quickstart.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "d5e91243",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-08T20:41:26.879421Z",
     "start_time": "2021-06-08T20:41:26.876140Z"
    }
   },
   "source": [
    "# Fairness Sandox Skeleton - QUICKSTART"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "00909876",
   "metadata": {},
   "source": [
    "Easy to Run/Compile Version of the Skeleton Notebook"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0a42ced1",
   "metadata": {},
   "source": [
    "### Setup\n",
    "\n",
    "Please run the code block below to install the necessary packages (if needed)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "4a692724",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-10T20:42:02.586362Z",
     "start_time": "2021-06-10T20:42:02.581026Z"
    }
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "from sklearn.metrics import confusion_matrix, accuracy_score, f1_score, roc_curve, auc\n",
    "from collections import Counter\n",
    "\n",
    "import fairlearn\n",
    "from fairlearn.metrics import *\n",
    "import aif360\n",
    "\n",
    "import copy"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "424b1320-f08f-480f-a185-29526c1f9d59",
   "metadata": {},
   "source": [
    "## Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "996906d7",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-10T20:42:06.040700Z",
     "start_time": "2021-06-10T20:42:06.035368Z"
    }
   },
   "outputs": [],
   "source": [
    "class Dataset:\n",
    "    def __init__(self, short_name = '', path = '', cat_cols = [], num_cols = []):\n",
    "        self.short_name = short_name\n",
    "        self.path = path\n",
    "        self.cat_cols = cat_cols\n",
    "        self.num_cols = num_cols\n",
    "        self.df = pd.read_csv(path, sep = ';')\n",
    "        \n",
    "# each dataset is a dictionary where keys = short name, values = Dataset object\n",
    "\n",
    "datasets = dict()\n",
    "\n",
    "def add_dataset(dataset):\n",
    "    if not isinstance(dataset, Dataset):\n",
    "        print(\"Error! Please enter a valid Dataset object\")\n",
    "    else:\n",
    "        if dataset.short_name not in datasets.keys():\n",
    "            datasets[dataset.short_name] = dataset\n",
    "\n",
    "# example - adding a dataset\n",
    "path_adult_income = 'Datasets/adult.csv'\n",
    "cat_cols = ['workclass', 'education','marital-status', 'occupation', 'relationship', 'race',\n",
    "            'gender', 'native-country','income']\n",
    "num_cols = ['age', 'fnlwgt', 'educational-num', 'capital-gain', 'capital-loss', 'hours-per-week']\n",
    "adult_income = Dataset('adult_income', path_adult_income, cat_cols, num_cols)\n",
    "\n",
    "add_dataset(adult_income)\n",
    "\n",
    "cat = ['school', 'sex', 'address','famsize','Pstatus','Mjob','Fjob','reason',\n",
    "       'guardian','schoolsup','famsup','paid', 'activities','nursery','higher', 'internet','romantic']\n",
    "num = ['age', 'Medu', 'Fedu','traveltime','studytime','failures', 'famrel',\n",
    "       'freetime','goout','Dalc','Walc','health','absences','G1', 'G2', 'G3']\n",
    "\n",
    "add_dataset(Dataset(\"student_mat\", path='Datasets/student-mat.csv', cat_cols=cat, num_cols=num))\n",
    "add_dataset(Dataset(\"student_por\", path='Datasets/student-por.csv', cat_cols=cat, num_cols=num))\n",
    "\n",
    "df_por = datasets['student_por'].df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "199704ec-0811-49fc-aba3-eb4db92faf1e",
   "metadata": {},
   "source": [
    "## EDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "3e5f3bce",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-10T20:42:07.426895Z",
     "start_time": "2021-06-10T20:42:07.269718Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "def plot_counts(df, attr):\n",
    "    if attr in df.columns:\n",
    "        df[attr].value_counts(normalize=True).plot.barh()\n",
    "    else:\n",
    "        print(\"Error! Please enter a valid feature.\")\n",
    "\n",
    "# example\n",
    "# plot_counts(df_por, 'sex')\n",
    "\n",
    "def group_by_plot(df, attr1, attr2):\n",
    "    for val in list(df[attr1].unique()):\n",
    "        print(val)\n",
    "        temp = df[df[attr1] == val]\n",
    "        sns.displot(temp[attr2])\n",
    "\n",
    "# example\n",
    "# group_by_plot(df_por, 'sex', 'G3')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3f42c626",
   "metadata": {},
   "source": [
    "### Problem Formulation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "6ce01c20",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-10T20:42:08.045526Z",
     "start_time": "2021-06-10T20:42:08.023712Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "def threshold(df, g_1=0.3, g_2=0.3, g_3=0.4, threshold=11):\n",
    "    \"\"\"\n",
    "    Added \"pass/fail\" to make problem binary classification\n",
    "    \"\"\"\n",
    "    assert g_1 + g_2 + g_3 == 1, \"The sum of the percentages should be 1\"\n",
    "    assert 0 < threshold < 20, \"Threshold needs to be between 0 and 20\"\n",
    "    df['pass'] = df.apply(lambda row: 1\n",
    "                                 if g_1*row['G1'] + g_2*row['G2'] + g_3*row['G3'] >= threshold\n",
    "                                 else 0, axis=1)\n",
    "threshold(df_por, threshold=14)\n",
    "\n",
    "sens_attrs = [df_por['sex'], df_por['address']]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0d385bc3-625f-4726-a2b3-7dac659fda04",
   "metadata": {},
   "source": [
    "### Data Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "547edc7c-4c9b-4f47-8a61-9398fc20f79f",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# format data\n",
    "X = df_por.iloc[:, :-2].values\n",
    "y = df_por.iloc[:, -1].values\n",
    "\n",
    "# OHE categorical features (prompt for user's choice here?)\n",
    "from sklearn.compose import ColumnTransformer\n",
    "from sklearn.preprocessing import OneHotEncoder\n",
    "\n",
    "# get indices of categorical columns\n",
    "def get_cat_cols(dataset):\n",
    "    df = dataset.df\n",
    "    res = []\n",
    "    for col in dataset.cat_cols:\n",
    "        res.append(df.columns.get_loc(col))\n",
    "    return res\n",
    "\n",
    "cat_cols = get_cat_cols(datasets['student_por'])\n",
    "ct = ColumnTransformer(transformers=[('encoder', OneHotEncoder(), cat_cols)], remainder='passthrough')\n",
    "X = np.array(ct.fit_transform(X))\n",
    "\n",
    "X_true = X\n",
    "y_true = df_por['pass']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "01ac6f6a-a6d0-49cb-922f-feb8d502d8a5",
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD4CAYAAAAXUaZHAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy86wFpkAAAACXBIWXMAAAsTAAALEwEAmpwYAAAK6ElEQVR4nO3db4il51nH8e/P3aYWCt0kOyxhd3ECXZAgmJYlRvpGsghJI25etCVF7FIW9k3ESiu6+kYFXyQIRgulsLjFbSmmoQrZ/gEJSYoUbHTSxpg0lI4hYXfZdKf5pyGtknr5Yu7SybizZ3bnnDnJle8Hhnme+7lnzjUwfPfw5JxJqgpJUi8/N+8BJEnTZ9wlqSHjLkkNGXdJasi4S1JDO+c9AMDu3btrcXFx3mNI0lvKY4899sOqWrjYtTdF3BcXF1laWpr3GJL0lpLkuY2ueVtGkhoy7pLUkHGXpIaMuyQ1ZNwlqSHjLkkNGXdJasi4S1JDxl2SGnpTvEP1rWLx+NfmPUIrz959+7xHkNrymbskNWTcJakh4y5JDRl3SWrIuEtSQ8Zdkhoy7pLUkHGXpIaMuyQ1ZNwlqSHjLkkNGXdJasi4S1JDxl2SGjLuktSQcZekhoy7JDVk3CWpIeMuSQ1tOu5JdiT5TpKvjvPrkzyaZDnJl5JcNdbfOc6Xx/XFGc0uSdrA5Txz/wTw9Jrze4B7q+q9wEvA0bF+FHhprN879kmSttGm4p5kH3A78DfjPMAtwJfHllPAHeP48DhnXD809kuStslmn7n/FfAHwP+O82uBl6vq9XF+Ftg7jvcCZwDG9VfGfknSNpkY9yS/AVyoqsem+cBJjiVZSrK0srIyzW8tSW97m3nm/gHgN5M8C9zH6u2YvwZ2Jdk59uwDzo3jc8B+gHH9PcAL679pVZ2oqoNVdXBhYWFLP4Qk6Y0mxr2q/qiq9lXVInAn8HBV/RbwCPChse0I8MA4Pj3OGdcfrqqa6tSSpEvayuvc/xD4ZJJlVu+pnxzrJ4Frx/ongeNbG1GSdLl2Tt7yM1X1DeAb4/gZ4KaL7Pkx8OEpzCZJukK+Q1WSGjLuktSQcZekhoy7JDVk3CWpIeMuSQ0Zd0lqyLhLUkPGXZIaMu6S1JBxl6SGjLskNWTcJakh4y5JDRl3SWrIuEtSQ8Zdkhoy7pLUkHGXpIaMuyQ1ZNwlqSHjLkkNGXdJasi4S1JDxl2SGjLuktSQcZekhoy7JDVk3CWpIeMuSQ0Zd0lqyLhLUkPGXZIaMu6S1JBxl6SGjLskNWTcJamhiXFP8vNJ/iXJvyV5KsmfjfXrkzyaZDnJl5JcNdbfOc6Xx/XFGf8MkqR1NvPM/b+BW6rql4EbgVuT3AzcA9xbVe8FXgKOjv1HgZfG+r1jnyRpG02Me616dZy+Y3wUcAvw5bF+CrhjHB8e54zrh5JkWgNLkibb1D33JDuSPA5cAB4E/gN4uapeH1vOAnvH8V7gDMC4/gpw7UW+57EkS0mWVlZWtvRDSJLeaFNxr6qfVNWNwD7gJuAXt/rAVXWiqg5W1cGFhYWtfjtJ0hqX9WqZqnoZeAT4VWBXkp3j0j7g3Dg+B+wHGNffA7wwjWElSZuzmVfLLCTZNY7fBfw68DSrkf/Q2HYEeGAcnx7njOsPV1VNcWZJ0gQ7J2/hOuBUkh2s/mNwf1V9Ncl3gfuS/DnwHeDk2H8S+EKSZeBF4M4ZzC1JuoSJca+qJ4D3XWT9GVbvv69f/zHw4alMJ0m6Ir5DVZIaMu6S1JBxl6SGjLskNWTcJakh4y5JDRl3SWrIuEtSQ8Zdkhoy7pLUkHGXpIaMuyQ1ZNwlqSHjLkkNGXdJasi4S1JDxl2SGjLuktSQcZekhoy7JDVk3CWpIeMuSQ0Zd0lqyLhLUkPGXZIaMu6S1JBxl6SGjLskNWTcJakh4y5JDRl3SWrIuEtSQ8Zdkhoy7pLUkHGXpIaMuyQ1ZNwlqaGJcU+yP8kjSb6b5Kkknxjr1yR5MMn3x+erx3qSfDrJcpInkrx/1j+EJOmNNvPM/XXgU1V1A3AzcFeSG4DjwENVdQB4aJwD3AYcGB/HgM9OfWpJ0iVNjHtVna+qb4/j/wKeBvYCh4FTY9sp4I5xfBj4fK36FrAryXXTHlyStLHLuueeZBF4H/AosKeqzo9LzwN7xvFe4MyaLzs71tZ/r2NJlpIsraysXO7ckqRL2HTck7wb+Hvg96rqP9deq6oC6nIeuKpOVNXBqjq4sLBwOV8qSZpg52Y2JXkHq2H/YlX9w1j+QZLrqur8uO1yYayfA/av+fJ9Y03SjCwe/9q8R2jl2btvn/cIW7aZV8sEOAk8XVV/uebSaeDIOD4CPLBm/WPjVTM3A6+suX0jSdoGm3nm/gHgt4F/T/L4WPtj4G7g/iRHgeeAj4xrXwc+CCwDrwEfn+bAkqTJJsa9qr4JZIPLhy6yv4C7tjiXJGkLfIeqJDVk3CWpIeMuSQ0Zd0lqyLhLUkPGXZIaMu6S1JBxl6SGjLskNWTcJakh4y5JDRl3SWrIuEtSQ8Zdkhoy7pLUkHGXpIaMuyQ1ZNwlqSHjLkkNGXdJasi4S1JDxl2SGjLuktSQcZekhoy7JDVk3CWpIeMuSQ0Zd0lqyLhLUkPGXZIaMu6S1JBxl6SGjLskNWTcJakh4y5JDRl3SWrIuEtSQxPjnuRzSS4keXLN2jVJHkzy/fH56rGeJJ9OspzkiSTvn+XwkqSL28wz978Fbl23dhx4qKoOAA+Nc4DbgAPj4xjw2emMKUm6HBPjXlX/BLy4bvkwcGocnwLuWLP++Vr1LWBXkuumNKskaZOu9J77nqo6P46fB/aM473AmTX7zo61/yfJsSRLSZZWVlaucAxJ0sVs+T+oVlUBdQVfd6KqDlbVwYWFha2OIUla40rj/oOf3m4Zny+M9XPA/jX79o01SdI2utK4nwaOjOMjwANr1j82XjVzM/DKmts3kqRtsnPShiR/B/wasDvJWeBPgLuB+5McBZ4DPjK2fx34ILAMvAZ8fAYzS5ImmBj3qvroBpcOXWRvAXdtdShJ0tb4DlVJasi4S1JDxl2SGjLuktSQcZekhoy7JDVk3CWpIeMuSQ0Zd0lqyLhLUkPGXZIaMu6S1JBxl6SGjLskNWTcJakh4y5JDRl3SWrIuEtSQ8Zdkhoy7pLUkHGXpIaMuyQ1ZNwlqSHjLkkNGXdJasi4S1JDxl2SGjLuktSQcZekhoy7JDVk3CWpIeMuSQ0Zd0lqyLhLUkPGXZIaMu6S1JBxl6SGZhL3JLcm+V6S5STHZ/EYkqSNTT3uSXYAnwFuA24APprkhmk/jiRpY7N45n4TsFxVz1TV/wD3AYdn8DiSpA3snMH33AucWXN+FviV9ZuSHAOOjdNXk3xvBrO8Xe0GfjjvISbJPfOeQHPg7+Z0/cJGF2YR902pqhPAiXk9fmdJlqrq4LznkNbzd3P7zOK2zDlg/5rzfWNNkrRNZhH3fwUOJLk+yVXAncDpGTyOJGkDU78tU1WvJ/kd4B+BHcDnquqpaT+OLsnbXXqz8ndzm6Sq5j2DJGnKfIeqJDVk3CWpIePeRJLFJE+uW/vTJL8/r5mkn0rykySPJ3kyyVeS7Jr3TN0Zd0nb4UdVdWNV/RLwInDXvAfqzrhL2m7/zOo72TVDxl3Sthl/WPAQvvdl5ox7Hxu9ptXXuurN4F1JHgeeB/YAD853nP6Mex8vAFevW7uGt8AfadLbwo+q6kZW/9BV8J77zBn3JqrqVeB8klsAklwD3Ap8c66DSWtU1WvA7wKfSjK3P1z4duA7VBsZ/1OUz/CzZ/B/UVVfnONIEgBJXq2qd685/wpwf1V9YY5jtWbcJakhb8tIUkPGXZIaMu6S1JBxl6SGjLskNWTcJakh4y5JDf0fRAMwlwg14S8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "count = pd.value_counts(df_por['address'], sort = True)\n",
    "count.plot(kind = 'bar', rot = 0)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f143a18a-0041-4399-9821-2250410920de",
   "metadata": {},
   "source": [
    "### Under-Sampling Process"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "144c9d4b-2b36-4f86-b2be-d0c8b5c442cc",
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# separate based on protected attribute\n",
    "sens_values = sens_attrs[1].unique()\n",
    "\n",
    "# TODO - add prompt for user to specify which value is favored and which is unfavored\n",
    "\n",
    "df_favored = df_por[df_por['address'] == 'U']\n",
    "df_unfavored = df_por[df_por['address'] == 'R']\n",
    "\n",
    "# under-sampling process\n",
    "df_undersampled = df_unfavored.sample(n=190, random_state=42)\n",
    "\n",
    "#print(df_favored.shape, df_unfavored.shape, df_undersampled.shape)\n",
    "\n",
    "# combine undersampled and original favored class to create dataset\n",
    "df_concat = pd.concat([df_favored,df_undersampled])\n",
    "df_concat.shape\n",
    "\n",
    "# for fairness measures later\n",
    "df_sens = df_concat['address']\n",
    "\n",
    "# format data\n",
    "X_bias = df_concat.iloc[:, :-2].values\n",
    "#print(X_undersampled.shape)\n",
    "y_bias = df_concat.iloc[:, -1].values\n",
    "\n",
    "# OHE\n",
    "ct = ColumnTransformer(transformers=[('encoder', OneHotEncoder(), cat_cols)], remainder='passthrough')\n",
    "X_bias_true = np.array(ct.fit_transform(X_bias))\n",
    "y_bias_true = df_concat['pass']\n",
    "\n",
    "#print(X_true.shape, X_bias_true.shape)\n",
    "#print(y_true.value_counts(), \"\\n\\n\", y_bias_true.value_counts())\n",
    "\n",
    "favored = len(df_favored)\n",
    "true_unfavored = len(df_por[df_por['address'] == 'R'])\n",
    "bias_unfavored = len(df_undersampled)\n",
    "\n",
    "x_vals = ['Favored', \"Unfavored\"]\n",
    "y_vals_true = [favored, true_unfavored]\n",
    "y_vals_bias = [favored, bias_unfavored]\n",
    "\n",
    "plt.subplot(1,2,1)\n",
    "plt.bar(x_vals, y_vals_true)\n",
    "plt.title(\"Ground Truth\")\n",
    "plt.ylabel(\"Count\")\n",
    "\n",
    "plt.subplot(1,2,2)\n",
    "plt.bar(x_vals, y_vals_bias)\n",
    "plt.title(\"Under-Sampling\")\n",
    "plt.ylim([0,500])\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "25f6827f",
   "metadata": {},
   "source": [
    "### Model Selection + Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "72d79618",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-10T20:42:09.057618Z",
     "start_time": "2021-06-10T20:42:08.942574Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy of Ground Truth Model on Ground Truth Data:  0.9845916795069337\n",
      "Accuracy of Biased Model on Biased Data:  0.9844236760124611\n",
      "Accuracy of Biased Model on Ground Truth Data:  0.9830508474576272\n"
     ]
    }
   ],
   "source": [
    "# modularize and add data struct of different ml techniques\n",
    "\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "\n",
    "classifier = DecisionTreeClassifier(min_samples_leaf = 10, max_depth = 4)\n",
    "\n",
    "classifier_true = classifier.fit(X_true, y_true)\n",
    "y_pred_truth = classifier_true.predict(X_true)\n",
    "\n",
    "classifier_bias = classifier.fit(X_bias_true, y_bias_true)\n",
    "y_pred_bias = classifier_bias.predict(X_bias_true)\n",
    "y_pred_bias_on_true = classifier_bias.predict(X_true)\n",
    "\n",
    "print(\"Accuracy of Ground Truth Model on Ground Truth Data: \", accuracy_score(y_pred_truth, y_true))\n",
    "print(\"Accuracy of Biased Model on Biased Data: \", accuracy_score(y_pred_bias, y_bias_true))\n",
    "print(\"Accuracy of Biased Model on Ground Truth Data: \", accuracy_score(y_pred_bias_on_true, y_true))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b00743e0-c0b9-4886-9eae-2d319faf1f00",
   "metadata": {},
   "source": [
    "Thus, we can see that under-sampling decreases model performance on the ground truth data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "17d049d3",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-10T21:16:04.515563Z",
     "start_time": "2021-06-10T21:16:04.499948Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Overall Accuracy:  0.9845916795069337\n",
      "Group Accuracy :  address\n",
      "R    0.979695\n",
      "U    0.986726\n",
      "Name: accuracy_score, dtype: object\n",
      "\n",
      "\n",
      "Overall Selection Rate:  0.2110939907550077\n",
      "Group Selection Rate :  address\n",
      "R    0.177665\n",
      "U    0.225664\n",
      "Name: selection_rate, dtype: object\n"
     ]
    }
   ],
   "source": [
    "# Ground Truth Model on Ground Truth Data\n",
    "\n",
    "gm_true = MetricFrame(accuracy_score, y_true, y_pred_truth, sensitive_features = sens_attrs[1])\n",
    "print(\"Overall Accuracy: \", gm_true.overall)\n",
    "print(\"Group Accuracy : \", gm_true.by_group)\n",
    "\n",
    "print(\"\\n\")\n",
    "\n",
    "sr_true = MetricFrame(selection_rate, y_true, y_pred_truth, sensitive_features = sens_attrs[1])\n",
    "print(\"Overall Selection Rate: \", sr_true.overall)\n",
    "print(\"Group Selection Rate : \", sr_true.by_group)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3c40158f",
   "metadata": {},
   "source": [
    "Thus, we can see that there is a **5% discrepancy** in the selection rate for Rural vs Urban students (with Urban students being preferred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "ec50ddde-b262-4c56-bda3-70c2df3988f4",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Overall Accuracy:  0.9844236760124611\n",
      "Group Accuracy :  address\n",
      "R    0.978947\n",
      "U    0.986726\n",
      "Name: accuracy_score, dtype: object\n",
      "\n",
      "\n",
      "Overall Selection Rate:  0.21495327102803738\n",
      "Group Selection Rate :  address\n",
      "R    0.168421\n",
      "U    0.234513\n",
      "Name: selection_rate, dtype: object\n"
     ]
    }
   ],
   "source": [
    "# Biased Model on Biased Data\n",
    "\n",
    "gm_bias = MetricFrame(accuracy_score, y_bias_true, y_pred_bias, sensitive_features = df_sens)\n",
    "print(\"Overall Accuracy: \", gm_bias.overall)\n",
    "print(\"Group Accuracy : \", gm_bias.by_group)\n",
    "\n",
    "print(\"\\n\")\n",
    "\n",
    "sr_bias = MetricFrame(selection_rate, y_bias_true, y_pred_bias, sensitive_features = df_sens)\n",
    "print(\"Overall Selection Rate: \", sr_bias.overall)\n",
    "print(\"Group Selection Rate : \", sr_bias.by_group)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b9877b0a-453a-4c0e-9ac5-c95dbb92aa0f",
   "metadata": {},
   "source": [
    "Thus, we can see that there is now a **7% discrepancy due to under-sampling** in the selection rate for Rural vs Urban students (with Urban students being preferred) on biased data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "97b1ea82-14b2-47f6-9030-16bf6f11814a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Overall Accuracy:  0.9830508474576272\n",
      "Group Accuracy :  address\n",
      "R    0.974619\n",
      "U    0.986726\n",
      "Name: accuracy_score, dtype: object\n",
      "\n",
      "\n",
      "Overall Selection Rate:  0.21571648690292758\n",
      "Group Selection Rate :  address\n",
      "R    0.172589\n",
      "U    0.234513\n",
      "Name: selection_rate, dtype: object\n"
     ]
    }
   ],
   "source": [
    "# Biased Model on Ground Truth Data\n",
    "\n",
    "gm_bias_on_true = MetricFrame(accuracy_score, y_true,\n",
    "                           y_pred_bias_on_true, sensitive_features = sens_attrs[1])\n",
    "print(\"Overall Accuracy: \", gm_bias_on_true.overall)\n",
    "print(\"Group Accuracy : \", gm_bias_on_true.by_group)\n",
    "\n",
    "print(\"\\n\")\n",
    "\n",
    "sr_bias_on_true = MetricFrame(selection_rate, y_true,\n",
    "                              y_pred_bias_on_true, sensitive_features = sens_attrs[1])\n",
    "print(\"Overall Selection Rate: \", sr_bias_on_true.overall)\n",
    "print(\"Group Selection Rate : \", sr_bias_on_true.by_group)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a828b712-e2f9-47b2-b308-9d41b356b2cc",
   "metadata": {},
   "source": [
    "Thus, we see that the biased model has lower overall and subgroup accuracy on the ground truth data. \n",
    "Also, we can observe that the selection rate disparity increased ever so slightly with the biased model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "10d7a575",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-10T21:16:54.784535Z",
     "start_time": "2021-06-10T21:16:54.725029Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ground Truth Model on Ground Truth Data\n",
      "EOD: 0.047169811320754707\n",
      "DPD: 0.04799874219486994\n",
      "EOR: 0.11994219653179189\n",
      "DPR: 0.7872996914501842\n",
      "\n",
      "\n",
      "Biased Model on Ground Truth Data\n",
      "EOD: 0.015425865310954804\n",
      "DPD: 0.061924441848973555\n",
      "EOR: 0.35982658959537567\n",
      "DPR: 0.7359448328704147\n"
     ]
    }
   ],
   "source": [
    "print(\"Ground Truth Model on Ground Truth Data\")\n",
    "\n",
    "print(\"EOD:\", equalized_odds_difference(y_true=y_true, y_pred = y_pred_truth, sensitive_features=sens_attrs[1]))\n",
    "print(\"DPD:\", demographic_parity_difference(y_true=y_true, y_pred = y_pred_truth, sensitive_features=sens_attrs[1]))\n",
    "\n",
    "print(\"EOR:\", equalized_odds_ratio(y_true=y_true, y_pred = y_pred_truth, sensitive_features=sens_attrs[1]))\n",
    "print(\"DPR:\", demographic_parity_ratio(y_true=y_true, y_pred = y_pred_truth, sensitive_features=sens_attrs[1]))\n",
    "\n",
    "print(\"\\n\")\n",
    "\n",
    "print(\"Biased Model on Ground Truth Data\")\n",
    "\n",
    "print(\"EOD:\", equalized_odds_difference(y_true=y_true, y_pred = y_pred_bias_on_true, sensitive_features=sens_attrs[1]))\n",
    "print(\"DPD:\", demographic_parity_difference(y_true=y_true, y_pred = y_pred_bias_on_true, sensitive_features=sens_attrs[1]))\n",
    "\n",
    "print(\"EOR:\", equalized_odds_ratio(y_true=y_true, y_pred = y_pred_bias_on_true, sensitive_features=sens_attrs[1]))\n",
    "print(\"DPR:\", demographic_parity_ratio(y_true=y_true, y_pred = y_pred_bias_on_true, sensitive_features=sens_attrs[1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "27629fa9",
   "metadata": {},
   "source": [
    "### Fairness Visualization (Keep this?)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "103a8164",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T17:51:41.613558Z",
     "start_time": "2021-06-09T17:51:41.586931Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "from fairlearn.widget import FairlearnDashboard\n",
    "FairlearnDashboard(sensitive_features = df_por['sex'],\n",
    "                   sensitive_feature_names = ['sex'],\n",
    "                   y_true = y_true,\n",
    "                   y_pred = {\"initial model\" : y_pred_truth})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "19ae6920",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T17:51:41.635383Z",
     "start_time": "2021-06-09T17:51:41.615653Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "from fairlearn.widget import FairlearnDashboard\n",
    "FairlearnDashboard(sensitive_features = df_por['address'],\n",
    "                   sensitive_feature_names = ['address'],\n",
    "                   y_true = y_true,\n",
    "                   y_pred = {\"initial model\" : y_pred_bias_on_true})"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8c85b935",
   "metadata": {},
   "source": [
    "# Fairness Intervention"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "3db7ace8-9024-47f7-9215-88463035949a",
   "metadata": {},
   "outputs": [],
   "source": [
    "from fairlearn.reductions import ExponentiatedGradient, DemographicParity\n",
    "np.random.seed(0)\n",
    "\n",
    "constraint = DemographicParity()\n",
    "mitigator_true = ExponentiatedGradient(classifier_true, constraint)\n",
    "mitigator_true.fit(X_true, y_true, sensitive_features = sens_attrs[1])\n",
    "y_pred_mitigated_true = mitigator_true.predict(X_true)\n",
    "\n",
    "constraint = DemographicParity()\n",
    "mitigator_bias = ExponentiatedGradient(classifier_bias, constraint)\n",
    "mitigator_bias.fit(X_bias_true, y_bias_true, sensitive_features = df_sens)\n",
    "y_pred_mitigated_bias = mitigator_bias.predict(X_bias_true)\n",
    "y_pred_mitigated_bias_on_true = mitigator_bias.predict(X_true)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2b7ea20d",
   "metadata": {},
   "source": [
    "# Evaluation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "40d85690-7232-405f-9256-d987abd01b22",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy of Ground Truth Model + Fairness Intervention on Ground Truth Data:  0.9691833590138675\n",
      "Accuracy of Biased Model + Fairness Intervention on Ground Truth Data:  0.975346687211094\n"
     ]
    }
   ],
   "source": [
    "print(\"Accuracy of Ground Truth Model + Fairness Intervention on Ground Truth Data: \",\n",
    "      accuracy_score(y_pred_mitigated_true, y_true))\n",
    "\n",
    "print(\"Accuracy of Biased Model + Fairness Intervention on Ground Truth Data: \",\n",
    "      accuracy_score(y_pred_mitigated_bias_on_true, y_true))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "ac90ed8c",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T17:51:42.986098Z",
     "start_time": "2021-06-09T17:51:42.973042Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Overall Accuracy:  0.9691833590138675\n",
      "Group Accuracy :  address\n",
      "R    0.928934\n",
      "U    0.986726\n",
      "Name: accuracy_score, dtype: object\n",
      "\n",
      "\n",
      "Overall Selection Rate:  0.22650231124807396\n",
      "Group Selection Rate :  address\n",
      "R    0.228426\n",
      "U    0.225664\n",
      "Name: selection_rate, dtype: object\n"
     ]
    }
   ],
   "source": [
    "# Ground Truth Model + Fairness Intervention on Ground Truth Data\n",
    "\n",
    "gm_mitigated = MetricFrame(accuracy_score, y_true, y_pred_mitigated_true, sensitive_features = sens_attrs[1])\n",
    "print(\"Overall Accuracy: \", gm_mitigated.overall)\n",
    "print(\"Group Accuracy : \", gm_mitigated.by_group)\n",
    "\n",
    "print(\"\\n\")\n",
    "\n",
    "sr_mitigated = MetricFrame(selection_rate, y_true, y_pred_mitigated_true, sensitive_features = sens_attrs[1])\n",
    "print(\"Overall Selection Rate: \", sr_mitigated.overall)\n",
    "print(\"Group Selection Rate : \", sr_mitigated.by_group)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "a3d5b09e-f879-4a03-ba44-9a1a66999204",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Overall Accuracy:  0.975346687211094\n",
      "Group Accuracy :  address\n",
      "R    0.949239\n",
      "U    0.986726\n",
      "Name: accuracy_score, dtype: object\n",
      "\n",
      "\n",
      "Overall Selection Rate:  0.22033898305084745\n",
      "Group Selection Rate :  address\n",
      "R    0.208122\n",
      "U    0.225664\n",
      "Name: selection_rate, dtype: object\n"
     ]
    }
   ],
   "source": [
    "# Biased Model + Fairness Intervention on Ground Truth Data\n",
    "\n",
    "gm_mitigated_bias_on_true = MetricFrame(accuracy_score, y_true, y_pred_mitigated_bias_on_true, sensitive_features = sens_attrs[1])\n",
    "print(\"Overall Accuracy: \", gm_mitigated_bias_on_true.overall)\n",
    "print(\"Group Accuracy : \", gm_mitigated_bias_on_true.by_group)\n",
    "\n",
    "print(\"\\n\")\n",
    "\n",
    "sr_mitigated_bias_on_true = MetricFrame(selection_rate, y_true, y_pred_mitigated_bias_on_true, sensitive_features = sens_attrs[1])\n",
    "print(\"Overall Selection Rate: \", sr_mitigated_bias_on_true.overall)\n",
    "print(\"Group Selection Rate : \", sr_mitigated_bias_on_true.by_group)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2a51941c",
   "metadata": {},
   "source": [
    "Hence, we observe an increase in overall and subgroup accuracy, but a decrease in disadvantaged subgroup selection rate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "35164d1d",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-06-09T17:51:43.021839Z",
     "start_time": "2021-06-09T17:51:42.998287Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "FairlearnDashboard(sensitive_features = sens_attrs[1],\n",
    "                   sensitive_feature_names = ['address'],\n",
    "                   y_true = y_true,\n",
    "                   y_pred = {\"initial model\" : y_pred_truth, \"mitigated model\": y_pred_mitigated_true})"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "95e393d0-149b-4da6-91ba-e74e5225b0e2",
   "metadata": {},
   "source": [
    "### Bias vs Accuracy vs Fairness Trade-Off"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "48687919-16ae-446a-8db3-7120e07fc5b7",
   "metadata": {},
   "outputs": [],
   "source": [
    "# if verbose, shows \"Finished iteration: ... \"\n",
    "# if apply_fairness, uses fairness intervention\n",
    "def tradeoff_visualization(classifier, apply_fairness = False, verbose = False):\n",
    "    \n",
    "    bias_amts = list(range(0,200,10))\n",
    "    accuracy_on_true = []\n",
    "    accuracy_on_biased = []\n",
    "    eod_on_true = []\n",
    "    eod_on_biased = []\n",
    "    dataset_size_true = np.full(shape=len(bias_amts), fill_value= X_true.shape[0]).tolist()\n",
    "    dataset_size_bias = []\n",
    "    table = []\n",
    "\n",
    "    classifier_true = classifier.fit(X_true, y_true)\n",
    "    y_pred_truth = classifier_true.predict(X_true)\n",
    "\n",
    "    df_undersampled = df_unfavored.sample(n=len(df_unfavored), random_state=42)\n",
    "\n",
    "    for i in range(20):\n",
    "        # under-sampling process\n",
    "        if i == 0:\n",
    "            df_undersampled = df_undersampled.sample(n=len(df_undersampled), random_state=42)\n",
    "        else:\n",
    "            df_undersampled = df_undersampled.sample(n=len(df_undersampled)-10, random_state=42)\n",
    "\n",
    "        # combine undersampled and original favored class to create dataset\n",
    "        df_concat = pd.concat([df_favored,df_undersampled])\n",
    "        df_concat.shape\n",
    "        df_sens = df_concat['address']\n",
    "\n",
    "        # format data\n",
    "        X_bias = df_concat.iloc[:, :-2].values\n",
    "        y_bias = df_concat.iloc[:, -1].values\n",
    "\n",
    "        # OHE\n",
    "        ct = ColumnTransformer(transformers=[('encoder', OneHotEncoder(), cat_cols)], remainder='passthrough')\n",
    "        X_bias_true = np.array(ct.fit_transform(X_bias))\n",
    "        y_bias_true = df_concat['pass']\n",
    "\n",
    "        dataset_size_bias.append(X_bias_true.shape[0])\n",
    "        classifier_bias = classifier.fit(X_bias_true, y_bias_true)\n",
    "        \n",
    "        if apply_fairness:\n",
    "            constraint = DemographicParity()\n",
    "            mitigator_bias = ExponentiatedGradient(classifier_bias, constraint)\n",
    "            mitigator_bias.fit(X_bias_true, y_bias_true, sensitive_features = df_sens)\n",
    "            y_pred_bias = classifier_bias.predict(X_bias_true)\n",
    "            y_pred_bias_on_true = mitigator_bias.predict(X_true)\n",
    "        \n",
    "        else:\n",
    "            y_pred_bias = classifier_bias.predict(X_bias_true)\n",
    "            y_pred_bias_on_true = classifier_bias.predict(X_true)\n",
    "\n",
    "        # model performance\n",
    "        acc_bias = accuracy_score(y_pred=y_pred_bias, y_true=y_bias_true)\n",
    "        accuracy_on_biased.append(acc_bias)\n",
    "\n",
    "        acc_bias_on_true = accuracy_score(y_pred=y_pred_bias_on_true, y_true=y_true)\n",
    "        accuracy_on_true.append(acc_bias_on_true)\n",
    "\n",
    "        # fairness performance\n",
    "        eod_true = equalized_odds_difference(y_true=y_bias_true, y_pred = y_pred_bias, sensitive_features=df_sens)\n",
    "        eod_on_true.append(eod_true)\n",
    "\n",
    "        eod_bias_on_true = equalized_odds_difference(y_true=y_true, y_pred = y_pred_bias_on_true, sensitive_features=sens_attrs[1])\n",
    "        eod_on_biased.append(eod_bias_on_true)\n",
    "\n",
    "        # table visualization \n",
    "        table_elem = [i*10, acc_bias, acc_bias_on_true]\n",
    "        table.append(table_elem)\n",
    "        \n",
    "        if verbose:\n",
    "            print(\"Finished Iteration: \", len(df_concat))\n",
    "\n",
    "    return bias_amts, dataset_size_true, dataset_size_bias, accuracy_on_biased, accuracy_on_true, eod_on_biased, eod_on_true"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "99bfd23f-6f40-45d1-a77f-e93d9b460459",
   "metadata": {},
   "outputs": [],
   "source": [
    "def visualizations(bias_amts, dataset_size_true, dataset_size_bias, accuracy_on_biased, accuracy_on_true,\n",
    "                   eod_on_true, eod_on_biased, fairness = False):\n",
    "    \n",
    "    if not fairness:\n",
    "        plt.figure(figsize=(17,7))\n",
    "\n",
    "        plt.subplot(1,2,1)\n",
    "        plt.plot(bias_amts, accuracy_on_true, label = 'Ground Truth')\n",
    "        plt.plot(bias_amts, accuracy_on_biased, label = 'Biased Data')\n",
    "        plt.xlabel(\"Amount of Bias (number of minority samples removed)\")\n",
    "        plt.ylabel(\"Accuracy Score\")\n",
    "        plt.axhline(y=accuracy_score(y_pred_truth, y_true), color = \"green\", label = \"Ground Truth Model Accuracy\", alpha = 0.5)\n",
    "        plt.title(\"Biased Model Accuracy\")\n",
    "        plt.ylim(0.92, 0.99)\n",
    "        plt.legend()\n",
    "\n",
    "        plt.subplot(1,2,2)\n",
    "        plt.plot(bias_amts, dataset_size_true, label = 'Ground Truth')\n",
    "        plt.plot(bias_amts, dataset_size_bias, label = 'Biased Data')\n",
    "        plt.xlabel(\"Amount of Bias (number of minority samples removed)\")\n",
    "        plt.ylabel(\"Dataset Size\")\n",
    "        plt.legend()\n",
    "\n",
    "        plt.show()\n",
    "    else:\n",
    "        \n",
    "        plt.plot(bias_amts, eod_on_true, label = 'Ground Truth')\n",
    "        plt.plot(bias_amts, eod_on_biased, label = 'Biased Data')\n",
    "        plt.xlabel(\"Amount of Bias (number of minority samples removed)\")\n",
    "        plt.ylabel(\"Equalized Odds Difference\")\n",
    "        plt.axhline(y=equalized_odds_difference(y_true=y_true, y_pred = y_pred_truth, sensitive_features=sens_attrs[1]), color = \"green\",\n",
    "                    label = \"Ground Truth EOD\", alpha = 0.5)\n",
    "        plt.legend()\n",
    "        plt.title(\"Biased Model Equalized Odds Difference\")\n",
    "        plt.show()\n",
    "        #plt.savefig(\"bias1.png\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "04f2ec56-f37c-49f3-8341-6933c11498b5",
   "metadata": {},
   "outputs": [],
   "source": [
    "bias_amts, dataset_size_true, dataset_size_bias, accuracy_on_biased, accuracy_on_true, eod_on_biased, eod_on_true = tradeoff_visualization(classifier, False, False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "bbe24539-8c4d-4d0e-ae9d-6a192b1525da",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1224x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "visualizations(bias_amts, dataset_size_true, dataset_size_bias, accuracy_on_biased, accuracy_on_true, eod_on_true, eod_on_biased, False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "d191369d-cee4-4b27-b260-2e1eca3e3a4a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "visualizations(bias_amts, dataset_size_true, dataset_size_bias, accuracy_on_biased, accuracy_on_true, eod_on_true, eod_on_biased, True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "abe7015c-a460-4e43-907c-0d19896145e7",
   "metadata": {},
   "outputs": [],
   "source": [
    "bias_amts, dataset_size_true, dataset_size_bias, accuracy_on_biased, accuracy_on_true, eod_on_biased, eod_on_true = tradeoff_visualization(classifier, True, False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "2b97fc83-fa23-407f-8246-c2cf9108aa9a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1224x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "visualizations(bias_amts, dataset_size_true, dataset_size_bias, accuracy_on_biased, accuracy_on_true, eod_on_true, eod_on_biased, False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "f67d72c5-c381-43be-a0a5-08019a5742b6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "visualizations(bias_amts, dataset_size_true, dataset_size_bias, accuracy_on_biased, accuracy_on_true, eod_on_true, eod_on_biased, True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bad30644-086e-4595-b159-2b158721c0fa",
   "metadata": {},
   "source": [
    "### Special Visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "e5fddfea-6572-460b-b360-a2eb250bb37d",
   "metadata": {},
   "outputs": [],
   "source": [
    "bias_amts = list(range(0,200,10))\n",
    "accuracy_on_true = []\n",
    "accuracy_on_biased = []\n",
    "accuracy_on_true_mitigated = []\n",
    "eod_on_true = []\n",
    "eod_on_biased = []\n",
    "eod_on_true_mitigated = []\n",
    "dataset_size_true = np.full(shape=len(bias_amts), fill_value= X_true.shape[0]).tolist()\n",
    "dataset_size_bias = []\n",
    "table = []\n",
    "\n",
    "classifier_true = classifier.fit(X_true, y_true)\n",
    "y_pred_truth = classifier_true.predict(X_true)\n",
    "\n",
    "df_undersampled = df_unfavored.sample(n=len(df_unfavored), random_state=42)\n",
    "\n",
    "for i in range(20):\n",
    "    # under-sampling process\n",
    "    if i == 0:\n",
    "        df_undersampled = df_undersampled.sample(n=len(df_undersampled), random_state=42)\n",
    "    else:\n",
    "        df_undersampled = df_undersampled.sample(n=len(df_undersampled)-10, random_state=42)\n",
    "\n",
    "    # combine undersampled and original favored class to create dataset\n",
    "    df_concat = pd.concat([df_favored,df_undersampled])\n",
    "    df_concat.shape\n",
    "    df_sens = df_concat['address']\n",
    "\n",
    "    # format data\n",
    "    X_bias = df_concat.iloc[:, :-2].values\n",
    "    y_bias = df_concat.iloc[:, -1].values\n",
    "\n",
    "    # OHE\n",
    "    ct = ColumnTransformer(transformers=[('encoder', OneHotEncoder(), cat_cols)], remainder='passthrough')\n",
    "    X_bias_true = np.array(ct.fit_transform(X_bias))\n",
    "    y_bias_true = df_concat['pass']\n",
    "\n",
    "    dataset_size_bias.append(X_bias_true.shape[0])\n",
    "    classifier_bias = classifier.fit(X_bias_true, y_bias_true)\n",
    "\n",
    "    constraint = DemographicParity()\n",
    "    mitigator_bias = ExponentiatedGradient(classifier_bias, constraint)\n",
    "    mitigator_bias.fit(X_bias_true, y_bias_true, sensitive_features = df_sens)\n",
    "    y_pred_mitigated_bias = classifier_bias.predict(X_bias_true)\n",
    "    y_pred_mitigated_bias_on_true = mitigator_bias.predict(X_true)\n",
    "\n",
    "    y_pred_bias = classifier_bias.predict(X_bias_true)\n",
    "    y_pred_bias_on_true = classifier_bias.predict(X_true)\n",
    "\n",
    "    # model performance\n",
    "    acc_bias = accuracy_score(y_pred=y_pred_bias, y_true=y_bias_true)\n",
    "    accuracy_on_biased.append(acc_bias)\n",
    "\n",
    "    acc_bias_on_true = accuracy_score(y_pred=y_pred_bias_on_true, y_true=y_true)\n",
    "    accuracy_on_true.append(acc_bias_on_true)\n",
    "    \n",
    "    acc_bias_mitigated_on_true = accuracy_score(y_pred=y_pred_mitigated_bias_on_true, y_true=y_true)\n",
    "    accuracy_on_true_mitigated.append(acc_bias_mitigated_on_true)\n",
    "\n",
    "    # fairness performance\n",
    "    eod_true = equalized_odds_difference(y_true=y_bias_true, y_pred = y_pred_bias, sensitive_features=df_sens)\n",
    "    eod_on_true.append(eod_true)\n",
    "\n",
    "    eod_bias_on_true = equalized_odds_difference(y_true=y_true, y_pred = y_pred_bias_on_true, sensitive_features=sens_attrs[1])\n",
    "    eod_on_biased.append(eod_bias_on_true)\n",
    "    \n",
    "    eod_bias_true_mitigated = equalized_odds_difference(y_true=y_true, y_pred = y_pred_mitigated_bias_on_true, sensitive_features=sens_attrs[1])\n",
    "    eod_on_true_mitigated.append(eod_bias_true_mitigated)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "f6297243-9839-471a-b22a-eecba0414cc9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1224x504 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(17,7))\n",
    "\n",
    "plt.subplot(1,2,1)\n",
    "plt.plot(bias_amts, accuracy_on_biased, label = 'On Biased Data + No Fairness Intervention', color = \"red\")\n",
    "plt.plot(bias_amts, accuracy_on_true, label = 'On Ground Truth + No Fairness Intervention', color = \"blue\")\n",
    "plt.plot(bias_amts, accuracy_on_true_mitigated, label = 'On Ground Truth + Fairness Intervention', color = \"purple\")\n",
    "plt.xlabel(\"Amount of Bias (number of minority samples removed)\")\n",
    "plt.ylabel(\"Accuracy Score\")\n",
    "plt.axhline(y=accuracy_score(y_pred_truth, y_true), color = \"green\", label = \"Ground Truth Model On Ground Truth Data\", alpha = 0.5)\n",
    "plt.title(\"Accuracy of Biased Model (trained on biased data) \\n\\nNote: ground truth model trained on ground truth data means \\na barebones DT classifier trained on unbiased data\")\n",
    "plt.ylim(0.92, 0.99)\n",
    "plt.legend()\n",
    "\n",
    "plt.subplot(1,2,2)\n",
    "plt.plot(bias_amts, dataset_size_true, label = 'Ground Truth')\n",
    "plt.plot(bias_amts, dataset_size_bias, label = 'Biased Data')\n",
    "plt.xlabel(\"Amount of Bias (number of minority samples removed)\")\n",
    "plt.ylabel(\"Dataset Size\")\n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "38888381-710d-407b-9db1-541bcabcdd60",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(bias_amts, eod_on_biased, label = 'On Biased Data + No Fairness Intervention', color = \"red\")\n",
    "plt.plot(bias_amts, eod_on_true, label = 'On Ground Truth + No Fairness Intervention', color = \"blue\")\n",
    "plt.plot(bias_amts, eod_on_true_mitigated, label = 'On Ground Truth + Fairness Intervention', color = \"purple\")\n",
    "plt.xlabel(\"Amount of Bias (number of minority samples removed)\")\n",
    "plt.ylabel(\"Equalized Odds Difference\")\n",
    "plt.axhline(y=equalized_odds_difference(y_true=y_true, y_pred = y_pred_truth, sensitive_features=sens_attrs[1]), color = \"green\",\n",
    "            label = \"Ground Truth EOD\", alpha = 0.5)\n",
    "plt.legend()\n",
    "plt.title(\"Biased Model Equalized Odds Difference\")\n",
    "plt.show()\n",
    "#plt.savefig(\"bias1.png\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}