304]\nclass = 0", fillcolor="#e58139c3"] ;
+225 -> 247 ;
+248 [label="X <= 1.33\ngini = 0.3\nsamples = 966\nvalue = [1241, 273]\nclass = 0", fillcolor="#e58139c7"] ;
+247 -> 248 ;
+249 [label="X <= 1.32\ngini = 0.32\nsamples = 654\nvalue = [814, 204]\nclass = 0", fillcolor="#e58139bf"] ;
+248 -> 249 ;
+250 [label="X <= 1.09\ngini = 0.31\nsamples = 637\nvalue = [798, 192]\nclass = 0", fillcolor="#e58139c2"] ;
+249 -> 250 ;
+251 [label="gini = 0.13\nsamples = 39\nvalue = [51, 4]\nclass = 0", fillcolor="#e58139eb"] ;
+250 -> 251 ;
+252 [label="gini = 0.32\nsamples = 598\nvalue = [747, 188]\nclass = 0", fillcolor="#e58139bf"] ;
+250 -> 252 ;
+253 [label="X <= 1.32\ngini = 0.49\nsamples = 17\nvalue = [16, 12]\nclass = 0", fillcolor="#e5813940"] ;
+249 -> 253 ;
+254 [label="gini = 0.0\nsamples = 2\nvalue = [0, 5]\nclass = 1", fillcolor="#399de5ff"] ;
+253 -> 254 ;
+255 [label="gini = 0.42\nsamples = 15\nvalue = [16, 7]\nclass = 0", fillcolor="#e581398f"] ;
+253 -> 255 ;
+256 [label="X <= 1.37\ngini = 0.24\nsamples = 312\nvalue = [427, 69]\nclass = 0", fillcolor="#e58139d6"] ;
+248 -> 256 ;
+257 [label="X <= 1.34\ngini = 0.12\nsamples = 80\nvalue = [121, 8]\nclass = 0", fillcolor="#e58139ee"] ;
+256 -> 257 ;
+258 [label="gini = 0.28\nsamples = 17\nvalue = [20, 4]\nclass = 0", fillcolor="#e58139cc"] ;
+257 -> 258 ;
+259 [label="gini = 0.07\nsamples = 63\nvalue = [101, 4]\nclass = 0", fillcolor="#e58139f5"] ;
+257 -> 259 ;
+260 [label="X <= 1.37\ngini = 0.28\nsamples = 232\nvalue = [306, 61]\nclass = 0", fillcolor="#e58139cc"] ;
+256 -> 260 ;
+261 [label="gini = 0.0\nsamples = 1\nvalue = [0, 2]\nclass = 1", fillcolor="#399de5ff"] ;
+260 -> 261 ;
+262 [label="gini = 0.27\nsamples = 231\nvalue = [306, 59]\nclass = 0", fillcolor="#e58139ce"] ;
+260 -> 262 ;
+263 [label="X <= 1.52\ngini = 0.46\nsamples = 59\nvalue = [57, 31]\nclass = 0", fillcolor="#e5813974"] ;
+247 -> 263 ;
+264 [label="gini = 0.0\nsamples = 2\nvalue = [0, 6]\nclass = 1", fillcolor="#399de5ff"] ;
+263 -> 264 ;
+265 [label="X <= 1.56\ngini = 0.42\nsamples = 57\nvalue = [57, 25]\nclass = 0", fillcolor="#e581398f"] ;
+263 -> 265 ;
+266 [label="X <= 1.52\ngini = 0.4\nsamples = 55\nvalue = [57, 22]\nclass = 0", fillcolor="#e581399d"] ;
+265 -> 266 ;
+267 [label="gini = 0.0\nsamples = 5\nvalue = [7, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+266 -> 267 ;
+268 [label="gini = 0.42\nsamples = 50\nvalue = [50, 22]\nclass = 0", fillcolor="#e581398f"] ;
+266 -> 268 ;
+269 [label="gini = 0.0\nsamples = 2\nvalue = [0, 3]\nclass = 1", fillcolor="#399de5ff"] ;
+265 -> 269 ;
+270 [label="X <= 1.83\ngini = 0.14\nsamples = 582\nvalue = [849, 71]\nclass = 0", fillcolor="#e58139ea"] ;
+224 -> 270 ;
+271 [label="X <= 1.83\ngini = 0.2\nsamples = 304\nvalue = [426, 54]\nclass = 0", fillcolor="#e58139df"] ;
+270 -> 271 ;
+272 [label="X <= 1.63\ngini = 0.2\nsamples = 303\nvalue = [426, 53]\nclass = 0", fillcolor="#e58139df"] ;
+271 -> 272 ;
+273 [label="X <= 1.58\ngini = 0.09\nsamples = 92\nvalue = [133, 7]\nclass = 0", fillcolor="#e58139f2"] ;
+272 -> 273 ;
+274 [label="X <= 1.58\ngini = 0.23\nsamples = 22\nvalue = [33, 5]\nclass = 0", fillcolor="#e58139d8"] ;
+273 -> 274 ;
+275 [label="gini = 0.15\nsamples = 21\nvalue = [33, 3]\nclass = 0", fillcolor="#e58139e8"] ;
+274 -> 275 ;
+276 [label="gini = 0.0\nsamples = 1\nvalue = [0, 2]\nclass = 1", fillcolor="#399de5ff"] ;
+274 -> 276 ;
+277 [label="X <= 1.6\ngini = 0.04\nsamples = 70\nvalue = [100, 2]\nclass = 0", fillcolor="#e58139fa"] ;
+273 -> 277 ;
+278 [label="gini = 0.11\nsamples = 25\nvalue = [32, 2]\nclass = 0", fillcolor="#e58139ef"] ;
+277 -> 278 ;
+279 [label="gini = 0.0\nsamples = 45\nvalue = [68, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+277 -> 279 ;
+280 [label="X <= 1.63\ngini = 0.23\nsamples = 211\nvalue = [293, 46]\nclass = 0", fillcolor="#e58139d7"] ;
+272 -> 280 ;
+281 [label="gini = 0.0\nsamples = 1\nvalue = [0, 1]\nclass = 1", fillcolor="#399de5ff"] ;
+280 -> 281 ;
+282 [label="X <= 1.66\ngini = 0.23\nsamples = 210\nvalue = [293, 45]\nclass = 0", fillcolor="#e58139d8"] ;
+280 -> 282 ;
+283 [label="gini = 0.35\nsamples = 33\nvalue = [45, 13]\nclass = 0", fillcolor="#e58139b5"] ;
+282 -> 283 ;
+284 [label="gini = 0.2\nsamples = 177\nvalue = [248, 32]\nclass = 0", fillcolor="#e58139de"] ;
+282 -> 284 ;
+285 [label="gini = 0.0\nsamples = 1\nvalue = [0, 1]\nclass = 1", fillcolor="#399de5ff"] ;
+271 -> 285 ;
+286 [label="X <= 2.08\ngini = 0.07\nsamples = 278\nvalue = [423, 17]\nclass = 0", fillcolor="#e58139f5"] ;
+270 -> 286 ;
+287 [label="X <= 1.95\ngini = 0.03\nsamples = 127\nvalue = [195, 3]\nclass = 0", fillcolor="#e58139fb"] ;
+286 -> 287 ;
+288 [label="gini = 0.0\nsamples = 64\nvalue = [92, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+287 -> 288 ;
+289 [label="X <= 1.95\ngini = 0.06\nsamples = 63\nvalue = [103, 3]\nclass = 0", fillcolor="#e58139f8"] ;
+287 -> 289 ;
+290 [label="gini = 0.0\nsamples = 1\nvalue = [0, 1]\nclass = 1", fillcolor="#399de5ff"] ;
+289 -> 290 ;
+291 [label="X <= 1.96\ngini = 0.04\nsamples = 62\nvalue = [103, 2]\nclass = 0", fillcolor="#e58139fa"] ;
+289 -> 291 ;
+292 [label="gini = 0.28\nsamples = 5\nvalue = [5, 1]\nclass = 0", fillcolor="#e58139cc"] ;
+291 -> 292 ;
+293 [label="gini = 0.02\nsamples = 57\nvalue = [98, 1]\nclass = 0", fillcolor="#e58139fc"] ;
+291 -> 293 ;
+294 [label="X <= 2.09\ngini = 0.11\nsamples = 151\nvalue = [228, 14]\nclass = 0", fillcolor="#e58139ef"] ;
+286 -> 294 ;
+295 [label="gini = 0.0\nsamples = 1\nvalue = [0, 1]\nclass = 1", fillcolor="#399de5ff"] ;
+294 -> 295 ;
+296 [label="X <= 2.31\ngini = 0.1\nsamples = 150\nvalue = [228, 13]\nclass = 0", fillcolor="#e58139f0"] ;
+294 -> 296 ;
+297 [label="X <= 2.3\ngini = 0.19\nsamples = 70\nvalue = [100, 12]\nclass = 0", fillcolor="#e58139e0"] ;
+296 -> 297 ;
+298 [label="gini = 0.07\nsamples = 68\nvalue = [100, 4]\nclass = 0", fillcolor="#e58139f5"] ;
+297 -> 298 ;
+299 [label="gini = 0.0\nsamples = 2\nvalue = [0, 8]\nclass = 1", fillcolor="#399de5ff"] ;
+297 -> 299 ;
+300 [label="X <= 2.55\ngini = 0.02\nsamples = 80\nvalue = [128, 1]\nclass = 0", fillcolor="#e58139fd"] ;
+296 -> 300 ;
+301 [label="gini = 0.03\nsamples = 41\nvalue = [63, 1]\nclass = 0", fillcolor="#e58139fb"] ;
+300 -> 301 ;
+302 [label="gini = 0.0\nsamples = 39\nvalue = [65, 0]\nclass = 0", fillcolor="#e58139ff"] ;
+300 -> 302 ;
+}
\ No newline at end of file
diff --git a/analysis_and_scripts/tree.png b/analysis_and_scripts/tree.png
new file mode 100644
index 0000000000000000000000000000000000000000..ebfc7534aaecc522bbd0dd5f4043c3c9be8cf6dd
Binary files /dev/null and b/analysis_and_scripts/tree.png differ
diff --git a/figures/tulos_kuva_placeholder_en.png b/figures/tulos_kuva_placeholder_en.png
new file mode 100644
index 0000000000000000000000000000000000000000..9ee6778167653a150cc23e0474718b5daeccc876
Binary files /dev/null and b/figures/tulos_kuva_placeholder_en.png differ
diff --git a/figures/valikoitumis_iso.jpg b/figures/valikoitumis_iso.jpg
new file mode 100644
index 0000000000000000000000000000000000000000..6953c10e643b324b236fc11cb11c9104e3438ee7
Binary files /dev/null and b/figures/valikoitumis_iso.jpg differ
diff --git a/figures/valikoitumisharha.png b/figures/valikoitumisharha.png
new file mode 100644
index 0000000000000000000000000000000000000000..c3549722c72ab791cbea7f340a282b4594394c5f
Binary files /dev/null and b/figures/valikoitumisharha.png differ
diff --git a/figures/valikoitumisharha_kaaavio.drawio b/figures/valikoitumisharha_kaaavio.drawio
new file mode 100644
index 0000000000000000000000000000000000000000..540943174f9f93d70c0e852e09b85d37f903e685
--- /dev/null
+++ b/figures/valikoitumisharha_kaaavio.drawio
@@ -0,0 +1 @@
+<mxfile modified="2019-04-21T15:58:55.500Z" host="www.draw.io" agent="Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:66.0) Gecko/20100101 Firefox/66.0" etag="NAxnU8qYrcP3EQhgaOI9" version="10.6.3" type="device"><diagram id="3qIiofentZYx9hMsOPwP" name="Page-1">7VpZc+MoEP41qpp9yJaErvjRcTKT2a05qrJXHlmLsRhLwoOQj/31CxJIQsfasWXF2fJLLBpoaPi+7gZi2LN4+4HCVfiJBCgygBlsDfveAMAyLZP/CMmukLieWwgWFAeyUSV4wv8g1VNKMxygVGvICIkYXunCOUkSNGeaDFJKNnqzbyTSR13BBWoJnuYwakv/xAELC+kt8Cv5I8KLUI1seZOiJoaqsbQkDWFANjWR/WDYM0oIK77i7QxFYvHUuhT93vfUlhOjKGGHdPjds25+QRb+uHx+frynn39zl3/cSC1rGGXSYAN4Edd3941wtXzWbCeXwvuREVVxk+YbNeUNLH+1rSr510L+5lr+VoKvxsw2pk7xl4mfO2+Z8qmjJf5eSBNDWFD0w6rfdyXhhuFKuxmiZImjQo9U3ewqbK+prCtYknSJ1lCsX8e80q6JMCVJs9IoXNNdE/dMuVwT2tGwq3OxBUoMtN0AlGRJgMTeWrx6E2KGnlZwLmo3nIpcFrI4ktVC03sY40iwcEbiVcYQFSDlfKVJqb4OJoUMRBna1kQSXB8QiRGjO95E1toK6JLp1q0sbyreAEfKwhpnPCmDkqqLUnWFZv4hAf0CcIMzgfuxhr1O0OIeBLAsIp3g2p0KrjeKGeABDTO2BVqYsdwOzDjnwoxzxcyFY8ZRGcTFYMbtwExjFVESTEU2wkvzCKYpnusLh7aY/cW/zZ9dWXqu1dwL001V2KlCwidfdPJdVX6uV1b98pLq2LsJKcnoHO1PFxikC8T2e14UaMlVe0trW9a1Y0pGUQQZXuspWdc2yhG+EpxzVUWmiR6ZWkgo7Ja96ilUU5HTUDRpKCoWpqUoR1Vp9vFA814daODCgOZcNtDcY4HW9HHmuEDze6NggNcHBUGPB8E8yDTi4KedjIEM4QQl/Rl4GZ3yEfWYdeokarH0BC3v9CPBCZp6w7aZqzOtn+pnhwEM70kEtKVuOBYenZnuSlJGyRLNSEQolyQkQWImOIoaIhjhRSL8kTjtcfmdiPWYH7GnsiLGQSCG6cwuqvzDlIbKSwLLe6UkFbgHJhxN9g+WcNyejZ4vTnb7FP2KUXQwxfUEdAzmn2zfW3Mh5kgu5JCtvFjP8p+epOl6BjnK9OQLr+ZZJoN4lhZoviQJTlkWF/5gKIIPQMo2I3fHMHKciH4N54OEc1snne13hHPTH5F06gJjaNY9yLhLruwbIBhe2TfIK4Ht72efZY/JPvtM7BM8COEaYsayOt6GznPfMBEto/t9rhx8BZPjR2cQ15aiHLrn4tzsmWs+0c/TPWwvZjrY+l8d1Gs5qNbzQpeDGjUntwZ9XwBH3fuOdO0rDdt77SuPKZdy7asSOAUZdUH00mtf4OrYc0BD0Zmvfa1BHxj8/wPS1NZeCtT8yTBQs4GO2dGh1nXTMO5b1khIcw5Fmn1RSHPdxjHZOxJprms3AupQb1m8WP37XNG8+idE++Ff</diagram></mxfile>
\ No newline at end of file
diff --git a/viitteet.bib b/viitteet.bib
index 31e1d8224066520dd6310eabd42866cbe05b847c..b51b30e246f98fababfce1c120e195349115ca4f 100644
--- a/viitteet.bib
+++ b/viitteet.bib
@@ -140,4 +140,33 @@
year = "2016",
language={finnish},
note = {viitattu 5.4.2019}
-}
\ No newline at end of file
+}
+
+@article{madras18,
+ title={Fairness Through Causal Awareness: Learning Latent-Variable Models for Biased Data},
+ author={Madras, David and Creager, Elliot and Pitassi, Toniann and Zemel, Richard},
+ journal={arXiv preprint arXiv:1809.02519},
+ year={2018},
+ language={finnish}
+}
+
+@booklet{tira,
+ author = "Jyrki Kivinen",
+ title = "Tietorakenteet ja algoritmit",
+ year = "2018",
+ month = "Kevät",
+ note = "Samannimisen kurssin kurssimateriaali",
+ language={finnish}
+}
+
+@article{scikit-learn,
+ title={Scikit-learn: Machine Learning in {P}ython},
+ author={Pedregosa, F. and Varoquaux, G. and Gramfort, A. and Michel, V.
+ and Thirion, B. and Grisel, O. and Blondel, M. and Prettenhofer, P.
+ and Weiss, R. and Dubourg, V. and Vanderplas, J. and Passos, A. and
+ Cournapeau, D. and Brucher, M. and Perrot, M. and Duchesnay, E.},
+ journal={Journal of Machine Learning Research},
+ volume={12},
+ pages={2825--2830},
+ year={2011}
+}
\ No newline at end of file