pydata_nyc_2018_building_fair_AI_models.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Tutorial on Building Fair AI Models\n",
    "\n",
    "## Presenter: Moninder Singh, IBM Research AI\n",
    "\n",
    "## PyData New York, 2018"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"toc\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Table of Contents"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[1. Summary](#summary)<br>\n",
    "[2. Data Used](#data_used)<br>\n",
    "[3. Learning model from MEPS dataset to predict utilization](#2015-data)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;[3.1. Load data & create splits for learning/validating/testing model](#2015-data-load)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;[3.2. Learning model from original data](#original-2015-data)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[3.2.1. Metrics for original data](#original-2015-metrics)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[3.2.2. Learning Logistic Regression (LR) classifier from original data](#lr_orig)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[3.2.2.1. Training LR model from original data](#lr-train)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[3.2.2.2. Validating LR model from original data](#lr-validate)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[3.2.2.3. Testing LR model from original data](#lr-test)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;[3.3 Bias Mitigation using pre-processing technique - Reweighing](#reweighing-2015)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[3.3.1. Transform data](#reweighing-2015-transform)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[3.3.2. Metrics for transformed data](#reweighing-2015-metrics)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[3.3.3. Learning Logistic Regression (LR) classifier from data transformed by reweighing](#lr_transf)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[3.3.3.1. Training LR model after reweighing](#lr-rw-train)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[3.3.3.2. Validating LR model after reweighing](#lr-rw-validate)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[3.3.3.3. Testing LR model after reweighing](#lr-rw-test)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;[3.4. Bias Mitigation using in-processing technique - Prejudice Remover (PR)](#kamishima)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[3.4.1. Training PR model](#ks-train)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[3.4.2. Validating PR model](#ks-validate)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[3.4.3. Testing PR model](#ks-test)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;[3.5. Bias Mitigation using pre-processing technique - Disparate Impact Remover (DIR)](#disparate-impact)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[3.5.1. Training DIR model](#di-train)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[3.5.2. Validating DIR model](#di-validate)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[3.5.3. Testing DIR model](#di-test)<br>\n",
    "&nbsp;&nbsp;&nbsp;&nbsp;[3.6. Summary of MEPS Model Learning Results](#summary-2015-learning)<br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"summary\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Summary"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[Back to TOC](#toc)<br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The notebook demonstrates how a variety of fairness metrics and bias mitigation algorithms can used to detect and reduce bias when learning classifiers.\n",
    "\n",
    "Additional information is available in the accompanying [presentation](https://github.com/monindersingh/pydata2018_fairAI_models_tutorial/blob/master/pydata_nyc_2018_building_fair_AI_models.pdf).\n",
    "\n",
    "An open source Python toolkit, [AI Fairness 360 (AIF360)](https://github.com/IBM/AIF360), is used for this demonstration. This toolkit provides additional [examples](https://github.com/IBM/AIF360/tree/master/examples) of bias detection/mitigation as well as [guidance](http://aif360.mybluemix.net/resources#guidance) figuring out what metrics and algorithms will be most apropriate for a given problem.\n",
    "\n",
    "Bias detection is demonstrated using several metrics, including disparate impact, average odds difference, statistical parity difference, equal opportunity difference, and Theil index.\n",
    "\n",
    "Bias mitigation is explored via a variety of methods, including reweighing (pre-processing algorithm), prejudice remover (in-processing algorithm), and disparate impact remover (pre-processing technique).\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"use_case\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"data_used\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Data used"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[Back to TOC](#toc)<br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Medical Expenditure Panel Survey (MEPS) data\n",
    "\n",
    "The specific data used is the [2015 Full Year Consolidated Data File](https://meps.ahrq.gov/mepsweb/data_stats/download_data_files_detail.jsp?cboPufNumber=HC-181).\n",
    "\n",
    "This file contains data from rounds 3,4,5 of panel 19 (2014) and rounds 1,2,3 of panel 20 (2015). The For this demonstration, a dataset was constructed from panel 19, round 5.\n",
    "\n",
    "The sensitive attribute is 'RACE' constructed as follows: 'Whites' (privileged class) defined by the features RACEV2X = 1 (White) and HISPANX = 2 (non Hispanic); 'Non-Whites' that includes everyone else. \n",
    "\n",
    "To measure utilization, a composite feature, 'UTILIZATION', was created to measure the total number of trips requiring some sort of medical care by summing up the following features: OBTOTV15, the number of office based visits;  OPTOTV15, the number of outpatient visits; ERTOT15, the number of ER visits;  IPNGTD15, the number of inpatient nights, and  + HHTOTD16, the number of home health visits.\n",
    "\n",
    "The model classification task is to predict whether a person would have 'high' utilization (defined as UTILIZATION >= 10, roughly the average utilization for the considered population). Such people can be selected for case management and provided additional resources. High utilization respondents constituted around 17% of each dataset.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load all necessary packages\n",
    "import sys\n",
    "sys.path.append(\"../\")\n",
    "import numpy as np\n",
    "\n",
    "#Datasets\n",
    "from aif360.datasets.meps_dataset_panel19_fy2015 import MEPSDataset19\n",
    "from aif360.datasets.meps_dataset_panel20_fy2015 import MEPSDataset20\n",
    "from aif360.datasets.meps_dataset_panel21_fy2016 import MEPSDataset21\n",
    "\n",
    "#Fairness Metrics\n",
    "from aif360.metrics import BinaryLabelDatasetMetric\n",
    "from aif360.metrics import ClassificationMetric\n",
    "\n",
    "#Scalers\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "from sklearn.metrics import accuracy_score\n",
    "\n",
    "#Classifiers\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "#Bias Mitigation Techniques\n",
    "from aif360.algorithms.preprocessing.reweighing import Reweighing\n",
    "from aif360.algorithms.inprocessing.prejudice_remover import PrejudiceRemover\n",
    "from aif360.algorithms.preprocessing import DisparateImpactRemover\n",
    "\n",
    "\n",
    "from IPython.display import Markdown, display\n",
    "%matplotlib inline\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "#LIME\n",
    "from aif360.datasets.lime_encoder import LimeEncoder\n",
    "import lime\n",
    "import lime.lime_tabular\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"2015-data\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Learning models from MEPS data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[Back to TOC](#toc)<br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"2015-data-load\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.1. Load data & create splits for learning/validating/testing model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[Back to TOC](#toc)<br>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Get the dataset and split into train, validate, and test\n",
    "\n",
    "np.random.seed(1)\n",
    "\n",
    "\n",
    "meps_orig = MEPSDataset19()\n",
    "meps_orig_train, meps_orig_validate, meps_orig_test = \\\n",
    "                                    meps_orig.split([0.5,0.8], shuffle=True)\n",
    "\n",
    "sens_attr = meps_orig_train.protected_attribute_names[0]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Show MEPS dataset details**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/markdown": [
       "#### Training Dataset shape"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(7915, 138)\n"
     ]
    },
    {
     "data": {
      "text/markdown": [
       "#### Validation Dataset shape"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(4749, 138)\n"
     ]
    },
    {
     "data": {
      "text/markdown": [
       "#### Test Dataset shape"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3166, 138)\n"
     ]
    },
    {
     "data": {
      "text/markdown": [
       "#### Favorable and unfavorable labels"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.0 0.0\n",
      ">= 10 Visits < 10 Visits\n"
     ]
    },
    {
     "data": {
      "text/markdown": [
       "#### Protected attribute names"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['RACE']\n"
     ]
    },
    {
     "data": {
      "text/markdown": [
       "#### Privileged and unprivileged protected attribute values"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[array([1.])] [array([0.])]\n",
      "White Non-White\n"
     ]
    },
    {
     "data": {
      "text/markdown": [
       "#### Dataset feature names"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['MCS42', 'AGE', 'PCS42', 'RACE', 'K6SUM42', 'REGION=1', 'REGION=2', 'REGION=3', 'REGION=4', 'SEX=1', 'SEX=2', 'MARRY=1', 'MARRY=2', 'MARRY=3', 'MARRY=4', 'MARRY=5', 'MARRY=6', 'MARRY=7', 'MARRY=8', 'MARRY=9', 'MARRY=10', 'FTSTU=-1', 'FTSTU=1', 'FTSTU=2', 'FTSTU=3', 'ACTDTY=1', 'ACTDTY=2', 'ACTDTY=3', 'ACTDTY=4', 'HONRDC=1', 'HONRDC=2', 'HONRDC=3', 'HONRDC=4', 'RTHLTH=-1', 'RTHLTH=1', 'RTHLTH=2', 'RTHLTH=3', 'RTHLTH=4', 'RTHLTH=5', 'MNHLTH=-1', 'MNHLTH=1', 'MNHLTH=2', 'MNHLTH=3', 'MNHLTH=4', 'MNHLTH=5', 'HIBPDX=-1', 'HIBPDX=1', 'HIBPDX=2', 'CHDDX=-1', 'CHDDX=1', 'CHDDX=2', 'ANGIDX=-1', 'ANGIDX=1', 'ANGIDX=2', 'MIDX=-1', 'MIDX=1', 'MIDX=2', 'OHRTDX=-1', 'OHRTDX=1', 'OHRTDX=2', 'STRKDX=-1', 'STRKDX=1', 'STRKDX=2', 'EMPHDX=-1', 'EMPHDX=1', 'EMPHDX=2', 'CHBRON=-1', 'CHBRON=1', 'CHBRON=2', 'CHOLDX=-1', 'CHOLDX=1', 'CHOLDX=2', 'CANCERDX=-1', 'CANCERDX=1', 'CANCERDX=2', 'DIABDX=-1', 'DIABDX=1', 'DIABDX=2', 'JTPAIN=-1', 'JTPAIN=1', 'JTPAIN=2', 'ARTHDX=-1', 'ARTHDX=1', 'ARTHDX=2', 'ARTHTYPE=-1', 'ARTHTYPE=1', 'ARTHTYPE=2', 'ARTHTYPE=3', 'ASTHDX=1', 'ASTHDX=2', 'ADHDADDX=-1', 'ADHDADDX=1', 'ADHDADDX=2', 'PREGNT=-1', 'PREGNT=1', 'PREGNT=2', 'WLKLIM=-1', 'WLKLIM=1', 'WLKLIM=2', 'ACTLIM=-1', 'ACTLIM=1', 'ACTLIM=2', 'SOCLIM=-1', 'SOCLIM=1', 'SOCLIM=2', 'COGLIM=-1', 'COGLIM=1', 'COGLIM=2', 'DFHEAR42=-1', 'DFHEAR42=1', 'DFHEAR42=2', 'DFSEE42=-1', 'DFSEE42=1', 'DFSEE42=2', 'ADSMOK42=-1', 'ADSMOK42=1', 'ADSMOK42=2', 'PHQ242=-1', 'PHQ242=0', 'PHQ242=1', 'PHQ242=2', 'PHQ242=3', 'PHQ242=4', 'PHQ242=5', 'PHQ242=6', 'EMPST=-1', 'EMPST=1', 'EMPST=2', 'EMPST=3', 'EMPST=4', 'POVCAT=1', 'POVCAT=2', 'POVCAT=3', 'POVCAT=4', 'POVCAT=5', 'INSCOV=1', 'INSCOV=2', 'INSCOV=3']\n"
     ]
    }
   ],
   "source": [
    "# print out some labels, names, etc.\n",
    "display(Markdown(\"#### Training Dataset shape\"))\n",
    "print(meps_orig_train.features.shape)\n",
    "display(Markdown(\"#### Validation Dataset shape\"))\n",
    "print(meps_orig_validate.features.shape)\n",
    "display(Markdown(\"#### Test Dataset shape\"))\n",
    "print(meps_orig_test.features.shape)\n",
    "display(Markdown(\"#### Favorable and unfavorable labels\"))\n",
    "print(meps_orig_train.favorable_label, meps_orig_train.unfavorable_label)\n",
    "print(meps_orig_train.metadata['label_maps'][0][meps_orig_train.favorable_label],\\\n",
    "      meps_orig_train.metadata['label_maps'][0][meps_orig_train.unfavorable_label])\n",
    "display(Markdown(\"#### Protected attribute names\"))\n",
    "print(meps_orig_train.protected_attribute_names)\n",
    "display(Markdown(\"#### Privileged and unprivileged protected attribute values\"))\n",
    "print(meps_orig_train.privileged_protected_attributes, \n",
    "      meps_orig_train.unprivileged_protected_attributes)\n",
    "print(meps_orig_train.metadata['protected_attribute_maps'][0][meps_orig_train.privileged_protected_attributes[0][0]],\n",
    "      meps_orig_train.metadata['protected_attribute_maps'][0][meps_orig_train.unprivileged_protected_attributes[0][0]])\n",
    "\n",
    "\n",
    "display(Markdown(\"#### Dataset feature names\"))\n",
    "print(meps_orig_train.feature_names)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"original-2015-data\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.2. Learning model from original data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[Back to TOC](#toc)<br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"original-2015-metrics\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.2.1. Metrics for original data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/markdown": [
       "#### Original training dataset"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Difference in mean outcomes between unprivileged and privileged groups = -0.136594\n",
      "Disparate impact (ratio of unprivileged favorable mean to privileged favorable mean) = 0.482305\n"
     ]
    }
   ],
   "source": [
    "# Metric for the original dataset\n",
    "sens_idx = meps_orig_train.protected_attribute_names.index(sens_attr)\n",
    "privileged_groups =  [{sens_attr:meps_orig_train.privileged_protected_attributes[sens_idx][0]}]\n",
    "unprivileged_groups = [{sens_attr:meps_orig_train.unprivileged_protected_attributes[sens_idx][0]}]\n",
    "metric_meps_orig_train = BinaryLabelDatasetMetric(meps_orig_train, \n",
    "                                             unprivileged_groups=unprivileged_groups,\n",
    "                                             privileged_groups=privileged_groups)\n",
    "display(Markdown(\"#### Original training dataset\"))\n",
    "print(\"Difference in mean outcomes between unprivileged and privileged groups = %f\" % \\\n",
    "                                                  metric_meps_orig_train.mean_difference())\n",
    "print(\"Disparate impact (ratio of unprivileged favorable mean to privileged favorable mean) = %f\" % \\\n",
    "                                                  metric_meps_orig_train.disparate_impact())\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"lr_orig\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.2.2. Learning Logistic Regression (LR) classifier from original data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"lr-train\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.2.2.1. Training LR model from original data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\envs\\aif360\\lib\\site-packages\\sklearn\\linear_model\\logistic.py:432: FutureWarning: Default solver will be changed to 'lbfgs' in 0.22. Specify a solver to silence this warning.\n",
      "  FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "#Train model on given dataset\n",
    "\n",
    "dataset = meps_orig_train  # data to train on\n",
    "\n",
    "scale = StandardScaler().fit(dataset.features)   # remember the scale\n",
    "\n",
    "model = LogisticRegression()       # model to learn\n",
    "\n",
    "X_train = scale.transform(dataset.features)      #apply the scale\n",
    "y_train = dataset.labels.ravel()\n",
    "\n",
    "\n",
    "model.fit(X_train, y_train,\n",
    "        sample_weight=dataset.instance_weights)\n",
    "y_train_pred = model.predict(X_train)\n",
    "\n",
    "#save model\n",
    "meps_orig_lr = model\n",
    "meps_orig_lr_scale = scale"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"lr-validate\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.2.2.2. Validating LR model from original data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|█████████████████████████████████████████| 50/50 [00:00<00:00, 167.17it/s]\n"
     ]
    }
   ],
   "source": [
    "#Validate model on given dataset and find threshold for best balanced accuracy\n",
    "import numpy as np\n",
    "from tqdm import tqdm\n",
    "thresh_arr = np.linspace(0.01, 0.5, 50)\n",
    "\n",
    "scale = meps_orig_lr_scale\n",
    "\n",
    "model = meps_orig_lr                  #model to validate\n",
    "dataset = meps_orig_validate        #data to validate on\n",
    "\n",
    "X_validate = scale.transform(dataset.features)   #apply the same scale as applied to the training data\n",
    "y_validate = dataset.labels.ravel()\n",
    "y_validate_pred_prob = model.predict_proba(X_validate)\n",
    "\n",
    "\n",
    "bal_acc_arr = []\n",
    "disp_imp_arr = []\n",
    "avg_odds_diff_arr = []\n",
    "stat_par_diff = []\n",
    "eq_opp_diff = []\n",
    "theil_ind = []\n",
    "    \n",
    "for thresh in tqdm(thresh_arr):\n",
    "    y_validate_pred = (y_validate_pred_prob[:,1] > thresh).astype(np.double)\n",
    "\n",
    "    dataset_pred = dataset.copy()\n",
    "    dataset_pred.labels = y_validate_pred\n",
    "\n",
    "    classified_metric = ClassificationMetric(dataset, \n",
    "                                                 dataset_pred,\n",
    "                                                 unprivileged_groups=unprivileged_groups,\n",
    "                                                 privileged_groups=privileged_groups)\n",
    "    metric_pred = BinaryLabelDatasetMetric(dataset_pred,\n",
    "                                                 unprivileged_groups=unprivileged_groups,\n",
    "                                                 privileged_groups=privileged_groups)\n",
    "    \n",
    "    bal_acc = 0.5*(classified_metric.true_positive_rate() + classified_metric.true_negative_rate())\n",
    "    \n",
    "    acc = accuracy_score(y_true=dataset.labels,\n",
    "                            y_pred=dataset_pred.labels)\n",
    "    bal_acc_arr.append(bal_acc)\n",
    "    avg_odds_diff_arr.append(classified_metric.average_odds_difference())\n",
    "    disp_imp_arr.append(metric_pred.disparate_impact())\n",
    "    stat_par_diff.append(classified_metric.statistical_parity_difference())\n",
    "    eq_opp_diff.append(classified_metric.equal_opportunity_difference())\n",
    "    theil_ind.append(classified_metric.theil_index())\n",
    "\n",
    "    \n",
    "thresh_arr_best_ind = np.where(bal_acc_arr == np.max(bal_acc_arr))[0][0]\n",
    "thresh_arr_best = np.array(thresh_arr)[thresh_arr_best_ind]\n",
    "\n",
    "best_bal_acc = bal_acc_arr[thresh_arr_best_ind]\n",
    "disp_imp_at_best_bal_acc = np.abs(1.0-np.array(disp_imp_arr))[thresh_arr_best_ind]\n",
    "\n",
    "avg_odds_diff_at_best_bal_acc = avg_odds_diff_arr[thresh_arr_best_ind]\n",
    "\n",
    "stat_par_diff_at_best_bal_acc = stat_par_diff[thresh_arr_best_ind]\n",
    "eq_opp_diff_at_best_bal_acc = eq_opp_diff[thresh_arr_best_ind]\n",
    "theil_ind_at_best_bal_acc = theil_ind[thresh_arr_best_ind]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Plot balanced accuracy, abs(1-disparate impact)\n",
    "\n",
    "fig, ax1 = plt.subplots(figsize=(10,7))\n",
    "ax1.plot(thresh_arr, bal_acc_arr)\n",
    "ax1.set_xlabel('Classification Thresholds', fontsize=16, fontweight='bold')\n",
    "ax1.set_ylabel('Balanced Accuracy', color='b', fontsize=16, fontweight='bold')\n",
    "ax1.xaxis.set_tick_params(labelsize=14)\n",
    "ax1.yaxis.set_tick_params(labelsize=14)\n",
    "\n",
    "\n",
    "ax2 = ax1.twinx()\n",
    "ax2.plot(thresh_arr, np.abs(1.0-np.array(disp_imp_arr)), color='r')\n",
    "ax2.set_ylabel('abs(1-disparate impact)', color='r', fontsize=16, fontweight='bold')\n",
    "\n",
    "ax2.axvline(np.array(thresh_arr)[thresh_arr_best_ind], \n",
    "            color='k', linestyle=':')\n",
    "ax2.yaxis.set_tick_params(labelsize=14)\n",
    "ax2.grid(True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Plot average odds difference\n",
    "fig, ax1 = plt.subplots(figsize=(10,7))\n",
    "ax1.plot(thresh_arr, bal_acc_arr)\n",
    "ax1.set_xlabel('Classification Thresholds', fontsize=16, fontweight='bold')\n",
    "ax1.set_ylabel('Balanced Accuracy', color='b', fontsize=16, fontweight='bold')\n",
    "ax1.xaxis.set_tick_params(labelsize=14)\n",
    "ax1.yaxis.set_tick_params(labelsize=14)\n",
    "\n",
    "\n",
    "ax2 = ax1.twinx()\n",
    "ax2.plot(thresh_arr, avg_odds_diff_arr, color='r')\n",
    "ax2.set_ylabel('avg. odds diff.', color='r', fontsize=16, fontweight='bold')\n",
    "\n",
    "ax2.axvline(np.array(thresh_arr)[thresh_arr_best_ind], color='k', linestyle=':')\n",
    "ax2.yaxis.set_tick_params(labelsize=14)\n",
    "ax2.grid(True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Threshold corresponding to Best balanced accuracy: 0.1900\n",
      "Best balanced accuracy: 0.7627\n",
      "Corresponding abs(1-disparate impact) value: 0.6066\n",
      "Corresponding average odds difference value: -0.1831\n",
      "Corresponding statistical parity difference value: -0.2643\n",
      "Corresponding equal opportunity difference value: -0.1608\n",
      "Corresponding Theil index value: 0.0936\n"
     ]
    }
   ],
   "source": [
    "lr_thresh_arr_orig_meps_best = thresh_arr_best\n",
    "print(\"Threshold corresponding to Best balanced accuracy: %6.4f\" % lr_thresh_arr_orig_meps_best)\n",
    "lr_best_bal_acc_arr_orig_meps = best_bal_acc\n",
    "print(\"Best balanced accuracy: %6.4f\" % lr_best_bal_acc_arr_orig_meps)\n",
    "lr_disp_imp_at_best_bal_acc_orig_meps = disp_imp_at_best_bal_acc\n",
    "print(\"Corresponding abs(1-disparate impact) value: %6.4f\" % lr_disp_imp_at_best_bal_acc_orig_meps)\n",
    "lr_avg_odds_diff_at_best_bal_acc_orig_meps = avg_odds_diff_at_best_bal_acc\n",
    "print(\"Corresponding average odds difference value: %6.4f\" % lr_avg_odds_diff_at_best_bal_acc_orig_meps)\n",
    "lr_stat_par_diff_at_best_bal_acc_orig_meps = stat_par_diff_at_best_bal_acc\n",
    "print(\"Corresponding statistical parity difference value: %6.4f\" % lr_stat_par_diff_at_best_bal_acc_orig_meps)\n",
    "lr_eq_opp_diff_at_best_bal_acc_orig_meps = eq_opp_diff_at_best_bal_acc\n",
    "print(\"Corresponding equal opportunity difference value: %6.4f\" % lr_eq_opp_diff_at_best_bal_acc_orig_meps)\n",
    "lr_theil_ind_at_best_bal_acc_orig_meps = theil_ind_at_best_bal_acc\n",
    "print(\"Corresponding Theil index value: %6.4f\" % lr_theil_ind_at_best_bal_acc_orig_meps)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"lr-test\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.2.2.3. Testing LR model from original data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Evaluate performance of a given model with a given threshold on a given dataset\n",
    "\n",
    "scale = meps_orig_lr_scale\n",
    "\n",
    "dataset = meps_orig_test   #apply model to this data\n",
    "model = meps_orig_lr               #this is the model applied\n",
    "\n",
    "thresh_arr = lr_thresh_arr_orig_meps_best  \n",
    "\n",
    "\n",
    "X_data = scale.transform(dataset.features)\n",
    "y_data = dataset.labels.ravel()\n",
    "y_data_pred_prob = model.predict_proba(X_data)  \n",
    "\n",
    "    \n",
    "y_pred = (y_data_pred_prob[:,1] > thresh_arr).astype(np.double)\n",
    "\n",
    "dataset_pred = dataset.copy()\n",
    "dataset_pred.labels = y_pred\n",
    "\n",
    "classified_metric = ClassificationMetric(dataset, \n",
    "                                        dataset_pred,\n",
    "                                        unprivileged_groups=unprivileged_groups,\n",
    "                                        privileged_groups=privileged_groups)\n",
    "metric_pred = BinaryLabelDatasetMetric(dataset_pred,\n",
    "                                        unprivileged_groups=unprivileged_groups,\n",
    "                                        privileged_groups=privileged_groups)\n",
    "    \n",
    "TPR = classified_metric.true_positive_rate()\n",
    "TNR = classified_metric.true_negative_rate()\n",
    "bal_acc = 0.5*(TPR+TNR)\n",
    "    \n",
    "acc = accuracy_score(y_true=dataset.labels,\n",
    "                    y_pred=dataset_pred.labels)\n",
    "\n",
    "#get results\n",
    "best_bal_acc = bal_acc\n",
    "disp_imp_at_best_bal_acc = np.abs(1.0-metric_pred.disparate_impact())\n",
    "\n",
    "avg_odds_diff_at_best_bal_acc = classified_metric.average_odds_difference()\n",
    "\n",
    "stat_par_diff_at_best_bal_acc = classified_metric.statistical_parity_difference()\n",
    "eq_opp_diff_at_best_bal_acc = classified_metric.equal_opportunity_difference()\n",
    "theil_ind_at_best_bal_acc = classified_metric.theil_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Threshold corresponding to Best balanced accuracy: 0.1900\n",
      "Best balanced accuracy: 0.7759\n",
      "Corresponding abs(1-disparate impact) value: 0.5738\n",
      "Corresponding average odds difference value: -0.2057\n",
      "Corresponding statistical parity difference value: -0.2612\n",
      "Corresponding equal opportunity difference value: -0.2228\n",
      "Corresponding Theil index value: 0.0921\n"
     ]
    }
   ],
   "source": [
    "lr_thresh_arr_orig_meps_best_test = thresh_arr_best\n",
    "print(\"Threshold corresponding to Best balanced accuracy: %6.4f\" % lr_thresh_arr_orig_meps_best_test)\n",
    "lr_best_bal_acc_arr_orig_meps_best_test = best_bal_acc\n",
    "print(\"Best balanced accuracy: %6.4f\" % lr_best_bal_acc_arr_orig_meps_best_test)\n",
    "lr_disp_imp_at_best_bal_acc_orig_meps_best_test = disp_imp_at_best_bal_acc\n",
    "print(\"Corresponding abs(1-disparate impact) value: %6.4f\" % lr_disp_imp_at_best_bal_acc_orig_meps_best_test)\n",
    "lr_avg_odds_diff_at_best_bal_acc_orig_meps_best_test = avg_odds_diff_at_best_bal_acc\n",
    "print(\"Corresponding average odds difference value: %6.4f\" % lr_avg_odds_diff_at_best_bal_acc_orig_meps_best_test)\n",
    "\n",
    "lr_stat_par_diff_at_best_bal_acc_orig_meps_best_test = stat_par_diff_at_best_bal_acc\n",
    "print(\"Corresponding statistical parity difference value: %6.4f\" % lr_stat_par_diff_at_best_bal_acc_orig_meps_best_test)\n",
    "lr_eq_opp_diff_at_best_bal_acc_orig_meps_best_test = eq_opp_diff_at_best_bal_acc\n",
    "print(\"Corresponding equal opportunity difference value: %6.4f\" % lr_eq_opp_diff_at_best_bal_acc_orig_meps_best_test)\n",
    "lr_theil_ind_at_best_bal_acc_orig_meps_best_test = theil_ind_at_best_bal_acc\n",
    "print(\"Corresponding Theil index value: %6.4f\" % lr_theil_ind_at_best_bal_acc_orig_meps_best_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For all the fairness metrics displayed above, the value should be close to '0' for fairness.\n",
    "\n",
    "abs(1-disparate impact) is typically desired to be < 0.2 for classifier predictions to be fair.\n",
    "\n",
    "However, for a logistic regression classifier trained with original training data, at the best classification rate, this is quite high. This implies unfairness.\n",
    "\n",
    "Similarly, average odds difference = 0.5((FPR_unpriv-FPR_priv)+(TPR_unpriv-TPR_priv)) must be close to zero for the classifier to be fair.\n",
    "\n",
    "For a logistic regression classifier trained with original training data, at the best classification rate, this is still high. This still implies unfairness."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"rf_orig\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"rf-train\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"reweighing-2015\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.3. Bias Mitigation using pre-processing technique - Reweighing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[Back to TOC](#toc)<br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"reweighing-2015-transform\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.3.1. Transform data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "RW = Reweighing(unprivileged_groups=unprivileged_groups,\n",
    "               privileged_groups=privileged_groups)\n",
    "RW.fit(meps_orig_train)\n",
    "meps_trans_train = RW.transform(meps_orig_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"reweighing-2015-metrics\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.3.2. Metrics for transformed data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/markdown": [
       "#### Transformed training dataset"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Difference in mean outcomes between privileged and unprivileged groups = 0.000000\n"
     ]
    }
   ],
   "source": [
    "metric_meps_transf_train = BinaryLabelDatasetMetric(meps_trans_train, \n",
    "                                         unprivileged_groups=unprivileged_groups,\n",
    "                                         privileged_groups=privileged_groups)\n",
    "display(Markdown(\"#### Transformed training dataset\"))\n",
    "print(\"Difference in mean outcomes between privileged and unprivileged groups = %f\" % \\\n",
    "                                                  metric_meps_transf_train.mean_difference())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"lr_transf\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.3.3. Learning Logistic Regression (LR) classifier from data transformed by reweighing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"lr-rw-train\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.3.3.1. Training LR model after reweighing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\envs\\aif360\\lib\\site-packages\\sklearn\\linear_model\\logistic.py:432: FutureWarning: Default solver will be changed to 'lbfgs' in 0.22. Specify a solver to silence this warning.\n",
      "  FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "#Train model on given dataset\n",
    "\n",
    "dataset = meps_trans_train  # data to train on\n",
    "\n",
    "scale = StandardScaler().fit(dataset.features)   # remember the scale\n",
    "\n",
    "model = LogisticRegression()       # model to learn\n",
    "\n",
    "X_train = scale.transform(dataset.features)      #apply the scale\n",
    "\n",
    "y_train = dataset.labels.ravel()\n",
    "\n",
    "\n",
    "model.fit(X_train, y_train,\n",
    "        sample_weight=dataset.instance_weights)\n",
    "y_train_pred = model.predict(X_train)\n",
    "\n",
    "#save model\n",
    "meps_trans_lr = model\n",
    "meps_trans_lr_scale = scale"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"lr-rw-validate\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.3.3.2. Validating  LR model after reweighing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|█████████████████████████████████████████| 50/50 [00:00<00:00, 151.65it/s]\n"
     ]
    }
   ],
   "source": [
    "#validate model on given dataset and find threshold for best balanced accuracy\n",
    "import numpy as np\n",
    "from tqdm import tqdm\n",
    "thresh_arr = np.linspace(0.01, 0.5, 50)\n",
    "\n",
    "scale = meps_trans_lr_scale\n",
    "\n",
    "model = meps_trans_lr                  #model to validate\n",
    "dataset = meps_orig_validate        #data to validate on\n",
    "\n",
    "X_validate = scale.transform(dataset.features)   #apply the same scale as applied to the training data\n",
    "y_validate = dataset.labels.ravel()\n",
    "y_validate_pred_prob = model.predict_proba(X_validate)\n",
    "\n",
    "\n",
    "bal_acc_arr = []\n",
    "disp_imp_arr = []\n",
    "avg_odds_diff_arr = []\n",
    "stat_par_diff = []\n",
    "eq_opp_diff = []\n",
    "theil_ind = []\n",
    "    \n",
    "for thresh in tqdm(thresh_arr):\n",
    "    y_validate_pred = (y_validate_pred_prob[:,1] > thresh).astype(np.double)\n",
    "\n",
    "    dataset_pred = dataset.copy()\n",
    "    dataset_pred.labels = y_validate_pred\n",
    "\n",
    "    classified_metric = ClassificationMetric(dataset, \n",
    "                                                 dataset_pred,\n",
    "                                                 unprivileged_groups=unprivileged_groups,\n",
    "                                                 privileged_groups=privileged_groups)\n",
    "    metric_pred = BinaryLabelDatasetMetric(dataset_pred,\n",
    "                                                 unprivileged_groups=unprivileged_groups,\n",
    "                                                 privileged_groups=privileged_groups)\n",
    "    \n",
    "    bal_acc = 0.5*(classified_metric.true_positive_rate() + classified_metric.true_negative_rate())\n",
    "    \n",
    "    acc = accuracy_score(y_true=dataset.labels,\n",
    "                            y_pred=dataset_pred.labels)\n",
    "    bal_acc_arr.append(bal_acc)\n",
    "    avg_odds_diff_arr.append(classified_metric.average_odds_difference())\n",
    "    disp_imp_arr.append(metric_pred.disparate_impact())\n",
    "    stat_par_diff.append(classified_metric.statistical_parity_difference())\n",
    "    eq_opp_diff.append(classified_metric.equal_opportunity_difference())\n",
    "    theil_ind.append(classified_metric.theil_index())\n",
    "\n",
    "    \n",
    "thresh_arr_best_ind = np.where(bal_acc_arr == np.max(bal_acc_arr))[0][0]\n",
    "thresh_arr_best = np.array(thresh_arr)[thresh_arr_best_ind]\n",
    "\n",
    "best_bal_acc = bal_acc_arr[thresh_arr_best_ind]\n",
    "disp_imp_at_best_bal_acc = np.abs(1.0-np.array(disp_imp_arr))[thresh_arr_best_ind]\n",
    "\n",
    "avg_odds_diff_at_best_bal_acc = avg_odds_diff_arr[thresh_arr_best_ind]\n",
    "\n",
    "stat_par_diff_at_best_bal_acc = stat_par_diff[thresh_arr_best_ind]\n",
    "eq_opp_diff_at_best_bal_acc = eq_opp_diff[thresh_arr_best_ind]\n",
    "theil_ind_at_best_bal_acc = theil_ind[thresh_arr_best_ind]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Plot balanced accuracy, abs(1-disparate impact)\n",
    "\n",
    "fig, ax1 = plt.subplots(figsize=(10,7))\n",
    "ax1.plot(thresh_arr, bal_acc_arr)\n",
    "ax1.set_xlabel('Classification Thresholds', fontsize=16, fontweight='bold')\n",
    "ax1.set_ylabel('Balanced Accuracy', color='b', fontsize=16, fontweight='bold')\n",
    "ax1.xaxis.set_tick_params(labelsize=14)\n",
    "ax1.yaxis.set_tick_params(labelsize=14)\n",
    "\n",
    "\n",
    "ax2 = ax1.twinx()\n",
    "ax2.plot(thresh_arr, np.abs(1.0-np.array(disp_imp_arr)), color='r')\n",
    "ax2.set_ylabel('abs(1-disparate impact)', color='r', fontsize=16, fontweight='bold')\n",
    "\n",
    "ax2.axvline(np.array(thresh_arr)[thresh_arr_best_ind], \n",
    "            color='k', linestyle=':')\n",
    "ax2.yaxis.set_tick_params(labelsize=14)\n",
    "ax2.grid(True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Plot average odds difference\n",
    "fig, ax1 = plt.subplots(figsize=(10,7))\n",
    "ax1.plot(thresh_arr, bal_acc_arr)\n",
    "ax1.set_xlabel('Classification Thresholds', fontsize=16, fontweight='bold')\n",
    "ax1.set_ylabel('Balanced Accuracy', color='b', fontsize=16, fontweight='bold')\n",
    "ax1.xaxis.set_tick_params(labelsize=14)\n",
    "ax1.yaxis.set_tick_params(labelsize=14)\n",
    "\n",
    "\n",
    "ax2 = ax1.twinx()\n",
    "ax2.plot(thresh_arr, avg_odds_diff_arr, color='r')\n",
    "ax2.set_ylabel('avg. odds diff.', color='r', fontsize=16, fontweight='bold')\n",
    "\n",
    "ax2.axvline(np.array(thresh_arr)[thresh_arr_best_ind], color='k', linestyle=':')\n",
    "ax2.yaxis.set_tick_params(labelsize=14)\n",
    "ax2.grid(True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Threshold corresponding to Best balanced accuracy: 0.2200\n",
      "Best balanced accuracy: 0.7581\n",
      "Corresponding abs(1-disparate impact) value: 0.2939\n",
      "Corresponding average odds difference value: -0.0084\n",
      "Corresponding statistical parity difference value: -0.0992\n",
      "Corresponding equal opportunity difference value: 0.0242\n",
      "Corresponding Theil index value: 0.0938\n"
     ]
    }
   ],
   "source": [
    "lr_thresh_arr_transf_meps_best = thresh_arr_best\n",
    "print(\"Threshold corresponding to Best balanced accuracy: %6.4f\" % lr_thresh_arr_transf_meps_best)\n",
    "lr_best_bal_acc_arr_transf_meps = best_bal_acc\n",
    "print(\"Best balanced accuracy: %6.4f\" % lr_best_bal_acc_arr_transf_meps)\n",
    "lr_disp_imp_at_best_bal_acc_transf_meps = disp_imp_at_best_bal_acc\n",
    "print(\"Corresponding abs(1-disparate impact) value: %6.4f\" % lr_disp_imp_at_best_bal_acc_transf_meps)\n",
    "lr_avg_odds_diff_at_best_bal_acc_transf_meps = avg_odds_diff_at_best_bal_acc\n",
    "print(\"Corresponding average odds difference value: %6.4f\" % lr_avg_odds_diff_at_best_bal_acc_transf_meps)\n",
    "\n",
    "lr_stat_par_diff_at_best_bal_acc_transf_meps = stat_par_diff_at_best_bal_acc\n",
    "print(\"Corresponding statistical parity difference value: %6.4f\" % lr_stat_par_diff_at_best_bal_acc_transf_meps)\n",
    "lr_eq_opp_diff_at_best_bal_acc_transf_meps = eq_opp_diff_at_best_bal_acc\n",
    "print(\"Corresponding equal opportunity difference value: %6.4f\" % lr_eq_opp_diff_at_best_bal_acc_transf_meps)\n",
    "lr_theil_ind_at_best_bal_acc_transf_meps = theil_ind_at_best_bal_acc\n",
    "print(\"Corresponding Theil index value: %6.4f\" % lr_theil_ind_at_best_bal_acc_transf_meps)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"lr-rw-test\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.3.3.3. Testing  LR model after reweighing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Evaluate performance of a given model with a given threshold on a given dataset\n",
    "\n",
    "scale = meps_trans_lr_scale\n",
    "\n",
    "dataset = meps_orig_test   #apply model to this data\n",
    "model = meps_trans_lr               #this is the model applied\n",
    "                                   #lr_transf_panel19 is LR model learned from Panel 19\n",
    "                                   #transformed data\n",
    "thresh_arr = lr_thresh_arr_transf_meps_best  # lr_thresh_arr_transf_panel19_best wass threshold for LR\n",
    "                                            # model with highest balanced accuracy\n",
    "\n",
    "\n",
    "X_data = scale.transform(dataset.features)\n",
    "y_data = dataset.labels.ravel()\n",
    "y_data_pred_prob = model.predict_proba(X_data)  \n",
    "\n",
    "    \n",
    "y_pred = (y_data_pred_prob[:,1] > thresh_arr).astype(np.double)\n",
    "\n",
    "dataset_pred = dataset.copy()\n",
    "dataset_pred.labels = y_pred\n",
    "\n",
    "classified_metric = ClassificationMetric(dataset, \n",
    "                                        dataset_pred,\n",
    "                                        unprivileged_groups=unprivileged_groups,\n",
    "                                        privileged_groups=privileged_groups)\n",
    "metric_pred = BinaryLabelDatasetMetric(dataset_pred,\n",
    "                                        unprivileged_groups=unprivileged_groups,\n",
    "                                        privileged_groups=privileged_groups)\n",
    "    \n",
    "TPR = classified_metric.true_positive_rate()\n",
    "TNR = classified_metric.true_negative_rate()\n",
    "bal_acc = 0.5*(TPR+TNR)\n",
    "    \n",
    "acc = accuracy_score(y_true=dataset.labels,\n",
    "                    y_pred=dataset_pred.labels)\n",
    "\n",
    "#get results\n",
    "best_bal_acc = bal_acc\n",
    "disp_imp_at_best_bal_acc = np.abs(1.0-metric_pred.disparate_impact())\n",
    "\n",
    "avg_odds_diff_at_best_bal_acc = classified_metric.average_odds_difference()\n",
    "\n",
    "stat_par_diff_at_best_bal_acc = classified_metric.statistical_parity_difference()\n",
    "eq_opp_diff_at_best_bal_acc = classified_metric.equal_opportunity_difference()\n",
    "theil_ind_at_best_bal_acc = classified_metric.theil_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Threshold corresponding to Best balanced accuracy: 0.2200\n",
      "Best balanced accuracy: 0.7539\n",
      "Corresponding abs(1-disparate impact) value: 0.2482\n",
      "Corresponding average odds difference value: -0.0151\n",
      "Corresponding statistical parity difference value: -0.0872\n",
      "Corresponding equal opportunity difference value: -0.0035\n",
      "Corresponding Theil index value: 0.0966\n"
     ]
    }
   ],
   "source": [
    "lr_thresh_arr_transf_meps_best_test = thresh_arr_best\n",
    "print(\"Threshold corresponding to Best balanced accuracy: %6.4f\" % lr_thresh_arr_transf_meps_best_test)\n",
    "lr_best_bal_acc_arr_transf_meps_best_test = best_bal_acc\n",
    "print(\"Best balanced accuracy: %6.4f\" % lr_best_bal_acc_arr_transf_meps_best_test)\n",
    "lr_disp_imp_at_best_bal_acc_transf_meps_best_test = disp_imp_at_best_bal_acc\n",
    "print(\"Corresponding abs(1-disparate impact) value: %6.4f\" % lr_disp_imp_at_best_bal_acc_transf_meps_best_test)\n",
    "lr_avg_odds_diff_at_best_bal_acc_transf_meps_best_test = avg_odds_diff_at_best_bal_acc\n",
    "print(\"Corresponding average odds difference value: %6.4f\" % lr_avg_odds_diff_at_best_bal_acc_transf_meps_best_test)\n",
    "\n",
    "lr_stat_par_diff_at_best_bal_acc_transf_meps_best_test = stat_par_diff_at_best_bal_acc\n",
    "print(\"Corresponding statistical parity difference value: %6.4f\" % lr_stat_par_diff_at_best_bal_acc_transf_meps_best_test)\n",
    "lr_eq_opp_diff_at_best_bal_acc_transf_meps_best_test = eq_opp_diff_at_best_bal_acc\n",
    "print(\"Corresponding equal opportunity difference value: %6.4f\" % lr_eq_opp_diff_at_best_bal_acc_transf_meps_best_test)\n",
    "lr_theil_ind_at_best_bal_acc_transf_meps_best_test = theil_ind_at_best_bal_acc\n",
    "print(\"Corresponding Theil index value: %6.4f\" % lr_theil_ind_at_best_bal_acc_transf_meps_best_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The fairness metrics for the logistic regression model learnt after reweighing are fairly improved, and thus the model is much more fair relative to the logistic regression model learnt from the original data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"rf_transf\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"kamishima\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### 3.4. Bias Mitigation using in-processing technique - Prejudice Remover (PR)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[Back to TOC](#toc)<br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"ks-train\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.4.1. Training PR model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = PrejudiceRemover(sensitive_attr=sens_attr, eta = 25.0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Train model on given dataset\n",
    "\n",
    "dataset = meps_orig_train  # data to train on\n",
    "\n",
    "tr_dataset = dataset.copy(deepcopy=True)\n",
    "scale = StandardScaler().fit(tr_dataset.features)   # remember the scale\n",
    "\n",
    "tr_dataset.features = scale.transform(tr_dataset.features)\n",
    "        \n",
    "\n",
    "\n",
    "model.fit(tr_dataset)\n",
    "\n",
    "#save model\n",
    "meps_orig_pr = model\n",
    "meps_orig_pr_scale = scale"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"ks-validate\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.4.2. Validating PR model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|█████████████████████████████████████████| 50/50 [00:00<00:00, 196.31it/s]\n"
     ]
    }
   ],
   "source": [
    "#validate model on given dataset and find threshold for best balanced accuracy\n",
    "import numpy as np\n",
    "from tqdm import tqdm\n",
    "thresh_arr = np.linspace(0.01, 0.5, 50)\n",
    "\n",
    "dataset = meps_orig_validate        #data to validate on\n",
    "\n",
    "model = meps_orig_pr\n",
    "scale = meps_orig_pr_scale\n",
    "\n",
    "te_dataset= dataset.copy(deepcopy=True)\n",
    "te_dataset.features = scale.transform(te_dataset.features)\n",
    "pred_dataset = model.predict(te_dataset)\n",
    "\n",
    "y_validate_pred_prob = pred_dataset.scores\n",
    "\n",
    "\n",
    "bal_acc_arr = []\n",
    "disp_imp_arr = []\n",
    "avg_odds_diff_arr = []\n",
    "stat_par_diff = []\n",
    "eq_opp_diff = []\n",
    "theil_ind = []\n",
    "    \n",
    "for thresh in tqdm(thresh_arr):\n",
    "    y_validate_pred = (y_validate_pred_prob[:,1] > thresh).astype(np.double)\n",
    "\n",
    "    dataset_pred = dataset.copy()\n",
    "    dataset_pred.labels = y_validate_pred\n",
    "\n",
    "    classified_metric = ClassificationMetric(dataset, \n",
    "                                                 dataset_pred,\n",
    "                                                 unprivileged_groups=unprivileged_groups,\n",
    "                                                 privileged_groups=privileged_groups)\n",
    "    metric_pred = BinaryLabelDatasetMetric(dataset_pred,\n",
    "                                                 unprivileged_groups=unprivileged_groups,\n",
    "                                                 privileged_groups=privileged_groups)\n",
    "    \n",
    "    bal_acc = 0.5*(classified_metric.true_positive_rate() + classified_metric.true_negative_rate())\n",
    "    \n",
    "    acc = accuracy_score(y_true=dataset.labels,\n",
    "                            y_pred=dataset_pred.labels)\n",
    "    bal_acc_arr.append(bal_acc)\n",
    "    avg_odds_diff_arr.append(classified_metric.average_odds_difference())\n",
    "    disp_imp_arr.append(metric_pred.disparate_impact())\n",
    "    stat_par_diff.append(classified_metric.statistical_parity_difference())\n",
    "    eq_opp_diff.append(classified_metric.equal_opportunity_difference())\n",
    "    theil_ind.append(classified_metric.theil_index())\n",
    "\n",
    "    \n",
    "thresh_arr_best_ind = np.where(bal_acc_arr == np.max(bal_acc_arr))[0][0]\n",
    "thresh_arr_best = np.array(thresh_arr)[thresh_arr_best_ind]\n",
    "\n",
    "best_bal_acc = bal_acc_arr[thresh_arr_best_ind]\n",
    "disp_imp_at_best_bal_acc = np.abs(1.0-np.array(disp_imp_arr))[thresh_arr_best_ind]\n",
    "\n",
    "avg_odds_diff_at_best_bal_acc = avg_odds_diff_arr[thresh_arr_best_ind]\n",
    "\n",
    "stat_par_diff_at_best_bal_acc = stat_par_diff[thresh_arr_best_ind]\n",
    "eq_opp_diff_at_best_bal_acc = eq_opp_diff[thresh_arr_best_ind]\n",
    "theil_ind_at_best_bal_acc = theil_ind[thresh_arr_best_ind]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Plot balanced accuracy, abs(1-disparate impact)\n",
    "\n",
    "fig, ax1 = plt.subplots(figsize=(10,7))\n",
    "ax1.plot(thresh_arr, bal_acc_arr)\n",
    "ax1.set_xlabel('Classification Thresholds', fontsize=16, fontweight='bold')\n",
    "ax1.set_ylabel('Balanced Accuracy', color='b', fontsize=16, fontweight='bold')\n",
    "ax1.xaxis.set_tick_params(labelsize=14)\n",
    "ax1.yaxis.set_tick_params(labelsize=14)\n",
    "\n",
    "\n",
    "ax2 = ax1.twinx()\n",
    "ax2.plot(thresh_arr, np.abs(1.0-np.array(disp_imp_arr)), color='r')\n",
    "ax2.set_ylabel('abs(1-disparate impact)', color='r', fontsize=16, fontweight='bold')\n",
    "\n",
    "ax2.axvline(np.array(thresh_arr)[thresh_arr_best_ind], \n",
    "            color='k', linestyle=':')\n",
    "ax2.yaxis.set_tick_params(labelsize=14)\n",
    "ax2.grid(True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Plot average odds difference\n",
    "fig, ax1 = plt.subplots(figsize=(10,7))\n",
    "ax1.plot(thresh_arr, bal_acc_arr)\n",
    "ax1.set_xlabel('Classification Thresholds', fontsize=16, fontweight='bold')\n",
    "ax1.set_ylabel('Balanced Accuracy', color='b', fontsize=16, fontweight='bold')\n",
    "ax1.xaxis.set_tick_params(labelsize=14)\n",
    "ax1.yaxis.set_tick_params(labelsize=14)\n",
    "\n",
    "\n",
    "ax2 = ax1.twinx()\n",
    "ax2.plot(thresh_arr, avg_odds_diff_arr, color='r')\n",
    "ax2.set_ylabel('avg. odds diff.', color='r', fontsize=16, fontweight='bold')\n",
    "\n",
    "ax2.axvline(np.array(thresh_arr)[thresh_arr_best_ind], color='k', linestyle=':')\n",
    "ax2.yaxis.set_tick_params(labelsize=14)\n",
    "ax2.grid(True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Threshold corresponding to Best balanced accuracy: 0.1200\n",
      "Best balanced accuracy: 0.6836\n",
      "Corresponding abs(1-disparate impact) value: 0.2268\n",
      "Corresponding average odds difference value: 0.0254\n",
      "Corresponding statistical parity difference value: -0.0830\n",
      "Corresponding equal opportunity difference value: 0.1172\n",
      "Corresponding Theil index value: 0.1119\n"
     ]
    }
   ],
   "source": [
    "pr_thresh_arr_orig_meps_best = thresh_arr_best\n",
    "print(\"Threshold corresponding to Best balanced accuracy: %6.4f\" % pr_thresh_arr_orig_meps_best)\n",
    "pr_best_bal_acc_arr_orig_meps = best_bal_acc\n",
    "print(\"Best balanced accuracy: %6.4f\" % pr_best_bal_acc_arr_orig_meps)\n",
    "pr_disp_imp_at_best_bal_acc_orig_meps = disp_imp_at_best_bal_acc\n",
    "print(\"Corresponding abs(1-disparate impact) value: %6.4f\" % pr_disp_imp_at_best_bal_acc_orig_meps)\n",
    "pr_avg_odds_diff_at_best_bal_acc_orig_meps = avg_odds_diff_at_best_bal_acc\n",
    "print(\"Corresponding average odds difference value: %6.4f\" % pr_avg_odds_diff_at_best_bal_acc_orig_meps)\n",
    "\n",
    "pr_stat_par_diff_at_best_bal_acc_orig_meps = stat_par_diff_at_best_bal_acc\n",
    "print(\"Corresponding statistical parity difference value: %6.4f\" % pr_stat_par_diff_at_best_bal_acc_orig_meps)\n",
    "pr_eq_opp_diff_at_best_bal_acc_orig_meps = eq_opp_diff_at_best_bal_acc\n",
    "print(\"Corresponding equal opportunity difference value: %6.4f\" % pr_eq_opp_diff_at_best_bal_acc_orig_meps)\n",
    "pr_theil_ind_at_best_bal_acc_orig_meps = theil_ind_at_best_bal_acc\n",
    "print(\"Corresponding Theil index value: %6.4f\" % pr_theil_ind_at_best_bal_acc_orig_meps)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"ks-test\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.4.3. Testing PR model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Test model on given dataset and find threshold for best balanced accuracy\n",
    "import numpy as np\n",
    "from tqdm import tqdm\n",
    "thresh_arr = np.linspace(0.01, 0.75, 75)\n",
    "\n",
    "dataset = meps_orig_test        #data to test on\n",
    "scale = meps_orig_pr_scale\n",
    "model = meps_orig_pr\n",
    "\n",
    "thresh_arr = pr_thresh_arr_orig_meps_best\n",
    "\n",
    "te_dataset= dataset.copy(deepcopy=True)\n",
    "te_dataset.features = scale.transform(te_dataset.features)\n",
    "pred_dataset = model.predict(te_dataset)\n",
    "\n",
    "y_data_pred_prob = pred_dataset.scores\n",
    "\n",
    "\n",
    "y_data_pred = (y_data_pred_prob[:,1] > thresh_arr).astype(np.double)\n",
    "\n",
    "dataset_pred = dataset.copy()\n",
    "dataset_pred.labels = y_data_pred\n",
    "\n",
    "classified_metric = ClassificationMetric(dataset, \n",
    "                                            dataset_pred,\n",
    "                                            unprivileged_groups=unprivileged_groups,\n",
    "                                            privileged_groups=privileged_groups)\n",
    "metric_pred = BinaryLabelDatasetMetric(dataset_pred,\n",
    "                                            unprivileged_groups=unprivileged_groups,\n",
    "                                            privileged_groups=privileged_groups)\n",
    "    \n",
    "bal_acc = 0.5*(classified_metric.true_positive_rate() + classified_metric.true_negative_rate())\n",
    "    \n",
    "acc = accuracy_score(y_true=dataset.labels,\n",
    "                    y_pred=dataset_pred.labels)\n",
    "\n",
    "best_bal_acc = bal_acc\n",
    "disp_imp_at_best_bal_acc = np.abs(1.0-metric_pred.disparate_impact())\n",
    "\n",
    "avg_odds_diff_at_best_bal_acc = classified_metric.average_odds_difference()\n",
    "\n",
    "stat_par_diff_at_best_bal_acc = classified_metric.statistical_parity_difference()\n",
    "eq_opp_diff_at_best_bal_acc = classified_metric.equal_opportunity_difference()\n",
    "theil_ind_at_best_bal_acc = classified_metric.theil_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Threshold corresponding to Best balanced accuracy: 0.1200\n",
      "Best balanced accuracy: 0.6880\n",
      "Corresponding abs(1-disparate impact) value: 0.1588\n",
      "Corresponding average odds difference value: 0.0523\n",
      "Corresponding statistical parity difference value: -0.0566\n",
      "Corresponding equal opportunity difference value: 0.1479\n",
      "Corresponding Theil index value: 0.1108\n"
     ]
    }
   ],
   "source": [
    "pr_thresh_arr_orig_meps_best_test = thresh_arr_best\n",
    "print(\"Threshold corresponding to Best balanced accuracy: %6.4f\" % pr_thresh_arr_orig_meps_best_test)\n",
    "pr_best_bal_acc_arr_orig_meps_best_test = best_bal_acc\n",
    "print(\"Best balanced accuracy: %6.4f\" % pr_best_bal_acc_arr_orig_meps_best_test)\n",
    "pr_disp_imp_at_best_bal_acc_orig_meps_best_test = disp_imp_at_best_bal_acc\n",
    "print(\"Corresponding abs(1-disparate impact) value: %6.4f\" % pr_disp_imp_at_best_bal_acc_orig_meps_best_test)\n",
    "pr_avg_odds_diff_at_best_bal_acc_orig_meps_best_test = avg_odds_diff_at_best_bal_acc\n",
    "print(\"Corresponding average odds difference value: %6.4f\" % pr_avg_odds_diff_at_best_bal_acc_orig_meps_best_test)\n",
    "\n",
    "pr_stat_par_diff_at_best_bal_acc_orig_meps_best_test = stat_par_diff_at_best_bal_acc\n",
    "print(\"Corresponding statistical parity difference value: %6.4f\" % pr_stat_par_diff_at_best_bal_acc_orig_meps_best_test)\n",
    "pr_eq_opp_diff_at_best_bal_acc_orig_meps_best_test = eq_opp_diff_at_best_bal_acc\n",
    "print(\"Corresponding equal opportunity difference value: %6.4f\" % pr_eq_opp_diff_at_best_bal_acc_orig_meps_best_test)\n",
    "pr_theil_ind_at_best_bal_acc_orig_meps_best_test = theil_ind_at_best_bal_acc\n",
    "print(\"Corresponding Theil index value: %6.4f\" % pr_theil_ind_at_best_bal_acc_orig_meps_best_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As in the case of re-weighing, prejudice remover has resulted in a fair model. However, it hs come at an expense of relative lower balanced accuracy."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"disparate-impact\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.5. Bias mitigation using pre-processing technique - Disparate Impact (DIR) Remover"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[Back to TOC](#toc)<br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"di-train\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.5.1. Training DIR model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\envs\\aif360\\lib\\site-packages\\BlackBoxAuditing\\audit_reading.py:7: UserWarning: matplotlib.pyplot as already been imported, this call will have no effect.\n",
      "  matplotlib.use('Agg') # Set the back-end\n",
      "C:\\ProgramData\\Anaconda3\\envs\\aif360\\lib\\site-packages\\BlackBoxAuditing\\consistency_graph.py:5: UserWarning: matplotlib.pyplot as already been imported, this call will have no effect.\n",
      "  matplotlib.use('Agg') # Set the back-end\n",
      "C:\\ProgramData\\Anaconda3\\envs\\aif360\\lib\\site-packages\\sklearn\\linear_model\\logistic.py:432: FutureWarning: Default solver will be changed to 'lbfgs' in 0.22. Specify a solver to silence this warning.\n",
      "  FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "#Train model on given dataset\n",
    "\n",
    "dataset = meps_orig_train  # data to train on\n",
    "\n",
    "tr_dataset = dataset.copy(deepcopy=True)\n",
    "scale = MinMaxScaler().fit(tr_dataset.features)   # remember the scale\n",
    "\n",
    "tr_dataset.features = scale.transform(tr_dataset.features)\n",
    "index = tr_dataset.feature_names.index(sens_attr)        \n",
    "\n",
    "di = DisparateImpactRemover(repair_level=1.0)\n",
    "train_repd = di.fit_transform(tr_dataset) #repair training dataset\n",
    "\n",
    "X_tr = np.delete(train_repd.features, index, axis=1)\n",
    "    \n",
    "\n",
    "y_tr = train_repd.labels.ravel()\n",
    "    \n",
    "model = LogisticRegression(random_state = 1, class_weight='balanced')\n",
    "model.fit(X_tr, y_tr)\n",
    "\n",
    "meps_orig_dir_scale = scale\n",
    "meps_orig_dir = model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"di-validate\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.5.2. Validating DIR model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|█████████████████████████████████████████| 75/75 [00:00<00:00, 187.41it/s]\n"
     ]
    }
   ],
   "source": [
    "#validate model on given dataset and find threshold for best balanced accuracy\n",
    "import numpy as np\n",
    "from tqdm import tqdm\n",
    "\n",
    "scale = meps_orig_dir_scale\n",
    "model = meps_orig_dir\n",
    "\n",
    "thresh_arr = np.linspace(0.01, 0.75, 75)\n",
    "\n",
    "dataset = meps_orig_validate        #data to validate on\n",
    "\n",
    "te_dataset= dataset.copy(deepcopy=True)\n",
    "te_dataset.features = scale.transform(te_dataset.features)\n",
    "\n",
    "validate_repd = di.fit_transform(te_dataset)  #repair validate dataset\n",
    "\n",
    "X_te = np.delete(validate_repd.features, index, axis=1)\n",
    "    \n",
    "\n",
    "y_te_pred_prob = model.predict_proba(X_te)\n",
    "\n",
    "\n",
    "bal_acc_arr = []\n",
    "disp_imp_arr = []\n",
    "avg_odds_diff_arr = []\n",
    "stat_par_diff = []\n",
    "eq_opp_diff = []\n",
    "theil_ind = []\n",
    "    \n",
    "for thresh in tqdm(thresh_arr):\n",
    "    y_te_pred = (y_te_pred_prob[:,1] > thresh).astype(np.double)\n",
    "\n",
    "    validate_repd_pred = validate_repd.copy()\n",
    "    validate_repd_pred.labels = y_te_pred\n",
    "\n",
    "    classified_metric = ClassificationMetric(validate_repd, \n",
    "                                                 validate_repd_pred,\n",
    "                                                 unprivileged_groups=unprivileged_groups,\n",
    "                                                 privileged_groups=privileged_groups)\n",
    "    metric_pred = BinaryLabelDatasetMetric(validate_repd_pred,\n",
    "                                                 unprivileged_groups=unprivileged_groups,\n",
    "                                                 privileged_groups=privileged_groups)\n",
    "    \n",
    "    bal_acc = 0.5*(classified_metric.true_positive_rate() + classified_metric.true_negative_rate())\n",
    "    \n",
    "    acc = accuracy_score(y_true=validate_repd.labels,\n",
    "                            y_pred=validate_repd_pred.labels)\n",
    "    bal_acc_arr.append(bal_acc)\n",
    "    avg_odds_diff_arr.append(classified_metric.average_odds_difference())\n",
    "    disp_imp_arr.append(metric_pred.disparate_impact())\n",
    "    stat_par_diff.append(classified_metric.statistical_parity_difference())\n",
    "    eq_opp_diff.append(classified_metric.equal_opportunity_difference())\n",
    "    theil_ind.append(classified_metric.theil_index())\n",
    "    \n",
    "thresh_arr_best_ind = np.where(bal_acc_arr == np.max(bal_acc_arr))[0][0]\n",
    "thresh_arr_best = np.array(thresh_arr)[thresh_arr_best_ind]\n",
    "\n",
    "best_bal_acc = bal_acc_arr[thresh_arr_best_ind]\n",
    "disp_imp_at_best_bal_acc = np.abs(1.0-np.array(disp_imp_arr))[thresh_arr_best_ind]\n",
    "\n",
    "avg_odds_diff_at_best_bal_acc = avg_odds_diff_arr[thresh_arr_best_ind]\n",
    "\n",
    "stat_par_diff_at_best_bal_acc = stat_par_diff[thresh_arr_best_ind]\n",
    "eq_opp_diff_at_best_bal_acc = eq_opp_diff[thresh_arr_best_ind]\n",
    "theil_ind_at_best_bal_acc = theil_ind[thresh_arr_best_ind]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Plot balanced accuracy, abs(1-disparate impact)\n",
    "\n",
    "fig, ax1 = plt.subplots(figsize=(10,7))\n",
    "ax1.plot(thresh_arr, bal_acc_arr)\n",
    "ax1.set_xlabel('Classification Thresholds', fontsize=16, fontweight='bold')\n",
    "ax1.set_ylabel('Balanced Accuracy', color='b', fontsize=16, fontweight='bold')\n",
    "ax1.xaxis.set_tick_params(labelsize=14)\n",
    "ax1.yaxis.set_tick_params(labelsize=14)\n",
    "\n",
    "\n",
    "ax2 = ax1.twinx()\n",
    "ax2.plot(thresh_arr, np.abs(1.0-np.array(disp_imp_arr)), color='r')\n",
    "ax2.set_ylabel('abs(1-disparate impact)', color='r', fontsize=16, fontweight='bold')\n",
    "\n",
    "ax2.axvline(np.array(thresh_arr)[thresh_arr_best_ind], \n",
    "            color='k', linestyle=':')\n",
    "ax2.yaxis.set_tick_params(labelsize=14)\n",
    "ax2.grid(True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAroAAAG2CAYAAACQ1yX/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvqOYd8AAAIABJREFUeJzs3Xd4VNXWx/HvSiWBEEroRapA6AakSXuFK4IUQQS9oIgCiuViL1e99ooKFhSxIIKKDRQVRFSKgDSld0IJJZQEAklInf3+cSYaQyAzkMmZsj7PkzvJafOLXsliZ+29xRiDUkoppZRS/ibI7gBKKaWUUkp5gha6SimllFLKL2mhq5RSSiml/JIWukoppZRSyi9poauUUkoppfySFrpKKaWUUsovaaGrlFJKKaX8kha6SimllFLKL2mhq5RSSiml/FKI3QHsFhQUZCIiIuyOoZRSSilVpPT0dGOM0YFKFwV8oRsREUFaWprdMZRSSnmxxMREAKpWrWpzEhXoROS03Rl8ScAXukoppVRRhg4dCsDChQvtDaKUcosWukoppVQRHnroIbsjKKXOgxhj7M5gq9KlSxttXVBKKaWULxCRdGNMabtz+AptZlZKKaWKkJCQQEJCgt0xlFJu0tYFpZRSqgjDhw8HtEdXKV+jha5SSilVhEcffdTuCEqp86A9utqjq5RSSikfoT267tEeXaWUUqoI8fHxxMfH2x1DKeUmHdHVEV2llFJF6NatG6A9usp+OqLrHu3RVUoppYrw5JNP2h1BKXUedERXR3SVUkop5SN0RNc92qOrlFJKFWHbtm1s27bN7hhKKTdpoauUUkoVYcyYMYwZM8buGEq5TETGishuEckQkTUi0rmI67s6r8sQkXgRufVCn+kNtEdXKaWUKsJzzz1ndwSlXCYiQ4CJwFjgN+frXBGJNcbsK+T6usAPwAfAMOAyYJKIHDXGfHU+z/QW2qNbEj26y5dD27YQon+vUEopX5SSnk2Ow0HFMuF2R1EBzpUeXRFZAaw3xozKd2wH8KUx5uFCrn8RGGiMaZjv2HtAU2NMh/N5prfQysvTdu6ELl1g0CCYPl2LXaWU8iGZOblMWRzP+M8WkJXroFqdi2lYpQwXV4miYZUoLq5choZVoigfGYqI2B1XKUQkDIgDxhc4NR/oeJbbOjjP5/cjcKOIhAJyHs/0CgFfdVWoUMHj6yLWuuUW6r/zDoePHmXrI49ggoM9+n5KKaUuXGpmDgdPZGBycolY9jalRbjnqhfIyEkmM+0oJ3YZVu6ClVhVQFCQECRCsPM1SCA4SBCEvBpYnP8jWOcqlgknJEgLZOWWEBFZne/rd40x7+b7OgYIBg4XuO8w0OMsz6wKLCjk+hDn8+Q8nukVAr7QTU5O/mshcI/p1g3q1qXKgw9SpVo1+Ogj0GJXKaU8xhjD0dRMdh1JY9fRVHYdTSX+qPW5w2G4rGEM3RpVplODGKIjQv9xb2JKBk9/v5nv1x+iTsXSPNm/GdM+fB+Atm3b/vX8QykZbD98ip1HUjmRnk1qZo71kZHz9+eZOeQ6zD8/jPWacjqb6Igcnu7fjD4tqpX4PyPls3KMMW1cuK5gb6oUcqyo6/OOyzmu8eoe2IAvdEvMAw+AwwEPPwxBQfDhh1rsKqVUMcvMyeWNn3cybfkeTmbk/HU8IjSY+pVLc0nt8uQ4HMzdmMjnq/cTHCRcUrsc3RpVpuvFlVi+K4kJC7aT4zDc0/NiRnepR6nQYKDSP95HRKheLoLq5SLo1qjyeWXdcfgU936xjts/+YMfNlbj6f7NqFA67AK+e6UAOAbkYo3S5leZM0dk8ySe5focIAmroHX3mV5BJ6OV9IYRzz0H//0v3HADfPCBFrtKKVVM1u8/wf1frGfb4VP0bl6VS+tUoH7lMtSvVIaqZUsRlK9FICfXwZ8JJ1i07SgLtx9h44GTf527vHFl/te3KbUrRv51bO3atQC0atWqWDPn5DqYvDieCQu2Ex0RyjMDmtOrWcFaQqm/uTEZbZ0xZnS+Y9uBr84xGW2AMaZRvmPvAs0LTEZz+ZneQgtdO3ZGe+YZeOwxGDEC3n/fGuFVSil1XjJzcnn95x28syiemDJhvDCwBd0buzfKeuRUBkt3HiOmTDidG1Y643xei5un5nRsTTzJvZ+vY9PBkwxoVZ0n+jWlXKSO7qozuVjoDgE+xloCbClwK3Az1ioKe0VkGoAx5gbn9XWBjcAUYDLQCZgEXFdgebGzPrO4v8/iooWuXVsAP/kkPPEE3HQTvPeeFrtKKXUeNuxP4b4v1rHt8CmuiavJY1fFntFzWxw8NaKbX3aug0m/7uKNX3YQGRZMV2c7RZeGMVQuW8pj76t8i6tbAIvIWOABoBpWEXu3MWax89xCAGNMt3zXdwVeA5oCB4EXjTHvuPpMb6WFrl2FLsDjj8PTT8Mnn8B119mTQSmlfNCBE6f5dMU+3l60i5gyYTw/sDn/17iK3bGKxaaDKby/ZDeLdxzjWGomAE2qlbWK3otjaHNRBcJCdHAkULla6CqLFrp2FroOBzRoAHXrws8/25NBKaU8JCk1k51HUtnpXPGgVvkI+reqQXk3J1wZY9h5JJWVe5JZtTuZVXuOc+DEaQBrFLdPLNGRxT+Km9+qVauAv1ddKAkOh2FL4kkWbT/Kom1HWbP3ODkOQ41yEbw97BJa1CxXYlmU99BC1z1a6NpZ6AI8+yw8+qi1sUT9+vblUEqpC7R05zG+W3/QKm6PpHI8Pfuvc+EhQWTmOAgLDqJnbBUGt6lJ54aVCC5kDdlch2HLoZP8Hp/Eyt3JrNqT/NezKkWFc2mdCrStU54O9WNoVDWqRL43T/fouuJURjZLdx7j6e+2cDQ1k2cGNOPaNrVsy6PsoYWue7TQtbvQPXAAate2lh175hn7ciil1AX4fHUCD321njLhITSqGkUD52oHDZw7h1UrW4qtiaf4Yk0Cs/88wPH0bKqWLcU1cTW5+pIanMrIYUV8Eiuche0p59JgtStEcmndCtZHnQpcVDHSlh3INm7cCECzZs1K/L0LSk7L4s5P/2DpziT+3a42/+vbVFsZAogWuu7RQtfuQhfgqqvgzz9h717dIlgp5XPeWxLPM99voXPDGCYPjyMy7Nx/jmXm5PLzliN8vjqBxduP4sj3Y6heTGna1atAu7oVaVevAtWiIzyc3jfl5DoYP3877yzaReva5Xj733FUjdYJa4FAC133aKHrDYXurFkwcCB89x306WNvFqWUcpExhld/2s4bv+ykd/OqvDakFeEh7q0NnpiSwdyNh6hYJpz2dSt47eoCy5YtA6Bjx442J/mnHzYc4r4v1hEZFsJb17emXb2KdkdSHqaFrnu00PWGQjc7G2rWhI4draJXKaW8nMNheGLOJqYt38uQNrV4bmDzQvtt/YU39OiezY7Dpxjz8Rr2Jafz2FWx3NDhIlvaO1TJ0ELXPVroekOhC9YWwa++Cvv3Q1XdFUcp5b2ycx3c/8U6Zq89yOgu9Xj4ysZ+X1ht27YNgEaNGhVxpT1OZmRzz8x1LNhymGHtrb7d0GDt2/VHWui6Rwtdbyl0t22Dxo3hxRetolcppWyUkZ1LjsOQW+AjO9fBk3M2sWDLEe6/ohFju9X3+yLXVzgchhd/3MrkRfF0bhjDm9df4pHNM5S9tNB1jxa63lLoAnTuDEeOwNatoD84lFIlLCk1kznrDjJr7UHWJZw463Ui8HT/Zgxrf1EJprPXokWLAOjatavNSYr2+aoEHpm1gToxpfngxrbUrhhpdyRVjLTQdY8Wut5U6H70EYwYAYsXW0WvUkp5WEZ2Lgu2HGbWHwdYtP0oOQ5Dk2pl6RlbhTLhwQQHBRESJAQFCSFBQnCQ0KByGS6pXd7u6CXKm3t0C7N8VxK3zViDAJOHt+HSuhXsjqSKiRa67tFC15sK3bQ0qFYNrr7aKnqVUspDdh5J5b0l8Xy//hCnMnOoUjacAa1qcPUlNWhctazd8bxOfHw8APXq1bM5iet2H0vj5qmrSDiezgsDWzAorqbdkVQx0ELXPVroelOhC3DrrTBtGhw6BNHRdqdRSvmZ/cfTmbhgB1/9sZ/wkGCubF6Vga1r0qF+Rb9eNSFQpaRnc9uMNSzblUTbOuVpUDmK+pVKU79SGepVKk3N8pH6793HaKHrHi10va3QXbUKLr0U3n7bKnqVUqoYHDmVwVu/7OSTlfsQEYa3v4jbutUnpky43dF8woIFCwDo0aOHzUncl53r4I2fd7BsVxK7jv5za+aw4CDqxETyr9iqXN+uNtXL6QYd3k4LXfdooettha4x0LIlhIdbRa9SShUhIzuXjOxcgoOE0OCgv3ppRYQT6VlMXhzPh0t3k51ruLZNLe66vIHuOOYmX+vRPZfktCzij6YSfzSNXUdT2XzoJEt3HgOgZ2wVhrevQ6cGFXU1DS+lha57bCl0RWQscD9QDdgEjDPGLDnLtVOBGws59de/aBHpBvxayDVNjDFbz5XF6wpdgNdfh//8x9oWuFUru9MopbxUamYO7y7axZQluzmdnXvG+ZAgwQAOY+jXsjp397iYOjH68/F8JCQkAFCrVi2bk3hGQnI6n6zcx8xVCSSnZVGvUmmGt7+IQXE1KVvKtSXKcnId7Dyayvr9KeTkGga3qalr+XqAFrruKfFCV0SGANOBscBvztebgFhjzL5Cro8GCg49LAUWG2Nucl7TDavQbQok57vuqDHmzD/98/HKQjcpCapXh9Gj4Y037E6jlPIy2bkOPluVwMQF2zmWmkWfFtVoc1F5cnINOQ5DTq7DenU4ALiqRXWaVNMJZqpoGdm5/LDhENOW72VtwgnCQ4KoG1Oa6uUiqF6uFNXLRVCjXATVy0VQJjyErYknWZeQwoYDKWw6mEJGtuOvZ7WoGc3Eoa2pq3+5KlZa6LrHjkJ3BbDeGDMq37EdwJfGmIdduL8TVoHcyRizzHmsG1ahW8kYc8ydPF5Z6AJcdx3MmwcHD0KE/opRKQXGGOZvPsyL87YSfzSNS+tU4JE+TWhVq5zd0fzevHnzAOjVq5fNSUrOhv0pfLP2AHuS0jl44jQHU05zIl9/b56I0GCa1ShL8xrlaFEzmuY1o9mWeIqHv95Adq6DJ/o1ZXBcTW2FKCZa6LqnRAtdEQkD0oHrjDFf5Dv+FtDMGFPkStzOVoY2xphm+Y51wyp09wLhwGbgGWNMYe0M/+C1he7ChdC9O7zzDowZY3capZTN/th3nOd/2MKqPcepX6k0D13ZhB5NKmvxUEL8qUf3QqRl5nAo5TQHTmSQcjqbRlWiaFC5TKErNxw8cZp7Pl/L7/HJ9Glejeeubk50pO7UdqG00HVPSRe61YEDQFdjzOJ8xx8H/m2MOecm4s42hoPAI8aYifmONwK6A6uAMGA4cCvQLf/75Lt+NDAaICwsLC4zM/NCv7XiZwy0a2e1MWzbBiEhdidSStlg44EUXv1pO79sPUJMmXDu7tmQIW1qEaK9jyUqMTERgKpVq9qcxLfkOgyTF+/i1fnbqRwVzmtDWtGuXkW7Y/k0LXTdY1eh2yX/5DMR+R/WKG/jIu6/HXgFqG6MSS7i2h+AHGNMv3Nd57UjugCzZsHAgfDppzB0qN1plFIlaGviSV77aTs/bjpMdEQoo7vUY0THOpQO17/0Kt+zLuEE//nsT/Ymp3Nvz4u54/8a2h3JZ2mh6x6fal0QkbXAJmPMv114r/8BQ40xTc51nVcXug4HxMZCqVLWCgz6K0ql/N7OI6d4bcEOvl9/iKjwEG7uXJeRl9V1eea78ow5c+YA0LdvX5uT+K60zBwembWBb9Ye5OVrWjC4jX+uYOFpWui6p0SHBowxWSKyBugJfJHvVE/gq3PdKyLtgJbAOBffrhVw6Hxyeo2gIHjwQRg5En78EQJoEoRSgcYYw4vztvHu4l2UCg3m9u71GdW5HuUiw+yOpoBXXnkF0EL3QpQOD+GVwS05eiqT/87eyMVVomipEymVh9m1vNjHWMuKLcXqpb0ZaGqM2Ssi0wCMMTcUuO89oAvQyBQILSLjgD1Ya/KGAcOAh4BBxpivz5XHq0d0AbKyoF49aNDAmqCmlPJLkxbu5KV52xgcV5OHrmxMRd2xzKscO2Yt6BMTE2NzEt+XnJZF3zd+w2EMc+68THfnc5OO6LqnxGczGGNmYo3KPgqsBS4Dehtj9jovqe38+IuIRAFDgfcKFrlOYcB4YD2wxPnMPkUVuT4hLAzuvRcWLYLly+1Oo5TygK//2M9L87bRt2V1XhzUQotcLxQTE6NFbjGpUDqMycPjSE7LYuyMP8jOdRR9k1LnSbcA9vYRXYDUVLjoIrjsMvjmG7vTKKWK0W87jjHiw5W0rVOBqSPbEh4SbHckVYivv7bGTQYOHGhzEv8x+88DjJu5lhEd6/BEv6Z2x/EZOqLrHl2fxheUKQN33gnffgubNtmdRilVTDYfPMmt09dQv1IZ3hkep0WuF3v99dd5/fXX7Y7hVwa0rsHNl9Vl6rI9fLlmv91xlJ/SEV1fGNEFOHbMGtW95hr46CO70yilLtCBE6cZOGkpgvD12I5UL6c7IHqzlJQUAKKjo21O4l9ych3c8MFKVu89zpe3dqBFTZ2cVhQd0XWPjuj6ipgYGDUKPvkE9u4t+nqllNdKSc9mxAcrSc/MZerItlrk+oDo6Ggtcj0gJDiIN65rTaUy4Yz5eA3HUr1wAyfl07TQ9SX33mu9vvqqvTmUUuctIzuXUR+vZm9SOpNviKNx1bJ2R1IumDlzJjNnzrQ7hl+qWCb8r8lp/d9cyntL4kk5nW13LOUntHXBV1oX8tx0E8ycCfv2WaO8SimfcfhkBnd9+icrdifz+nWt6deyut2RlIu6desGwEJd5tFjlu9K4rWftrNyTzKRYcEMjqvJjR3rUK9SGbujeRVtXXCPFrq+Vuhu2WLtlvbYY/DUU3anUUq56NetR7j3i3WczsrluYHNuLp1TbsjKTekp6cDEBkZaXMS/7fxQAofLt3DnHUHycp10L1RJW7qVJe4i8oTHCSEBgcRHBS4O4VqoeseLXR9rdAFuPpq+OUX+O476NzZ7jRKqXPIynHw8o9bmbJkN02qleXN61tTX0eolCrS0VOZzFixl+m/7zujd1cEQoOsgjciLJh7el7MsPYX2ZS0ZGmh6x4tdH2x0N2zB664AnbvhnffhREj7E6klCrE3qQ07vz0T9bvT+GGDhfxSO8mlArVJcR80fTp0wEYNmyYzUkCT2ZOLvM3HeZQymmycw05uYZch4NshyEn18H6/Sms2J3MS4NacG3bWnbH9TgtdN2jha4vFroAx4/DtdfCggVw//3w/PMQrD9AlSoppzKy+X79IYKChLKlQogqFUrZUqFElQohqlQIS3cl8cjXGwgSeOmalvRqVtXuyOoCaI+u98rMyeWWj1bz285jTBzq/73vWui6RwtdXy10AbKzYdw4mDQJ+vaFGTMgKsruVEr5vdV7krn787UkJJ8+53VxF5Vn4tBW1CyvfZ2+LjvbWgUgNDTU5iSqMKezcrnxw5X8sfc4bw+Lo2dsFbsjeYwWuu7RQteXC908b70F//mPNUltzhxrYwmlVLHLznXw+s87eOvXndQsH8lL17SgRrkITmXkcDIjm1MZOZxyvkaEBXN16xqEBusqjkqVhNTMHP793gq2HDzJeze2ocvFleyO5BFa6LpHC11/KHQBfvoJBg+G8HCYPRs6dLA7kVJ+Jf5oKnfPXMu6/SkMjqvJ//o1pUx4iN2xVAmZOnUqACN0ToRXS0nPZuiU39l9LJWPbrqUdvUq2h2p2Gmh6x4tdP2l0AXYuhWuugrS0qx1dvVXbEpdMGMMn65M4OnvNhMWEsTzA5vTu3k1u2OpEqY9ur7jWGomQyYvJzElgxmj2tOqln9tK6yFrnu00PWnQhfg22+hf3+rheGqq+xOo5RPO3wyg//O2siCLYe5rEEM4we3pGp0KbtjKaWKkJiSwbWTl3MiPYvZt3fyq00ntNB1jxa6/lboZmdDjRrQtSt88YXdaZTySbkOw/Tf9/Lyj9vIznVw/xWNGNmpLkEBvEi9Ur4mITmdPq8voXG1snw2qr3f/Perha57dJaEvwkNheuvt0Z2k5PtTqOUz9l0MIWBk5byv2830bp2OX4c14VbOtfzmx+S6vxMmTKFKVOm2B1DuaFWhUge7RPLyt3JfLYqwe44yiZa6PqjG2+ErCz47DO7kyjlM9Iyc3jmu830e3MpB06cZuLQVkwbeSl1YnTgRMHMmTOZOXOm3TGUmwa3qUnH+hV5/octJKZk2B1H2UBbF/ytdQHAGGjVCkqVghUr7E6jlFdISs1k97E0chyGXIdxvjrIyTUkp2Xx+s87OJiSwfXtavPgFY2JjtTJnEr5gz3H0rhiwmK6XFyJd4fHIeLbv53R1gX36No4/kjEGtW9915rJYbGje1OpJSt1iac4Ib3V3AyI+es1zSqEsVX17cm7qIKJZhMKeVpdWJKc0/Pi3l+7lbmbUzkSl01JaDoiK4/jugCHD5sTUrL2x5YqQC1ek8yIz5cRYXSYTzRL5aI0BBCgoXgICFYrNfQ4CDqVypNiG7uoM5i0qRJAIwdO9bmJOp85OQ66P/WUo6cymTB3V19+jc2OqLrHv1T3V9VqQK9esHHH0Nurt1plLLF8l1J3PDBSipHhfP5mA78X+MqdKhfkbZ1KnBJ7fK0rFWOZjWiaVQ1SotcdU5z5sxhzpw5dsdQ5ykkOIgXB7UgOS2L5+dusTuOVxGRcBF5Q0SOiUiaiHwrIjVduG+siOwWkQwRWSMinfOdq+B85lYROS0iCSLytoiU+A4e+ie7P7vxRjhwAH7+2e4kSpW4JTuOctPUldQoF8FnY9rr+rfqgsydO5e5c+faHUNdgGY1ormlc10+W5XAsl3H7I7jTSYAg4DrgM5AWeA7EQk+2w0iMgSYCDwHtAaWAXNFpLbzkupADeABoDkwDOgCfOqh7+GstHXBX1sXADIyoFo16N0bZsywO41SJebXrUcYM30N9SuVYfrNl1KxTLjdkZRSXuB0Vi69Ji5GgHnjulAq9Ky1nNcqztYFEYkGjgI3GWNmOI/VAvYCVxpjfjzLfSuA9caYUfmO7QC+NMY8fJZ7egPfAeWMMSeLI78rdETXn5UqBUOHwqxZcLLE/j+llK1+3JTI6I9X06hKFJ+OaqdFrioWEydOZOLEiXbHUBcoIiyY569uzp6kdCYs2GF3HG8QB4QC8/MOGGMSgC1Ax8JuEJEw533zC5yaf7Z7nMoCmUD6BeR1W8CvulChQgW/3rs8qnlz4k6fZuvTT5PYp4/dcZTyqJOns9mXfJqHWgVTJyabtSuX2R1J+YnPP/8cgJYtW9qcRBWHZ9sFcTx5O7PnJlAuwucmpoWIyOp8X79rjHn3PJ9VFcgFCvZyHHaeK0wMEOy8puA9PQq7QUTKAU8DU4wxZ1/+xgO0dcGfWxfAWlO3SROoXBkWL7Y7jVIeY+1otoxmNaL5aOSllAkP+L/HK6XO4lRGNrd8tJoVu5O5/4pGjO1W32fW13WldUFEngH+W8SjumP10k4DQk2+glBEfgW2GWNuLeTZ1YEDQBdjzJJ8x/8HXGeMaVzg+tLAPKyCupcxpkR37tDWBX+Xt6bukiWwa5fdaZTyiJT0bG6b/gflI8N4Z1icFrlKqXOKKhXKtJsvpV/L6rz84zYenb2RnFyH3bGK0wSgSREfK4FErNHZmAL3V+bMEds8x7CK1oIjvmfcIyJlgLxZnFeVdJELWugGhuHDrYJ32jS7kyhV7BwOw7iZf3Io5TSThl1CpSjtyVXFb/z48YwfP97uGKoYhYcEM2FIK27rVp8ZK/Yx5uM1pGeV6G/VPcYYc8wYs7WIj3RgDZAN9My717m0WBOslRQKe3aW876eBU71zH+PiERhjeQGA72NManF+T26SgvdQFCzJlx+uVXoOvzqb6xKMfHnHfy67SiP923KJbXL2x1H+anly5ezfPlyu2OoYhYUJDzYqzFPD2jGr9uOcN27v3P0VKbdsUqMMSYFeB94WUR6iEhr4GNgPbAg7zrnerh35Lv1VWCEiNwiIk1EZCJWG8Q7zuujsCanlQdGAKVFpKrzI6wkvre/smuPrp/36OaZPt0a2V24ELp2tTuNUsXil62HGTl1NYMuqcn4wS18psdOKeV9Fmw+zJ2f/klMVBgf3XQp9SqVsTtSoYp7ZzQRKQW8DFwPRAA/A2Odqy/kXWOAJ40xT+Q7NhZrndxqwEbgbmPMYue5bsCvZ3nL7saYhcWVvyha6AZKoZuWBlWrwqBBMHWq3WmUumB7k9Lo+8Zv1CwfyddjO/rkephKKe+yLuEEI6euIivHwa3d6nNTpzpEhnlXz79uAewebV0IFKVLw7Bh8NlncPhs/eVK+YbTWbncOv0PRIR3hsVpkas87oUXXuCFF16wO4bysJa1yjH79k60r1+Rl3/cRteXFzL9971k+9dEtYCihW4gGTcOMjNh0iS7kyh13owxPDJrA1sTTzJhaCtqV4y0O5IKAGvXrmXt2rV2x1AloFaFSKbc0IYvb+1AnYqRPDp7I/96bTHfrT9IoP8W3Bdp60KgtC7k6dcPli+HffsgIsLuNEq57GRGNou3H+WHDYf4YUMid/e4mP/0aGh3LKWUHzPG8MvWI7w0bxvbDp+ieY1oHu3ThHb1KtqWSVsX3KOFbqAVugsXQvfuMHkyjB5tdxqlzikhOZ0FWw6zYMthVsQnk+MwlI8MZdAlNXmkdxOCgnTymVLK83Idhtl/HuDVn7ZzMOU09/3Lvk0mtNB1jxa6gVboGgNt2kB6OmzaBEHavaK8z6o9yTw6ayPbDp8CoEHlMvRoUoUeTSrTunZ5grXAVSXs6aefBuCxxx6zOYmy0+msXB76ej3frD3IFU2rMH5wS6JKlewWwlrouse7phIqzxOBe+6xJqbNnQt9+tidSKl/2HzwJCM/XEWFMmE8dlUsPZpU5qKK+me6ste2bdvsjqC8QESYtclEy5rlePaHLfR/aynvDo+jQeUou6Ops9AR3UAb0QXIzoa6daFRI/j5Z7vTKPWXhOR0Br69jJAg4avbOlLGv0hAAAAgAElEQVS9nPaRK6W80+/xSdzxyR+czsrllWtb0atZwR1xPUNHdN2jv7cORKGhcNdd8MsvoLOIlZc4lprJ8PdXkJXjYNrIS7XIVUp5tfb1KjLnzstoWCWKW6ev4aV5W8l1BPbgoTfSQjdQjR5tra376qt2J1GK1MwcRk5dReLJDD4Y0ZaGVfTXgMq7PP744zz++ON2x1Beplp0BDPHtOe6S2szaeEubp2+Rpcg8zLaoxuoypWDm2+21tR9/nmoUcPuRCpAZeU4uG36GjYdPMm7w+OIu6i83ZGUOkNCQkLRF6mAFB4SzPMDm9OyZjThoUG6FbmX0R7dQOzRzRMfDw0bwgMPWMWuUiXM4TCMm7mWb9cd5OVrWjC4TS27IymllFfTHl33aOtCIKtXD66+Gt55B1JT7U6jAowxhqe/38y36w7yYK/GWuQqpZQqdlroBrp77oETJ2DqVLuTqACSmZPLw19v4MOlexjZqS63dq1ndySlzunhhx/m4YcftjuGUspNWugGuo4doX17eO01yM21O40KAIkpGQyZ/DufrUrg9u71ebRPE+1pU14vKSmJpKQku2MopdykPbqB3KOb54sv4Npr4auvYOBAu9MoP7Z6TzK3zfiDtMwcXr22Jb2aVbM7klJK+RTt0XWPFrpa6EJODjRtCllZsG4dlC1rdyLlZ4wxTF+xjye/3UTN8hG8e0MbLtYlxJRSym1a6LpHWxcUhITAhx/Cvn1w9912p1F+JjMnl4e+2sBjszfSuWEM39xxmRa5yufcd9993HfffXbHUEq5SdfRVZaOHeHBB61lxvr1g/797U6kfITDYXjxx61sPXQKgLx2WwFEhH3J6ew8ksqd/9eAcT0uJjhI+3GV7zl9+rTdEZRS50FbF7R14W9ZWdCuHRw4ABs3QuXKdidSPmDWn/u5e+Y6GleNIjzE+iVR3p8qxkBIsDCmS/0S2wdeKaX8mbYuuEcLXS10/2njRoiLg9694euv/x6eU6oQ6Vk5/N/4RVQpG86ssZ0I0tFapZTyKC103aM9uuqfmjWDZ5+F2bPho4/sTqO83DsLd5F4MoPH+8Zqkav82rhx4xg3bpzdMZRSbtJCV53p7ruhSxe46y7Ys8fuNMpLHThxmsmL4+nbsjpxF1WwO45SSil1BlsKXREZKyK7RSRDRNaISOdzXDtVREwhH2kFruvqfFaGiMSLyK2e/078VHDw36O5I0aAw2FrHOWdXpi7FRF46MrGdkdRyuMmTJjAhAkT7I6hlHJTiRe6IjIEmAg8B7QGlgFzRaT2WW75D1CtwEc88Hm+Z9YFfnA+qzXwPPCGiAzy0Lfh/+rUgYkTYdEia9c0pfJZvSeZOesOMrpLfWqUi7A7jlJKKVWoEp+MJiIrgPXGmFH5ju0AvjTGFLmRuIh0An4DOhljljmPvQgMNMY0zHfde0BTY0yHcz1PJ6OdgzEwYADMmwd//gmxsXYnUl7A4TD0f2spR09l8st9XYkM01UKlf+7/fbbAXjrrbdsTqICnU5Gc0+JjuiKSBgQB8wvcGo+0NHFx4wCNuUVuU4dCnnmj0AbEQk9n6wKa8WFKVMgKAgmT7Y7jfISX/2xnw0HUnjwykZa5KqAERERQUSE/vZCKV/jUqErQrtier8YIBg4XOD4YaDIRTZFJBoYDEwpcKrqWZ4Z4nzPgs8ZLSKrRWR1Tk6Oi9EDVOXK0LOntQpDgC9FpyAtM4eXftxGq1rl6N+yht1xlCox48ePZ/z48XbHUEq5ydUR3eUirBfhThHKF8P7FqyYpJBjhRmGVSh/7OIzCzuOMeZdY0wbY0ybkBAdkSrSgAHW9sBr19qdRNls0sKdHD2VqcuJKaWU8gnutC40BSYAB0SYIUL383i/Y0AuZ47eVubMEdnCjAK+MsYkFzieeJZn5gBJ55FT5de3r9W+MHu23UmUjRKS05myZDcDWlXnktrF8fddpXzH6NGjGT16tN0xlFJucrXQfQ3YjzVKWgoYCiwQYYcID4kU3XYAYIzJAtYAPQuc6om1YsJZiUg7oCVnti0ALAd6FPLM1caYbFeyqXOoVAk6ddJCN4AZY3jm+80Ei/CgLiemAlDFihWpWLGi3TGUUm5ya9UFES4DrgMGYY2YgtUakAt8AzxrDOf8/bZzebGPgbHAUuBW4GasFRL2isg0AGPMDQXuew/oAjQyBUI7lxfbiFUETwY6AZOA64wxX50rj6664KJXX4V774X4eKhb1+40qoR9/cd+7vl8Hfdf0YjbuzewO45SSgUsXXXBPW6tumAMvxnD7UBbYFG+UyHAQGCFCP3P/QwzExgHPAqsBS4Dehtj9jovqe38+IuIRGGNIr9XsMh1PnM30BurEF4L/Be4q6giV7mhv/Nf6zff2JtDlbiE5HQe/2YTbeuU59au9e2Oo5RSSrnM3RHdnlgjsH2xJoXlzUb5EygL1Ac2G0OzYs7pMTqi64bmzaFiRVi40O4kqoTk5DoY8u7vbE88xdxxnalZPtLuSErZ4qabbgLgww8/tDmJCnQ6ouseV5cXu1+EHcA8YADWCK4BZgNdjSEOaAWcBC72UFZltwEDYMkSOHbM7iSqhLz16y7W7D3OM1c30yJXBbRatWpRq1Ytu2Mopdzk0oiuCA6swlawitkPgNeNYU+B67YCDY0huPijeoaO6LphzRpo0wY+/BBGjLA7jfKwP/YdZ/A7y+nbohoThra2O45SSil0RNdd7hS68cAbwPvGkHqW66oDocawt7Dz3kgLXTcYA7VrW8XurFl2p1EelJqZQ++JS8h1GOaO60zZUrrBoFJKeQMtdN3j6m4JVwPfGnPuTR2M4eCFR1JeS8SalPbBB5CeDpH6q2x/9cS3m9h/PJ3PRnfQIlcpYNiwYQBMnz7d5iRKKXe4uurCQqCWyD+30xUhRoTaIkQXezLlnQYMgNOn4aef7E6iPOT79Yf4cs1+bu/egEvrVrA7jlJeoVGjRjRq1MjuGEopN7nauvAV1iS0u43h9XzH7wAmArOM4RqPpfQgbV1wU3a2tYHE1VdbvbrKrxw8cZpeExZTt1IZvry1A6HBbq1AqJRSysO0dcE9rv4Ua+d8Lbgu7ddYE9TaoQJDaChcdRXMmQM5OXanUcUoJ9fB3TPXkuMwTBjSSotcpZRSPs/Vn2SVnK8nChxPKXBeBYIBAyApCZadc9dm5WMm/ryDFbuTeap/M+rG6GCBUvkNHTqUoUOH2h1DKeUmVwvdU87XfxU4nvd1oaswKD91xRUQFgazZ9udRBWTJTuO8uavO7kmribXxNW0O45SXqdVq1a0atXK7hhKKTe52qM7H+iBNYL7CrAFaALcA0QDC4zhCg/m9Bjt0T1PffrAli2wa5e1GoPyWYdPZtB74hIqlA7jmzs6ERnm6mIsSimlSpr26LrH1RHdd5yvZYEngc+dr+Wcx98u5lzK2w0YALt3w4YNdidRFyDXYfjPZ3+SnpXLpH9fokWuUkopv+JSoWsMXwOvYk08y/8B8Iox6O+wA03fvtZIrrYv+LSJP+/g9/hknurflIZVouyOo5TXGjRoEIMGDbI7hlLKTS5PqzaG+7BWV3gWeM/52s4YHvBQNuXNqlaF9u3hm2/sTqLO0287jvHGLzsYdElNBrepZXccpbxahw4d6NChg90xlCp2IhIuIm+IyDERSRORb0WkyMkaIjJWRHaLSIaIrBGRzme5TkRknogYESnxpWhd6tH1Z9qjewFeegkefBD27IGLLrI7jXLDkVMZ9J74G+UiQ/lW+3KVUspnFHeProi8DfQHbgSSsH6DXw6IM8bknuWeIcB0YCzwm/P1JiDWGLOvwLX3Ad2B3sBgY8yXxZXdFS4XuiKEYIVsBEQUPG8MTxVvtJKhhe4FWLcOWrWCGTPg+uvtTqNclOswDHtvBX8mHOfbOy7jYm1ZUEopn1Gcha6IRANHgZuMMTOcx2oBe4ErjTE/nuW+FcB6Y8yofMd2AF8aYx7Od6wNMAuIAw5zIYWuSF7RbTDG5dEZly4UoTLWNsDn2v/QJwtddQEaN4agIGv1BeUTUjNzeHz2RpbHJ/HSNS20yFXKRf369QPg22+/tTmJUsUqDggF5ucdMMYkiMgWoCNwRqErImHO+8YXODXfeU/edVHAp8AYY8wRufAVms7rAa5WxE8Cjc9xPrD7HwJVeDg0aACbN9udRLlgzd7j3D1zLQnH0/nP5Q0ZrOvlKuWyyy+/3O4ISnlCVSAXOFbg+GHnucLEAMHOawre0yPf1+8A84wxPxRDToB9nEe96Wqh+y/nw6di9WAY4D/Anc7PX3D3jb1FhQoVWLhwod0xfFbTypWJXL2aVfrP0GsZ4MipTI6ezODftYOo1SqSyNCDLFp00O5oSvmMli1bAujPC+UNQkRkdb6v3zXGvJv/AhF5BvhvEc/pfo5zQtFFZcHzf90jIsOBlkCbIp7hOmPqnM9trha6NZyvD2EVuhjDmyL8CmwAfHZoKDk5mW7dutkdw3d16QIvvUS3jh2t3dKUV9l9LI1xM9eyLiGNgZdcxBP9mlK2VKjdsZRSSp2/HGNMUQXkBKzJYueyD2iPNTobg9Wrm6cysPgs9x3DGgUuOOJbmb9HeS8HYoHUAi0LM0VkuTHmsn/cKfIqVu/tvYjcAIAx04rI7xJXd0ZLA0ph9XGcxiqQqzo/PwnsN4baxRGopOlktAs0YwYMGwabNkFsrN1plJMxhs9WJfDUnM2EhQTx7NXNuKpFdbtjKeWzrrzySgDmzp1rcxIV6Dw0GW2EMeYT57GaWEVwUZPR1hljRuc7th34yhjzsIjUAMoXuG0D1o663xhj4gs80AE4MCbkH58XA1cfkoQ1qhsNJGKN4M4AMpznC34zKlDkFbebN2uh60UmL47nhblb6dSgIuMHt6Ra9BkLpSil3NC3b1+7IyhV7IwxKSLyPvCyiBzh7+XF1gML8q4Tka3Am8aYN52HXgU+FpGVwFLgVqA6zp10jTEHgAP538s5sptwRpFrcQCCVXjDeU48K4yrhe42rEK3PtZQ9r+xhqXB6sf4o7gCKR/TqJG1Q5pOSPMa2w+f4tX527miaRXe/nccQUHF9ueFUgFr7NixdkdQylPuBnKAmVjLx/4M3FBgDd1GWO0NABhjZopIReBRoBqwEehtjNl7nhmOAFWAv4tgkcIKYrBaHOq7+mBXWxeuxWpanoE1orsUqOQ8fRToZQx/uvqm3kRbF4pB/frQti189pndSQJedq6DgZOWceDEaebf3YWYMuF2R1JKKVWMinvDCK8gMgO4zsWrDcYEu/pol0Z0jeFz4PO/89AQq/DNAZYawwlX31D5odhYHdH1EpMX7WLDgRTeuv4SLXKVKkY9elirJi1YsKCIK5VS5+FurElxlwANsLoF9p3zDhcVWeiKEA7kVTF9jGGrMZwEvimOAMoPxMbC/PmQkwMhupWsXbYcOsnEn3dwVYtq9GlRze44SvmVIUOG2B1BKf9lzBFgKJA3MQ2MqVscjy6yKjGGTBEqAlHk751QKk9sLGRlQXw8XHyx3WkCUnaug3s/X0d0RChP9W9mdxyl/M6oUaOKvkgpdX7yLy/2934NxSLIxevyflfTsrjeWPmR/CsvKFu89etONh86yTMDmlOhtK5nrJRSyqeMw9qIDOBD4IPierCrhe4EIBn4VIQhIjQSoXb+j+IKpHxQY+fu0Fro2mLjgRTe/GUnA1pVp1ezs+3YqJS6EN26ddPNhZTyHNuXF1uMNYxcAfikkPPGjWcpfxMVBbVra6Frg6wcB/d9sY7ypcN4ol9Tu+Mo5bdGjBhhdwSl/JnHlhdzpzjVxTjV2enKC7Z445cdbE08xXs3tKFcpLYsKOUpWugq5VG/Yi0vlrcBmQB1znKtW/27rha6H7nzUBWAmjSBhQshNxeCXV7eTl2AzQdPMmnhLgZdUpMesVXsjqOUX8vOzgYgNDTU5iRK+aX8y4vljdaWzPJiAMZwU3G8mfJjsbGQkQF790K9enan8XvGGJ75fjNRpUJ4/CrdelkpT+vZsycACxcutDeIUv7ozOXFTIktL6aUS/KvvKCFrsf9svUIy3Yl8b++sURH6giTUp52yy232B1BqUDRvTgf5lKhK1LkMg/GGG4uhjzKVzVpYr1u3gxXXWVvFj+XnevguR+2UC+mNMPaX2R3HKUCwrBhw+yOoJT/ErFW7zJmH7D7H8cKY13nEldHdEdw9uZfcZ7TQjeQlS8P1arBli12J/F7n63cx66jabw7PI7QYFdXCFRKXYj09HQAIiMjbU6ilF/ag7XEWIjz83NNOHNrpS9ddUEVH115weNOZmTz2oIdtKtbgZ46AU2pEtO7d29Ae3SV8iA5y+cXxNVCt2BDcAhQD3gMaA3o76qVVeh++CEYA6J/L/KESb/uIjkti0f7xCL6z1ipEnPbbbfZHUEpfzaNv0dx839+wVxddWFvIYd3ibAcOAbcBiwqrlDKR8XGQmoq7N8PtWrZncbvJCSn88HS3QxsXYPmNaOLvkEpVWyGDBlidwSl/JcxIwr9vBhcaINfCFbV3asYsihfl3/lBVXsXv5xGwLcd0Uju6MoFXBSUlJISUmxO4ZSyk0XsupCKaATEA7of/3qn4XuFVfYm8XP/LnvON+uO8id/9eA6uUi7I6jVMDp378/oD26SnmESFGre+VnMMblBRAudNWFvCbBH1x9Q+XHYmKgUiUd0S1m1uYQW4gpE86Yri5v762UKkZ33XWX3RGU8mcjcK0v1+2Vvi501YVM4FNgnBvPUf5MV14odnM3JrJm73GeH9icMuG6x4tSdhg4cKDdEZTydx6ZYX2+qy4AZBpDYnGGUX4gNhY+/VRXXigmGdm5vDB3K42qRHFtG53gp5Rdjh07BkBMTIzNSZTyS/l3Q4sCJgMngFeA/UBN4F4gBhjlzoMvZNUFpc4UGwsnTkBiorWBhLogr/+8g33J6Uy/uR3BQfoXB6Xscs011wDao6uURxjz98pdIpOAqsBlGLM73/FFwA6gL/Ctq492dTJaL+BS4E9jmJPveD+gFbDSGOa5+qbKj+WfkKaF7gXZsD+FyYvjGRxXk8sa6iiSUna699577Y6gVKC41vl6usDxvK8H4saorqutC48D7YArCxxPBZ4AloMWuop/FrqXX25vFh+WlePg/i/XUbF0GI/2ibU7jlIBr2/fvnZHUCpQhDtfv0Lkef5uXXjIeTzUnYe5Wug2dr4uL3B8pfO1iTtvqvxYlSpQvrxOSLtAby/cxdbEU0y5oQ3RkW79N62U8oDERGtKStWqVW1OopTf+xFr1LY98E2Bc8Z53mWuFrqRztcywKl8x6MKnFeBTkRXXrhAWxNP8uavO+jXsjo9Y6vYHUcpBQwdOhTQHl2lSsCdQFOgsN2RtgJurfXnaqF7CKgN/Be4I9/xR5yvB915U+XnYmNh1iy7U/iknFwHD3y5nrKlQnmiX1O74yilnB566KGiL1JKXThjDiHSGrgB+D+gInAM+BWYhjEZ7jzO1UJ3AdbivLeJ8C9gG1alXR9rGHmBO2+q/FxsLEyZAkePWhtIKJe999tu1u9P4c3rW1OhdJjdcZRSTr166U73SpUYq5h91/lxQYJcvO4FrIlnYBW3vZ2vAqQ5zytlyT8hTbls19FUXv1pO1c0rUKf5rpihVLeJCEhgYSEBLtjKKXc5FKhawy7gH9h9UZIvo/NwL+MId5jCZXvySt0t2yxN4cPyXUYHvhyPRGhwTw9oBmim20o5VWGDx/O8OHD7Y6hlHKTy/uJGsPvQFMR6gNVgMPOAlipf6pRA6KidETXDdOW72HN3uO8em1LKkeVsjuOUqqARx991O4ISqnz4Grrwl+MYZcxLLuQIldExorIbhHJEJE1ItK5iOvDROQp5z2ZIrJPRO7Kd36EiJhCPrRisIOuvOCWjQdSeHHeVro3qsTVrWvYHUcpVYgePXrQo0cPu2MopdzkUqErwnQRckV4vMDxR53Hp7v6hiIyBJgIPAe0BpYBc0Wk9jlu+xToBYzGmgQ3GFhf4Jp0oFr+D+PmzDxVjJo00ULXBUdOZTBq2moqRIbx0jUttWVBKS8VHx9PfLx26Snla1wd0e3kfP24wPHpWL26nXDdPcBUY8wUY8wWY8ydWMuX3VbYxSLyL6AH0NsY85MxZo8xZoUxZmGBS40xJjH/hxuZVHGLjYVDhyA52e4kXiszJ5dbP17DifRsptzYhkpR4UXfpJSyxciRIxk5cqTdMZQKTCKCyHkt4+RqoZs3Bbxg8XjY+erSVjEiEgbEAfMLnJoPdDzLbQOAVcA9IrJfRHaIyOsiUqbAdREistd5zXdircGm7BIXZ73+/ru9ObyUMYb/ztrIH/tOMH5wS5pWj7Y7klLqHJ588kmefPJJu2Mo5f9ErkTkJUQGOr8eDpwEEhFZjUhldx7naqGb1wLQocDxDgXOFyUGCObvAjnPYc5eLNcDLgNaAoOwNqzoBUzNd802YCTQH7jOmWepiDQs7IEiMlpEVovI6pycHBejK7d06ABhYfDrr3Yn8UofLN3Dl2v2c9flDenTQpcSU8rbde3ala5du9odQ6lAMBa4FyiNSATwFlAaq4OgNfCUOw9zddWFDVjtCVNFeATYAjQBnsXaMGKDO2/qvCc/KeRYniDnueuNMSkAInIH8KOIVDHGHDbGLAeW//UwkWXAWqxt5M7YKs4Y89cixKVLlz7b+6oLEREB7duDbpd5hsXbj/Ls95vp1bQq4y4v9O9iSikvs23bNgAaNSpsV1KlVDFq4Xz9DbgUKINVd+4CrgKucOdhro7oTnW+1gA+AlY6X2sVOF+UY0AuZ47eVubMUd48h4ADeUWuU94CrYVOYDPG5AKrAa0i7NS9O/zxB6SkFH1tgIg/msodn/zBxVWieOXalgQF6eQzpXzBmDFjGDNmjN0xlAoEeb24BwDnwvxMAG50fl7dnYe5umHE+8BX/HOziLyf0F8awweuPcdkAWuAngVO9cRafaEwS4HqBXpyL3a+7i3sBrGmrrfAKpKVXbp1A4cDFi+2O4lXOJmRzS3TVhMSHMSUG9pQOtzlZayVUjZ77rnneO655+yOoVQgyHK+VsSq5QxWi2pGgfMucWfDiMEiXAv0xblhBPCtMXzhzhsCrwIfi8hKrCL2Vqzq/B0AEZlmvZ+5wXn9J8BjwIci8gRQDmt5si+NMUec9/wP+B3YAZTFaldowVlWclAlpH17CA+32hf69rU7je0e+GI9+5LSmXFLO2pViLQ7jlLKDR07nm2+tFKqmO0DmmK1LlTn7xbZvIXmj7jzMLeGlIzhc+Dz/MdEKAMMMoaPXHuGmSkiFYFHsVZz2Ii1dFje6GztAtenikgP4A2s1ReOA7OBh/JdVg6r57YqkAL8CXQxxqx05/tTxaxUKWtSmk5IY83e48zblMj9VzSiXb2KdsdRSrlp48aNADRr1szmJEr5vU+w9lqo6/x6AcYcR6S/8+s/3HmYGOP+XCwRgrBWPhgG9ANKGeNe0ewtSpcubdLS0uyO4b+eegqeeAKSkqB8ebvT2Gb4+yvYfPAkSx7sTmSYT/6nolRA69atGwALdYKt/5s2Dd54A5YutVYP8jIikm6MKW13Do+x2k/vAzoDu4GnMCYJkVuAdsBXGDPP5ce5U+iK0BaruB2KtVQYOFdMMIZglx/kRbTQ9bAlS6BLF5g9G/r3L/p6P7RqTzKD31nOf3s3YVSXenbHUUqdh1WrVgHQtm1bm5MojzLG2tlz2zb47jvo08fuRGfw+0K3mBU5GU2Eus6tfrdi9cHegTUjLm9C2mngM4+mVL7r0kutFoYAbl947aftxJQJZ1j7i+yOopQ6T23bttUiNxAsWWIVuQCffmpvFlUszvo7VBHGAMP55yYRBddCMkAVY0j1QDblD8LDoVOngF1P9/f4JJbtSuLRPk2ICPPJX3oopYC1a9cC0KpVK5uTKI+aPBmio60J1LNmQXo6RLo4efjxx6FcObjnHs9m9EciuW5cbTDG5R7Ac43ovo1V5OaN3GYDPwA3A93+ejctclVRuneHdeusPt0AYozh1Z+2UylKR3OV8nXjxo1j3LhxdsdQnpSUBF9+CcOHw8iRkJZmtS+44sQJeOUV2LzZsxn9V8Hla4v6cJkrFbEBPgDuN4YTACI0dedNVIBzTuJg0SIYONDWKCVp+a4kVu5O5n99YykVqqO5SvmyCRMm2B1Bedq0aZCVBaNHQ2wsVKsGn3wC115b9L1Tp1qjv7ff7vGYfmof/9whtyLWjmjZQJLz61AgHTeXF3N1Z7SRwFYR3hahh/PNlHJN27bWr34CqH3BGMNrC7ZTpWw4111a6AZ+Sikf0qpVK21b8GfGwLvvWuu/N28OwcEwZAjMnWuN1p6LwwGTJlnLabZuXTJ5i5GIhIvIGyJyTETSRORbEanpwn1jRWS3iGSIyBoR6VzINZeKyE8ikioip0RkmYjEnPEwY+pgTF2MqQtcg1X0vgJEY0x1IBp4zXn1v935/s5V6L4IJPD3MHFlYDTwI9Yivkq5JizM6tMNoAlpv+08xqo9x7m9ewMdzVXKD6xateqvlReUH1qyBLZuhfzbPF93nTXC+/XX5753wQLYsQPuuMOzGT1nAjAIuA5rSa+ywHcictYfXiIyBGvzrueA1li7284Vkdr5rmkHzAcWAu2BOGA81ihtUXnKYC0rZu2GZr0+AUQ6n+GyIpcXE6EL1qS0QVgbM+TJu/EgMN0YHnbnjb2FLi9WQp5/Hh55BI4cgUqVir7ehxljGPT2Mg6lZLDw/m6Eh2ihq5Sv03V0/dywYVY/7sGDf08+MwYaNoS6deGnn85+b//+8PvvsG+fNQHbw4pzeTERiQaOAjcZY2Y4j9UC9gJXGmN+PMt9K4D1xphR+Y7twNq19mHn18uAX40x/3UzVDoQDvwLY37Od7wHVuGcgTEuby9aZOuCMXMC3fUAACAASURBVCw2hlFYu44NBr7F2mc4b6S3BvCAG9+CCkTdu1uvixbZm6MELNp+lD/2neD27g20yFXKT7z55pu8+eabdsdQnpA3CW3YsH+usCACQ4fCL79AYmLh9+7ZYxXIo0aVSJHrAXFY7ajz8w4YYxKALUCh+16LSJjzvvkFTs3Pu0dEKmMtaHBIRH4TkcMiskRELnch01Hn6xxEvvx/9u48zub6e+D4621fs4bsW6GkiZB9+VmKUCi0SJJvolJRtK9SUSgtFJKtskZkqxnZstQkRPbdIPsMZnv//jh3MsbM3Htn7szn3jvn+XjM45rP/dzPPXfSOPd9z/scjBmFMTOR/NMmut8jHrdnsJZoYBYwyxiKIEMjHuTK9mMBp2jRovoOPROY2Fga58nD0SlT2FH86vKcYLLreCRDQ+Ipc3EPoaF7nA5HKeVD+u9F8Ck7cyZVL11ife3aRCb575uvalXqxcezY9gwDiWzmbryuHGUA9bWqsWlzPu7kcMYsyHR9+OstePSeK1SQBxwIsnxCNd9ySkOZHedk/QxrVx/TpiO9AYwGPgDWSxdbIypY639M5WYPkNKInID9yQ6bpBEd2wqj71KmkYAX3EBQ2XgAWt5K10XcoiWLmSiO++Uj3a2bHE6kgzz87YIek/awLudb9ZNaEoFkdWrVwPQsGGyi1wqUFkrHRYKF4Y1a5I/55ZbIH9+cP0d+M/Fi1C2rEz/dFfH60OelC4YY94G3JUMtABKA5OBnDZRQmiM+QXYbq19PJlrlwYOAU2ttb8mOv4a0MNaW90Y0xBYBbxrrX0x0TmrgT+ttf3cvMjXkWqBPImOXgTew9o33LyuK3i8opsSa9kNgZnkqkzWogW88AJEREDJkk5H43OXYuN4d+E2yhXNS9c6bjesKqUCyIsvyr/VuqIbZFaulE1oEyakfE6PHjB0qJQpVKx4+fh330nZg3+2FBsFTHFzzn5kk1h2ZJU2cUlACWBFCo87gawCJ13xLcHlVd4jrtukjYX/BtyvAln7OsZ8hFQNFHM951qsPeP2sUl42l5MqfRL6KcbpP9QfB66mx3HzvNmx5rkzK7/aykVTL744gu++OILp8NQvjZuHFxzTeq9crt3l9sZM648PnYsVK8OLVtmXHxpZK09Ya3d5uYrCtiIdEFonfBYV2uxGkgnheSuHe16XOskd7VO9Ji9SLOCaknOuQHZ6ObJiziDtT9h7VSsXZyWJBd8sKKrlMdq14aCBSXR7dbN6Wh8auexc4z9ZSd31bqOFtVLOB2OUsrHqlVL+u+1CngnT8L330OfPlKakJKKFaVH7vTpMGSIHNuwAdatg48/lk1rAcpae8YY8xXwgTHmGDKc4UNgE7As4TxjzDbgE2ttwo7MD4FvjDHrkBKFx5EyiM9d17XGmA+AN4wxm5Aa3fuQFeSr+7AZ86qXgb/p6ama6KrMkyMHNGkSdP104+MtQ2f/Rd5c2Xmtgw4NVCoYhbk6xjRr1szhSJTPTJ4Mly5JxwR3evSAp56SPSY33SSruQUKQM+eGR9nxnsGiAW+BfICy4Ge1tq4ROdUQ8obALDWfmuMKQa8DFwHbAbaWWv3JTpnlKtDw0ik/GAL0rIsuY1or3PlZDR3PE50070ZLdDpZrRMNmIEDB4svQqvu87paHxi6m/7eGnOZt7vUov76pZzOhylVAbQPrpBxlpJWK+5RnrguhMRAaVLS63uM89AmTLQu7dMRMtkvuyj6zeMiffibIu1Hvfu1BVdlbkS+umGhso75AAXcfYiwxduo0HlYtx7m25AUypYTUhts5IKPD//DH//DV995dn5JUtKLe706ZIcX7rkr5vQAtUjif6cE2lLZoAvgYNAWaAPsnHuZW8unOKKrmsimsesTXF3nl/TFd1MFhcHxYpJ4f+4tLb98x/9pmxk+bZjLB7YlErFg+sNtlJKBaUzZyAkRGprN2++ckhEaiZMgEcflb0mtWs7trE6KFd0E5PWaEOB2iQuczDmVmQT3Ais9XhQWWoruqF4Xi9h3VxLKZE9u/QcDII63SVbjrJo81EGt62mSa5SQW7ZMtmX06pVKzdnqkxx5oxMJJs1CxYvhpdekjHz7lgL/frBgQPSWszTJBegc2d57LlzMODq/VTKZ3q7bvcnOZ5Q//sQXkzkddcDyXjxpZRnWreGnTvhn3+cjiTNzl2M4dV5W6heqiB9m1Z2/wClVEB7++23efvtt50OI2s7eRImToS77oISJWRk72+/Qa1akuhOmuT+GlOmSPnBG2/A7bd79/yFC0OnTlC+vNyqjFLYdTseY2piTGGMqQmMdx2/xpuLpVa6MDHJoTZIc+BVXK6XaIS0olhg7X8ZeEDR0gUHHDggvyiGDZPC/gD0ytzNTPltH3OeaERIucLuH6CUCmgHDhwAoFw53XCaqayVetqPPpKV29hYqFABunSBrl2hfn051r69lBIsWABt2yZ/rV27pGShdm25ZnaP9zNddu4cXLggibZDskDpwhJklHBKVQXLsDaF/8jJXM6TrgvG8AAyIq6btcxMdPw+YDrQ11o8rOj2L5roOqR+fanX3bDB/bl+ZuO+U3T9fDUPN6jI6x21nZhSSvlcTIxMHhsxAsLDZTNYr16S3Napc3Xv2rNnpX3l7t2wYgXceuvV12vcWD5J3LQJAvgNSxZIdKshU9muTebeY0AzrN3u6eU8Hd+UsMPtpyTHFyJlC4M9fUKlAHk3vnGjjFQMMO8t2kaJgrkZ1FYbyCuVVfz000/89FPSfwKVz509CyNHQuXKUppw6ZJ0Rti3D4YPh9tuS35AwzXXwKJFUKQItGsn5yf2+usy4GH8+IBOcrMESWJrAu8B64BdwG/AcOBmb5Jc8DzRrei6fSLJ8YTeGhW8eVKl6NJFbmfPdjYOL63d/S/r9p6kX7MqFMit+y+VyiqGDx/O8OHDnQ4juCUkoYMGwfXXw48/SleE3r0hd273jy9dWpLdixfhzjvh1Ck5HhoK774rHRO6ds3Ql6B8xNrjWDsUa2/H2uuxtgHWvoi1x729lKelC38i2TXACeAIMgkjYUrGZmu5xdsn9wdauuCgkBAZu7hqldOReOyBL9ey/eh5Vr7Qgjw501DfpZQKSEePHgWgVKlSDkcSpGJipO61enX45BMpT0irsDBo00Y2m82YAfXqSXeFjRtlmlmAC/rShQTG3A60A0ogJQsLsHadt5fxdEX3JSAeKVMoDtzsujVIsbAHPT2USqJLF1i9WqakBYCN+06yaue//K9pZU1ylcpiSpUqpUluRgoLg9OnYciQ9CW5AM2ayXjfFSvgxhtlqtm0aUGR5GYZxnyGND94CXjMdbsGY8Z6eymPEl1rWQDcgdRIWC4nuGuBNtbyo7dPrNR/5Qtz5jgbh4fGLN9J0fy5eOD28k6HopTKZPPnz2f+/PlOhxG85s6VVdc2bXxzvW7dZCPb6dPS4Se9ybPKPMb0Av5H8q1sH8eYnl5dzpPShSufn3xAEeCUtUR59WA/pKULDqtRA667Tlq9+LE/D5ym09hVPH9HNZ5oXtXpcJRSmax58+YAhDo0DSuoxcdLy8l69Xy/b2PfPrl2chvYAlTQly4Ysxq4HRkQ8ZHrtjzwDLJnbA3WNvL0cl7tpjGGHEitbjFrWeTNY5VKVpcuskng+HG4NrlOIv7h4593UihvTno2qOh0KEopB8ycOdP9SSptNmyAQ4fgnnt8f+0Kulc+ANVEqgY6YO3m/44a8wuwict7xjziaY0uxnAvcAhYA8x3HVtuDLuNwUefNagsp0sXeTc/b57TkaRoy+EzLPs7gkcbV9JOC0plUcWLF6d48eLuT1TemztXhje0b+90JMo/5HLdHkxy/GCS+z3iUaJrDE2QwRAJG9ASPgP4EVlG1n4dKm1CQqBSJZlX7qc++XknBXPn4OGGFZ0ORSnlkNmzZzM7wNohBow5c6B5cyha1OlIlH844LodgTEyetSYQsAHSe73iKcrukNd5yZt0rvMddvAmydV6j/GyKru8uWyacDPbD96jkWbj9KrUUUK5c3pdDhKKYeMGTOGMWPGOB1G8Nm2Tb4yomxBBaoFyILqI8C/GHMGOAn0RkoavNoV6mmiezsJ9RJX2u26LePNkyp1hS5dpIeiH+5o/uSXneTPlZ3ejSo5HYpSykHz5s1jnh+XWAWsuXPltlMnZ+NQ/uRtYD+XKwgKJvrzPuAdby7maaKbsLtvf5LjeZPcKuW9evWgTBm/K1/Ydfw8CzYd5qEGFSmS36uSIKVUkClUqBCFChVyOozgM2cO1K0LZcs6HYnyF9b+C9QHvkIGlMW6bscDDbD2pDeX83RnzSFkzG/SEoVBrtukBcNKeS5bNujcWcY/nj/vN029x/6yk9w5stGnia7mKpXVffvttwB069bN4UiCyKFDsG6d9LlVKjFrI5BBEenm6YruYmTJeG7CAWPYhiS61nW/UmnXpYvMJ1+40OlIANj/bxTzwg/zQP0KFC/gwYx1pVRQ++yzz/jss8+cDiO4JJSC3H23s3GooObRwAhjKAOEA8WQxPa/u4B/gRBrOZQhEWYwHRjhJ+LioHRp2XnrWjlx0lsLtjJ5zV5WvtCSktfkcTocpZTDoqJkPlK+fPkcjiSItG4NBw7IZjTlsaAfGOFjno4APgQ0ApYA8UiCG+/6vkmgJrnKj2TPLu/qf/wRLlxwNJSLMXHM/v0gbW4spUmuUgqQBFeTXB86dQpCQ3U1V2U4jwdGWMs/1nIHsvutLFDQWu6wFn0rpnyjSxeIjIQlSxwNY/GWo5yKiqFHvfKOxqGU8h9TpkxhypQpTocRPH78EWJjta2YynCeDowoZAzljaG4tVy0lsPWctEYiruO61ZUlX4tWkCRIo53X5i+bj/li+ajYZVijsahlPIfX375JV9++aXTYQSPuXOlXK1uXacjUUHO0xXdCcAe4P4kx7u7jn/ly6BUFpUzp7y7nz0bzpxxJITdx8+zdvdJutcrR7Zsxv0DlFJZwtKlS1m6dKnTYQSHCxdg0SLpnZvN4w+WlUoTT/+G1XfdJl1qm43U69ZHKV944gkpX5g40ZGnn7H+ADmyGbrW0Z6OSqnLcubMSc6cOh3RJ5YuhagoLVtQ3jGm/H9fXvA00b3WdZt0RuuZJPcrlT516kCjRvDxx9KJIRNdio1j5saDtKpRkhIFdROaUuqySZMmMWnSJKfDCA5z50KhQtCsmdORqMCyF6ki2O3mvCt4muiec922SXI84fvz3jypUql6+mnYvTvTe+ou3RrBychoetTXTWhKqStpousjsbHwww9w112QSydOKq8ljAL2mKeT0X4HWgETjOEm4G+gBvAs0ld3ozdPqlSq7r5bxkGOHg0dOmTa005ft58yhfPSpGrxTHtOpVRgCA0NdTqE4LByJfz7r5YtqLRYwZWzHDziaaL7OZLoXgO8kei4cT2pjotRvpMzJ/TvD0OHwpYtcNNNGf6Ue09Esmrnvwxqc4NuQlNKqYwyezbkzg1t2zodiQo01jZPy8M8HRgxG/iQy0vGiZeOR1p7eTSwUj7Rpw/kySO1uplgxvoDZM9muPe2cpnyfEqpwDJ+/HjGjx/vdBiB7dIlmDZNPqkrUMDpaFQW4emKLtYyyBi+BToCJYEI4AdrWZ9RwaksrHhxeOABmDwZhg2DokUz7KmiY+OZufEALauX0EloSqlkfesaTf7YY485HEkAmztXyhb0Z6hSY0zTVO61wL9Yu9Xjy1nrdblDUMmfP7+NjIx0OgyVnE2b4JZb4P33YfDgDHuaRX8dod/U35nYqy4tqpfIsOdRSqksrXVr2LFDNhtr/9w0M8ZEWWvzOx1HhjEmHve1uIeBx7H2R3eX83hF1xgKAu2ACsBVy17W8qan11LKI7VqQfPm8Mkn8MwzkMPjv65embZuP6UL5aHpDdolTymlMsSePbBsGbz5pia5yhPuNsuUAWZjTF2s3ZTaiR5lDsZQF1gIpPb5sSa6yveeflp25/7wA3Tu7PPLHzgZxa87TvBMqxvIrpvQlFIp+PTTTwF44oknHI4kQE2YIAlur15OR6L839dAa6A0sBrYD5QDGgJHgD+QBgm5kO5fvVK7mKdvq0YBxbh6M5rX/cyU8kqHDlCxIowZkyGXn7F+P9kM3FdXJ6EppVI2f/585s+f73QYgSk2VqZd3nEHlNMNv8qt5cB1QHesbYy192NtE+AB1/FvgXuQ/NPt1BFPPwuuhdRLhCFjgCNJQy8zpbyWPTsMGACDBkF4OISE+OzSMXHxfLfhIC2qleC6Qnl9dl2lVPBZtGiR0yEErsWL4dChTOuiowLey67bpFOjFiDJ7YtYeyPGnAFKubuYp4nuaSAf0Nnaq8YAK5WxeveGV1+VX5JffeWzy87aeJDj5y5xv05CU0qpjPPll1CihExDU8q9Cq7bpzFmGJe7Jjzuuq3kuj2HB3msp6ULk123NT08P1XGmCeMMXuMMReNMRuNMU3cnJ/LGPOm6zGXjDH7jTFPJTmnizFmq+v+rcYYHbsSLIoUgZ49YepUOHHCJ5c8HRXNez9to17ForTUTgtKKTdGjx7N6NGjnQ4j8Bw9CvPnS21uzpxOR6MCw3bX7ZvAMYwJx5gI4D2kmmA7xmRHWt0edncxTxPdvcAZYJ4xvG8MjxpDz8RfnkZvjOkGjAaGAbcihcaLjDGpLatNB+4A+gLVgHuB/3bZGWMaIDUbU4EQ1+33xpj6nsal/NyTT0qzcdeGkPQasWQ7Zy/G8kanmzBGy8yVUqlbvnw5y5cvdzqMwPP11xAXB48+6nQkKnC8CMQjZQpFgZuB4q7v44ChQEsgJ7DK3cU86qNrDO56mllrPe3gYH4DNllrH0t0bAcw01o7NJnz2wDfA1Wstcku5xljvgWKWmtbJzq2DDhure2RWjzaRzeAdO4s3RfmzJFNamm0+dAZOnyykocbVOT1jhk/XlgppbIka+GGG6B0aQgLczqaoBH0fXQBjGkOvAPURxZl44G1wEtYG4YxOYDcwCWsjU3tUt40s0up44LHnReMMbmAOsCSJHctQdpGJOduYD3wrDHmoDFmhzFmjDEm8fzABslcc3Eq11SB6OuvoU4duPde+OWXNF0iPt7yyrzNFMufi2da3+DjAJVSSv1nxQrYuVNGuivlDWtDsbYRUBBpLVbQ1YEhzHV/LNZGuktywfPNaI+kOdgrFQeyI+ODE4tAeqIlpzLQGLgEdAEKAx8j/dW6us4plcI1k92NZ4zpi5RBkCtXLq9egHJQwYKwcCE0awYdO8LPP0Pdul5dYtbvB/lj/2k+6FqLQnm1Xkwp5ZkRI0YAMGjQIIcjCSDjx0OhQtCli9ORqEBiTCjwFTATay8Ah9JzOY8SXWv5Oj1Pktwlk3xvkjmWIJvrvvuttWcAjDEDgMXGmJLW2oQE1+NrWmvHAeNAShe8D185plgxWLIEGjeWnoy//go33ujRQ89ciGH4om3ULl+YLrW1b65SynNr1qxxOoTAcuoUzJwpq7n58jkdjQosTYEmwMdIaeoErP0trRfL7Dl8J5BC4qQrrSW4ekU2wRHgUEKS6/K36zZhA9tRL6+pAlnp0rB0KeTKJbPT9+zx6GEfLf2HU1HRvNmpJtl0CppSyguzZs1i1qxZTocROKZOlQ3EWragvBeNLFZeA/QBVmPMFox5FmOu9fZiHie6xvCQMfxuDJHGEJfky22NBIC1NhrYiIx2S6w10n0hOauA0klqchOKK/e5btd4eU0V6KpUkWT3wgVJdo8cSfX0rYfPMnnNXh6oX4GaZQplToxKKZUVWStlC3Xq+HTIj8oYxpjcxpiPjTEnjDGRxpgfjDFuP/Z01yrWGFPKGPONMeao67p/GmMe8CCkksCjyIS0hO4LNYAPgIMYM9ub1+dRomsM9yGzh28B8pK+McAfAr2MMX2MMTWMMaORetvP5bnMZGPM5ETnTwP+BSYaY24yxjRC2pPNtNYec50zGmhpjBlqjKlujBkKtEBGF6tgVbMmLFokfRrbtIGTJ5M9zVrLaz9splDenDzXRjegKaW8N3z4cIYPH+50GIFh40bYtElXcwPHKGQPVA+kZOAaYIGRXrXJ8rBV7GQkQe2EtAibDHxjjGmaajTWnsHaiUgnrbLAQOA3JNfM6bqexzxd0e3vur2QEAaSfIJMTdt31SNSYK39Fgn6ZSAc2WjWzlqbcI3yXC5JwFp7HtmoVgjpvvAdMoq4d6JzVgPdgYeR/ro9gW42HTUdKkDUrw/z5sH27fDWW8meMjf8EOv3nuKFO6pTOJ9uPlRKeS88PJzw8HCnwwgMX38NefJAj1S7eyo/YIwphKyeDrbWLrXW/g48BNQi5SYBAM8Ck6y14621f1trn0RKTfslOqchMNZa+5u1dre1diRwAKjnRYjngZPAKaT01Wue9tE9hWT4jZCs3VpLdmN4BRgAtLSWLWkJwGnaRzdItG4NERGyipBIVHQszT4IpXThvMzp11Brc5VSKiPFxkKZMtC0KXz/vdPRBCVf9tE1xrRESgRKWGuPJzq+Bfnk/LVkHpMLiAJ6WGu/T3R8LFDTWtvM9f1PyMLog0ii2gEZANbQWpvyu0ZjcgLtgPuBu4A8Cfe4rrcCa1t4+ho9bS+W8AP93fUkGEN2YCTwBjAG+D9Pn9SfFC1alNDQUKfDUOlUoXx5Ki1bxsp584gtdLkG92RkND0rXqDKtYYVK7RhuVJKZaTCGzcScuwYm2vW5IT+25pRchhjNiT6fpyrm1RalEJWSpMO5EqxRSuet4q9D5jhunYs0ia2R6pJ7uXrJPxDnrA6dRApfZiAtbvdPP4Knia6Z4Eiric8hzTwvRMZCwwyuSIgnTx5kubNmzsdhkqv7NlhwgQaWwuu/57x8ZbWH4WRP3c+BtzfSEf9KqXS7C1XadQrr7zicCR+bupUKFCAms8/D3nzOh1NsIq11t6W2gnGmLeBl9xcJ7VV0dTaviZw19b1bSQpboUku3cDk40xTa21f6Zy3cKu20vAD8AEYAmelCAkw9NE9zCS6JZAWnvVA+Yluj/5XUBKZZZ69aQmLCwM7r4bgBU7jrPreCSjuoVokquUSpft27c7HYL/i46G2bPld7AmuU4bBUxxc85+4HZkdbY4cDzRfSWAFSk8zm2rWGNMFeBJICRRUvunqzPDk0jbsJSEAxOBKVh7ys1rcMvTRPcPoCaycjuZq1dwfT1QQinv5M4NDRpcMU99wqq9lCiYm3Y3X+dgYEqpYDBlirucQbFsmXS/6dbN6UiyPGvtCa4uR7iKMWYjEIO0ZJ3mOlYW6ZaQbItWa22063GtgcSF2K2BhGbTCVNCkm4gi8NdIwRra7uL2xueJrpPAM8D56wlyhgKAd2Qmos5wHu+DEqpNGnWDN54A06fZsel7Kz45ziD2txArhyZPRdFKaWyoBkzoHBhafeoAoK19owx5ivgA2PMMaSj1odIB6tlCecZY7YBn1hrP3Ed+hBpFbYOmXfwOIlaxQLbgJ3Ap8aYQa7r3o0kw+7bgxmTA9mQVg1pa5s08Dc9fY2ejgCOBCITfT8c0IaCyr80ayaNyleuZGJMBXLnyEaPeuXdP04ppdx49dVXAXjzTY//fc1aLlyAuXPhvvtkaqUKJM8gC5ffIknlcqCntTbxamw1pLwBkFaxxphiSKvY64DNJGoVa62NMca0Q3LF+UABJPF9xFo7P9VojCkBhLqeMyXpT3SNwasMwVr2e3O+Uj5Xvz7kysXFpcuZfU0b7rm1DMUK5HY6KqVUEDhw4IDTIfi3RYvg3Dno3t3pSJSXrLUXkbrZJ1M556qNLtbaT4FPU3nMDmQQhbfeAKqncr9Xm9JSW9Hd68XFrJtrKZXx8uaF+vU5u3g5F+9uxSONKjkdkVIqSEycONHpEPzbjBlQosR/XW+USoc2SF45CXjE9eenkUTc4mVFgbvixeRG/ab0pZTj4po0pdg/W2hdJi/VShV0OhyllAp+58/DggXQtSvk0DUvlW5lXLdD/jsitcGdgRuQscAeS+1vpHZSUAFnXfmaNLDx9M911P3J4eHQsiUsXQp16mR8cEqpgDV06FAA3n33XYcj8UPz50uNrpYtKN+IA3IiG9higBwYcy2wz3V/X6RHr0dSTHSt5ZF0BKmUI0ZHFadutuzcsvtPZHpgaiePhlOn4PPPYfz4TIlPKRWY/v33X6dD8F8zZsjY30aNnI5EBYd/kVXdQsBRZAV3KnDRdX8Rby5m0jhoImjkz5/fRkZGuj9R+b3f95+i86erWffjq5QokAvWrEn55H//hbJlZS573rxw9Cjky5fy+Uoppa526hSULAlPPgkjRzodTZZgjImy1uZ3Oo4MY8xSoCUys+Fp4AGu3DO2EmubeXo5jxuMGkM1Y/jQGH40hp+TfC339DpKZZSJq/ZSME8OCt/RCjZsgNTewEyaBBcvwocfyk7hOXMyLU6llAoac+dCTIwOiVC+NB4YB+RBOjAc5/J+sBPAQG8u5tGKrjHUQXqaJbfkZQBrLdm9eWJ/oSu6weHImQs0fu8XejeqyEvZ98Odd8KSJdC69dUnx8dDtWpQqpRMUqtaFapUkVpdpZRKxqBBgwAYMWKEw5H4mTvugH/+gV27QEetZ4qgX9FNyphrgBZIr99VWHvam4d7uqL7IpAf7bag/NTkNfuw1tKzQUWpE8ue/YpxwFdYtgx27oR+/SBbNnj4YVi+HPZrK2ilVPIuXLjAhQsXnA7Dvxw/Lr9Pu3fXJFdlHGvPYu08rP3R2yQXPE90GyL1Ef0SnhaoBfwA/AP4dC6xUt44dPoCU9buo82NpShXNB8ULChdFFJKdD/9FK69Frq4+lj37CkT1SZPzryglVIBZezYsYwdO9bpMDLXwYNQubJMO/vxR9nTkNisWRAXp90WlF/zNNEt5rqdmnDAWjYjLR5uQMbHKZXpYuLieWr6H1gLQ9slMfq2nwAAIABJREFUGqTSrBmsWwdRUVc+4MABaYXTpw/kdk1Nq1QJWrSQut0svjlTKaX+s2gR7NkjZV133QXlysHgwbB5s9w/YwbUqAE33+xsnEqlwtNEN+HzmosJfzaGakifM4COPo5LKY98tPQfNu47xbDON1OhWKKSpWbNIDoa1q698gHjxkky27fvlcd79ZIas1WrMjxmpVTgGThwIAMHerUHJvCFhUlHhYgI2bBbvz6MGiWJbZ06sGKFbELTsgXlxzxNdI+5bosio4EBfgES+jfF+zAmpTzy647jfBa2i+51y9HxltJX3tm4sdTfJi5fiI6GL7+E9u2hYsUrz+/SBQoUkFVdpZTK6qyV35/NmkGuXHD33dJh4fBhSXbj4+X4Aw84HalSqfK068JsoBPQFmgNDObKnmbfWuuuO79/0q4LgenYuYu0G/0rRfPnYl7/xuTNlUzTjzp1pF43NFS+/+47WX1YuFC6MiTVuzfMnAlHjkD+rLOhVSmlrrJ7t3SjGTsWnngi+XMuXbpcAqYyTZbrupBOnq7ovoGMmdqLjF1bwuWOC8uRhr5KZYr4eMuz3/7J+UuxfHJ/7eSTXJCViLVrpV8uyCa0SpWgbdvkz+/VS3vqKqUUXP40rFkqffk1yVUBwKNE11r+tJZvrWWntZyzljuQMoZC1tLGWo5nbJhKXfZZ2C5W7jzB6x1u4oaSBVM+sVkzWXFYtw62bpVf3I8/LiUNyWncWHYYT5yYegAHDsDKlWl/AUqpgNO/f3/69+/vdBiZZ8UKKFZMNpspFcA8noyWjFzAJV8FopQnNuw9yYdL/6HjLaXpVrdc6ic3aSKbJMLC4LPPZPWhd++Uz0/oqfvzz7BvX/LnhIbCrbdC8+YyRlgplSXkzZuXvHnzOh1G5gkLg6ZNU14YUCpApPo32BhqG8P7xjDGGFq6jvUxhuPAEeC0MeiYGJUpTkVG89T0PyhbJC/v3FMT426nb9Gisjt44UL4+mu4914oXjz1xzz8sNwm11P3889l0lru3NI7csmStL0QpVTAGTFiRNaZinbggLQVS61sQakAkWKiawyNka4KzwH9gaXGMByZP1wUqdHNCzxjDI9nQqwqC7PWMnjmJo6fv8QnPWpTME9O9w+Cy3W6586lvKEisQoVoGXLK3vqxsRA//4ySa1NG+khee210kBdKaWCjSf1uUoFiNRWdAcjfXITj/sd7LrPACcS/fmhjApQKYCJq/ay7O8Iht5Zg5vLFvL8gQm/qENC4PbbPXtMr16y43jlSilPuOMO2cg2eDD88AMUKSLHfvpJVnaVUkGvb9++9E3afztYhYVB4cI6CEIFhdQS3duQFmKLgSeARUhSa4Ee1lICSGigd2NGBqmytj8PnObdRX/T+saSPNKooncPbt4cChWSJNXTpuadO0tP3XfekQbpK1dK6cP770N2V4eH9u0lCf7tN+/iUUoFpGLFilGsWDH3JwaDsDDZnJs9hY42SgWQFPvoGsMlIAdQxFrOGkMh4BSS6OaxlhhjyIVMS4u3lhyZFbQvaR9d/3b2Ygztx/xKfDz8+FRjCufL5f1F4uK8/4X96KMwYYJMBZozBxo0uPL+06el3nfIEHj7be9jUkopf3TkCJQuDR98AIMGOR2NSob20fVOaiu6OQGs5azr9kzCHdYS47qNdh3S+X/K56y1DJm1icOnLzKmx61pS3IhbasSQ4ZIsrt+/dVJLsjHeo0aaZ2uUiq4rFght1qfq4KE21VYY3jVk2NK+drU3/az8K+jDLmzOnUqFMncJ7/+ehkXnJp27SQhPnQIypTJnLiUUo545JFHAJjors92oAsLk4mSt97qdCRK+YQn5QavJfqzTeaYUj635fAZ3lywlWY3XEvfJpWdDid57dtLortwITz2mNPRKKUyULlybvp2B4uwMPm0KkdAViMqdZXUanTjvbiOtZaArFrXGl3/c/5SLB0/XklkdCwLn2pCsQJ+OmbSWqhYEWrX1rHBSqnAd/w4lCgBw4bB0KFOR6NSoDW63kntLdsbmRaFUi7WWl6e8xd7/41k2mO3+2+SC9LFoV07+OYbGTWcEXPfL1yQ/r2DBkHNmr6/vlJKJdD6XBWEUkx0rdVEV2W+Wb8fYm74YZ5tfQO3Vw6AVj7t28vEtBUrZGqar02aJK3NTp6UHr5KKUc8+OCDAEyZMsXhSDLQihWQNy/cdpvTkSjlMzrEWvmNqOhYhi/6m9sqFKF/i6pOh+OZli0hT56M6b4QFwcjR0rXiPnzYetW3z+HUsoj1apVo1q1ak6HkbHCwqBhQ8iVxg43SvkhTXSV35i4ai8nzkcztF0NsmcLkI51+fJBixayIc3X5s6FXbtg7Fh5nhEjfP8cSimPvPLKK7zyyitOh5FxTp2CTZu0bEEFHU10lV84cyGGL8J20apGicxvJZZe7drBjh3y5SvWyiS2KlWgTx/p6TtlirQyU0opX/v1V/m907Sp05Eo5VOa6Cq/MH7Fbs5ejOXZ1gH40WD79nLry/KFX3+Fdevg2WeldOHZZyE+HkaN8t1zKKU81r17d7p37+50GBknLEw21Nav73QkSvmUJrrKccfPXWLCqj10uKU0N5a+xulwvFepEtSo4dvyhQ8+kBHDvXrJ9xUrwn33wRdfyPhhpVSmCgkJISQkxOkwMk5YmCS5efI4HYlSPqWJrnLcp6E7uRQbzzOtrnc6lLRr317+oTh/Pv3X2roVFiyA/v2lNjfB88/DuXOS7CqlMtWQIUMYMmSI02Gk3datsG9f8vedOQN//KH1uSooaaKrHHXo9AWmrt1P19plqXxtAafDSbt27SA6GpYtS/+1Ro6UVZX+/a88HhICbdpI+cKlS+l/HqVU1nDhAjRoIDX/vXrB9u1X3r96tZRGaaKrgpAmuspRHy+XDVxPBfJqLkDjxnDNNemv0z1yRDad9e4N11579f3PPw9Hj8qQCqVUpunSpQtdunRxOoy0WbAAzp6Fjh3hu++k1KpbN/jzT7k/LExG/t5+u7NxKpUBNNFVjtlzIpLvNx7kgdvLU6ZwXqfDSZ+cOWW1deFC2bmcVmPGQGysbD5LTsuWMnL4gw9kBUYplSkaNGhAgwYNnA4jbaZNg1Kl4PvvYe9eGDIEFi2ST4k6doR586BuXcivU2VV8NFEVznmo6X/kCt7Np5oHiDDIdxp3x4OH768SuKtc+fgs8+gc2f5iDE5xsiq7j//6KQ0pTLRoEGDGDRokNNheO/0aXkD3q2bdHApUQKGDZN63TfegFWrYNs2LVtQQSvFEcBKZaSth8/yw5+H6d+iCtcWzO10OL5xxx1yO2eO1Nju2gU7d16+3b1bEtjnnpMhEybJUIzx42VTyODBqT9Ply5QuTK89x506nT1dZRSKsGcObJ/oEePK48XKQKvvgrPPCPDadq1cyY+pTKYsen5mDUI5M+f30ZGRjodRpbT5+v1rNtzkl+fb0mhfDmdDsd36taFDRuuPFawoCS4lSrJ6smxY1J+MHgwdO0qtXExMZfPCQtz/zyffiqb1VasgCZNMua1KKX+07FjRwB+CLRPUlq3ljfZO3fqm+IgYYyJstZqnYmHdEVXZbrf959i2d/HGNy2WnAluSA1tkuWSNJapQpUrSr9cBP+gbl4UTaSjRghKyxDh8qKSu7ccOCAlC54olcveO01mZ6mia5SGe7//u//nA7Be0ePws8/y+8ZTXJVFqUrurqim6ni4i2dP13FodMXCRvcnPy5s+h7rfh4mD9fNpWtWiXHbrwR/voLsnlYOv/mm5Lsrlmju6WVUlf7+GN46inYskV+v6igoCu63tFEVxPdTDVl7T5enruZ0d1D6BRSxulw/MOaNTBuHDzwALRq5fnjTp+GW2+VsocNG2RXtVJKJWjQAKKi0r5BVvklTXS9o10XVKY5fu4S7/+0jYZVitHxltJOh+M/GjSAiRO9S3IBCheWTSQnT8oGtejojIlPKcWdd97JnXfe6XQYntuzB9auvXoTmlJZjCa6KtO8u/BvLsTE8WanmhitF/ONW26RJHn1avmIUimVITp06ECHDh2cDsNzM2bIbffuzsahlMO0dEFLFzLF2t3/0n3cWga0qMqgttWcDif4DB0Kw4fD55/D//7ndDRKKafdfLNMa0zYA6CChpYueEdXdFWGi46N5+W5mylXNC8DWgbJcAh/8/bb0sf3ySdh5Uqno1FKOWnzZvnSsgWlnEl0jTFPGGP2GGMuGmM2GmNS7I9kjGlujLHJfFVPdE6vFM7JkzmvSKXmy5W72XnsPG90vIk8ObM7HU5wyp5dxnxWqCC9eQ8edDoipYJKq1ataOVtHb1Tpk+X3wn33ut0JCoAGGNyG2M+NsacMMZEGmN+MMaUdfOYpq7zDrnyrV7JnGOMMa8bYw4bYy4YY0KNMTdl2AtJQaYnusaYbsBoYBhwK7AaWGSMKe/moTcB1yX62pHk/qgk919nrb3ow9BVGhw8FcWY5Ttoe1NJWlYv6XQ4wa1IEZlZHxkpY4Qv6l9/pXylW7dudOvWzekw3LNWEt3/+z8oqb9zlUdGAV2AHkAT4BpggTEmtZWpAsBm4GngQgrnPA88BzwJ1AWOAUuNMQV9FLdHMr1G1xjzG7DJWvtYomM7gJnW2qHJnN8c+AW41lp7IoVr9gI+sdYW8DYerdHNWH2+3sDqXSdY9mwzShfO63Q4WcO8eXD33bKa87//QcWKUK4c5MrldGRKqYy2dq10cpk0CR5+2OloVAbwZY2uMaYQcBx4xFo71XWsHLAPuNNau9iDa5wHBlhrJyU6ZoDDSG72jutYXiTZHWSt/cIX8XsiU1d0jTG5gDrAkiR3LQEaunn4BmPMEWPMcmNMi2Tuz2uM2WeMOWiMWWCMudUXMau0W7o1gmV/R/D0/12vSW5m6tQJ3noLvv9eWpZVrQp58kDZstCokfTrHTfO6SiVUhlh2jSZtHjPPU5HogJDHSAnifIya+0B4G/c52WpqQSUSnLdC8CKdF7Xa5k9lqo4kB2ISHI8Akip+OkI0A9YD+QCHgKWG2OaW2tXuM7ZDvQG/gQKIkvpq4wxt1hrk5Y4YIzpC/QFKFOmDKGhoel5TSoZ8RZ2R5zjpdqGqvYAoaEHnA4pa2ncmNzTp5P3yBHyRESQ5+hR8hw9Su6ICPL99BO5p01jTaFCXNKPNlUwspZsly4Rn8d32zQGDhwIwKhRo3x2TV8zcXE0mDKFM/XqseX3350OR2WcHMaYDYm+H2etTevqRSkgDkj6iXmE6760Snhscvlepk6Lcmr+atJ6CZPMMTnR2u1IIptgjTGmIjAIeWeAtXYNsOa/ixmzGghH6kKuai7q+gsxDqR0oXnz5ml7FSpFwxb+zbjwSL77XwPqVSrqdDgqsT17oHJlGuzeDYFQc6iUt95+G0aMgH37oFAhn1wyIdH1638vli2DU6e49umn/TtOlV6x1trbUjvBGPM28JKb6yT36fh/lyCFvMxLHud7GSWzN6OdQN45JH2XUIKrs/7U/AZcn9Kd1to4YENq56iMs+ngab78dTc96pXXJNcfVaoEjRvDN9/IxhWlgsm2bfDmm3DmDPz8s88u26tXL3r16uWz62WIadOgYEFo187pSJTzRgE13HytA44in7QXT/J4b/OypI66btOb76Vbpia61tpoYCPQOsldrZHuC54KQUoakuUqgq6V2jkqY8TExfP8zE0UL5Cboe2qu3+AcsaDD8Lff0N4uNORKOU71sLjj0OBApA/PyxJuh0k7WJiYoiJifHZ9XzuwgWYNUs6ruTVPRFZnbX2hLV2m5uvKCQniyFRXuZqLVYD7/KypPYgyW7i6+ZBujqk57pec6KP7odAL2NMH2NMDWPMaKA08DmAMWayMWZywsnGmIHGmLuNMdcbY24yxrwL3A18kuic14wxbY0xlY0xIcBXSKL7eWa+MAXjVuxm29FzvHV3Ta7Jk9PpcFRK7r0XcuaEKVOcjkQp3/n6awgLg/ffh5YtYfFin31q0bp1a1q3TrpG40d++AHOnoWHHnI6EhVArLVnkJzpA2NMK9dG/m+ATcCyhPOMMduMMQMSfV/AGBPiyrmyAeVd35d3Xdciq8pDjDGdjTE1gUnAeWBaJr08wIEaXWvtt8aYYsDLSL/bzUA7a+0+1ylJ++nmAkYgxcsXgC1Ae2vtwkTnFEZqbksBZ4A/gKbW2nUZ9kLUVXYdP8/o5Ttod3Mp2t6Unhp2leGKFoX27eWjzvffl+bySgWyEydg0CDpLNK7N1y6BPPnw86dcH36q9j69OnjgyAz0DffSGcVrc1V3nsGiAW+BfICy4GerjLQBNW4srzhNqT1a4I3XF9fA71cx953XW8sUAQpO21jrT3n+5eQskzvo+tvtI+ub8THW7qNW8M/EedZ+mxTShTUoXR+b9YsmaK2eDG0aeN0NEqlzyOPyCcU4eFw002XE9yPP4YBA9w/PpBFRECZMpLoDx/udDQqg/myj25W4MgIYBV8pq7bz/q9p3ipfQ1NcgNF+/ZQuLCWL6jAFxoqAxIGD5YkF6R/dOXK8kbOB6KiooiKivLJtXxuxgyIi9OyBaWSoSu6uqKbbkfOXKD1hysIKVeYbx6th+wFVAGhb18pX4iIkM07SgWaS5fgllsgOho2b4Z8+S7f98QTMHkynDyZ7smACe26/LLv+m23SS3yxo1OR6Iyga7oekdXdFW6WGt5ec5m4uItw+65WZPcQPPggxAZKWODlQpE778P27fDp59emeQCtG0rf79XrUr30/Tr149+/fql+zo+t3WrJLi6mqtUsjTRVekyf9MRlm87xnNtbqB8sXzuH6D8S+PGUL68li+owLRjB7zzjgw+ueOOq+9v0QJy5PBJ+UK3bt3o5o8DVr75RjaT9ujhdCRK+SUtXdDShTQ7GRlN6w/DKFskL7OfaET2bLqaG5BefFFWxQ4dAh0JrPyJtfDnn7IqmzOnfOXKdfnPjz0G69fLkIjrrkv+Gs2aSdutP/5IVyhnzpwBoJCPJq35RHw8VKwINWvCwoVuT1fBQUsXvKMruipNrLW8MGsT5y7G8l7XWprkBrIHH5SNLDNmOB2JUpcdPQpdusCtt8onD/XrQ+3aktRVqyYbzZYvh3ffTTnJBSlfCA+XOvR06NSpE506dUrXNXwuLAwOHICePZ2ORCm/lel9dFVwmLH+AEu3RvBSuxpUL3WN0+Go9LjxRkkmpkyBp592OhqV1VkLU6fCU09BVBQMGwZ16kBMzOWv6Gi5LVoU7ror9eu1bQsvvQRLl8qbujR66qmn0vzYDPPNNzLy198ScKX8iJYuaOmC13YfP0/7MSupXaEw3/SuTzZdzQ18H30Ezz4rY4Gr6+hm5ZDDh+F//4MFC6BBA5gwIf1/H+PjoVQpSXi/+cY3cfqDqCgpNbr3Xvk5qSxDSxe8o6ULyisxcfE88204uXJkY8S9t2iSGyy6d4ds2WQlTanMZi1MnCifLixfDh9+CL/+6ps3XdmyQevWsGSJJL1pdOLECU6cOJH+eHxl3jw4f167LSjlhia6yiujl+3gz4NneLfzzVxXKK/T4Shfue46aNVKyhf8/VOe0aOl1MLf41Se+fNP6ZjQuzfUqgWbNsEzz/h2LHXbtnDsmDxXGnXt2pWuXbv6Lqb0+uYbKFdONtsppVKkia7y2Pq9J/k0dCdd65Sl3c2pbP5QgenBB2HvXp/0HM1QixbJ5qK9e52ORKXH9u3ySUJICKxbJ6N6Q0NlopmvJYy4Tkebseeee47nnnvORwGl09Gj8loefFBWrJVSKdL/Q5RHzl6MYeCMcMoWycfrHW9yOhyVEe65Rxru+3sdY3i43P72m7NxqLTZuxceeUTKFBYsgJdfhj17YMCAjEvaSpWS6WnpSHQ7dOhAhw4dfBhUOkyfLmUYWraglFua6CqPvDZvC0fPXmRU9xAK5NZmHUGpQAHo2lVGAv/7r9PRJO/IkcttotatczYW5Z3Dh2Uk7w03SKI2cKAkuG+9BYULZ/zzt20rn1acP5+mhx89epSjR4/6OKg0+uYbGftbo4bTkSjl9zTRVW798Odh5vxxiCdbVqV2+SJOh6My0uDBkgiMGeN0JMlLWM3Nl09XdAPJgQPSA3f8eOjTB3btgpEj4dprMy+Gtm2lJdkvv6Tp4d27d6d79+4+DioNNm+W4Re6mquUR3RpTqXq2NmLvDznL2qXL8yAFhlQO6f8S82aUsIwZoy0G/OnKVBwOdHt3l1WnmNiZEKW8l8XLsDdd0s7rN9/h5tvdiaORo3kDdLixZCGEoQhQ4ZkQFBeslamGGbPLv8PKKXc0hVdlao35m/lYmw8I+8LIUd2/euSJbz0Epw+DZ9+6nQkVwsPh0qVZHPRxYvw119OR5T1nDwJW7d6dq61Mqb3jz+kdZ1TSS5A7tzQvLm0GfOWtdyxZg13OFm6EB8P/ftL2cLgwVCihHOxKBVANHNRKfp5WwQ//nWEp1pWpVJx7U2dZdSpA+3aSS9TfxumEh4uu/Tr15fvtXwhc61fLy3AatWSVX93Ld4+/FAS3DffTNMqqs+1bQs7dkhtsDc+/JADb77JgUcekdZncXEZE19K4uNlkMZnn8Hzz8u0OKWURzTRVcmKio7llblbuL5EAfo2reJ0OCqzvfwynDgB48Y5Hcll589LkhISAhUqSH2nJrqZ5+uvoUkTKRVp00bGRT/2GFy6lPz5S5dKUtali3xK4A/atpVbb7ovrFgBL7zAQ8WL81CZMjBqlIzcPXs2Y2JMKi5Oegx/+aX8fzl8OBgd1KOUpzTRVckavWwHh05fYFjnm8mVQ/+aZDkNGkDLlvDBB1Ii4A82bZIVxJAQ+Ye+fn3tvOBOfLz8jE6eTPs1YmKkQ0KvXlLnun69tAV76SX46iv4v/+TYQyJ7doF3bpJC7FJk/wnMbvhBnmT9NNPnp1/5Ii8jipVePnLL3l50iQp6fnpJ/lZ7NuXoeESGws9e8qbjDfekA4V/vKzVCpAaAajrrL18Fm+XLmHHvXKUbdiUafDUU55+WX5h37iRKcjEQkb0UJC5LZ+fdi2Dc6ccS4mf7ZxIzRuLD+nEiVkDO7nn8uwAU+dOCGroKNHS7K7eDEULy79bt9+G2bMkA1mt912+b/P+fOy+Qxg7lxpW+cvjJHY5s2D119PfSRwTAzcd5+s3M6eTatOnWjVqhX06ydDSw4cgHr1YO3ajIk1Jgbuv182XQ4bBq++mjHPo1Sws9Zm6a98+fJZdVlsXLzt9MlKW+etJfZU5CWnw1FOio+3tmFDa8uXtzY62ulorH3sMWuLFpW4rLV28WJrwdply5yNy99ERFjbp4+1xlhbooS1H39s7ZAh1l5/vfy8jLG2cWNrP/zQ2h07rD13ztq4uKuvEx5ubcWK1ubObe3XX6f8fBs2WFu2rLX58ln77bfWdu5sbbZs1i5dmnGvMT2ioqzt2VN+Fh07WnvmTPLnPfusnDNtmrXW2l27dtldu3Zdvn/rVmsrV5afz/Tpvo/xnnvk+UeM8O21VcADIq0f5E+B8uV4AE5/aaJ7pcmr99gKLyywc/846HQoyh8sXCi/JiZMcDoSa+vWtbZly8vfnzwpsb3zjnMx+ZPoaGtHj7a2UCFrc+SQRO306cv3x8db+9df1r7+urW1asnPLvFX/vzWlixpbZUq1t5yi7V581pbpoy169a5f+4jR6xt0ODytUaOzLjX6Qvx8fKzyp7d2urVrd227cr7v/tOXseTT/53qFmzZrZZs2ZXnnf8uLVNmsi5/frJ38m0xvPXX/Jza9PG2jx55JqjR6fteiqoaaLr3ZeRn1nWlT9/fhvpbzvLHRJx9iKtRoYRUr4wk3vXw2gtmLJWPpY+e1bKBLJndyaO2FgoWFAma40cefl4tWpQvbp8FJ2VLV8um8O2bJEShdGj3U/N2rFDhiecOSPlBom/IiOlh/J778n4XE9cugQvvCCb1d5/PzBqSUND4d57IToapkyRzhDbtkHdutIKLTQUcuUCICwsDIBmzZpdeY2E1/3xx1CsmLz2nj3djzM+d05qnZcska/Dh+V4jRqy2e+eeyDpcykFGGOirLXaCslDmuhqovuf/lN/Z9nfESx5pikViun/Q8plzhzo3FlqBXv0cCaGLVtkmMXkyVdOhOrZU5KEI0cCI7HytY0b4cUX5WdQsSJ89JF0BMiKP4u02r9fksrff5c62Jkz4fhx+b5sWc+vEx4ufW5Xr4aGDWHs2Mv15AliY2HZMumFO2eODNMoUkTenLRpI7fly/v29amgo4mud3QzmgIS9cz9v+s1yVVX6tQJbroJ3nkn9c07GSnpRrQE9etDRIRsDAoGcXGyuujO9u2yEnnbbbBhA4wYAX//LRutNMn1TvnysHKlvIF6801Z0Z0x46okd/v27Wzfvj3l64SEwK+/yubNHTukH/VTT8nwlfBwmTRYtizceadsZnv4YTn/+HH49lt49FFNcpXKALqiqyu6HDgZxb2fr+GavDlY8GQTbSemrjZ9uuwAnz1bVr8y2+DB8tHwuXNXjvxdv152vn//PXTtmvlx+VJ8vHQ4CAuTgQx168prq1tXPs7Onl0S+jfekJZdefJI8vTcc/43qjkQWSs/17x5kx2v27x5cwBCQ0PdX+vUKXjlFRnwkDOnlDfkzAnt28unEO3ayaQ2pdJAV3S9o4luFk909/0bSY9xa4mMjmP6Y7dzY+lrnA5J+aO4OEm2ChaUFcTMXjVs3VqShw0brjweHS0xPfWU9PwNZJ99JjXI3bvLKt/69ZeHEuTPL8nv779LQtavn5Qs6BjYTLN69WoAGjZs6PmDfv9d+u7WqSOtyooVy6DoVFaiia53NNHNwonu3hOR9Bi/losxcUzpU5+bSuuqkErFhAny8eqiRXDHHZn3vNZKQtepk0yHSur222XD0IoVmReTr+3bJzXIDRpIr1pjZIV3xw4Z+LDnPw7aAAAfLUlEQVR+vSRN1avLSmGFCk5HrJRyiCa63tFEN4smuntOyEpudFw8U/vUp8Z1upKr3IiOhqpVpY7w118zb1X34EEoV05KFwYMuPr+p5+WBPjMGciRI3Ni8iVrpWRhzRrYvFmTWD+1efNmAGrWrOlwJCqr00TXO1qMmQXtOn6e7uPWEB0Xz7THNMlVHsqVC55/HlatytzV05Q2oiWoXx+ioqQzQyCaOBGWLpVWXprk+q0BAwYwILk3Wkopv6YrullsRXfnsfPcP34tcfGWaY/dTrVSBZ0OSQWSCxegUiWpF12yJHOe8+235eP6s2elHjepnTvh+uvhiy+gb9/MiclXDh2Sjha33CI9bd31XlWOWb9+PQB169Z1OBKV1emKrnf0t2oWsvPYOXqMX0u8hRl9NclVaZA3r+z0X7pU6kYzQ3i4lEwkl+QCVKkCRYtKLWsgsRYef1xKQr76SpNcP1e3bl1NcpUKQPqbNYv4dcdxuny2BmthRt/6XF9Sk1yVRv36SZP7d97JnOcLD0+5bAGkVrhePfjtt8yJx1emTZPJWG+/LYm88mvh4eGEJ5TRKKUChia6Qc5ay7gVu3h4wjquK5SHWf0aULWEJrkqHQoWlA1g8+bBX39l7HOdOQO7dqWe6ILU6W7ZIn12A0FEhLREu/12+Vkqvzdw4EAGDhzodBhKKS9pohvELkTH8fSMcIYt3MYdNUsxq19DnXqmfOPJJ6FAAXj33Yx9nk2b5NaTRNdaGYkbCAYMgPPnpWVb9uxOR6M8MGrUKEaNGuV0GEopLwVgLx7liYOnoug7eSN/Hz3L4LbVeKJ5FYyOBlW+UrSoDDcYMUImdV1/fcY8T8JHxbfemvp5CbWTv/0GrglWfiMmRvrk7twpX+HhMHMmDBsmQzhUQAhx92ZLKeWXtOtCEHZdWL3rBAOm/UFMXDxjetxKi2o6PUllgIgIqFhRRgN/9VXGPMejj8L8+fJc7t6oVa0q3QtmzfL+ecaNg6ZNZSCDL2zeDC+8ANu3w969MlkuQf780jd3xowrxxkrv6ZdF5S/0K4L3tFEN4gSXWstE1btZdjCv6lUPD/je95GpeL6/4LKQE8+CZ9/LnW05cv7/vp16sjYVE9amd1/v/T3PXjQu+fYvl0S3NtukxXh9HY/iImB2rXh8GEZXVy16pVfJUtm/ghllW7NXZ8UhIaGOhqHUproekdLF4JEVHQsQ2b9xQ9/HqbNjSX5sFsIBXLrf16VwQYPlkT3gw9kcpkvxcTIyqinm7Xq14fp06U3bZkynj/PtGlyu2GD/PnBB72PNbEPPpC4582Djh3Tdy3lNz755BOnQ1BKpYGu6AbBiu6eE5E8/s1Gdhw7x3NtqtGvWRWyZdMVI5VJHn1UEsS9e2W10lc2bZJShKlTZbXWnbVroUEDmD0b7rnHs+ewVuqLK1SQDg8REbLCmy9f2mLesQNuvhk6dIDvv0/bNZRSKhW6ousd7boQ4JZtjaDjJys5du4iX/euR/8WVTXJVZlryBAZevDBB769rqcb0RKEhEjNqzf9dNetk7KLBx+Ejz6SsoeRI72PFSRp/t//IE8eGDMmbddQfmv16tWsXr3a6TCUUl7SRDdAxcVbRi7ZTp/JG6hYLD/zn2xMk+uvdToslRVdf70kimPHStmAr4SHyyS2G27w7Pw8eWQFeOVKz59j6lTInRs6d4YmTaBLFxg+XOprvTVpkozxfe89uO467x+v/NqLL77Iiy++6HQYSikvaelCAJYunI6K5ukZ4YT9c5x765Tlrbtrkien9uJUDtqzB6pVg969pWbXF1q2hMhI71Zo330XXnwR/vjDfe/d2Fip5W3a9HKZwa5d0vLroYe86yRx7JhsaLvpJggL03G+QWj79u0AVKtWzeFIVFanpQve0d/GAWbzoTPc9fFKVu86wTv31OT9rrU0yVXOq1QJ+vaV5HDnzvRfz1r3o3+T06+fTG577z335y5bJgnqAw9cPlalikwsmzhRkmVPPfOMDIAYN06T3CBVrVo1TXKVCkD6GzmAfLt+P50/W01cvOW7/zXggfoVdAiE8h8vvyw1sq+9lv5r7d8Pp055n+gWLgyPPw7ffSers6mZOlXOv/POK4+//LIMxHjuOUm43Vm0SDbjvfiiDoAIYmFhYYSFhTkdhlLKS5roBoCLMXG8MHMTL8z6i3oVi7LgycbcWr6I02EpdaVSpaQV2PTpl0f3plXCph9PN6IlNnAg5MiR+qayqCiYMwe6dpUa3cQKF5Zpb7/8IsMqUhMZKavI1avD0KHex6oCxmuvvcZrvngTp5SfMcbkNsZ8bIw5YYyJNMb8YIwp6+YxTV3nHTLGWGNMryT35zTGvGeM2eS65hFjzDRjTAY0XE+d1uj6eY3ugZNR9Ju6kc2HzjKgRVWeaX0D2bWrgvJXp05JGUOTJu6TxJTMmyftxEqWhC1bZEOat/r2hcmTZfRuci3PZsyAHj0kmU1uZHBMDNSqJRPNNm+GXLmSf55BgyShXrFCXrMKWrt37wagcuXKDkeisjpf1+gaYz4DOgEPA/8CHwKFgTrW2rgUHtMOaAz8DkwGnrDWTkp0fyFgJvAlEA4UAkYCxYBa1tpYX8Xvjia6fpzohm4/xsBvw4mLt3x0XwitbvRhj1KlMsqwYfDSS7BqFTRs6PnjrJUWX4MGyZSyH36QVeK02LFDNscNGSLxJNWhg9QA79uXck3tjz/CXXfB6NFStwtw7pwMlvjtN/n64Qfo0+f/27vzcDnqOt/j7w9kYwkS1gQIRCRgWBwcFkFAAgNX1IEHwiAREaIOXMxFBkbGDZeAKODINYALi/MYQkQiIQyOw7DeREQwENRAEMJiFsAECQmGhJDl5Hv/+NUhlT6nz+nu08vpPp/X89Rzqqt+VfXtb+qcfLv6V7+CG26oLE4zszJVs9DNCtLXgE9HxM+yZcOBhcBHIuLeEvaxEjg/X+gWabcv8DSp0H2qp7GXyoVuLyx017VtYOIDz/GjmS+yz86Duf7MgxjhR/las1i1CvbcM/VXnTGjtMfdrluXHid8ww1piK/Jkyt/aEO7006D++9P/X232Wbj8qVL0/BfF10E3/1u8e0j4MMfToXtKaekwvZPf9rYb3evveDoo9MV3Xe9q2exWq/3wAMPAHDcccc1OBLr66pc6B4LPAjsFBGv5ZY/DUyLiG7765RR6B4GPAoMj4gyn9VeuT7/jNjtttuuVz27fF3bBhYtW82gteu5/NAB7LLtBhbMfZwFjQ7MrAy7nn46I6+7jjnf+x7LDzmky7abr1zJfpdeynazZ7PwjDOY/9nPpgc59NDg447joGnTePGLX+SlsWPfWb7LXXex9/r1zN57b1Z287u/1dixHDRzJm3TprFi1ChWnH02b44axYp99mF9e3FbzugM1rQuvvhiACZOnNjgSMzoJ2l27vWNEXFjhfsaCrQBSwuWv5qtqwpJA0hdF/6rnkUu+Ipur7qie9cfX+Frd84FwRVjDuAf37dLo0Myq8yaNanrwA47wOOPF7+qu2BB6h4wb166mvuZz1Q3juOPT/1858/feNPZkUemvsRz55Z2tXnVqnR12SOc9GkvvfQSAMOHD29wJNbXlXJFV9LlwCXd7OoYYBdSH9v+kSsIJc0A5kXEeSXE0+UVXUn9gFuB/YAPRcTr3e2zmjzqQi+wcs16vvCLOfzLbX9k76GDufuCo1zkWnMbODCNXPDEE3DHHZuuW7483XB20UVw6KHpaWr33Vf9IhfgS1+CxYvhllvS6wULUt/hT36y9MJ1q61c5BrDhw93kWvNZCIwqpvpMWAJsDmwQ8H2O5Gu6vZIVuT+HHgf8A/1LnLBV3QbfkX3qZf/xud//nsWLXuL848dyQXH7kW/zf35w1pAWxsccABs2JD6ws6YATNnwpw5qZ/roEHpqWTXXJOG56qFCDjkEFixAp55Jj1I4pJL0hXeESNqc0xrSffccw8AJ5xwQoMjsb6uRjejjYuIW7NluwGL6OHNaJL6A7cB+wOjI2JxNWIuV0MKXUnjgX8DhpHuwLswIn5TpO1oYEYnq0ZFxLO5dqcC3wLeA7wIXBIRd3YXSyMK3Q0bgkf//Dq3zlrEvU8vYcfBA/n+6Qdy2J7b1zUOs5qbPj3dXAbpKu8HPwjHHJOG9Dr00I5j2NbCtGnpxrTbb08PsxgyBB5+uPbHtZYyOhuGrjfd02F9U42GFzuJTYcXG0JueDFJzwI/iIgfZK+3BvbKdvEIcCXwS2BZRCzKruROAw4BTgT+kjvk3yJidbXi707dC11JpwNTgPHAw9nPTwP7RsSiTtqPJhW6+wHLcqtey/0DHA78BvgmMB0YA1wKHBERs7qKp56F7rJVa7njiZe59bFFzF+6im237M9pB+3G+NF7MWSrIuN0mjWziPSUsmHDUmE7aFD9Y2hrSyNArF4NL78MP/whjB9f/zisqS1ZsgSAoZUOeWdWJTUodAcB/w6cAWxBGoVhfES8lGsTwKURMSF7PZrOL0LeHBHjJI0A5hc55Ke7G6GhmhpR6M4CnoyIc3LLnicNY9Hh0UK5ZO4YEYV3Bba3mQpsFxHH55Y9QCqGP9FVPLUudCOC2QuX87PfLeTup5awtm0DB+8xhE8etjsf2X8Yg/pvXrNjm1nmppvSQyT69Ut9dnco7I5mZtYcql3otrq6Di+WDS9xEPC9glX3Ad2NLD9b0kDgT8DlEZH/JHE4cF1B+3uB83sQblW88sZqTrv+UQYP7McnDh3OGR/Yg32GDm50WGZ9y1lnpZvjDj7YRa5V5L+yJ/2deOKJDY7EzMpR73F0dyDd3Vd4J9+rQLFRuBcDnwMeBwYAnwIelDQ6Ih7K2gwtss9Ov2OSdC5wLsCAYo/2rJLdhmzJT8cdwgf23I4tB/T5YYvNGmPgwDTMWU8fQmF91tVXXw240DVrNo2qvAr7S6iTZalhxDxgXm7Ro1nfj4uBh/JNy9jnjcCNkLoulBp0pY557061PoSZdWfYsEZHYE1s2rRpjQ7BzCpQ70J3KekJHIVXWssdr20WMDb3ekkV9mlmZtapHdzlxawp1XXA1ohYCzwBHF+w6njS8BSlOpDUpaHdo1XYp5mZWaemT5/O9OnTGx2GmZWpEV0X/i9wi6THgN8C55EeQXc9gKTJABFxVvb6QmABabzdAcCZwMnAqbl9XgM8JOkrwJ3AKaRH2x1Z+7djZmat7tprrwVgzJgxDY7EzMpR90I3IqZK2h74GumBEXOBj0bEwqzJ7gWbDCCN0rArsJpU8H4sIu7O7fMRSWOBy0nj574InN7dGLpmZmaluOuuuxodgplVwI8AbvAjgM3MzMxK5XF0y1PXPrpmZmbNaOrUqUydOrXRYZhZmXxF11d0zcysG6NHjwZg5syZDY3DzFd0y+NC14WumZl146233gJgSz90xBrMhW55/KguMzOzbrjANWtO7qNrZmbWjSlTpjBlypRGh2FmZXLXBXddMDOzbriPrvUW7rpQHhe6LnTNzKwb69atA6B///4NjsT6Ohe65XEfXTMzs264wDVrTu6ja2Zm1o1JkyYxadKkRodhZmVy1wV3XTAzs264j671Fu66UJ4+X+hK2gCsrtLu+gHrq7SvVuB8dOScbMr56Mg56cg52ZTz0VFfyskWEeFv5EvU5wvdapI0OyIObnQcvYXz0ZFzsinnoyPnpCPnZFPOR0fOiRXjTwRmZmZm1pJc6JqZmZlZS3KhW103NjqAXsb56Mg52ZTz0ZFz0pFzsinnoyPnxDrlPrpmZmZm1pJ8RdfMzMzMWpILXTMzMzNrSS50yyBpvKT5kt6W9ISko7ppf3TW7m1Jf5Z0Xr1irYdy8iFpmKRbJT0rqU3SpDqGWjdl5mSMpPskvSbpTUmzJJ1Uz3hrrcx8HC3pEUmvS1qdnSsX1zPeeij370huuyMlrZc0t9Yx1lOZ58hoSdHJ9N56xlxrFfxfM0DSZdk2ayQtknRBveKttTLPkUlFzhE/GaqPcqFbIkmnA9cA3wHeDzwC/I+k3Yu0fzdwd9bu/cAVwHWSTq1PxLVVbj6AgcBS4EpgVl2CrLMKcnI08P+Aj2Xt7wbuLLXw6e0qyMdK4FrgQ8C+wOXApZLG1yHcuqggJ+3bDQEmAw/WPMg6qjQfwH7AsNz0fC3jrKcKc/Jz4ATgXGAf4DTgyRqHWhcV5ONf2PTcGAb8GfhF7aO13sg3o5VI0izgyYg4J7fseWBaRHylk/ZXAWMiYmRu2U+A/SLi8HrEXEvl5qNg218BSyNiXG2jrK+e5CTX/jHgNxHxhRqFWTdVysd0YE1EfKJGYdZVpTnJ8jAHEPBPEbF/zYOtgwr+ro4GZgA7RsTSugVaRxXk5H8BtwPvacWc9PTviKQjgIeBIyLikdpFar2Vr+iWQNIA4CDgvoJV9wEfLLLZ4Z20vxc4WFL/6kZYXxXmo6VVMSeDgeXViqtRqpEPSe/P2v66utE1RqU5ya5oDyVd4W4ZPTxHZktaLOlBScfUJMAGqDAnJwOPA/8q6WVJz0u6VtLWNQy1Lqr0d/Uc4GkXuX2XC93S7ABsDrxasPxV0n9AnRlapH2/bH/NrJJ8tLoe50TS/wF2A26pbmgNUXE+sv+s1wCzgR9FxPW1CbHuys6JpAOAbwKfjIi22oZXd5WcI4uBzwGnAmOAecCDkj5UqyDrrJKc7AkcCfwdKS/nk7oxTKpNiHXVo7+rkt5F6sZxU/VDs2bRr9EBNJnCfh7qZFl37Ttb3qzKzUdfUFFOsr7b/w6MjYiFtQisQSrJx1HA1sBhwFWS5kdEKxT/7UrKiaSBwG3AxRExvx6BNUjJ50hEzCMVt+0elTQCuBh4qBbBNUg5vzebZevOiIi/AUg6H7hX0s4RUVgkNqNK/685k1Qot9LfDyuTC93SLAXa6PgJcic6ftJst6RI+/XA61WNrv4qyUerqzgnWZF7C3BWRPyyNuHVXcX5yBV1T0naGZhAa/xHVW5OhpFuyvuppJ9myzYDJGk98NGIKPxKt5lU6+/ILGBstYJqsEpyshh4pb3IzTyT/dy9i+2aQU/PkXOAOyJiWbUDs+bhrgsliIi1wBPA8QWrjifdAdqZR4HjOmk/OyLWVTfC+qowHy2t0pxI+jgwBRgXEdNqF2F9VfEc2Yw0YkfTqyAnrwAHAAfmpuuBF7L5pv5dq+I5ciCp2Gt6Febkt8AuBX1y985+NvW3Qz05RyR9gNSdw90W+rqI8FTCBJwOrAX+GRhFGu5kJbBHtn4yMDnX/t3AKmBi1v6fs+1PbfR7aUQ+smXt/1k/BPwym9+30e+lgefIWGAdaTicoblpu0a/lwbl4/PAPwIjs+mzwArgyka/l0blpJPtJwBzG/0+GniOXEi6+WokaYixK0hfYY9p9HtpYE62Bl4ijbywH3AEMBe4vdHvpRH5yG33E+A5stGlPPXdyV0XShQRUyVtD3yN9JXiXNJXh+2fmHcvaD9f0keB75NunvgLcEFE3FHHsGum3Hxk/lDw+kTSFYcRtYqznirIyXmk7kMTs6ndr4HRtY229irIx+bAVaTzYT3wIvBl0lXMllDh703LqiAfA4DvAbsCq4GngY9FxN11CrnmKvi/ZqWk44DrSKMvLAf+k/S70/Qq+Z2RNJh0IeGyiOjr9430eR5H18zMzMxakvvompmZmVlLcqFrZmZmZi3Jha6ZmZmZtSQXumZmZmbWklzompmZmVlLcqFrZmZmZi3Jha5ZLydppKQfSHpG0kpJb0p6VtJNkg7LtVsgKSQtaGC47bFMymIJSSNyy3eW9DNJiyW1ZesnShqRaz+phnFtK2lCNp1catz1Iml07vjdTROybdpfz6x3vN2p5b9rOf9WBXmtahxm1rv5gRFmvZikTwM/puNjcPfJph1JT4pqFteQnnTUKNsC38zmbyYNrG9mZi3Kha5ZLyXpWNJjLDcjPeb028ANwF+BPYB/YuMz7XuViBgHjOtk1UHZzzeAPSNieW6dahxWt7qIu17Hn0kuD5LGAT/NXt6cxVd1kgZFxNu12LeZWSO564JZ73UFG39Hr42Ir0fEyxGxNiKej4grgHO62oGkAyVNl/SCpBWS1klaki07uKDtuyVNlrRI0tuS3pA0N/uKeKdcu3MkzZa0TNIaSa9Iul/S2bk2m3yt3P7VMbBX1mRbYFm2flxXX3FL+ntJP8+Os1bSUkkzJB2ard9a0s2SnpL0evYe35D0kKTTc/uZAMzP7frswmN20eViK0mXSnpa0mpJb0n6g6R/ldQv126T9yHprCyHq5W6npxNDUk6VtLvsuO9KOmLkvKF84RcfKdI+g9JS0mP021vM0rSLbl8/1XSNEnvKzhWSedLwTYfl/RkV/mQdJSkX0p6LXe+3lZ4/C5ysEsW78rsfPgxMLhI27Lfg5k1mYjw5MlTL5uAnUhXcdunXUvYZkHWdkFu2diC/eSnVcCoXNunu2i7f9bmtC7aTMvta1Ju+QhgdBfbjcvatL+elNvPKcC6YttlbYZ2se8Azs7aTeiizaTO4s6WbQU80cW2dwObZW3z72N5kfZHlnEejOssLwVt2tcvLZKrM3NtJxS0f6ddtv5I4K0ica8GjirzfMnnY0l3+QDOBNqKtHsbGF3sHMuWbQE808m2f+ksj6W8B0+ePDX35Cu6Zr3TiNz8ioh4pcL9/B74MDCM1M93G+Bz2botgf8NIGl7YN9s+bWk4m474BDg68DfsnUfyn6uJPURHkjqRvFx4J5iQUTEzIgQsDBbtDAilE2TOttG0hbATWzsYvUNYGdgB1LB/eds+Zukfr8jsvc0CPggqWADuCiLYQLw7twhbs7FMK5Y7MCFwN9n8/eScrknKbcAHyF9oCi0LTA++3lVbvmnujhWT2wPfBcYApxfwvEEnEDKWfvV0ptIxeJCUjeTgcD7gddIef0hlHW+5O1MF/mQtBVwHelbjPWkDznbAOdl7QaSuu505Szgvdn874DdSN8ivNHhzVf2HsysybiPrllrWwJ8FphIKgS3KFi/T/ZzOakY2JZUuL1JujI2JyIuz7Wfn/3cCvga6UrnM8B9EVHtwuAIUvEGMDMivpVbNy03/xap+J0KjCJ9TZ3v77sPPfOx3PxXImIJgKTL2Hgz20eBWwu2eyIifpy1nQJ8KVu+Rw/jKeZV4BsR0SbpZuAH3Rzv6oi4N5t/StJINhaJe5D+bQsdIGkoqZ94KedLXnf5OCLbH8DdEdGe2xsknQccCOwtaa+IeKHIMY7NzV/R/gFR0tWk/u55pZ7zZtbEfEXXrHdakJvfRtIuFe7nF8AXSQVgYZFL+7KI2EC6svYyMBK4BJhCKoCekjQ8a/8j4Hagvf1E0lXOVyV9ucIYi9k5N/+nLtp9iXSl8QOkK4CFN7UN6mEcO+bmF+XmF+bmO+vPOS83v6qK8RTzYkS0lXG8PxS8LrVP6vZlnC953eWjWJ6h+1y/E1tu/uUi80BZ57yZNTEXuma9UET8FXgst+jfOmuXvxGqk3VDSN0WIF3t2w/YnI1fUxce81fA7qQroCcBl5H6S+5PunpLRLwdER8nfcV7JPAZYBbpa+XvSNq1tHdYkldz86O6aJfvNnAyMDDrJvF6J22jgjhey83vXmT+r51st66Hxy3XO8eLiFKOt7rgdf493J/r1vHOROqL/HR2jG7Pl2Lx0Xk+iuW58HVnuW63NDe/W5H5jUGU/x7MrMm40DXrvS4hXTkFuCC7Y34XSf2VHiLxVVKfymLWs7GgWA+sIH3F/63OGku6DvgHUv/be4A7gDXZ6t2zNqdKOh/YFZhDuro7p30XFCkoKvRbNharx0j6qqQdJQ2RdLKk9v7C63PbvAH0l/R1Nr261y5f/I7M+oV251e5+W8rPfRiBKnPcLv/LmE/vVpEPA88l708XtKFSg/Y2FbSwZK+AdzW3r6U86VMvyV1JwD4iKSTlEbUOIfUTxhgXhfdFgBm5Oa/LGlXSe8BvtBZ4xq8BzPrZVzomvVSEfEA6WaxtaTf1W8Cr2SvnyONqzuki+3fBB7MXu4KvES6SrpvkU0+B9yfO8Yc0o1KkLonQLqyeh2pK8Gb2XRutm4x8GQZb7FLEbGaNHxaeyH7bdLVvGXAnaQbwsjm280kFS0X0MkNSBGxknSnPaQb1lZmQ22N6yKUa9j0xrMlpL7K7WMC/w+pf3ArOJc0ugHA90mF53LgceBSNu1OUsr5UrKIWAV8nvThrj9wF+n8ujFrsoaNN6YVMxl4Nps/nNQt4QU27RaRV9X3YGa9jwtds14sIn4C/B2pb+xzpK+bV5H6O/4HcGU3uziTVIQtJ91FPoXiTya7EniYVEyuJ93k9XtS0XhN1uZB0k1XL5AKyjZSgXsbcHRWnFZNRNxJ6nt7G2mIqPWkQvfXbOy3exXwHVKxsjpbdyzF75r/FPAQ6Qp3KTGsIo02cRnpZqU1pGLwj8DFwElZf8+mFxG/JhXwk0lF4jpSvp8kfcD5aq55KedLucf/GWkoul+Rrr6vJ304+wVwaKQHanS1/WrgOGA66ffkDdIDN4qNN13192BmvYtK68plZmZmZtZcfEXXzMzMzFqSC10zMzMza0kudM3MzMysJbnQNTMzM7OW5ELXzMzMzFqSC10zMzMza0kudM3MzMysJbnQNTMzM7OW5ELXzMzMzFrS/wfaH+lcGD1F/wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 720x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Plot average odds difference\n",
    "fig, ax1 = plt.subplots(figsize=(10,7))\n",
    "ax1.plot(thresh_arr, bal_acc_arr)\n",
    "ax1.set_xlabel('Classification Thresholds', fontsize=16, fontweight='bold')\n",
    "ax1.set_ylabel('Balanced Accuracy', color='b', fontsize=16, fontweight='bold')\n",
    "ax1.xaxis.set_tick_params(labelsize=14)\n",
    "ax1.yaxis.set_tick_params(labelsize=14)\n",
    "\n",
    "\n",
    "ax2 = ax1.twinx()\n",
    "ax2.plot(thresh_arr, avg_odds_diff_arr, color='r')\n",
    "ax2.set_ylabel('avg. odds diff.', color='r', fontsize=16, fontweight='bold')\n",
    "\n",
    "ax2.axvline(np.array(thresh_arr)[thresh_arr_best_ind], color='k', linestyle=':')\n",
    "ax2.yaxis.set_tick_params(labelsize=14)\n",
    "ax2.grid(True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Threshold corresponding to Best balanced accuracy: 0.5500\n",
      "Best balanced accuracy: 0.7696\n",
      "Corresponding abs(1-disparate impact) value: 0.4911\n",
      "Corresponding average odds difference value: -0.1083\n",
      "Corresponding statistical parity difference value: -0.1881\n",
      "Corresponding equal opportunity difference value: -0.0944\n",
      "Corresponding Theil index value: 0.0916\n"
     ]
    }
   ],
   "source": [
    "dir_thresh_arr_orig_meps_best = thresh_arr_best\n",
    "print(\"Threshold corresponding to Best balanced accuracy: %6.4f\" % dir_thresh_arr_orig_meps_best)\n",
    "dir_best_bal_acc_arr_orig_meps = best_bal_acc\n",
    "print(\"Best balanced accuracy: %6.4f\" % dir_best_bal_acc_arr_orig_meps)\n",
    "dir_disp_imp_at_best_bal_acc_orig_meps = disp_imp_at_best_bal_acc\n",
    "print(\"Corresponding abs(1-disparate impact) value: %6.4f\" % dir_disp_imp_at_best_bal_acc_orig_meps)\n",
    "dir_avg_odds_diff_at_best_bal_acc_orig_meps = avg_odds_diff_at_best_bal_acc\n",
    "print(\"Corresponding average odds difference value: %6.4f\" % dir_avg_odds_diff_at_best_bal_acc_orig_meps)\n",
    "\n",
    "dir_stat_par_diff_at_best_bal_acc_orig_meps = stat_par_diff_at_best_bal_acc\n",
    "print(\"Corresponding statistical parity difference value: %6.4f\" % dir_stat_par_diff_at_best_bal_acc_orig_meps)\n",
    "dir_eq_opp_diff_at_best_bal_acc_orig_meps = eq_opp_diff_at_best_bal_acc\n",
    "print(\"Corresponding equal opportunity difference value: %6.4f\" % dir_eq_opp_diff_at_best_bal_acc_orig_meps)\n",
    "dir_theil_ind_at_best_bal_acc_orig_meps = theil_ind_at_best_bal_acc\n",
    "print(\"Corresponding Theil index value: %6.4f\" % dir_theil_ind_at_best_bal_acc_orig_meps)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"di-test\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.5.3. Testing DIR model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Test model on given dataset and find threshold for best balanced accuracy\n",
    "import numpy as np\n",
    "from tqdm import tqdm\n",
    "thresh_arr = np.linspace(0.01, 0.75, 75)\n",
    "\n",
    "dataset = meps_orig_test        #data to test on\n",
    "scale = meps_orig_dir_scale\n",
    "model = meps_orig_dir\n",
    "\n",
    "te_dataset= dataset.copy(deepcopy=True)\n",
    "te_dataset.features = scale.transform(te_dataset.features)\n",
    "\n",
    "di = DisparateImpactRemover(repair_level=1.0)\n",
    "\n",
    "test_repd = di.fit_transform(te_dataset)  #repair test dataset\n",
    "\n",
    "index = te_dataset.feature_names.index(sens_attr)        \n",
    "X_te = np.delete(test_repd.features, index, axis=1)\n",
    "    \n",
    "\n",
    "y_te_pred_prob = model.predict_proba(X_te)\n",
    "\n",
    "thresh_arr = dir_thresh_arr_orig_meps_best    \n",
    "\n",
    "y_te_pred = (y_te_pred_prob[:,1] > thresh).astype(np.double)\n",
    "\n",
    "test_repd_pred = test_repd.copy()\n",
    "test_repd_pred.labels = y_te_pred\n",
    "\n",
    "classified_metric = ClassificationMetric(test_repd, \n",
    "                                             test_repd_pred,\n",
    "                                             unprivileged_groups=unprivileged_groups,\n",
    "                                             privileged_groups=privileged_groups)\n",
    "metric_pred = BinaryLabelDatasetMetric(test_repd_pred,\n",
    "                                             unprivileged_groups=unprivileged_groups,\n",
    "                                             privileged_groups=privileged_groups)\n",
    "    \n",
    "bal_acc = 0.5*(classified_metric.true_positive_rate() + classified_metric.true_negative_rate())\n",
    "    \n",
    "acc = accuracy_score(y_true=test_repd.labels,\n",
    "                        y_pred=test_repd_pred.labels)\n",
    "\n",
    "best_bal_acc = bal_acc\n",
    "disp_imp_at_best_bal_acc = np.abs(1.0-metric_pred.disparate_impact())\n",
    "\n",
    "avg_odds_diff_at_best_bal_acc = classified_metric.average_odds_difference()\n",
    "\n",
    "stat_par_diff_at_best_bal_acc = classified_metric.statistical_parity_difference()\n",
    "eq_opp_diff_at_best_bal_acc = classified_metric.equal_opportunity_difference()\n",
    "theil_ind_at_best_bal_acc = classified_metric.theil_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Threshold corresponding to Best balanced accuracy: 0.5500\n",
      "Best balanced accuracy: 0.7233\n",
      "Corresponding abs(1-disparate impact) value: 0.5133\n",
      "Corresponding average odds difference value: -0.0898\n",
      "Corresponding statistical parity difference value: -0.1213\n",
      "Corresponding equal opportunity difference value: -0.1282\n",
      "Corresponding Theil index value: 0.1126\n"
     ]
    }
   ],
   "source": [
    "dir_thresh_arr_orig_meps_best_test = thresh_arr_best\n",
    "print(\"Threshold corresponding to Best balanced accuracy: %6.4f\" % dir_thresh_arr_orig_meps_best_test)\n",
    "dir_best_bal_acc_arr_orig_meps_best_test = best_bal_acc\n",
    "print(\"Best balanced accuracy: %6.4f\" % dir_best_bal_acc_arr_orig_meps_best_test)\n",
    "dir_disp_imp_at_best_bal_acc_orig_meps_best_test = disp_imp_at_best_bal_acc\n",
    "print(\"Corresponding abs(1-disparate impact) value: %6.4f\" % dir_disp_imp_at_best_bal_acc_orig_meps_best_test)\n",
    "dir_avg_odds_diff_at_best_bal_acc_orig_meps_best_test = avg_odds_diff_at_best_bal_acc\n",
    "print(\"Corresponding average odds difference value: %6.4f\" % dir_avg_odds_diff_at_best_bal_acc_orig_meps_best_test)\n",
    "\n",
    "dir_stat_par_diff_at_best_bal_acc_orig_meps_best_test = stat_par_diff_at_best_bal_acc\n",
    "print(\"Corresponding statistical parity difference value: %6.4f\" % dir_stat_par_diff_at_best_bal_acc_orig_meps_best_test)\n",
    "dir_eq_opp_diff_at_best_bal_acc_orig_meps_best_test = eq_opp_diff_at_best_bal_acc\n",
    "print(\"Corresponding equal opportunity difference value: %6.4f\" % dir_eq_opp_diff_at_best_bal_acc_orig_meps_best_test)\n",
    "dir_theil_ind_at_best_bal_acc_orig_meps_best_test = theil_ind_at_best_bal_acc\n",
    "print(\"Corresponding Theil index value: %6.4f\" % dir_theil_ind_at_best_bal_acc_orig_meps_best_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "While disparate impact remover resulted in a model that is relatively fairer than, for examaple, the logistic regression model learnt from original data, it is quite unfair with an bs(1-disparate impact) value much higher than the typically desired value of < 0.2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"summary-2015-learning\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.6. Summary of MEPS Model Learning Results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[Back to TOC](#toc)<br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Logistic Regression classifier 'without' bias mitigation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Threshold corresponding to Best balanced accuracy: 0.1900\n",
      "Best balanced accuracy: 0.7759\n",
      "Corresponding abs(1-disparate impact) value: 0.5738\n",
      "Corresponding average odds difference value: -0.2057\n",
      "Corresponding statistical parity difference value: -0.2612\n",
      "Corresponding equal opportunity difference value: -0.2228\n",
      "Corresponding Theil index value: 0.0921\n"
     ]
    }
   ],
   "source": [
    "print(\"Threshold corresponding to Best balanced accuracy: %6.4f\" % lr_thresh_arr_orig_meps_best_test)\n",
    "print(\"Best balanced accuracy: %6.4f\" % lr_best_bal_acc_arr_orig_meps_best_test)\n",
    "print(\"Corresponding abs(1-disparate impact) value: %6.4f\" % lr_disp_imp_at_best_bal_acc_orig_meps_best_test)\n",
    "print(\"Corresponding average odds difference value: %6.4f\" % lr_avg_odds_diff_at_best_bal_acc_orig_meps_best_test)\n",
    "print(\"Corresponding statistical parity difference value: %6.4f\" % lr_stat_par_diff_at_best_bal_acc_orig_meps_best_test)\n",
    "print(\"Corresponding equal opportunity difference value: %6.4f\" % lr_eq_opp_diff_at_best_bal_acc_orig_meps_best_test)\n",
    "print(\"Corresponding Theil index value: %6.4f\" % lr_theil_ind_at_best_bal_acc_orig_meps_best_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Logistic Regression classifier using Pre-Processing Bias Mitigation - ReWeighing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Threshold corresponding to Best balanced accuracy: 0.2200\n",
      "Best balanced accuracy: 0.7539\n",
      "Corresponding abs(1-disparate impact) value: 0.2482\n",
      "Corresponding average odds difference value: -0.0151\n",
      "Corresponding statistical parity difference value: -0.0872\n",
      "Corresponding equal opportunity difference value: -0.0035\n",
      "Corresponding Theil index value: 0.0966\n"
     ]
    }
   ],
   "source": [
    "print(\"Threshold corresponding to Best balanced accuracy: %6.4f\" % lr_thresh_arr_transf_meps_best_test)\n",
    "print(\"Best balanced accuracy: %6.4f\" % lr_best_bal_acc_arr_transf_meps_best_test)\n",
    "print(\"Corresponding abs(1-disparate impact) value: %6.4f\" % lr_disp_imp_at_best_bal_acc_transf_meps_best_test)\n",
    "print(\"Corresponding average odds difference value: %6.4f\" % lr_avg_odds_diff_at_best_bal_acc_transf_meps_best_test)\n",
    "print(\"Corresponding statistical parity difference value: %6.4f\" % lr_stat_par_diff_at_best_bal_acc_transf_meps_best_test)\n",
    "print(\"Corresponding equal opportunity difference value: %6.4f\" % lr_eq_opp_diff_at_best_bal_acc_transf_meps_best_test)\n",
    "print(\"Corresponding Theil index value: %6.4f\" % lr_theil_ind_at_best_bal_acc_transf_meps_best_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Model learnt using In-Processing Bias Mitigation - Prejudice Remover"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Threshold corresponding to Best balanced accuracy: 0.1200\n",
      "Best balanced accuracy: 0.6880\n",
      "Corresponding abs(1-disparate impact) value: 0.1588\n",
      "Corresponding average odds difference value: 0.0523\n",
      "Corresponding statistical parity difference value: -0.0566\n",
      "Corresponding equal opportunity difference value: 0.1479\n",
      "Corresponding Theil index value: 0.1108\n"
     ]
    }
   ],
   "source": [
    "print(\"Threshold corresponding to Best balanced accuracy: %6.4f\" % pr_thresh_arr_orig_meps_best_test)\n",
    "print(\"Best balanced accuracy: %6.4f\" % pr_best_bal_acc_arr_orig_meps_best_test)\n",
    "print(\"Corresponding abs(1-disparate impact) value: %6.4f\" % pr_disp_imp_at_best_bal_acc_orig_meps_best_test)\n",
    "print(\"Corresponding average odds difference value: %6.4f\" % pr_avg_odds_diff_at_best_bal_acc_orig_meps_best_test)\n",
    "print(\"Corresponding statistical parity difference value: %6.4f\" % pr_stat_par_diff_at_best_bal_acc_orig_meps_best_test)\n",
    "print(\"Corresponding equal opportunity difference value: %6.4f\" % pr_eq_opp_diff_at_best_bal_acc_orig_meps_best_test)\n",
    "print(\"Corresponding Theil index value: %6.4f\" % pr_theil_ind_at_best_bal_acc_orig_meps_best_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Model learnt using Pre-Processing Bias Mitigation - Disparate Impact Remover"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Threshold corresponding to Best balanced accuracy: 0.5500\n",
      "Best balanced accuracy: 0.7233\n",
      "Corresponding abs(1-disparate impact) value: 0.5133\n",
      "Corresponding average odds difference value: -0.0898\n",
      "Corresponding statistical parity difference value: -0.1213\n",
      "Corresponding equal opportunity difference value: -0.1282\n",
      "Corresponding Theil index value: 0.1126\n"
     ]
    }
   ],
   "source": [
    "print(\"Threshold corresponding to Best balanced accuracy: %6.4f\" % dir_thresh_arr_orig_meps_best_test)\n",
    "print(\"Best balanced accuracy: %6.4f\" % dir_best_bal_acc_arr_orig_meps_best_test)\n",
    "print(\"Corresponding abs(1-disparate impact) value: %6.4f\" % dir_disp_imp_at_best_bal_acc_orig_meps_best_test)\n",
    "print(\"Corresponding average odds difference value: %6.4f\" % dir_avg_odds_diff_at_best_bal_acc_orig_meps_best_test)\n",
    "print(\"Corresponding statistical parity difference value: %6.4f\" % dir_stat_par_diff_at_best_bal_acc_orig_meps_best_test)\n",
    "print(\"Corresponding equal opportunity difference value: %6.4f\" % dir_eq_opp_diff_at_best_bal_acc_orig_meps_best_test)\n",
    "print(\"Corresponding Theil index value: %6.4f\" % dir_theil_ind_at_best_bal_acc_orig_meps_best_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Of all the models, the logistic regression model after reweighing gives the best balance in terms of balanced accuracy and fairness. While the model learnt by prejudice remover is slightly fairer, it has much lower accuracy. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"deployment-2015-2015\"></a>"
   ]
  }
 ],
 "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": "3.5.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}