03. Fair Classification.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Ensuring Fairness in Classification Tasks\n",
    "\n",
    "This tutorial will focus on a real-world learning-with-constraint application.\n",
    "\n",
    "We will use the Moving Targets' APIs to train a neural classifier while imposing a fairness constraint on the whole population, namely the [**Disparate Impact Discrimination Index** ($\\mbox{DIDI}$)](https://arxiv.org/abs/1903.10598).\n",
    "\n",
    "> NOTE: since we will leverage on the [Gurobi](https://www.gurobi.com/) solver to deal with the master problem, you will need to install it separately. A free academic license for the solver can be obtained [here](https://www.gurobi.com/academia/academic-program-and-licenses/). Otherwise, you may change the line of code involving the backend object and use one among the other supported backends."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import tensorflow as tf\n",
    "\n",
    "np.random.seed(0)\n",
    "tf.random.set_seed(0)\n",
    "sns.set_style('whitegrid')\n",
    "sns.set_context('notebook')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## The Adult Dataset\n",
    "\n",
    "This time, we will benchmark our approach on the **Adult** dataset, whose original version can be found on the [UCI Repository](https://archive.ics.uci.edu/ml/datasets/Adult) but, as well as in the previous tutorial, we will use a preprocessed version that can be found in the official Moving Targets' [github repository](https://github.com/moving-targets/moving-targets).\n",
    "\n",
    "The task consists in predicting the **income** class (*0* = income $\\geq 50k$, *1* otherwise) based on the personal attributes of an individual. In particular, we will focus on learning the best classifier while imposing a fairness constraint on the (one-hot encoded) protected feature **race**, which indicates five different protected ethnicity groups: *American-Indian/Eskimo*, *Asian-Pacific Islander*, *Black*, *White*, and *Other*."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>age</th>\n",
       "      <th>fnlwgt</th>\n",
       "      <th>education-num</th>\n",
       "      <th>capital-gain</th>\n",
       "      <th>capital-loss</th>\n",
       "      <th>hours-per-week</th>\n",
       "      <th>workclass_Federal-gov</th>\n",
       "      <th>workclass_Local-gov</th>\n",
       "      <th>workclass_Private</th>\n",
       "      <th>workclass_Self-emp-inc</th>\n",
       "      <th>...</th>\n",
       "      <th>relationship_Unmarried</th>\n",
       "      <th>relationship_Wife</th>\n",
       "      <th>sex_Female</th>\n",
       "      <th>sex_Male</th>\n",
       "      <th>race_Amer-Indian-Eskimo</th>\n",
       "      <th>race_Asian-Pac-Islander</th>\n",
       "      <th>race_Black</th>\n",
       "      <th>race_Other</th>\n",
       "      <th>race_White</th>\n",
       "      <th>income</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>30162.000000</td>\n",
       "      <td>3.016200e+04</td>\n",
       "      <td>30162.000000</td>\n",
       "      <td>30162.000000</td>\n",
       "      <td>30162.000000</td>\n",
       "      <td>30162.000000</td>\n",
       "      <td>30162.000000</td>\n",
       "      <td>30162.000000</td>\n",
       "      <td>30162.000000</td>\n",
       "      <td>30162.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>30162.000000</td>\n",
       "      <td>30162.000000</td>\n",
       "      <td>30162.000000</td>\n",
       "      <td>30162.000000</td>\n",
       "      <td>30162.000000</td>\n",
       "      <td>30162.000000</td>\n",
       "      <td>30162.000000</td>\n",
       "      <td>30162.000000</td>\n",
       "      <td>30162.00000</td>\n",
       "      <td>30162.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>38.437902</td>\n",
       "      <td>1.897938e+05</td>\n",
       "      <td>10.121312</td>\n",
       "      <td>1092.007858</td>\n",
       "      <td>88.372489</td>\n",
       "      <td>40.931238</td>\n",
       "      <td>0.031265</td>\n",
       "      <td>0.068530</td>\n",
       "      <td>0.738877</td>\n",
       "      <td>0.035608</td>\n",
       "      <td>...</td>\n",
       "      <td>0.106492</td>\n",
       "      <td>0.046615</td>\n",
       "      <td>0.324315</td>\n",
       "      <td>0.675685</td>\n",
       "      <td>0.009482</td>\n",
       "      <td>0.029673</td>\n",
       "      <td>0.093396</td>\n",
       "      <td>0.007659</td>\n",
       "      <td>0.85979</td>\n",
       "      <td>0.248922</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>13.134665</td>\n",
       "      <td>1.056530e+05</td>\n",
       "      <td>2.549995</td>\n",
       "      <td>7406.346497</td>\n",
       "      <td>404.298370</td>\n",
       "      <td>11.979984</td>\n",
       "      <td>0.174035</td>\n",
       "      <td>0.252657</td>\n",
       "      <td>0.439254</td>\n",
       "      <td>0.185313</td>\n",
       "      <td>...</td>\n",
       "      <td>0.308471</td>\n",
       "      <td>0.210816</td>\n",
       "      <td>0.468126</td>\n",
       "      <td>0.468126</td>\n",
       "      <td>0.096915</td>\n",
       "      <td>0.169687</td>\n",
       "      <td>0.290991</td>\n",
       "      <td>0.087179</td>\n",
       "      <td>0.34721</td>\n",
       "      <td>0.432396</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>1.376900e+04</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>28.000000</td>\n",
       "      <td>1.176272e+05</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>37.000000</td>\n",
       "      <td>1.784250e+05</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>47.000000</td>\n",
       "      <td>2.376285e+05</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>45.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>90.000000</td>\n",
       "      <td>1.484705e+06</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>4356.000000</td>\n",
       "      <td>99.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 48 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                age        fnlwgt  education-num  capital-gain  capital-loss  \\\n",
       "count  30162.000000  3.016200e+04   30162.000000  30162.000000  30162.000000   \n",
       "mean      38.437902  1.897938e+05      10.121312   1092.007858     88.372489   \n",
       "std       13.134665  1.056530e+05       2.549995   7406.346497    404.298370   \n",
       "min       17.000000  1.376900e+04       1.000000      0.000000      0.000000   \n",
       "25%       28.000000  1.176272e+05       9.000000      0.000000      0.000000   \n",
       "50%       37.000000  1.784250e+05      10.000000      0.000000      0.000000   \n",
       "75%       47.000000  2.376285e+05      13.000000      0.000000      0.000000   \n",
       "max       90.000000  1.484705e+06      16.000000  99999.000000   4356.000000   \n",
       "\n",
       "       hours-per-week  workclass_Federal-gov  workclass_Local-gov  \\\n",
       "count    30162.000000           30162.000000         30162.000000   \n",
       "mean        40.931238               0.031265             0.068530   \n",
       "std         11.979984               0.174035             0.252657   \n",
       "min          1.000000               0.000000             0.000000   \n",
       "25%         40.000000               0.000000             0.000000   \n",
       "50%         40.000000               0.000000             0.000000   \n",
       "75%         45.000000               0.000000             0.000000   \n",
       "max         99.000000               1.000000             1.000000   \n",
       "\n",
       "       workclass_Private  workclass_Self-emp-inc  ...  relationship_Unmarried  \\\n",
       "count       30162.000000            30162.000000  ...            30162.000000   \n",
       "mean            0.738877                0.035608  ...                0.106492   \n",
       "std             0.439254                0.185313  ...                0.308471   \n",
       "min             0.000000                0.000000  ...                0.000000   \n",
       "25%             0.000000                0.000000  ...                0.000000   \n",
       "50%             1.000000                0.000000  ...                0.000000   \n",
       "75%             1.000000                0.000000  ...                0.000000   \n",
       "max             1.000000                1.000000  ...                1.000000   \n",
       "\n",
       "       relationship_Wife    sex_Female      sex_Male  race_Amer-Indian-Eskimo  \\\n",
       "count       30162.000000  30162.000000  30162.000000             30162.000000   \n",
       "mean            0.046615      0.324315      0.675685                 0.009482   \n",
       "std             0.210816      0.468126      0.468126                 0.096915   \n",
       "min             0.000000      0.000000      0.000000                 0.000000   \n",
       "25%             0.000000      0.000000      0.000000                 0.000000   \n",
       "50%             0.000000      0.000000      1.000000                 0.000000   \n",
       "75%             0.000000      1.000000      1.000000                 0.000000   \n",
       "max             1.000000      1.000000      1.000000                 1.000000   \n",
       "\n",
       "       race_Asian-Pac-Islander    race_Black    race_Other   race_White  \\\n",
       "count             30162.000000  30162.000000  30162.000000  30162.00000   \n",
       "mean                  0.029673      0.093396      0.007659      0.85979   \n",
       "std                   0.169687      0.290991      0.087179      0.34721   \n",
       "min                   0.000000      0.000000      0.000000      0.00000   \n",
       "25%                   0.000000      0.000000      0.000000      1.00000   \n",
       "50%                   0.000000      0.000000      0.000000      1.00000   \n",
       "75%                   0.000000      0.000000      0.000000      1.00000   \n",
       "max                   1.000000      1.000000      1.000000      1.00000   \n",
       "\n",
       "             income  \n",
       "count  30162.000000  \n",
       "mean       0.248922  \n",
       "std        0.432396  \n",
       "min        0.000000  \n",
       "25%        0.000000  \n",
       "50%        0.000000  \n",
       "75%        0.000000  \n",
       "max        1.000000  \n",
       "\n",
       "[8 rows x 48 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "df = pd.read_csv('https://raw.githubusercontent.com/moving-targets/moving-targets/main/res/adult.csv')\n",
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAscAAAE/CAYAAACq327HAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAAxq0lEQVR4nO3deVxU9f7H8ffIprmRBJbXJfcNxVJLxCX3Fc3d3Oumt0xNuy6kuNzM3TLNNL23vGVuaCOgeNHU1BBzQU25uf1KKMrAJXFhG2B+f3CcJBABR0Hv6/l4+Hg4Z875ns/MfBjenPmeMyar1WoVAAAAABUp6AIAAACAwoJwDAAAABgIxwAAAICBcAwAAAAYCMcAAACAgXAMAAAAGAjHAOwmJiZGtWvXVvfu3W3/unXrpk2bNhV0aXrllVd05cqVAq0hLCxMrVq1Uq9evZSUlJTpvsGDBys0NDRP44WGhmrw4MGSpMWLFyswMDDH9bt3765r165lWf7JJ5/Iz88vT/u+ndls1t/+9rd8b28vsbGx8vPzk6+vr7p166Y+ffpo586dtvtr1qxZ4D0AoPBzLOgCADxaihYtqqCgINvt2NhYde3aVZ6enqpVq1aB1bV///4C2/ctISEh6tOnj0aOHGn3sd988827rnP76/KouXLlivr3768333xTc+bMkclk0unTp/Xyyy+rWLFi8vHxKegSATwkCMcA7quyZcuqUqVKioqKUq1atbRx40atW7dO6enpcnV11dSpU1W1alX5+fnp6tWr+vnnn/XCCy9o5MiRevfdd3X06FE5ODiobdu2GjdunCwWixYuXKjDhw8rLS1NderUkb+/v0qUKKHWrVurR48eOnDggC5cuKBOnTpp4sSJevvttyVJQ4cO1cqVK3X69GmtWLFCKSkpunLlil588UWNHTtWkrRy5Upt2rRJxYsXV6NGjbRr1y7t3r1bKSkpd9zv7SwWi+bOnasDBw7IwcFB9evX19tvv63169dr165dcnFx0fXr1zVp0qQ7Pmd3ehxSxhHiLVu2yNXVVZUqVbJt4+fnp+rVq6tEiRLavXu3VqxYIUn64YcfNGzYMO3Zs0d16tTRgQMHVLJkSb377rsKDw+Xm5ub3NzcVLJkSUkZR7AHDhyojh07Zrm9adMmbdiwQRaLRfHx8Ro+fLgGDBiQqfYdO3Zo+fLlMplMcnBw0MSJE9W4ceNM65jNZoWEhCg9PV2xsbEqW7as5s6dq7Jly+r69euaNWuWzp49K4vFIm9vb02cOFGOjo7y9PRUmzZtdPr0aS1cuFD16tWzjbl27Vo9++yzevHFF23LatWqpQ8//ND22G5JSEjQjBkzFBUVpfj4eBUvXlwLFy5UlSpV7lh/bh4XgEcD0yoA3FfHjh3TTz/9JC8vLx06dEiBgYFas2aNAgMD9eqrr2r06NG2dZOSkhQSEqIJEyZoyZIlSk5O1rZt2xQYGKijR4/q0KFDWrlypRwcHGQ2mxUcHCwPDw8tXLjQNkZCQoLWrl2r9evX64svvtDPP/+sOXPmSJI+++wzPfnkk/r00081d+5cmc1mbdiwQStXrtSVK1f0zTffyGw2a9OmTTKbzbp586Zt3Lvt95bly5crLi5OQUFBCgoKUnp6uubPn69XX31VrVu31rBhw3IMxjk9jp07d2rHjh0KDAzU+vXrdePGjSzbdenSRREREbp48aKkjCDas2dPOTg42NZZu3atoqKiFBISok8//VQXLly4az03b97Uxo0btXLlSgUGBmrRokVasGBBlvXmz5+v6dOny2w2680339TBgwezHe/o0aOaNm2atm3bprp162rWrFmSpNmzZ6tu3boym80KDAzU77//rlWrVknK+MOjVatW2r59e6ZgLEmRkZF69tlns+ynUaNGqlmzZqZl+/btU6lSpRQQEKDt27fL09NTa9asybH+3D4uAA8/jhwDsKukpCR1795dkpSWlqbHH39cCxYs0FNPPaXVq1crOjpa/fv3t60fHx+vq1evSpIaNmxoWx4eHq63335bDg4OcnBw0BdffCFJWrBgga5fv67w8HBJGYHJzc3Ntl2bNm0kZRyxdnNzU3x8vCpUqGC732Qy6eOPP9aePXu0detW/fDDD7JarUpMTNTevXvVsWNHlSpVSpI0cOBAffvtt5KkPXv25LjfW/bt26dx48bJyclJUsaR1zfeeCPPz2N2j+PAgQNq166d7Wh1r169tHr16kzblShRQh06dFBwcLCGDRum4OBgrV27NtM6Bw4cUNeuXeXs7CxnZ2f5+vrqzJkzOdZTvHhxffzxx9q7d6+ioqJ0+vRpJSQkZFmvS5cuGjVqlFq2bCkfHx8NHz482/F8fHxUuXJlSVLfvn1tPbNnzx6dPHnSNk/9z3OzGzVqlO14JpNJVqs1x8dwS8eOHVWhQgVbPx46dEjPPPNMjvXn9nEBePgRjgHY1Z/nHN8uPT1d3bt314QJE2y34+LiVLp0aUnSY489ZlvX0dFRJpPJdvvChQsqWrSo0tPTNXnyZLVs2VJSxhHN5ORk23ouLi62/2cXmBISEtSjRw+1bdtWjRo1Uq9evbRz505ZrVY5OjpmWv/2o6132+/t6/35tsViyfb5yEl2j+PPj+f2+m7Xp08f23SVatWqZfrjIDt/Huf2fdyq/bffflO/fv3Ut29fNWzYUB07dtTXX3+dZaxx48apd+/eCgsLk9ls1sqVK2U2m1WkSOYPKv/83N66nZ6ersWLF6tq1aqSpGvXrmXqg9t75HYNGjTQ8ePHNWjQoEzL169fr8TERL388su2ZWvXrlVAQIAGDhwoX19fubq6KiYmJsf6c/u4ADz8+KkG8MD4+PgoJCREcXFxkqR169Zp6NCh2a7r7e2tzZs3Kz09XSkpKRozZowOHz6sZs2aac2aNUpJSVF6erqmTp2q999//677dnBwUGpqqqKjo3Xjxg2NHTtWrVu31qFDh2xjtWzZUjt27ND169clKdNVNnK73+bNm2v9+vWyWCxKT0/XmjVr7HYyWPPmzRUaGqpr164pPT39jn+ENGjQQJL00UcfqU+fPtmOExgYqOTkZNvUlVvKlCmjyMhISdJPP/1kO6IcGRmpMmXKaOTIkWrevLktGKelpdm2TU1NVevWrZWQkKCXXnpJ06dP1w8//KDU1NQsNXz77beKjY2VlBFgW7VqJSnjef73v/8tq9WqlJQUvf7667ZPDXLSr18/HTp0SMHBwbZwHxkZqSVLlqhGjRqZ1g0LC1OPHj3Up08fVa5cWbt371ZaWlqO9ef2cQF4+HHkGMAD07x5cw0fPlyvvPKKTCaTSpQooaVLl2Y6MnjLqFGjNGvWLHXv3l1paWnq3Lmz2rdvrxYtWmjevHnq0aOH0tLSVLt27Vxdhqxdu3YaMGCAli5dqhdeeEGdOnVSqVKlVLFiRVWrVk3R0dFq3ry5+vbtq379+qlo0aKqXr26ihUrJkkaOXJkrvb7+uuva968eXrxxReVmpqq+vXra+rUqff+5Elq2bKlzpw5o169eqlUqVKqVauWfv/992zX7dOnj5YtW6a2bdtmua9///766aef1LVr1ywn9r3++uvy8/PT3r17VaVKFds0Bh8fH23atEkdO3ZUsWLFVL9+fZUpU0bR0dG2bR0dHTV58mSNHz/eduR/9uzZcnZ2zlJD2bJlNWHCBF28eFHVqlXTO++8I0maMmWKZs2aJV9fX1ksFjVt2lSvvvrqXZ8bV1dXrV69WgsWLNCKFStUpEgRFStWTLNmzcryx8krr7yiadOmyWw2y8HBQXXr1tXZs2dzrD+3jwvAw89kze0kLQB4xJ08eVLHjh3TkCFDJEmrVq3Sd999pw8++KBgC3vEmM1mbd++3XZFDQAoTDhyDACGypUr65///KcCAgJkMpn01FNPaebMmQVdFgDgAeLIMQAAAGDghDwAAADAQDgGAAAADIVmznF6erpu3rwpJyenbM9cBwAAAOzBarXKYrGoePHiWa5XXmjC8c2bN3X27NmCLgMAAAD/I2rUqKGSJUtmWlZowvGtr1qtUaMG145UxsXrPT09C7oMFDL0BbJDXyA79AWyQ19kSElJ0dmzZ23583aFJhzfmkrh7Oyc6WtT/5fxPCA79AWyQ18gO/QFskNf/CG7qbyckAcAAAAYCMcAAACAodBMqwAAAEDhZLFYFBMTo6SkpIIuJU+KFi2q8uXLZzu3+E4IxwAAAMhRTEyMSpYsqaeffvqhueSu1WrV5cuXFRMTo8qVK+d6O6ZVAAAAIEdJSUlyc3N7aIKxlHGynZubW56PdhOOAQAAcFcPUzC+JT815yoc37hxQ127dlVMTIwkKTw8XL6+vmrfvr0WLVpkW+/UqVPq1auXOnTooClTpig1NTXPBQEAAAAF5a7h+LvvvtNLL72kqKgoSRmH1SdPnqxly5Zp27ZtioyM1N69eyVJEyZM0NSpU7V9+3ZZrVYFBATc1+IBAABQME6ePKkxY8YUdBl2d9dwHBAQoOnTp8vDw0OSdOLECVWqVEkVKlSQo6OjfH19FRoaql9++UVJSUlq0KCBJKlnz54KDQ29r8UDAACgYNSrV09Lliwp6DLs7q5Xq5g1a1am23FxcXJ3d7fd9vDwUGxsbJbl7u7uio2NzXNBkZGRed7G3urXqSGnYiXvvuJ91LBhwwLd/y2WxOs68f3Zgi6jwBWGnpAKR1/QE3+gL/5AX/yBvvgDffGHR7EvDh48qJkzZ8rT01MlSpTQmTNn9Ntvv6lKlSp6//33Vbx4cX333Xd69913lZiYKCcnJ02cOFHPP/+c9u8P1wcffKCkpCQ5OTlp5MiR8vHxUXBwsHbt2qXk5GT9+uuvevLJJ9WvXz9t2LBB0dHRGjRokAYPHixJCgwMVEBAgKxWq0qXLq1JkyZle1WKlJQURURE5Ppx5flSblarNcsyk8l0x+V55enpWTi+1nBG6YKuoFBwmhFfKN5gCwV6QhI9kQV9IYm+yIK+kERfZPEw90WHAOnXP1314fI5yZIoJVxW5Nn/6vP3p8pUxKS+I/0Vun6lurVrpjdeG6V3x4/QC97PKvLMj3r7H1P1+bqNmjRpkpYvXy4vLy+dO3dOgwYN0qZNm+Ti4qLjx49ry5YtKlu2rHx9fbVz506tXr1aZ8+eVd++fTVixAgdOXJE27Zt0/r161WsWDGFhYVp4sSJ2rZtW5bSnZ2d5eXllWlZcnLyHQ/I5jkcly1bVpcuXbLdjouLk4eHR5blFy9etE3FAAAAwKOr+XNecnbO+KKNGpUrKv76DZ398WcVKVJEL3g/K0nyrFlFWz5doL0nTqhixYq2wFq9enU9++yzOnTokEwmk+rVq6ennnpKklS+fHk1a9ZMRYoUUYUKFZScnKzExETt2bNH0dHR6t+/v62G+Ph4Xb16Va6urvf0WPIcjr28vHT+/HlFR0erfPny2rp1q3r16qW//OUvcnFxUUREhBo2bKjAwEC1aNHinooDAABA4VfUxdn2f5MpY6aBg0MR/XkSwdnzPys9Pev0EqvVqtTUVDk5OcnZ2TnTfY6OWeNqenq6unfvrgkTJthux8XFqXTpez86n+frHLu4uGju3LkaPXq0OnfurCpVqqhjx46SpIULF2rOnDnq1KmTEhMTNWTIkHsuEAAAAA+fKhXKyWQyaf+RE5Kk/549r6FvzVT9+vV1/vx5nTiRsfzcuXM6fPiwnnvuuVyP7ePjo5CQEMXFxUmS1q1bp6FDh9ql7lwfOd69e7ft/97e3goODs6yTq1atbRp0ya7FAYAAICHl7Ozkz585y3NXvq55n+8Rk5Ojvrwnbfk5uamxYsXa+bMmUpKSpLJZNKcOXNUuXJlHTt2LFdjN2/eXMOHD9crr7wik8mkEiVKaOnSpXb5ohKTNbsz6QrArYnRnJBXyMyIL+gKCg96IgM9kRl9kYG+yIy+yEBfZPYQ98WpDgGqXclO55KVe8Y+4+TSqVOnVLt27UzLcsqdfH00AAAAYCAcAwAAAAbCMQAAAGAgHAMAAAAGwjEAAABgIBwDAAAAhjx/Qx4AAACQ5FZHRe/D5XeTLGkq6uRg93Fzi3AMAACAPCvq4qKn/ULsPm7U3C65Wm/Lli1avny5LBaLhg0bpoEDB9pl/4RjAAAAPFRiY2O1aNEimc1mOTs7q3///nr++edVrVq1ex6bOccAAAB4qISHh6tJkyZydXXVY489pg4dOig0NNQuYxOOAQAA8FCJi4uTu7u77baHh4diY2PtMjbhGAAAAA8Vq9WaZZnJZLLL2IRjAAAAPFTKli2rS5cu2W7HxcXJw8PDLmMTjgEAAPBQadq0qQ4cOKArV64oMTFRO3bsUIsWLewyNlerAAAAQJ4lJSfn+rJreRo3F9c5Llu2rMaNG6chQ4bIYrGod+/eql+/vl32TzgGAABAnhW9/H3+Niz3TM7j5vILQHx9feXr65u/GnLAtAoAAADAwJFjAADsyGpJkmlGfEGXUShYLUkyORUt6DKAPCEcAwBgRyanovflK3UfRvdjPipwvzGtAgAAADAQjgEAAAAD4RgAAAAwMOcYAAAAeedWTXIpaf9xLUlSLk/kvHHjhvr376+PP/5Y5cuXt8vuCccAAADIO5eS0ozS9h83l1d7+e677+Tv76+oqCi77p5wnA0uw/MHLsMDAAAKo4CAAE2fPl0TJ06067iE42xwGZ4/cBkeAABQGM2aNeu+jMsJeQAAAICBI8cAkE9MwfoDU7AAPCoIxwCQT0zB+gNTsAA8KgjHAAAAyLvk67m+skSe5OFSbvcD4RgAAAB5d/n/8rdduWdyvj+PwXj37t35q+MOOCEPAAAAMBCOAQAAAAPhGAAAADAQjgEAAHAXVlmt1oIuIs/yUzPhGAAAADkqGv+jLt9MfagCstVq1eXLl1W0aN5O8ONqFQAAAMhR+aPzFKNJuli6iiTTvQ0Wf8ouNeVG0aJFVb58+TxtQzgGAABAjpxSrqryt2/bZ7BC/s2iTKsAAAAADIRjAAAAwEA4BgAAAAyEYwAAAMBwT+E4KChIXbp0UZcuXTRv3jxJ0qlTp9SrVy916NBBU6ZMUWpqql0KBQAAAO63fIfjxMREzZo1S6tXr1ZQUJCOHDmi8PBwTZgwQVOnTtX27dtltVoVEBBgz3oBAACA+ybf4TgtLU3p6elKTExUamqqUlNT5ejoqKSkJDVo0ECS1LNnT4WGhtqrVgAAAOC+yvd1jkuUKKE333xTnTp1UtGiRfXcc8/JyclJ7u7utnXc3d0VGxubp3EjIyPzW5LdNGzYsKBLKFQiIiIKuoQCR09kRk9koC8yoy8y0BeZ0RcZ6IvMCnNf5Dscnz59Wl9++aW+/vprlSxZUuPHj9f+/fuzrGcy5e1bVDw9PeXi4pLfsnAf8AONP6MnkB36AtmhL5Cdgu6L5OTkOx6Qzfe0irCwMHl7e8vNzU3Ozs7q2bOnDh48qEuXLtnWuXjxojw8PPK7CwAAAOCBync4rlWrlsLDw5WQkCCr1ardu3frueeek4uLi+1QeWBgoFq0aGG3YgEAAID7Kd/TKpo1a6bvv/9ePXv2lJOTk+rVq6cRI0aoXbt28vf3182bN1WnTh0NGTLEnvUCAAAA902+w7EkjRgxQiNGjMi0rFatWtq0adM9FQUAAAAUBL4hDwAAADAQjgEAAAAD4RgAAAAwEI4BAAAAA+EYAAAAMBCOAQAAAAPhGAAAADAQjgEAAAAD4RgAAAAwEI4BAAAAA+EYAAAAMDgWdAEAAACPOqslSaYZ8QVdRqFgtSTJ5FS0oMu4I8IxAADAfWZyKqqn/UIKuoxCIWpul4IuIUdMqwAAAAAMHDkGcoGPw/5Q2D8OAwDgXhCOgVzg47A/FPaPwwAAuBdMqwAAAAAMhGMAAADAQDgGAAAADIRjAAAAwEA4BgAAAAyEYwAAAMBAOAYAAAAMhGMAAADAQDgGAAAADIRjAAAAwEA4BgAAAAyEYwAAAMBAOAYAAAAMhGMAAADAQDgGAAAADIRjAAAAwEA4BgAAAAyEYwAAAMBAOAYAAAAMhGMAAADAQDgGAAAADIRjAAAAwEA4BgAAAAyEYwAAAMBAOAYAAAAMhGMAAADAQDgGAAAADPcUjnfv3q2ePXuqY8eOevfddyVJ4eHh8vX1Vfv27bVo0SK7FAkAAAA8CPkOxz///LOmT5+uZcuWacuWLfr++++1d+9eTZ48WcuWLdO2bdsUGRmpvXv32rNeAAAA4L7Jdzj+6quv1LlzZz355JNycnLSokWLVKxYMVWqVEkVKlSQo6OjfH19FRoaas96AQAAgPvGMb8bRkdHy8nJSX/961918eJFtWrVStWrV5e7u7ttHQ8PD8XGxuZp3MjIyPyWZDcNGzYs6BIKlYiIiIIuocDRE5nRExnoi8zoiwz0RWb0RQb6IrPC3Bf5DsdpaWk6cuSIVq9erccee0wjR45UsWLFsqxnMpnyNK6np6dcXFzyWxbuA36g8Wf0BLJDXyA79AWyU9B9kZycfMcDsvkOx0888YS8vb1VpkwZSVKbNm0UGhoqBwcH2zpxcXHy8PDI7y4AAACAByrfc45btWqlsLAwXbt2TWlpafrmm2/UsWNHnT9/XtHR0UpLS9PWrVvVokULe9YLAAAA3Df5PnLs5eWlV199VQMGDJDFYpGPj49eeuklValSRaNHj1ZycrJatmypjh072rNeAAAA4L7JdziWpN69e6t3796Zlnl7eys4OPieigIAAAAKAt+QBwAAABgIxwAAAICBcAwAAAAYCMcAAACAgXAMAAAAGAjHAAAAgIFwDAAAABgIxwAAAICBcAwAAAAYCMcAAACAgXAMAAAAGAjHAAAAgIFwDAAAABgIxwAAAICBcAwAAAAYCMcAAACAgXAMAAAAGAjHAAAAgIFwDAAAABgIxwAAAICBcAwAAAAYCMcAAACAgXAMAAAAGAjHAAAAgIFwDAAAABgIxwAAAICBcAwAAAAYCMcAAACAgXAMAAAAGAjHAAAAgIFwDAAAABgIxwAAAICBcAwAAAAYCMcAAACAgXAMAAAAGAjHAAAAgIFwDAAAABgIxwAAAICBcAwAAAAYCMcAAACAgXAMAAAAGAjHAAAAgIFwDAAAABjuORzPmzdPfn5+kqRTp06pV69e6tChg6ZMmaLU1NR7LhAAAAB4UO4pHB84cECbN2+23Z4wYYKmTp2q7du3y2q1KiAg4J4LBAAAAB6UfIfjq1evatGiRXrttdckSb/88ouSkpLUoEEDSVLPnj0VGhpqlyIBAACAB8ExvxtOmzZN48aN04ULFyRJcXFxcnd3t93v7u6u2NjYPI8bGRmZ35LspmHDhgVdQqESERFR0CUUOHoiM3oiA32RGX2Rgb7IjL7IQF9kVpj7Il/heOPGjXrqqafk7e0ts9ksSbJarVnWM5lMeR7b09NTLi4u+SkL9wk/0PgzegLZoS+QHfoC2SnovkhOTr7jAdl8heNt27bp4sWL6t69u+Lj45WQkCCTyaRLly7Z1rl48aI8PDzyVzEAAABQAPIVjletWmX7v9ls1qFDhzRnzhx17dpVERERatiwoQIDA9WiRQu7FQoAAADcb/mec5ydhQsXyt/fXzdv3lSdOnU0ZMgQew4PAAAA3Ff3HI579uypnj17SpJq1aqlTZs23XNRAAAAQEHgG/IAAAAAA+EYAAAAMBCOAQAAAAPhGAAAADAQjgEAAAAD4RgAAAAwEI4BAAAAA+EYAAAAMBCOAQAAAAPhGAAAADAQjgEAAAAD4RgAAAAwEI4BAAAAA+EYAAAAMBCOAQAAAAPhGAAAADAQjgEAAAAD4RgAAAAwEI4BAAAAA+EYAAAAMBCOAQAAAAPhGAAAADAQjgEAAAAD4RgAAAAwEI4BAAAAA+EYAAAAMBCOAQAAAAPhGAAAADAQjgEAAAAD4RgAAAAwEI4BAAAAA+EYAAAAMBCOAQAAAAPhGAAAADAQjgEAAAAD4RgAAAAwEI4BAAAAA+EYAAAAMBCOAQAAAAPhGAAAADAQjgEAAAAD4RgAAAAw3FM4Xrp0qbp06aIuXbpo/vz5kqTw8HD5+vqqffv2WrRokV2KBAAAAB6EfIfj8PBwhYWFafPmzQoMDNR///tfbd26VZMnT9ayZcu0bds2RUZGau/evfasFwAAALhv8h2O3d3d5efnJ2dnZzk5Oalq1aqKiopSpUqVVKFCBTk6OsrX11ehoaH2rBcAAAC4b/IdjqtXr64GDRpIkqKiorRt2zaZTCa5u7vb1vHw8FBsbOw9FwkAAAA8CI73OsC5c+f0t7/9TZMmTZKjo6POnz+f6X6TyZSn8SIjI++1pHvWsGHDgi6hUImIiCjoEgocPZEZPZGBvsiMvshAX2RGX2SgLzIrzH1xT+E4IiJCY8aM0eTJk9WlSxcdOnRIly5dst0fFxcnDw+PPI3p6ekpFxeXeykLdsYPNP6MnkB26Atkh75Adgq6L5KTk+94QDbf0youXLigN954QwsXLlSXLl0kSV5eXjp//ryio6OVlpamrVu3qkWLFvndBQAAAPBA5fvI8SeffKLk5GTNnTvXtqx///6aO3euRo8ereTkZLVs2VIdO3a0S6EAAADA/ZbvcOzv7y9/f/9s7wsODs53QQAAAEBB4RvyAAAAAAPhGAAAADAQjgEAAAAD4RgAAAAwEI4BAAAAA+EYAAAAMBCOAQAAAAPhGAAAADAQjgEAAAAD4RgAAAAwEI4BAAAAA+EYAAAAMBCOAQAAAAPhGAAAADAQjgEAAAAD4RgAAAAwEI4BAAAAA+EYAAAAMBCOAQAAAAPhGAAAADAQjgEAAAAD4RgAAAAwEI4BAAAAA+EYAAAAMBCOAQAAAAPhGAAAADAQjgEAAAAD4RgAAAAwEI4BAAAAA+EYAAAAMBCOAQAAAAPhGAAAADAQjgEAAAAD4RgAAAAwEI4BAAAAA+EYAAAAMBCOAQAAAAPhGAAAADAQjgEAAAAD4RgAAAAwEI4BAAAAA+EYAAAAMBCOAQAAAMN9CcdbtmxR586d1a5dO61Zs+Z+7AIAAACwO0d7DxgbG6tFixbJbDbL2dlZ/fv31/PPP69q1arZe1cAAACAXdk9HIeHh6tJkyZydXWVJHXo0EGhoaEaNWpUjttZrVZJUkpKir1Lyhf3xxwKuoRCITk5uaBLKDToiQz0RGb0RQb6IjP6IgN9kRl9kaEw9MWtvHkrf97OZM1u6T1YsWKFEhISNG7cOEnSxo0bdeLECc2cOTPH7a5fv66zZ8/asxQAAADgjmrUqKGSJUtmWmb3I8fZJnCT6a7bFS9eXDVq1JCTk1Ou1gcAAADyw2q1ymKxqHjx4lnus3s4Llu2rI4cOWK7HRcXJw8Pj7tuV6RIkSzJHQAAALgfihYtmu1yu1+tomnTpjpw4ICuXLmixMRE7dixQy1atLD3bgAAAAC7uy9HjseNG6chQ4bIYrGod+/eql+/vr13AwAAANid3U/IAwAAAB5WfEMeAAAAYCAcAwAAAAbCMQAAAGAgHAMAAAAGwjEAAABgeKTD8dmzZ1WzZk1t3779gexv8ODBOnjwYK7XP3jwoAYPHixJmjJlik6ePHnPNcTExMjT01Pdu3fP9G/NmjV33MbPz09msznTstjYWA0fPvye6ynsctsj69at07p16+yyz9atW6tz587q3r27unXrpp49e+rbb7+1y9g1a9a8675jYmLueT8ffvihPvzww3seB1kdPHhQzzzzjK0/OnXqpOXLl0vK/+uX1/cmFD43b97UP/7xD7Vr107dunXTgAEDdODAAUnShg0btHXrVknZv5/j0fTyyy/rq6++st2eN2+ennnmGaWkpNiWNWvWTK1atcr2feNW7rh+/bpGjhz5QGp+WNj9OseFidlsVocOHbR+/Xp16NChoMvJ0axZs+w2loeHh4KCgu5pjLJly+qf//ynnSoqvHLbIy+99JJd97ty5UqVL19ekvTNN99o3Lhx2rdvn5ycnOy6HzycPD09tXr1akkZoahz585q165dAVeFgmK1WvXaa6+pdu3aCgkJkbOzs77//nuNGDFC7733no4dO6bnnnuuoMvEA+bt7a2jR4/a3hvCw8Pl5eWliIgIeXt7Kzo6Wo899phSU1Oz3f5W7oiJidHp06cfWN0Pg0c2HKempio4OFhr1qxR//799dNPP6lixYpq3bq1OnXqpD179sjBwUFvvfWWPv30U0VHR2vSpEnq3LmzLl26pGnTpum3336TyWTS3//+dzVt2lQffvihjh8/rgsXLmjgwIEaOHDgHfffrFkzdejQQREREXJwcNAHH3ygChUqKCwsTHPmzJGLi4sqV65sW3/w4MEaNWqUGjZsqBkzZujcuXO6dOmSKleurKVLl+rSpUsaNWqUqlevrlOnTsnNzU2LFy+Wq6trrp8Ti8WiyZMn69y5c5KkAQMGqG/fvrb7ExMT9corr6hr165q2bKlhgwZot27d8vPz0/FihVTRESErl+/rsmTJysoKEinT59W27Zt5efnp/T0dM2ePVsHDhyQyWRSt27dNGLEiLy/cA/QnXpk3rx52r9/vxwcHNSmTRuNGjXKdpR09OjR+uKLLxQUFKTExESZTCZ98MEHqlq1qlq3bq1u3bopLCxMiYmJmjdvnjw9Pe9aR+PGjXXlyhVdv35dly5d0syZM5WQkKArV67o5Zdf1pAhQ3T16lVNmTJFP/74o5ydneXn5ydvb+87jvnbb79p/PjxSkhIUJEiReTv768GDRrY7r9x44YmT56s2NhYxcXFqVGjRpo/f74OHTqkFStWqGjRovrhhx9Us2ZNLVy4UM7OzvrXv/6lgIAAPf744ypVqpTty3327dunJUuWKDU1VeXLl9fMmTP1+OOPq3Xr1qpfv75OnTqltWvXys3N7d5esP9RSUlJcnBwUMmSJW3L7vT6SdLChQu1c+dOOTg4qF+/fho6dKhtu8uXL2vo0KEaO3as2rZt+8AfC/Ln0KFD+vXXX/X555/LZDJJkurUqaPXX39dw4YNU8mSJfXtt9/K3d1dkrRnzx6tXbtWly9f1muvvaZ+/frp5s2beuedd3Tu3DmlpaVp+PDh6tq1q8xmszZv3qyrV6+qVatWeuuttwryoSIPmjRpotmzZ0vK+LTX2dlZHTt2VFhYmLy9vXXkyBE1bdpU+/bt00cffaRTp04pMTFR8+fPl5eXly13rFq1SnFxcXrjjTf00UcfKTAwUJ999pnS09NVt25dTZ8+XS4uLgX8aB+sR3ZaxZ49e1SuXDlVrlxZbdu21fr16233eXh4KCQkRHXr1tXKlSv16aefasGCBVq5cqWkjL+mevXqJbPZrOXLl2vatGm6ceOGJCklJUXbtm3LMRhL0sWLF+Xt7a3AwEA1btxYa9asUUpKivz8/LRkyRKZzeZsv9P72LFjcnJy0oYNG/TVV18pOTlZe/fulSSdPn1aL7/8srZu3apSpUppy5Yt2e47Li4uy7SKM2fO6NixY4qPj1dgYKBWrVqlo0eP2raxWCwaNWqUOnTokO1ji4uLU3BwsMaMGaO3335b//jHPxQYGKiAgABdv35d69at04ULFxQcHKyNGzdqx44d2rNnT84vUgHLrkd++eUX7du3T8HBwVq/fr2ioqKUnJxs2+bGjRvauXOnVq9era1bt6pt27Zau3at7X5XV1dt2rRJ/fv314oVK3JVR1BQkJ5++mmVKVNGGzdu1MiRI/Xll1/q888/16JFiyRJixcvVsWKFfWf//xH8+fP1wcffJDjmJs2bdILL7wgs9msCRMmKCIiIstjr127tjZs2KDt27fr+PHj+u9//yspowenTZum//znP/r1118VFhamkydP6ssvv9TmzZu1atUq/fbbb5KkK1eu6L333tMnn3yiwMBANWvWTAsXLrTtp0WLFtq+fTvBOI8iIyPVvXt3+fr6qnXr1nruuefk4eFhu/9Or19oaKiOHj2qLVu2aOPGjTKbzbp48aIk6fr16xoxYoRGjRpFMH7InDx5Up6enrZgfEvjxo1VrFgxtW7dWmPGjFHz5s0lZfye2rhxo1asWGF7D1m+fLnq1q0rs9msNWvW6OOPP9bPP/8sKSNYbd68mWD8kKlbt65++uknJScnKywsTD4+PvLx8VFYWJgk6ciRI2rWrJkkqVq1agoMDNTgwYP1ySefZBrH399fHh4e+uijj3Tu3DkFBARo/fr1CgoKkpubW5b1/xc8skeOzWazunbtKknq3Lmzxo8fr7Fjx0rK+IUtSeXKlZOHh4ccHR1Vrlw5Xbt2TVLGRxM//vijlixZIinjCOOtN5G8fBX2rTeq6tWr68iRIzpz5ow8PDxUtWpVSVKPHj20ePHiTNs0btxYrq6uWrNmjX788UdFRUUpISFBkuTm5qY6derYxoyPj892v3eaVvH777/r/Pnz+utf/6oWLVpo/PjxtvsWL16sIkWKaOnSpdmOeftzVr16dVvYcXV1VXx8vA4ePKgePXrIwcFBxYoVk6+vrw4cOKAXXnght0/XA3enHnFxcVH//v3VqlUr2+1bSpQooffee08hISGKiorSN998o9q1a9vuv/0137Fjxx33PWLECDk5Ocliseipp56yhV0/Pz998803WrFihc6cOWN77Q8fPmwLnTVr1tSGDRtyfGze3t4aPXq0Tp06pZYtW2rQoEGZ7u/atatOnDihf//73/rxxx919epV276qV6+uJ598UpJUtWpVxcfH6/z582rZsqWKFy8uSerYsaPS09P13Xff6cKFCxoyZIgkKT09XaVLl7btx8vLK8c6kb0/T6t47bXXbH+8S3d+/Q4fPqxOnTrJ2dlZzs7Omd4Hpk+frieeeELt27d/4I8H98ZkMiktLS3LcovFku36bdq0kclkUvXq1fX7779Lyvi9lpSUpC+//FKSlJCQYPsUsU6dOnJ0fGTjwCPLwcFBXl5eOnnypMLCwjRw4EBVqFBBSUlJio+P17FjxzR58mRJsv1BXK1atRzPsTl48KCio6NtnypbLBZb7vhf8kj+NFy+fFn79u1TZGSkPv/8c1mtVl27ds0WVm6f15ndG0J6ero+++wz25SF2NhYPfHEE9q5c6ftaG9sbKxt2oCHh0e283NvhSqTySSr1SqTyaT09HTb/Q4ODlm22bVrl5YsWaIhQ4aoZ8+e+v3333XrG75vD2m3xry1vpRxsk6vXr3u+Lw8/vjjCgkJ0f79+7V371716NFDISEhkqQuXbooISFBS5Ys0aRJk7Jsm5vn7HZWqzXbN/PCIqce2bhxow4dOqR9+/apf//+tpAiSRcuXNDgwYM1aNAgtWjRQk888YROnTplu//211y6c5/cPuf4dmPHjlWpUqXUqlUrde7c2fb6/Pk5/+GHHxQVFZXptX/zzTdt9zds2FAhISHas2ePtm3bZjvie8vq1au1fft29e3bV02bNtXZs2dz7LM/966jo6NSUlKUlpamZ599Vh9//LEkKTk5WTdv3szyfCD/ihcvrrZt2yo8PNy27E6v35/7JCYmRmXKlJEkDR8+XHv37tW6devu+skXChcvLy+tXr1aFosl03vx8ePHVa9evSzr3/rdcvuR5vT0dC1YsEB169aVJF26dEmlS5fWli1bsv0UEw+HW/OOT5w4oQULFtiW7dq1S66urrbpWNn1RHbS0tLUqVMn+fv7S8r447ww/y6/Xx7JaRXBwcFq0qSJ9u3bp927d+vrr7/Wa6+9dtejbbc0adLE9lH5//3f/6lbt25KTEzMtE7ZsmUVFBSkoKCgXJ+4VrNmTV2+fNk28f1W8LndgQMH1KlTJ/Xq1UtPPPGEDh8+nGNjtmnTxlbH7eEoO7t27dL48eP1wgsvyN/fX4899pguXLggSapdu7YmTJigLVu2ZAp7udWkSRMFBgYqLS1NiYmJ2rJli55//vk8j/Og3KlHpk+frkGDBqlx48aaNGmSqlatqvPnz9u2O3nypCpVqqRhw4bJy8tL+/bty/H1yWuf7N+/X2PGjFHbtm11+PBhSRlvVo0aNdK2bdskZQTj4cOHq3Xr1nd87efPn6+goCD16NFD06ZN0/fff59lP/369VO3bt1kMpl0+vTpLH/g3M7b21t79uzR9evXlZycbDtD2svLS8ePH7c9R8uWLbPNfYV9pKWl6dChQ5mO3tzp9WvcuLG++uorWSwWJSYm6tVXX1VsbKykjJ/x6dOna+nSpbZleDg0atRI1apV0+zZs21HiyMjI7V8+XKNHDlSDg4Odw0wTZo0sV1xJy4uTt26dbO9/+Ph1aRJEwUFBalGjRq2P459fHy0atUq+fj45GoMR0dH20l7zz//vL766itdvnxZVqtVM2bM0GeffXbf6i+sHskjx2azWePGjcu0bMCAAfrXv/6lEiVK3HV7f39/TZs2Tb6+vpIygkZutrsbJycnvf/++5owYYIcHR2z/aiiT58+Gj9+vEJDQ+Xs7KwGDRrk+dJNt+Yc3+5W2Nu+fbu6dOkiFxcXtW/fPtOlv1xdXfX3v/9d/v7+tnlqudWvXz9FRUWpe/fuslgs6tatW6E+uz6nHnF3d1fXrl1VrFgx1a5dWy1atLDNx/Xx8dG6devUuXNnOTs7q379+raPJu1h9OjRGjBggEqVKqXKlSvrL3/5i2JiYjRmzBj5+/urW7ducnR01Pz583M8AjB48GD9/e9/1+bNm+Xg4KDp06dnun/o0KGaMWOGPv30UxUvXlzPPPOMYmJiVLFixWzHq127toYOHarevXurVKlSKleunCTJ3d1ds2fP1tixY5Wenq6yZcvajl4g/27NOZYyTpStV6+ehg8fbpsmcafXr0+fPoqMjFTPnj2Vnp6uIUOGZDrx9+mnn9bAgQP1zjvv6KOPPiqQx4b8Wbp0qRYtWqSuXbvKwcFBpUuX1oIFC/T888/r0qVLev/99zOdtPlno0aN0owZM9S1a1elpaVpwoQJqlixoo4cOfIAHwXsrUaNGrp69aoGDBhgW9akSRONHTs21+HYzc1N5cqV0+DBg7V69WqNGjVKQ4cOVXp6umrXrl3oT66/H0zWW5+lAgAAAP/jHslpFQAAAEB+EI4BAAAAA+EYAAAAMBCOAQAAAAPhGAAAADAQjgEAAAAD4RgAAAAw/D/wwBRF64JrLQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 864x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "x, y = df.drop(columns='income'), df['income'].astype('category').cat.codes\n",
    "\n",
    "data = x[[c for c in df.columns if c.startswith('race_')]].idxmax(axis=1).map(lambda c: c[5:])\n",
    "data = pd.concat((data, y), axis=1).rename(columns={0: 'race', 1: 'income'})\n",
    "data = data.groupby('race')['income'].value_counts(normalize=True).mul(100).unstack()\n",
    "fig = data.plot(kind='bar', stacked=True, rot=0, figsize=(12, 5))\n",
    "fig.set(xlabel='', ylabel='', title='Percentage of Individuals per Class');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The above barplot shows that while the *Asian-Pacific Islander* and *White* groups are over-represented in the positive class (*0*), the three remaining groups are under-represented. Since these classifiers are usually deployed to assist human in decision-marking processes in the financial and insurance fields, the over- and under-representation of some protected groups is likely to lead to discriminatory practices towards their members.\n",
    "\n",
    "Similarly to the previous tutorial, we will rely on the **Disparate Impact Discrimination Index**, a metric used to quantify the level of (statistical) discrimination which has been introduced [here](https://arxiv.org/abs/1903.10598). This time, we will use its definition for Classification Tasks ($\\mbox{DIDI}_c$), namely:\n",
    "\n",
    "$$\n",
    "\\mbox{DIDI}_c = \\sum_{p \\in P} \\sum_{c \\in C} \\mid \\bar{y^c} - \\bar{y_p^c} \\mid\n",
    "$$\n",
    "\n",
    "where $D$ is the dataset of the entire population and:\n",
    "* $P$ is the domain of the protected feature and $p$ represents each protected group (e.g., $P = \\{p_0=\\mbox{black}, p_1=\\mbox{hispanic}, p_2=\\mbox{asian}, ...\\}$)\n",
    "* $C$ is the domain of the output classes and $c$ represents each output class (e.g., $C = \\{c_0=\\mbox{accepted}, c_1=\\mbox{rejected}, ...\\}$)\n",
    "* $\\bar{y^c}$ is the average number of samples with class $c$ respectively to the whole population, i.e., $\\bar{y^c} = \\frac{\\mid D^c \\mid}{\\mid D \\mid}$\n",
    "* $\\bar{y_p^c}$ is the average number of samples with class $c$ and protected feature $p$ respectively to the whole protected population, i.e., $\\bar{y_p^c} = \\frac{\\mid D_p^c \\mid}{\\mid D_p \\mid}$\n",
    "\n",
    "Intuitively, the *DIDI* measures the sum of the deviations from the overall percentage number of samples targeted as $c$ respectively to the percentage number of samples targeted as $c$ within each protected group, for all the possible targets $c \\in C$. Thus, as it approaches a zero value, it means that we are reducing the level of discrimination. Let us see which is the violation level on our dataset using the `DIDI` class which can be found inside the Moving Targets' `metric` package. Notice that in case of one-hot encoded protected groups, it is enough to feed the object with a common prefix (i.e., **race**), and it will automatically compute the indicator matrix based on that:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "DIDI: 0.9000396751247404\n"
     ]
    }
   ],
   "source": [
    "from moving_targets.metrics import DIDI\n",
    "\n",
    "didi = DIDI(protected='race', classification=True, percentage=False)\n",
    "print(f'DIDI: {didi(x=x, y=None, p=y)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A pitfall of the *DIDI* is that it is an absolute metric, therefore we have no ways to know whether this is a large amount of discrimination or not. Nonetheless, we can compare different *DIDI* values to understand which of the possible target vectors is less discriminative, or even the level of discrimination of a target vector $y'$ with respect to a target vector $y$ by computing their ratio.\n",
    "\n",
    "> This is exactly what the `percentage` parameter does: it compares the *DIDI* computed on the predictions $p$ with the *DIDI* computed on the original targets $y$ and yields their ratio. Indeed, here we passed no original targets `y` because we wanted the absolute *DIDI* value, to get the relative value we should have done:\n",
    "> ```python\n",
    "> didi = DIDI(protected='race', classification=True, percentage=True)\n",
    "> didi(x=x, y=y, p=y)\n",
    "> ```\n",
    "> which in this case would yield $1.0$ because the two target vectors are the same.\n",
    "\n",
    "Now, say that we measured a certain $\\mbox{DIDI}(y)$ on our original targets. Forcing this value to be exactly zero may lead to excessively poor performances, thus we can formulate our constraint as *20% of the initial discrimination*, or rather:\n",
    "$$\n",
    "\\mbox{DIDI}(p) \\leq 0.2 * \\mbox{DIDI}(y)\n",
    "$$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Moving Targets\n",
    "\n",
    "Let us now split the data points into a train and test set in order to compute our reference *DIDI*, then prepare our custom `Master` class."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8856401259221229"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "x_train, x_test, y_train, y_test = train_test_split(x, y, shuffle=True)\n",
    "didi_train = didi(x=x_train, y=None, p=y_train)\n",
    "didi_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "from moving_targets.masters import ClassificationMaster\n",
    "\n",
    "class FairClassification(ClassificationMaster):\n",
    "    def __init__(self, protected, didi, backend='gurobi', loss='hd', alpha='harmonic', stats=False):\n",
    "        super().__init__(backend=backend, loss=loss, alpha=alpha, stats=stats, types='discrete', labelling=False)\n",
    "        self.protected = protected\n",
    "        self.didi = didi\n",
    "\n",
    "    def build(self, x, y, p):\n",
    "        # 1. RETRIEVE THE VARIABLES\n",
    "        z = super(FairClassification, self).build(x, y, p)\n",
    "        # additionally, we transform 1d variables (outputs of binary tasks) into a bi-dimensional matrix for compatibility\n",
    "        z_bi = np.transpose([1 - z, z]) if z.ndim == 1 else z\n",
    "\n",
    "        # 2. ADD THE CONSTRAINTS\n",
    "        #   > compute the indicator matrix\n",
    "        indicator_matrix = DIDI.get_indicator_matrix(x=x, protected=self.protected)\n",
    "        #   > retrieve the number of protected groups and output classes, respectively\n",
    "        groups, classes = len(indicator_matrix), len(np.unique(y))\n",
    "        #   > add a continuous variable for each pair of protected group and output class\n",
    "        deviations = self.backend.add_continuous_variables(groups, classes, lb=0.0, name='deviations')\n",
    "        #   > compute the percentage number of samples having a certain output class for the whole dataset\n",
    "        total_avg = [self.backend.mean(z_bi[:, c]) for c in range(classes)]\n",
    "        for g, protected_group in enumerate(indicator_matrix):\n",
    "            #   > retrieve the subset of the variables having <label> as protected feature (i.e., the protected group)\n",
    "            protected_vars = z_bi[protected_group]\n",
    "            if len(protected_vars) == 0:\n",
    "                continue\n",
    "            #   > compute the percentage number of samples having a certain output class for the protected group\n",
    "            protected_avg = [self.backend.mean(protected_vars[:, c]) for c in range(classes)]\n",
    "            #   > compute the partial deviation as the absolute value (which is linearized) of the difference between\n",
    "            #   > the total average samples and the average samples within the protected group\n",
    "            self.backend.add_constraints([deviations[g, c] >= total_avg[c] - protected_avg[c] for c in range(classes)])\n",
    "            self.backend.add_constraints([deviations[g, c] >= protected_avg[c] - total_avg[c] for c in range(classes)])\n",
    "        # > eventually compute the DIDI as the sum of this deviations and constrain it\n",
    "        didi = self.backend.sum(deviations)\n",
    "        self.backend.add_constraint(didi <= self.didi)\n",
    "\n",
    "        # 3. RETURN THE (UNMODIFIED) VARIABLES\n",
    "        return z"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This master is very similar to the one discussed in the previous tutorial, with the sole exception of the necessity to deal with categorical outputs (included the trick to handle binary tasks as if they were multi-class tasks) rather than continuous ones. Therefore, for a more detailed explanation of the code, we redirect to that tutorial.\n",
    "\n",
    "Now that we have formulated the master problem, we can build and fit the `MACS` object, with the usual custom callback to the `fit` method to visually address the level of discrimination during the iterations. However, instead of a simple *scikit-learn*-based logistic regression, we may find it useful to train a neural model due to the inner complexity of this predictive problem. Moving Targets' provides some predefined architectures `tensorflow` and `torch`. For the sake of this tutorial, we will rely on a Multilayer Perceptron architecture built upon the former.\n",
    "\n",
    "> Do not forget that, even though the input data may be in each acceptable format (e.g., lists, numpy, pandas), the Moving Targets' library always expects and returns a numpy array for the output data.\n",
    "\n",
    "Finally, since this dataset is bigger than the previous ones, we rely on the `verbose` parameter of the `MACS` object to get check for the progress of the training, as well as creating a custom instance of the `GurobiBackend` class to fix a time limit for the master problem."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Starting the MACS process...\n",
      "  > it. 00: 23.0191 s\n",
      "  > it. 01: 42.1254 s\n",
      "  > it. 02: 53.8674 s\n",
      "  > it. 03: 69.9787 s\n",
      "  > it. 04: 71.6825 s\n",
      "  > it. 05: 61.1946 s\n",
      "  > it. 06: 56.9003 s\n",
      "  > it. 07: 70.4109 s\n",
      "  > it. 08: 69.7178 s\n",
      "  > it. 09: 70.7259 s\n",
      "  > it. 10: 70.8994 s\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x1152 with 11 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "... process ended correctly!\n"
     ]
    }
   ],
   "source": [
    "from moving_targets import MACS\n",
    "from moving_targets.learners import TensorflowMLP\n",
    "from moving_targets.masters.backends import GurobiBackend\n",
    "from moving_targets.metrics import CrossEntropy, Accuracy, DIDI\n",
    "from util.fairness import ClassificationCallback\n",
    "\n",
    "master=FairClassification(protected='race', didi=0.2 * didi_train, backend=GurobiBackend(time_limit=10))\n",
    "learner=TensorflowMLP(loss='binary_crossentropy', output_activation='sigmoid', epochs=30, verbose=False, x_scaler='std')\n",
    "metrics=[CrossEntropy(eps=1e-6), Accuracy(), DIDI(protected='race', classification=True, percentage=True)]\n",
    "model = MACS(init_step='pretraining', master=master, learner=learner, metrics=metrics)\n",
    "\n",
    "callbacks = [ClassificationCallback(protected='race', figsize=(16, 16), tight_layout=True)]\n",
    "val_data = dict(test=(x_test, y_test.values))\n",
    "history = model.fit(x=x_train, y=y_train.values, iterations=10, val_data=val_data, callbacks=callbacks, verbose=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can now plot the training metrics and statistics via the `History` object and inspect them:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x648 with 9 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "history.plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Conclusions\n",
    "\n",
    "Again, using Moving Targets we could satisfy our requirement (at least on the training set, but the performance may rise on the test set as well using more computational power) without losing too much in accuracy.\n",
    "\n",
    "This was the last tutorial. Now, if you have been convinced about the potential of this approach, the last thing to do is to implement your own models and contribute to the development of the library!"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}