Jupyter_code.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "a5ae889e",
   "metadata": {},
   "source": [
    "## Title:  Discussion on the ethical concerns of COMPASS"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "00e36390",
   "metadata": {},
   "source": [
    "# Content of the document\n",
    " 1. Problem Statement \n",
    " 2. Data Collection\n",
    " 3. Data Preparation \n",
    " 4. Data Preprocessing   \n",
    " 5. Model Assessement (Fairness) & optimization\n",
    " 7. Conclusion\n",
    " 8. References "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6c622863",
   "metadata": {},
   "source": [
    "# 1. Problem Statement\n",
    "\n",
    "We are investigating COMPAS, a popular commercial algorithm for predicting reoffending likelihoods (recidivism) for criminal defendants. There has been evidence that the algorithm is biased towards white defendants and against black inmates. Next, use a fairness approach to improve the model.\n",
    "\n",
    "In order to accomplish this, I need first to explore the data and prepare it, then assess the fairness, then use one of the approaches to optimize it"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "91a43d3d",
   "metadata": {},
   "source": [
    "Here are the steps I will take in this project: \n",
    "    \n",
    "     1. Data Collection\n",
    "     2. Data Exploration: This will be done to identify the most important features and combine them in new ways.\n",
    "     3. Data Preprocessing: Lay out a pipeline of tasks for transforming data for use in my machine learning model.\n",
    "     4. Model Assessment: Determine the type of descrimination.\n",
    "     6. How to improve the fairness \n",
    "     7. Conclusion & recommendations     "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "52f8a0ec",
   "metadata": {},
   "source": [
    "# 2. Data Collection\n",
    "\n",
    "In this step I do two things: \n",
    "    \n",
    "  - Identify data sources\n",
    "  - Split the data into training and test sets\n",
    "\n",
    "\n",
    "Before starting, as a first step, I will call some libraries I need in order to build my model.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "be4cc83b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Libraries \n",
    "import seaborn as sns\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from IPython.display import set_matplotlib_formats\n",
    "%matplotlib inline\n",
    "import os\n",
    "import warnings\n",
    "from sklearn.model_selection import StratifiedShuffleSplit\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "from sklearn.datasets import make_regression, make_classification, make_blobs\n",
    "import sklearn.model_selection\n",
    "import sklearn.linear_model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "949550e7",
   "metadata": {},
   "source": [
    "#### Source of the data: (kaggle.com, n.d.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "bbdba879",
   "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>id</th>\n",
       "      <th>name</th>\n",
       "      <th>first</th>\n",
       "      <th>last</th>\n",
       "      <th>compas_screening_date</th>\n",
       "      <th>sex</th>\n",
       "      <th>dob</th>\n",
       "      <th>age</th>\n",
       "      <th>age_cat</th>\n",
       "      <th>race</th>\n",
       "      <th>...</th>\n",
       "      <th>v_type_of_assessment</th>\n",
       "      <th>v_decile_score</th>\n",
       "      <th>v_score_text</th>\n",
       "      <th>v_screening_date</th>\n",
       "      <th>in_custody</th>\n",
       "      <th>out_custody</th>\n",
       "      <th>priors_count.1</th>\n",
       "      <th>start</th>\n",
       "      <th>end</th>\n",
       "      <th>event</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>miguel hernandez</td>\n",
       "      <td>miguel</td>\n",
       "      <td>hernandez</td>\n",
       "      <td>14/08/2013</td>\n",
       "      <td>Male</td>\n",
       "      <td>18/04/1947</td>\n",
       "      <td>69</td>\n",
       "      <td>Greater than 45</td>\n",
       "      <td>Other</td>\n",
       "      <td>...</td>\n",
       "      <td>Risk of Violence</td>\n",
       "      <td>1</td>\n",
       "      <td>Low</td>\n",
       "      <td>14/08/2013</td>\n",
       "      <td>07/07/2014</td>\n",
       "      <td>14/07/2014</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>327</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>miguel hernandez</td>\n",
       "      <td>miguel</td>\n",
       "      <td>hernandez</td>\n",
       "      <td>14/08/2013</td>\n",
       "      <td>Male</td>\n",
       "      <td>18/04/1947</td>\n",
       "      <td>69</td>\n",
       "      <td>Greater than 45</td>\n",
       "      <td>Other</td>\n",
       "      <td>...</td>\n",
       "      <td>Risk of Violence</td>\n",
       "      <td>1</td>\n",
       "      <td>Low</td>\n",
       "      <td>14/08/2013</td>\n",
       "      <td>07/07/2014</td>\n",
       "      <td>14/07/2014</td>\n",
       "      <td>0</td>\n",
       "      <td>334</td>\n",
       "      <td>961</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3.0</td>\n",
       "      <td>michael ryan</td>\n",
       "      <td>michael</td>\n",
       "      <td>ryan</td>\n",
       "      <td>31/12/2014</td>\n",
       "      <td>Male</td>\n",
       "      <td>06/02/1985</td>\n",
       "      <td>31</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>Caucasian</td>\n",
       "      <td>...</td>\n",
       "      <td>Risk of Violence</td>\n",
       "      <td>2</td>\n",
       "      <td>Low</td>\n",
       "      <td>31/12/2014</td>\n",
       "      <td>30/12/2014</td>\n",
       "      <td>03/01/2015</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>457</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.0</td>\n",
       "      <td>kevon dixon</td>\n",
       "      <td>kevon</td>\n",
       "      <td>dixon</td>\n",
       "      <td>27/01/2013</td>\n",
       "      <td>Male</td>\n",
       "      <td>22/01/1982</td>\n",
       "      <td>34</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>African-American</td>\n",
       "      <td>...</td>\n",
       "      <td>Risk of Violence</td>\n",
       "      <td>1</td>\n",
       "      <td>Low</td>\n",
       "      <td>27/01/2013</td>\n",
       "      <td>26/01/2013</td>\n",
       "      <td>05/02/2013</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "      <td>159</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5.0</td>\n",
       "      <td>ed philo</td>\n",
       "      <td>ed</td>\n",
       "      <td>philo</td>\n",
       "      <td>14/04/2013</td>\n",
       "      <td>Male</td>\n",
       "      <td>14/05/1991</td>\n",
       "      <td>24</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>African-American</td>\n",
       "      <td>...</td>\n",
       "      <td>Risk of Violence</td>\n",
       "      <td>3</td>\n",
       "      <td>Low</td>\n",
       "      <td>14/04/2013</td>\n",
       "      <td>16/06/2013</td>\n",
       "      <td>16/06/2013</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>63</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 52 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    id              name    first       last compas_screening_date   sex  \\\n",
       "0  1.0  miguel hernandez   miguel  hernandez            14/08/2013  Male   \n",
       "1  2.0  miguel hernandez   miguel  hernandez            14/08/2013  Male   \n",
       "2  3.0      michael ryan  michael       ryan            31/12/2014  Male   \n",
       "3  4.0       kevon dixon    kevon      dixon            27/01/2013  Male   \n",
       "4  5.0          ed philo       ed      philo            14/04/2013  Male   \n",
       "\n",
       "          dob  age          age_cat              race  ...  \\\n",
       "0  18/04/1947   69  Greater than 45             Other  ...   \n",
       "1  18/04/1947   69  Greater than 45             Other  ...   \n",
       "2  06/02/1985   31          25 - 45         Caucasian  ...   \n",
       "3  22/01/1982   34          25 - 45  African-American  ...   \n",
       "4  14/05/1991   24     Less than 25  African-American  ...   \n",
       "\n",
       "   v_type_of_assessment  v_decile_score  v_score_text  v_screening_date  \\\n",
       "0      Risk of Violence               1           Low        14/08/2013   \n",
       "1      Risk of Violence               1           Low        14/08/2013   \n",
       "2      Risk of Violence               2           Low        31/12/2014   \n",
       "3      Risk of Violence               1           Low        27/01/2013   \n",
       "4      Risk of Violence               3           Low        14/04/2013   \n",
       "\n",
       "   in_custody  out_custody priors_count.1 start  end event  \n",
       "0  07/07/2014   14/07/2014              0     0  327     0  \n",
       "1  07/07/2014   14/07/2014              0   334  961     0  \n",
       "2  30/12/2014   03/01/2015              0     3  457     0  \n",
       "3  26/01/2013   05/02/2013              0     9  159     1  \n",
       "4  16/06/2013   16/06/2013              4     0   63     0  \n",
       "\n",
       "[5 rows x 52 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Load the data from Kaggle Repository\n",
    "initial_data = pd.read_csv('cox-violent-parsed.csv')\n",
    "\n",
    "# Examine date structure and return the top 5 rows of the data frame.\n",
    "initial_data.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b5930fe2",
   "metadata": {},
   "source": [
    "In the table above, the date columns are displayed in object format"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "8f58a913",
   "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>id</th>\n",
       "      <th>name</th>\n",
       "      <th>first</th>\n",
       "      <th>last</th>\n",
       "      <th>compas_screening_date</th>\n",
       "      <th>sex</th>\n",
       "      <th>dob</th>\n",
       "      <th>age</th>\n",
       "      <th>age_cat</th>\n",
       "      <th>race</th>\n",
       "      <th>...</th>\n",
       "      <th>v_type_of_assessment</th>\n",
       "      <th>v_decile_score</th>\n",
       "      <th>v_score_text</th>\n",
       "      <th>v_screening_date</th>\n",
       "      <th>in_custody</th>\n",
       "      <th>out_custody</th>\n",
       "      <th>priors_count.1</th>\n",
       "      <th>start</th>\n",
       "      <th>end</th>\n",
       "      <th>event</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>miguel hernandez</td>\n",
       "      <td>miguel</td>\n",
       "      <td>hernandez</td>\n",
       "      <td>2013-08-14</td>\n",
       "      <td>Male</td>\n",
       "      <td>1947-04-18</td>\n",
       "      <td>69</td>\n",
       "      <td>Greater than 45</td>\n",
       "      <td>Other</td>\n",
       "      <td>...</td>\n",
       "      <td>Risk of Violence</td>\n",
       "      <td>1</td>\n",
       "      <td>Low</td>\n",
       "      <td>2013-08-14</td>\n",
       "      <td>2014-07-07</td>\n",
       "      <td>2014-07-14</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>327</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>miguel hernandez</td>\n",
       "      <td>miguel</td>\n",
       "      <td>hernandez</td>\n",
       "      <td>2013-08-14</td>\n",
       "      <td>Male</td>\n",
       "      <td>1947-04-18</td>\n",
       "      <td>69</td>\n",
       "      <td>Greater than 45</td>\n",
       "      <td>Other</td>\n",
       "      <td>...</td>\n",
       "      <td>Risk of Violence</td>\n",
       "      <td>1</td>\n",
       "      <td>Low</td>\n",
       "      <td>2013-08-14</td>\n",
       "      <td>2014-07-07</td>\n",
       "      <td>2014-07-14</td>\n",
       "      <td>0</td>\n",
       "      <td>334</td>\n",
       "      <td>961</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3.0</td>\n",
       "      <td>michael ryan</td>\n",
       "      <td>michael</td>\n",
       "      <td>ryan</td>\n",
       "      <td>2014-12-31</td>\n",
       "      <td>Male</td>\n",
       "      <td>1985-06-02</td>\n",
       "      <td>31</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>Caucasian</td>\n",
       "      <td>...</td>\n",
       "      <td>Risk of Violence</td>\n",
       "      <td>2</td>\n",
       "      <td>Low</td>\n",
       "      <td>2014-12-31</td>\n",
       "      <td>2014-12-30</td>\n",
       "      <td>2015-03-01</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>457</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.0</td>\n",
       "      <td>kevon dixon</td>\n",
       "      <td>kevon</td>\n",
       "      <td>dixon</td>\n",
       "      <td>2013-01-27</td>\n",
       "      <td>Male</td>\n",
       "      <td>1982-01-22</td>\n",
       "      <td>34</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>African-American</td>\n",
       "      <td>...</td>\n",
       "      <td>Risk of Violence</td>\n",
       "      <td>1</td>\n",
       "      <td>Low</td>\n",
       "      <td>2013-01-27</td>\n",
       "      <td>2013-01-26</td>\n",
       "      <td>2013-05-02</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "      <td>159</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5.0</td>\n",
       "      <td>ed philo</td>\n",
       "      <td>ed</td>\n",
       "      <td>philo</td>\n",
       "      <td>2013-04-14</td>\n",
       "      <td>Male</td>\n",
       "      <td>1991-05-14</td>\n",
       "      <td>24</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>African-American</td>\n",
       "      <td>...</td>\n",
       "      <td>Risk of Violence</td>\n",
       "      <td>3</td>\n",
       "      <td>Low</td>\n",
       "      <td>2013-04-14</td>\n",
       "      <td>2013-06-16</td>\n",
       "      <td>2013-06-16</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>63</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 52 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    id              name    first       last compas_screening_date   sex  \\\n",
       "0  1.0  miguel hernandez   miguel  hernandez            2013-08-14  Male   \n",
       "1  2.0  miguel hernandez   miguel  hernandez            2013-08-14  Male   \n",
       "2  3.0      michael ryan  michael       ryan            2014-12-31  Male   \n",
       "3  4.0       kevon dixon    kevon      dixon            2013-01-27  Male   \n",
       "4  5.0          ed philo       ed      philo            2013-04-14  Male   \n",
       "\n",
       "         dob  age          age_cat              race  ...  \\\n",
       "0 1947-04-18   69  Greater than 45             Other  ...   \n",
       "1 1947-04-18   69  Greater than 45             Other  ...   \n",
       "2 1985-06-02   31          25 - 45         Caucasian  ...   \n",
       "3 1982-01-22   34          25 - 45  African-American  ...   \n",
       "4 1991-05-14   24     Less than 25  African-American  ...   \n",
       "\n",
       "   v_type_of_assessment  v_decile_score  v_score_text  v_screening_date  \\\n",
       "0      Risk of Violence               1           Low        2013-08-14   \n",
       "1      Risk of Violence               1           Low        2013-08-14   \n",
       "2      Risk of Violence               2           Low        2014-12-31   \n",
       "3      Risk of Violence               1           Low        2013-01-27   \n",
       "4      Risk of Violence               3           Low        2013-04-14   \n",
       "\n",
       "   in_custody  out_custody priors_count.1 start  end event  \n",
       "0  2014-07-07   2014-07-14              0     0  327     0  \n",
       "1  2014-07-07   2014-07-14              0   334  961     0  \n",
       "2  2014-12-30   2015-03-01              0     3  457     0  \n",
       "3  2013-01-26   2013-05-02              0     9  159     1  \n",
       "4  2013-06-16   2013-06-16              4     0   63     0  \n",
       "\n",
       "[5 rows x 52 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Convert date columns from object format to time format\n",
    "\n",
    "# attributes is the list of columns to be converted\n",
    "attributes= [\"in_custody\",\"out_custody\",\"v_screening_date\",\"compas_screening_date\", \"dob\",\"c_jail_in\",\"c_jail_out\",\"c_offense_date\",\n",
    "             \"screening_date\",\"vr_offense_date\",\"r_jail_out\",\"r_jail_in\",\"r_offense_date\",\"c_arrest_date\"]\n",
    "\n",
    "initial_data = pd.read_csv('cox-violent-parsed.csv',parse_dates=attributes)\n",
    "\n",
    "# Create a copy of the original data\n",
    "my_data = initial_data.copy()\n",
    "\n",
    "# Examine my data structure and return the top 5 rows of the data frame.\n",
    "my_data.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "64be8329",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 18316 entries, 0 to 18315\n",
      "Data columns (total 52 columns):\n",
      " #   Column                   Non-Null Count  Dtype         \n",
      "---  ------                   --------------  -----         \n",
      " 0   id                       11001 non-null  float64       \n",
      " 1   name                     18316 non-null  object        \n",
      " 2   first                    18316 non-null  object        \n",
      " 3   last                     18316 non-null  object        \n",
      " 4   compas_screening_date    18316 non-null  datetime64[ns]\n",
      " 5   sex                      18316 non-null  object        \n",
      " 6   dob                      18316 non-null  datetime64[ns]\n",
      " 7   age                      18316 non-null  int64         \n",
      " 8   age_cat                  18316 non-null  object        \n",
      " 9   race                     18316 non-null  object        \n",
      " 10  juv_fel_count            18316 non-null  int64         \n",
      " 11  decile_score             18316 non-null  int64         \n",
      " 12  juv_misd_count           18316 non-null  int64         \n",
      " 13  juv_other_count          18316 non-null  int64         \n",
      " 14  priors_count             18316 non-null  int64         \n",
      " 15  days_b_screening_arrest  17019 non-null  float64       \n",
      " 16  c_jail_in                17019 non-null  datetime64[ns]\n",
      " 17  c_jail_out               17019 non-null  datetime64[ns]\n",
      " 18  c_case_number            17449 non-null  object        \n",
      " 19  c_offense_date           14364 non-null  datetime64[ns]\n",
      " 20  c_arrest_date            3085 non-null   datetime64[ns]\n",
      " 21  c_days_from_compas       17449 non-null  float64       \n",
      " 22  c_charge_degree          17449 non-null  object        \n",
      " 23  c_charge_desc            17435 non-null  object        \n",
      " 24  is_recid                 18316 non-null  int64         \n",
      " 25  r_case_number            8417 non-null   object        \n",
      " 26  r_charge_degree          8417 non-null   object        \n",
      " 27  r_days_from_arrest       6359 non-null   float64       \n",
      " 28  r_offense_date           8417 non-null   datetime64[ns]\n",
      " 29  r_charge_desc            8277 non-null   object        \n",
      " 30  r_jail_in                6359 non-null   datetime64[ns]\n",
      " 31  r_jail_out               6359 non-null   datetime64[ns]\n",
      " 32  violent_recid            0 non-null      float64       \n",
      " 33  is_violent_recid         18316 non-null  int64         \n",
      " 34  vr_case_number           1339 non-null   object        \n",
      " 35  vr_charge_degree         1339 non-null   object        \n",
      " 36  vr_offense_date          1339 non-null   datetime64[ns]\n",
      " 37  vr_charge_desc           1339 non-null   object        \n",
      " 38  type_of_assessment       18316 non-null  object        \n",
      " 39  decile_score.1           18316 non-null  int64         \n",
      " 40  score_text               18293 non-null  object        \n",
      " 41  screening_date           18316 non-null  datetime64[ns]\n",
      " 42  v_type_of_assessment     18316 non-null  object        \n",
      " 43  v_decile_score           18316 non-null  int64         \n",
      " 44  v_score_text             18310 non-null  object        \n",
      " 45  v_screening_date         18316 non-null  datetime64[ns]\n",
      " 46  in_custody               17722 non-null  datetime64[ns]\n",
      " 47  out_custody              17722 non-null  datetime64[ns]\n",
      " 48  priors_count.1           18316 non-null  int64         \n",
      " 49  start                    18316 non-null  int64         \n",
      " 50  end                      18316 non-null  int64         \n",
      " 51  event                    18316 non-null  int64         \n",
      "dtypes: datetime64[ns](14), float64(5), int64(14), object(19)\n",
      "memory usage: 7.3+ MB\n"
     ]
    }
   ],
   "source": [
    "#Check the type data of my attributes\n",
    "my_data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "2a8405a2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "id                          7315\n",
       "days_b_screening_arrest     1297\n",
       "c_jail_in                   1297\n",
       "c_jail_out                  1297\n",
       "c_case_number                867\n",
       "c_offense_date              3952\n",
       "c_arrest_date              15231\n",
       "c_days_from_compas           867\n",
       "c_charge_degree              867\n",
       "c_charge_desc                881\n",
       "r_case_number               9899\n",
       "r_charge_degree             9899\n",
       "r_days_from_arrest         11957\n",
       "r_offense_date              9899\n",
       "r_charge_desc              10039\n",
       "r_jail_in                  11957\n",
       "r_jail_out                 11957\n",
       "violent_recid              18316\n",
       "vr_case_number             16977\n",
       "vr_charge_degree           16977\n",
       "vr_offense_date            16977\n",
       "vr_charge_desc             16977\n",
       "score_text                    23\n",
       "v_score_text                   6\n",
       "in_custody                   594\n",
       "out_custody                  594\n",
       "dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Check if there are null values in my dataset\n",
    "my_data_not_nut = my_data.isnull().sum()\n",
    "#display non null data\n",
    "my_data_not_nut[my_data_not_nut>0]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a74cb1a6",
   "metadata": {},
   "source": [
    "48% (25 from 51) of the attributes has null values!"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4137662a",
   "metadata": {},
   "source": [
    "The dataset I have has 52 attributes, to make my analysis more efficient I will drop the unecessary ones to my study"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "4696d51e",
   "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>id</th>\n",
       "      <th>name</th>\n",
       "      <th>dob</th>\n",
       "      <th>compas_screening_date</th>\n",
       "      <th>c_offense_date</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>age_cat</th>\n",
       "      <th>race</th>\n",
       "      <th>c_charge_degree</th>\n",
       "      <th>c_charge_desc</th>\n",
       "      <th>days_b_screening_arrest</th>\n",
       "      <th>decile_score</th>\n",
       "      <th>is_recid</th>\n",
       "      <th>r_offense_date</th>\n",
       "      <th>c_case_number</th>\n",
       "      <th>v_decile_score</th>\n",
       "      <th>is_violent_recid</th>\n",
       "      <th>vr_offense_date</th>\n",
       "      <th>score_text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>miguel hernandez</td>\n",
       "      <td>1947-04-18</td>\n",
       "      <td>2013-08-14</td>\n",
       "      <td>2013-08-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>69</td>\n",
       "      <td>Greater than 45</td>\n",
       "      <td>Other</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Aggravated Assault w/Firearm</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>13011352CF10A</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>miguel hernandez</td>\n",
       "      <td>1947-04-18</td>\n",
       "      <td>2013-08-14</td>\n",
       "      <td>2013-08-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>69</td>\n",
       "      <td>Greater than 45</td>\n",
       "      <td>Other</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Aggravated Assault w/Firearm</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>13011352CF10A</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3.0</td>\n",
       "      <td>michael ryan</td>\n",
       "      <td>1985-06-02</td>\n",
       "      <td>2014-12-31</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Male</td>\n",
       "      <td>31</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>Caucasian</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5</td>\n",
       "      <td>-1</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Medium</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.0</td>\n",
       "      <td>kevon dixon</td>\n",
       "      <td>1982-01-22</td>\n",
       "      <td>2013-01-27</td>\n",
       "      <td>2013-01-26</td>\n",
       "      <td>Male</td>\n",
       "      <td>34</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Felony Battery w/Prior Convict</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2013-05-07</td>\n",
       "      <td>13001275CF10A</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2013-05-07</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5.0</td>\n",
       "      <td>ed philo</td>\n",
       "      <td>1991-05-14</td>\n",
       "      <td>2013-04-14</td>\n",
       "      <td>2013-04-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>24</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Cocaine</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>2013-06-16</td>\n",
       "      <td>13005330CF10A</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18311</th>\n",
       "      <td>NaN</td>\n",
       "      <td>alexsandra beauchamps</td>\n",
       "      <td>1984-12-21</td>\n",
       "      <td>2014-12-29</td>\n",
       "      <td>2014-12-28</td>\n",
       "      <td>Female</td>\n",
       "      <td>31</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(M1)</td>\n",
       "      <td>Battery</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>14018106MM10A</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Medium</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18312</th>\n",
       "      <td>NaN</td>\n",
       "      <td>winston gregory</td>\n",
       "      <td>1958-01-10</td>\n",
       "      <td>2014-01-14</td>\n",
       "      <td>2014-01-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>57</td>\n",
       "      <td>Greater than 45</td>\n",
       "      <td>Other</td>\n",
       "      <td>(F2)</td>\n",
       "      <td>Aggravated Battery / Pregnant</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>14000581CF10A</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18313</th>\n",
       "      <td>NaN</td>\n",
       "      <td>farrah jean</td>\n",
       "      <td>1982-11-17</td>\n",
       "      <td>2014-09-03</td>\n",
       "      <td>2014-08-03</td>\n",
       "      <td>Female</td>\n",
       "      <td>33</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(M1)</td>\n",
       "      <td>Battery on Law Enforc Officer</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>14003308CF10A</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18314</th>\n",
       "      <td>NaN</td>\n",
       "      <td>florencia sanmartin</td>\n",
       "      <td>1992-12-18</td>\n",
       "      <td>2014-06-30</td>\n",
       "      <td>2014-06-28</td>\n",
       "      <td>Female</td>\n",
       "      <td>23</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>Hispanic</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Ethylone</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>2015-03-15</td>\n",
       "      <td>14008895CF10A</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18315</th>\n",
       "      <td>NaN</td>\n",
       "      <td>florencia sanmartin</td>\n",
       "      <td>1992-12-18</td>\n",
       "      <td>2014-06-30</td>\n",
       "      <td>2014-06-28</td>\n",
       "      <td>Female</td>\n",
       "      <td>23</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>Hispanic</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Ethylone</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>2015-03-15</td>\n",
       "      <td>14008895CF10A</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>18316 rows × 20 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        id                   name        dob compas_screening_date  \\\n",
       "0      1.0       miguel hernandez 1947-04-18            2013-08-14   \n",
       "1      2.0       miguel hernandez 1947-04-18            2013-08-14   \n",
       "2      3.0           michael ryan 1985-06-02            2014-12-31   \n",
       "3      4.0            kevon dixon 1982-01-22            2013-01-27   \n",
       "4      5.0               ed philo 1991-05-14            2013-04-14   \n",
       "...    ...                    ...        ...                   ...   \n",
       "18311  NaN  alexsandra beauchamps 1984-12-21            2014-12-29   \n",
       "18312  NaN        winston gregory 1958-01-10            2014-01-14   \n",
       "18313  NaN            farrah jean 1982-11-17            2014-09-03   \n",
       "18314  NaN    florencia sanmartin 1992-12-18            2014-06-30   \n",
       "18315  NaN    florencia sanmartin 1992-12-18            2014-06-30   \n",
       "\n",
       "      c_offense_date     sex  age          age_cat              race  \\\n",
       "0         2013-08-13    Male   69  Greater than 45             Other   \n",
       "1         2013-08-13    Male   69  Greater than 45             Other   \n",
       "2                NaT    Male   31          25 - 45         Caucasian   \n",
       "3         2013-01-26    Male   34          25 - 45  African-American   \n",
       "4         2013-04-13    Male   24     Less than 25  African-American   \n",
       "...              ...     ...  ...              ...               ...   \n",
       "18311     2014-12-28  Female   31          25 - 45  African-American   \n",
       "18312     2014-01-13    Male   57  Greater than 45             Other   \n",
       "18313     2014-08-03  Female   33          25 - 45  African-American   \n",
       "18314     2014-06-28  Female   23     Less than 25          Hispanic   \n",
       "18315     2014-06-28  Female   23     Less than 25          Hispanic   \n",
       "\n",
       "      c_charge_degree                   c_charge_desc  \\\n",
       "0                (F3)    Aggravated Assault w/Firearm   \n",
       "1                (F3)    Aggravated Assault w/Firearm   \n",
       "2                 NaN                             NaN   \n",
       "3                (F3)  Felony Battery w/Prior Convict   \n",
       "4                (F3)           Possession of Cocaine   \n",
       "...               ...                             ...   \n",
       "18311            (M1)                         Battery   \n",
       "18312            (F2)   Aggravated Battery / Pregnant   \n",
       "18313            (M1)   Battery on Law Enforc Officer   \n",
       "18314            (F3)          Possession of Ethylone   \n",
       "18315            (F3)          Possession of Ethylone   \n",
       "\n",
       "       days_b_screening_arrest  decile_score  is_recid r_offense_date  \\\n",
       "0                         -1.0             1         0            NaT   \n",
       "1                         -1.0             1         0            NaT   \n",
       "2                          NaN             5        -1            NaT   \n",
       "3                         -1.0             3         1     2013-05-07   \n",
       "4                         -1.0             4         1     2013-06-16   \n",
       "...                        ...           ...       ...            ...   \n",
       "18311                     -1.0             6         0            NaT   \n",
       "18312                     -1.0             1         0            NaT   \n",
       "18313                     -1.0             2         0            NaT   \n",
       "18314                     -2.0             4         1     2015-03-15   \n",
       "18315                     -2.0             4         1     2015-03-15   \n",
       "\n",
       "       c_case_number  v_decile_score  is_violent_recid vr_offense_date  \\\n",
       "0      13011352CF10A               1                 0             NaT   \n",
       "1      13011352CF10A               1                 0             NaT   \n",
       "2                NaN               2                 0             NaT   \n",
       "3      13001275CF10A               1                 1      2013-05-07   \n",
       "4      13005330CF10A               3                 0             NaT   \n",
       "...              ...             ...               ...             ...   \n",
       "18311  14018106MM10A               4                 0             NaT   \n",
       "18312  14000581CF10A               1                 0             NaT   \n",
       "18313  14003308CF10A               2                 0             NaT   \n",
       "18314  14008895CF10A               4                 0             NaT   \n",
       "18315  14008895CF10A               4                 0             NaT   \n",
       "\n",
       "      score_text  \n",
       "0            Low  \n",
       "1            Low  \n",
       "2         Medium  \n",
       "3            Low  \n",
       "4            Low  \n",
       "...          ...  \n",
       "18311     Medium  \n",
       "18312        Low  \n",
       "18313        Low  \n",
       "18314        Low  \n",
       "18315        Low  \n",
       "\n",
       "[18316 rows x 20 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#List of the attributes I am keeping\n",
    "needed_attributes = [\"id\",\"name\",\"dob\",\"compas_screening_date\",\"c_offense_date\",\"sex\",\"age\",\"age_cat\",\"race\",\"c_charge_degree\",\"c_charge_desc\", \n",
    "\"days_b_screening_arrest\", \"decile_score\", \"is_recid\",\"r_offense_date\", \"c_case_number\",\"v_decile_score\", \n",
    "\"is_violent_recid\",\"vr_offense_date\",\"score_text\"] \n",
    "\n",
    "#Copy my data in a new variable, to keep the original one untouched\n",
    "analysis = my_data.loc[:,needed_attributes].copy()\n",
    "analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "2b0eb7af",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>name</th>\n",
       "      <th>dob</th>\n",
       "      <th>compas_screening_date</th>\n",
       "      <th>c_offense_date</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>age_cat</th>\n",
       "      <th>race</th>\n",
       "      <th>c_charge_degree</th>\n",
       "      <th>c_charge_desc</th>\n",
       "      <th>days_b_screening_arrest</th>\n",
       "      <th>decile_score</th>\n",
       "      <th>is_recid</th>\n",
       "      <th>r_offense_date</th>\n",
       "      <th>c_case_number</th>\n",
       "      <th>v_decile_score</th>\n",
       "      <th>is_violent_recid</th>\n",
       "      <th>vr_offense_date</th>\n",
       "      <th>score_text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>miguel hernandez</td>\n",
       "      <td>1947-04-18</td>\n",
       "      <td>2013-08-14</td>\n",
       "      <td>2013-08-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>69</td>\n",
       "      <td>Greater than 45</td>\n",
       "      <td>Other</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Aggravated Assault w/Firearm</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>13011352CF10A</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>miguel hernandez</td>\n",
       "      <td>1947-04-18</td>\n",
       "      <td>2013-08-14</td>\n",
       "      <td>2013-08-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>69</td>\n",
       "      <td>Greater than 45</td>\n",
       "      <td>Other</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Aggravated Assault w/Firearm</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>13011352CF10A</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5.0</td>\n",
       "      <td>ed philo</td>\n",
       "      <td>1991-05-14</td>\n",
       "      <td>2013-04-14</td>\n",
       "      <td>2013-04-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>24</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Cocaine</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>2013-06-16</td>\n",
       "      <td>13005330CF10A</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6.0</td>\n",
       "      <td>ed philo</td>\n",
       "      <td>1991-05-14</td>\n",
       "      <td>2013-04-14</td>\n",
       "      <td>2013-04-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>24</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Cocaine</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>2013-06-16</td>\n",
       "      <td>13005330CF10A</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7.0</td>\n",
       "      <td>ed philo</td>\n",
       "      <td>1991-05-14</td>\n",
       "      <td>2013-04-14</td>\n",
       "      <td>2013-04-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>24</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Cocaine</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>2013-06-16</td>\n",
       "      <td>13005330CF10A</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18306</th>\n",
       "      <td>NaN</td>\n",
       "      <td>raheem smith</td>\n",
       "      <td>1995-06-28</td>\n",
       "      <td>2013-10-20</td>\n",
       "      <td>2013-10-19</td>\n",
       "      <td>Male</td>\n",
       "      <td>20</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Cocaine</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>13014650CF10A</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18307</th>\n",
       "      <td>NaN</td>\n",
       "      <td>raheem smith</td>\n",
       "      <td>1995-06-28</td>\n",
       "      <td>2013-10-20</td>\n",
       "      <td>2013-10-19</td>\n",
       "      <td>Male</td>\n",
       "      <td>20</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Cocaine</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>13014650CF10A</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18308</th>\n",
       "      <td>NaN</td>\n",
       "      <td>raheem smith</td>\n",
       "      <td>1995-06-28</td>\n",
       "      <td>2013-10-20</td>\n",
       "      <td>2013-10-19</td>\n",
       "      <td>Male</td>\n",
       "      <td>20</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Cocaine</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>13014650CF10A</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18314</th>\n",
       "      <td>NaN</td>\n",
       "      <td>florencia sanmartin</td>\n",
       "      <td>1992-12-18</td>\n",
       "      <td>2014-06-30</td>\n",
       "      <td>2014-06-28</td>\n",
       "      <td>Female</td>\n",
       "      <td>23</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>Hispanic</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Ethylone</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>2015-03-15</td>\n",
       "      <td>14008895CF10A</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18315</th>\n",
       "      <td>NaN</td>\n",
       "      <td>florencia sanmartin</td>\n",
       "      <td>1992-12-18</td>\n",
       "      <td>2014-06-30</td>\n",
       "      <td>2014-06-28</td>\n",
       "      <td>Female</td>\n",
       "      <td>23</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>Hispanic</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Ethylone</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>2015-03-15</td>\n",
       "      <td>14008895CF10A</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>11423 rows × 20 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        id                 name        dob compas_screening_date  \\\n",
       "0      1.0     miguel hernandez 1947-04-18            2013-08-14   \n",
       "1      2.0     miguel hernandez 1947-04-18            2013-08-14   \n",
       "4      5.0             ed philo 1991-05-14            2013-04-14   \n",
       "5      6.0             ed philo 1991-05-14            2013-04-14   \n",
       "6      7.0             ed philo 1991-05-14            2013-04-14   \n",
       "...    ...                  ...        ...                   ...   \n",
       "18306  NaN         raheem smith 1995-06-28            2013-10-20   \n",
       "18307  NaN         raheem smith 1995-06-28            2013-10-20   \n",
       "18308  NaN         raheem smith 1995-06-28            2013-10-20   \n",
       "18314  NaN  florencia sanmartin 1992-12-18            2014-06-30   \n",
       "18315  NaN  florencia sanmartin 1992-12-18            2014-06-30   \n",
       "\n",
       "      c_offense_date     sex  age          age_cat              race  \\\n",
       "0         2013-08-13    Male   69  Greater than 45             Other   \n",
       "1         2013-08-13    Male   69  Greater than 45             Other   \n",
       "4         2013-04-13    Male   24     Less than 25  African-American   \n",
       "5         2013-04-13    Male   24     Less than 25  African-American   \n",
       "6         2013-04-13    Male   24     Less than 25  African-American   \n",
       "...              ...     ...  ...              ...               ...   \n",
       "18306     2013-10-19    Male   20     Less than 25  African-American   \n",
       "18307     2013-10-19    Male   20     Less than 25  African-American   \n",
       "18308     2013-10-19    Male   20     Less than 25  African-American   \n",
       "18314     2014-06-28  Female   23     Less than 25          Hispanic   \n",
       "18315     2014-06-28  Female   23     Less than 25          Hispanic   \n",
       "\n",
       "      c_charge_degree                 c_charge_desc  days_b_screening_arrest  \\\n",
       "0                (F3)  Aggravated Assault w/Firearm                     -1.0   \n",
       "1                (F3)  Aggravated Assault w/Firearm                     -1.0   \n",
       "4                (F3)         Possession of Cocaine                     -1.0   \n",
       "5                (F3)         Possession of Cocaine                     -1.0   \n",
       "6                (F3)         Possession of Cocaine                     -1.0   \n",
       "...               ...                           ...                      ...   \n",
       "18306            (F3)         Possession of Cocaine                     -1.0   \n",
       "18307            (F3)         Possession of Cocaine                     -1.0   \n",
       "18308            (F3)         Possession of Cocaine                     -1.0   \n",
       "18314            (F3)        Possession of Ethylone                     -2.0   \n",
       "18315            (F3)        Possession of Ethylone                     -2.0   \n",
       "\n",
       "       decile_score  is_recid r_offense_date  c_case_number  v_decile_score  \\\n",
       "0                 1         0            NaT  13011352CF10A               1   \n",
       "1                 1         0            NaT  13011352CF10A               1   \n",
       "4                 4         1     2013-06-16  13005330CF10A               3   \n",
       "5                 4         1     2013-06-16  13005330CF10A               3   \n",
       "6                 4         1     2013-06-16  13005330CF10A               3   \n",
       "...             ...       ...            ...            ...             ...   \n",
       "18306             9         0            NaT  13014650CF10A               9   \n",
       "18307             9         0            NaT  13014650CF10A               9   \n",
       "18308             9         0            NaT  13014650CF10A               9   \n",
       "18314             4         1     2015-03-15  14008895CF10A               4   \n",
       "18315             4         1     2015-03-15  14008895CF10A               4   \n",
       "\n",
       "       is_violent_recid vr_offense_date score_text  \n",
       "0                     0             NaT        Low  \n",
       "1                     0             NaT        Low  \n",
       "4                     0             NaT        Low  \n",
       "5                     0             NaT        Low  \n",
       "6                     0             NaT        Low  \n",
       "...                 ...             ...        ...  \n",
       "18306                 0             NaT       High  \n",
       "18307                 0             NaT       High  \n",
       "18308                 0             NaT       High  \n",
       "18314                 0             NaT        Low  \n",
       "18315                 0             NaT        Low  \n",
       "\n",
       "[11423 rows x 20 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check if my dataset has any duplicates for the same name\n",
    "\n",
    "analysis.duplicated().sum()\n",
    "analysis[analysis.duplicated([\"name\",\"age\",\"sex\",\"race\",\"dob\"],keep=False)]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9d6043da",
   "metadata": {},
   "source": [
    "The data has many multiple duplicates. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "4e210062",
   "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>id</th>\n",
       "      <th>name</th>\n",
       "      <th>dob</th>\n",
       "      <th>compas_screening_date</th>\n",
       "      <th>c_offense_date</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>age_cat</th>\n",
       "      <th>race</th>\n",
       "      <th>c_charge_degree</th>\n",
       "      <th>c_charge_desc</th>\n",
       "      <th>days_b_screening_arrest</th>\n",
       "      <th>decile_score</th>\n",
       "      <th>is_recid</th>\n",
       "      <th>r_offense_date</th>\n",
       "      <th>c_case_number</th>\n",
       "      <th>v_decile_score</th>\n",
       "      <th>is_violent_recid</th>\n",
       "      <th>vr_offense_date</th>\n",
       "      <th>score_text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>miguel hernandez</td>\n",
       "      <td>1947-04-18</td>\n",
       "      <td>2013-08-14</td>\n",
       "      <td>2013-08-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>69</td>\n",
       "      <td>Greater than 45</td>\n",
       "      <td>Other</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Aggravated Assault w/Firearm</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>13011352CF10A</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.0</td>\n",
       "      <td>kevon dixon</td>\n",
       "      <td>1982-01-22</td>\n",
       "      <td>2013-01-27</td>\n",
       "      <td>2013-01-26</td>\n",
       "      <td>Male</td>\n",
       "      <td>34</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Felony Battery w/Prior Convict</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2013-05-07</td>\n",
       "      <td>13001275CF10A</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2013-05-07</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9.0</td>\n",
       "      <td>ed philo</td>\n",
       "      <td>1991-05-14</td>\n",
       "      <td>2013-04-14</td>\n",
       "      <td>2013-04-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>24</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Cocaine</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>2013-06-16</td>\n",
       "      <td>13005330CF10A</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10.0</td>\n",
       "      <td>marcu brown</td>\n",
       "      <td>1993-01-21</td>\n",
       "      <td>2013-01-13</td>\n",
       "      <td>2013-12-01</td>\n",
       "      <td>Male</td>\n",
       "      <td>23</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Cannabis</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>13000570CF10A</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>11.0</td>\n",
       "      <td>bouthy pierrelouis</td>\n",
       "      <td>1973-01-22</td>\n",
       "      <td>2013-03-26</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Male</td>\n",
       "      <td>43</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>Other</td>\n",
       "      <td>(F7)</td>\n",
       "      <td>arrest case no charge</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>12014130CF10A</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18310</th>\n",
       "      <td>NaN</td>\n",
       "      <td>malcolm simmons</td>\n",
       "      <td>1993-03-25</td>\n",
       "      <td>2014-01-02</td>\n",
       "      <td>2014-01-31</td>\n",
       "      <td>Male</td>\n",
       "      <td>23</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Leaving the Scene of Accident</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>14001422CF10A</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18311</th>\n",
       "      <td>NaN</td>\n",
       "      <td>alexsandra beauchamps</td>\n",
       "      <td>1984-12-21</td>\n",
       "      <td>2014-12-29</td>\n",
       "      <td>2014-12-28</td>\n",
       "      <td>Female</td>\n",
       "      <td>31</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(M1)</td>\n",
       "      <td>Battery</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>14018106MM10A</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Medium</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18312</th>\n",
       "      <td>NaN</td>\n",
       "      <td>winston gregory</td>\n",
       "      <td>1958-01-10</td>\n",
       "      <td>2014-01-14</td>\n",
       "      <td>2014-01-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>57</td>\n",
       "      <td>Greater than 45</td>\n",
       "      <td>Other</td>\n",
       "      <td>(F2)</td>\n",
       "      <td>Aggravated Battery / Pregnant</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>14000581CF10A</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18313</th>\n",
       "      <td>NaN</td>\n",
       "      <td>farrah jean</td>\n",
       "      <td>1982-11-17</td>\n",
       "      <td>2014-09-03</td>\n",
       "      <td>2014-08-03</td>\n",
       "      <td>Female</td>\n",
       "      <td>33</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(M1)</td>\n",
       "      <td>Battery on Law Enforc Officer</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>14003308CF10A</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18315</th>\n",
       "      <td>NaN</td>\n",
       "      <td>florencia sanmartin</td>\n",
       "      <td>1992-12-18</td>\n",
       "      <td>2014-06-30</td>\n",
       "      <td>2014-06-28</td>\n",
       "      <td>Female</td>\n",
       "      <td>23</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>Hispanic</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Ethylone</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>2015-03-15</td>\n",
       "      <td>14008895CF10A</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10310 rows × 20 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         id                   name        dob compas_screening_date  \\\n",
       "1       2.0       miguel hernandez 1947-04-18            2013-08-14   \n",
       "3       4.0            kevon dixon 1982-01-22            2013-01-27   \n",
       "8       9.0               ed philo 1991-05-14            2013-04-14   \n",
       "9      10.0            marcu brown 1993-01-21            2013-01-13   \n",
       "10     11.0     bouthy pierrelouis 1973-01-22            2013-03-26   \n",
       "...     ...                    ...        ...                   ...   \n",
       "18310   NaN        malcolm simmons 1993-03-25            2014-01-02   \n",
       "18311   NaN  alexsandra beauchamps 1984-12-21            2014-12-29   \n",
       "18312   NaN        winston gregory 1958-01-10            2014-01-14   \n",
       "18313   NaN            farrah jean 1982-11-17            2014-09-03   \n",
       "18315   NaN    florencia sanmartin 1992-12-18            2014-06-30   \n",
       "\n",
       "      c_offense_date     sex  age          age_cat              race  \\\n",
       "1         2013-08-13    Male   69  Greater than 45             Other   \n",
       "3         2013-01-26    Male   34          25 - 45  African-American   \n",
       "8         2013-04-13    Male   24     Less than 25  African-American   \n",
       "9         2013-12-01    Male   23     Less than 25  African-American   \n",
       "10               NaT    Male   43          25 - 45             Other   \n",
       "...              ...     ...  ...              ...               ...   \n",
       "18310     2014-01-31    Male   23     Less than 25  African-American   \n",
       "18311     2014-12-28  Female   31          25 - 45  African-American   \n",
       "18312     2014-01-13    Male   57  Greater than 45             Other   \n",
       "18313     2014-08-03  Female   33          25 - 45  African-American   \n",
       "18315     2014-06-28  Female   23     Less than 25          Hispanic   \n",
       "\n",
       "      c_charge_degree                   c_charge_desc  \\\n",
       "1                (F3)    Aggravated Assault w/Firearm   \n",
       "3                (F3)  Felony Battery w/Prior Convict   \n",
       "8                (F3)           Possession of Cocaine   \n",
       "9                (F3)          Possession of Cannabis   \n",
       "10               (F7)           arrest case no charge   \n",
       "...               ...                             ...   \n",
       "18310            (F3)   Leaving the Scene of Accident   \n",
       "18311            (M1)                         Battery   \n",
       "18312            (F2)   Aggravated Battery / Pregnant   \n",
       "18313            (M1)   Battery on Law Enforc Officer   \n",
       "18315            (F3)          Possession of Ethylone   \n",
       "\n",
       "       days_b_screening_arrest  decile_score  is_recid r_offense_date  \\\n",
       "1                         -1.0             1         0            NaT   \n",
       "3                         -1.0             3         1     2013-05-07   \n",
       "8                         -1.0             4         1     2013-06-16   \n",
       "9                          NaN             8         0            NaT   \n",
       "10                         NaN             1         0            NaT   \n",
       "...                        ...           ...       ...            ...   \n",
       "18310                     -1.0             3         0            NaT   \n",
       "18311                     -1.0             6         0            NaT   \n",
       "18312                     -1.0             1         0            NaT   \n",
       "18313                     -1.0             2         0            NaT   \n",
       "18315                     -2.0             4         1     2015-03-15   \n",
       "\n",
       "       c_case_number  v_decile_score  is_violent_recid vr_offense_date  \\\n",
       "1      13011352CF10A               1                 0             NaT   \n",
       "3      13001275CF10A               1                 1      2013-05-07   \n",
       "8      13005330CF10A               3                 0             NaT   \n",
       "9      13000570CF10A               6                 0             NaT   \n",
       "10     12014130CF10A               1                 0             NaT   \n",
       "...              ...             ...               ...             ...   \n",
       "18310  14001422CF10A               5                 0             NaT   \n",
       "18311  14018106MM10A               4                 0             NaT   \n",
       "18312  14000581CF10A               1                 0             NaT   \n",
       "18313  14003308CF10A               2                 0             NaT   \n",
       "18315  14008895CF10A               4                 0             NaT   \n",
       "\n",
       "      score_text  \n",
       "1            Low  \n",
       "3            Low  \n",
       "8            Low  \n",
       "9           High  \n",
       "10           Low  \n",
       "...          ...  \n",
       "18310        Low  \n",
       "18311     Medium  \n",
       "18312        Low  \n",
       "18313        Low  \n",
       "18315        Low  \n",
       "\n",
       "[10310 rows x 20 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# remove duplication of the same case, every case should be represented one time\n",
    "analysis_ND=analysis.drop_duplicates(subset='c_case_number',keep='last')\n",
    "analysis_ND"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "83930329",
   "metadata": {},
   "source": [
    "after applying drop duplicate, 1123 Rows deleted from the dataset, now I have only 10310 case remaining "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "73a77b9e",
   "metadata": {},
   "source": [
    "My next step would be prepare my data for analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "252eb70c",
   "metadata": {},
   "source": [
    "# 2. Preparing dataset for analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fe173cf6",
   "metadata": {},
   "source": [
    "Before starting to prepare my data, I remove both the attributes name and dob "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "d9c88aaf",
   "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>id</th>\n",
       "      <th>compas_screening_date</th>\n",
       "      <th>c_offense_date</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>age_cat</th>\n",
       "      <th>race</th>\n",
       "      <th>c_charge_degree</th>\n",
       "      <th>c_charge_desc</th>\n",
       "      <th>days_b_screening_arrest</th>\n",
       "      <th>decile_score</th>\n",
       "      <th>is_recid</th>\n",
       "      <th>r_offense_date</th>\n",
       "      <th>c_case_number</th>\n",
       "      <th>v_decile_score</th>\n",
       "      <th>is_violent_recid</th>\n",
       "      <th>vr_offense_date</th>\n",
       "      <th>score_text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>2013-08-14</td>\n",
       "      <td>2013-08-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>69</td>\n",
       "      <td>Greater than 45</td>\n",
       "      <td>Other</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Aggravated Assault w/Firearm</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>13011352CF10A</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.0</td>\n",
       "      <td>2013-01-27</td>\n",
       "      <td>2013-01-26</td>\n",
       "      <td>Male</td>\n",
       "      <td>34</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Felony Battery w/Prior Convict</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2013-05-07</td>\n",
       "      <td>13001275CF10A</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2013-05-07</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9.0</td>\n",
       "      <td>2013-04-14</td>\n",
       "      <td>2013-04-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>24</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Cocaine</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>2013-06-16</td>\n",
       "      <td>13005330CF10A</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10.0</td>\n",
       "      <td>2013-01-13</td>\n",
       "      <td>2013-12-01</td>\n",
       "      <td>Male</td>\n",
       "      <td>23</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Cannabis</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>13000570CF10A</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>11.0</td>\n",
       "      <td>2013-03-26</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Male</td>\n",
       "      <td>43</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>Other</td>\n",
       "      <td>(F7)</td>\n",
       "      <td>arrest case no charge</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>12014130CF10A</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18310</th>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-01-02</td>\n",
       "      <td>2014-01-31</td>\n",
       "      <td>Male</td>\n",
       "      <td>23</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Leaving the Scene of Accident</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>14001422CF10A</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18311</th>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-12-29</td>\n",
       "      <td>2014-12-28</td>\n",
       "      <td>Female</td>\n",
       "      <td>31</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(M1)</td>\n",
       "      <td>Battery</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>14018106MM10A</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Medium</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18312</th>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-01-14</td>\n",
       "      <td>2014-01-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>57</td>\n",
       "      <td>Greater than 45</td>\n",
       "      <td>Other</td>\n",
       "      <td>(F2)</td>\n",
       "      <td>Aggravated Battery / Pregnant</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>14000581CF10A</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18313</th>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-09-03</td>\n",
       "      <td>2014-08-03</td>\n",
       "      <td>Female</td>\n",
       "      <td>33</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(M1)</td>\n",
       "      <td>Battery on Law Enforc Officer</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>14003308CF10A</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18315</th>\n",
       "      <td>NaN</td>\n",
       "      <td>2014-06-30</td>\n",
       "      <td>2014-06-28</td>\n",
       "      <td>Female</td>\n",
       "      <td>23</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>Hispanic</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Ethylone</td>\n",
       "      <td>-2.0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>2015-03-15</td>\n",
       "      <td>14008895CF10A</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10310 rows × 18 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         id compas_screening_date c_offense_date     sex  age  \\\n",
       "1       2.0            2013-08-14     2013-08-13    Male   69   \n",
       "3       4.0            2013-01-27     2013-01-26    Male   34   \n",
       "8       9.0            2013-04-14     2013-04-13    Male   24   \n",
       "9      10.0            2013-01-13     2013-12-01    Male   23   \n",
       "10     11.0            2013-03-26            NaT    Male   43   \n",
       "...     ...                   ...            ...     ...  ...   \n",
       "18310   NaN            2014-01-02     2014-01-31    Male   23   \n",
       "18311   NaN            2014-12-29     2014-12-28  Female   31   \n",
       "18312   NaN            2014-01-14     2014-01-13    Male   57   \n",
       "18313   NaN            2014-09-03     2014-08-03  Female   33   \n",
       "18315   NaN            2014-06-30     2014-06-28  Female   23   \n",
       "\n",
       "               age_cat              race c_charge_degree  \\\n",
       "1      Greater than 45             Other            (F3)   \n",
       "3              25 - 45  African-American            (F3)   \n",
       "8         Less than 25  African-American            (F3)   \n",
       "9         Less than 25  African-American            (F3)   \n",
       "10             25 - 45             Other            (F7)   \n",
       "...                ...               ...             ...   \n",
       "18310     Less than 25  African-American            (F3)   \n",
       "18311          25 - 45  African-American            (M1)   \n",
       "18312  Greater than 45             Other            (F2)   \n",
       "18313          25 - 45  African-American            (M1)   \n",
       "18315     Less than 25          Hispanic            (F3)   \n",
       "\n",
       "                        c_charge_desc  days_b_screening_arrest  decile_score  \\\n",
       "1        Aggravated Assault w/Firearm                     -1.0             1   \n",
       "3      Felony Battery w/Prior Convict                     -1.0             3   \n",
       "8               Possession of Cocaine                     -1.0             4   \n",
       "9              Possession of Cannabis                      NaN             8   \n",
       "10              arrest case no charge                      NaN             1   \n",
       "...                               ...                      ...           ...   \n",
       "18310   Leaving the Scene of Accident                     -1.0             3   \n",
       "18311                         Battery                     -1.0             6   \n",
       "18312   Aggravated Battery / Pregnant                     -1.0             1   \n",
       "18313   Battery on Law Enforc Officer                     -1.0             2   \n",
       "18315          Possession of Ethylone                     -2.0             4   \n",
       "\n",
       "       is_recid r_offense_date  c_case_number  v_decile_score  \\\n",
       "1             0            NaT  13011352CF10A               1   \n",
       "3             1     2013-05-07  13001275CF10A               1   \n",
       "8             1     2013-06-16  13005330CF10A               3   \n",
       "9             0            NaT  13000570CF10A               6   \n",
       "10            0            NaT  12014130CF10A               1   \n",
       "...         ...            ...            ...             ...   \n",
       "18310         0            NaT  14001422CF10A               5   \n",
       "18311         0            NaT  14018106MM10A               4   \n",
       "18312         0            NaT  14000581CF10A               1   \n",
       "18313         0            NaT  14003308CF10A               2   \n",
       "18315         1     2015-03-15  14008895CF10A               4   \n",
       "\n",
       "       is_violent_recid vr_offense_date score_text  \n",
       "1                     0             NaT        Low  \n",
       "3                     1      2013-05-07        Low  \n",
       "8                     0             NaT        Low  \n",
       "9                     0             NaT       High  \n",
       "10                    0             NaT        Low  \n",
       "...                 ...             ...        ...  \n",
       "18310                 0             NaT        Low  \n",
       "18311                 0             NaT     Medium  \n",
       "18312                 0             NaT        Low  \n",
       "18313                 0             NaT        Low  \n",
       "18315                 0             NaT        Low  \n",
       "\n",
       "[10310 rows x 18 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Drop the attributes name and dob because they will not impact my study \n",
    "analysis_ND = analysis_ND.drop(\"dob\", axis=1)\n",
    "analysis_ND = analysis_ND.drop(\"name\", axis=1)\n",
    "analysis_ND"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d022655c",
   "metadata": {},
   "source": [
    "1. Keep only rows that has a case number and has an id "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "81557fcc",
   "metadata": {},
   "outputs": [],
   "source": [
    "# keep only rows with a case number \n",
    "analysis_ND = analysis_ND[analysis_ND[\"c_case_number\"] != \"NaN\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "63bac7a1",
   "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>id</th>\n",
       "      <th>compas_screening_date</th>\n",
       "      <th>c_offense_date</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>age_cat</th>\n",
       "      <th>race</th>\n",
       "      <th>c_charge_degree</th>\n",
       "      <th>c_charge_desc</th>\n",
       "      <th>days_b_screening_arrest</th>\n",
       "      <th>decile_score</th>\n",
       "      <th>is_recid</th>\n",
       "      <th>r_offense_date</th>\n",
       "      <th>c_case_number</th>\n",
       "      <th>v_decile_score</th>\n",
       "      <th>is_violent_recid</th>\n",
       "      <th>vr_offense_date</th>\n",
       "      <th>score_text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>2013-08-14</td>\n",
       "      <td>2013-08-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>69</td>\n",
       "      <td>Greater than 45</td>\n",
       "      <td>Other</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Aggravated Assault w/Firearm</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>13011352CF10A</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.0</td>\n",
       "      <td>2013-01-27</td>\n",
       "      <td>2013-01-26</td>\n",
       "      <td>Male</td>\n",
       "      <td>34</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Felony Battery w/Prior Convict</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2013-05-07</td>\n",
       "      <td>13001275CF10A</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2013-05-07</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9.0</td>\n",
       "      <td>2013-04-14</td>\n",
       "      <td>2013-04-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>24</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Cocaine</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>2013-06-16</td>\n",
       "      <td>13005330CF10A</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10.0</td>\n",
       "      <td>2013-01-13</td>\n",
       "      <td>2013-12-01</td>\n",
       "      <td>Male</td>\n",
       "      <td>23</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Cannabis</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>13000570CF10A</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>11.0</td>\n",
       "      <td>2013-03-26</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Male</td>\n",
       "      <td>43</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>Other</td>\n",
       "      <td>(F7)</td>\n",
       "      <td>arrest case no charge</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>12014130CF10A</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10989</th>\n",
       "      <td>10990.0</td>\n",
       "      <td>2013-03-18</td>\n",
       "      <td>2013-03-17</td>\n",
       "      <td>Male</td>\n",
       "      <td>34</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession Of Alprazolam</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>2013-05-20</td>\n",
       "      <td>13003878CF10A</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>2013-05-12</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10992</th>\n",
       "      <td>10993.0</td>\n",
       "      <td>2013-03-22</td>\n",
       "      <td>2013-03-22</td>\n",
       "      <td>Male</td>\n",
       "      <td>23</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Poss3,4 Methylenedioxymethcath</td>\n",
       "      <td>0.0</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>13004155CF10A</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10994</th>\n",
       "      <td>10995.0</td>\n",
       "      <td>2014-04-28</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Female</td>\n",
       "      <td>31</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F1)</td>\n",
       "      <td>arrest case no charge</td>\n",
       "      <td>-3.0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>14005707CF10A</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10996</th>\n",
       "      <td>10997.0</td>\n",
       "      <td>2013-08-02</td>\n",
       "      <td>2013-07-02</td>\n",
       "      <td>Male</td>\n",
       "      <td>30</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>Hispanic</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Poss Unlaw Issue Driver Licenc</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2014-07-06</td>\n",
       "      <td>13001942CF10A</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10997</th>\n",
       "      <td>10998.0</td>\n",
       "      <td>2014-02-24</td>\n",
       "      <td>2014-02-23</td>\n",
       "      <td>Male</td>\n",
       "      <td>52</td>\n",
       "      <td>Greater than 45</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Tampering With Physical Evidence</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>14002552CF10A</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>6204 rows × 18 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            id compas_screening_date c_offense_date     sex  age  \\\n",
       "1          2.0            2013-08-14     2013-08-13    Male   69   \n",
       "3          4.0            2013-01-27     2013-01-26    Male   34   \n",
       "8          9.0            2013-04-14     2013-04-13    Male   24   \n",
       "9         10.0            2013-01-13     2013-12-01    Male   23   \n",
       "10        11.0            2013-03-26            NaT    Male   43   \n",
       "...        ...                   ...            ...     ...  ...   \n",
       "10989  10990.0            2013-03-18     2013-03-17    Male   34   \n",
       "10992  10993.0            2013-03-22     2013-03-22    Male   23   \n",
       "10994  10995.0            2014-04-28            NaT  Female   31   \n",
       "10996  10997.0            2013-08-02     2013-07-02    Male   30   \n",
       "10997  10998.0            2014-02-24     2014-02-23    Male   52   \n",
       "\n",
       "               age_cat              race c_charge_degree  \\\n",
       "1      Greater than 45             Other            (F3)   \n",
       "3              25 - 45  African-American            (F3)   \n",
       "8         Less than 25  African-American            (F3)   \n",
       "9         Less than 25  African-American            (F3)   \n",
       "10             25 - 45             Other            (F7)   \n",
       "...                ...               ...             ...   \n",
       "10989          25 - 45  African-American            (F3)   \n",
       "10992     Less than 25  African-American            (F3)   \n",
       "10994          25 - 45  African-American            (F1)   \n",
       "10996          25 - 45          Hispanic            (F3)   \n",
       "10997  Greater than 45  African-American            (F3)   \n",
       "\n",
       "                          c_charge_desc  days_b_screening_arrest  \\\n",
       "1          Aggravated Assault w/Firearm                     -1.0   \n",
       "3        Felony Battery w/Prior Convict                     -1.0   \n",
       "8                 Possession of Cocaine                     -1.0   \n",
       "9                Possession of Cannabis                      NaN   \n",
       "10                arrest case no charge                      NaN   \n",
       "...                                 ...                      ...   \n",
       "10989          Possession Of Alprazolam                     -1.0   \n",
       "10992    Poss3,4 Methylenedioxymethcath                      0.0   \n",
       "10994             arrest case no charge                     -3.0   \n",
       "10996    Poss Unlaw Issue Driver Licenc                      0.0   \n",
       "10997  Tampering With Physical Evidence                     -1.0   \n",
       "\n",
       "       decile_score  is_recid r_offense_date  c_case_number  v_decile_score  \\\n",
       "1                 1         0            NaT  13011352CF10A               1   \n",
       "3                 3         1     2013-05-07  13001275CF10A               1   \n",
       "8                 4         1     2013-06-16  13005330CF10A               3   \n",
       "9                 8         0            NaT  13000570CF10A               6   \n",
       "10                1         0            NaT  12014130CF10A               1   \n",
       "...             ...       ...            ...            ...             ...   \n",
       "10989             9         1     2013-05-20  13003878CF10A               4   \n",
       "10992             9         0            NaT  13004155CF10A               9   \n",
       "10994             4         0            NaT  14005707CF10A               5   \n",
       "10996             2         1     2014-07-06  13001942CF10A               2   \n",
       "10997             1         0            NaT  14002552CF10A               1   \n",
       "\n",
       "       is_violent_recid vr_offense_date score_text  \n",
       "1                     0             NaT        Low  \n",
       "3                     1      2013-05-07        Low  \n",
       "8                     0             NaT        Low  \n",
       "9                     0             NaT       High  \n",
       "10                    0             NaT        Low  \n",
       "...                 ...             ...        ...  \n",
       "10989                 1      2013-05-12       High  \n",
       "10992                 0             NaT       High  \n",
       "10994                 0             NaT        Low  \n",
       "10996                 0             NaT        Low  \n",
       "10997                 0             NaT        Low  \n",
       "\n",
       "[6204 rows x 18 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# remove all rows with any NaN and NaT values\n",
    "#analysis_ND =  analysis_ND.dropna()\n",
    "analysis_ND = analysis_ND.dropna( how='any', subset=['id'])\n",
    "analysis_ND "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "404a0378",
   "metadata": {},
   "source": [
    "2. I want explore how long people take to be charge for a crime and see if I can filter my selection for cases"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "0dea0c1a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    5942.000000\n",
       "mean       -0.120498\n",
       "std        74.925514\n",
       "min      -578.000000\n",
       "25%        -1.000000\n",
       "50%        -1.000000\n",
       "75%        -1.000000\n",
       "max      1040.000000\n",
       "Name: days_b_screening_arrest, dtype: float64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Check how long arrested people take to be charged for a crime\n",
    "analysis_ND[\"days_b_screening_arrest\"].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6e743248",
   "metadata": {},
   "source": [
    "More than 75% have less than -1 and with -597 as minimum\n",
    "to keep only quality data, we will limit my analysis on crimes charger within 30 days of arrest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "ff41772a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Keep only crimes charged within 30 days of arrest.\n",
    "analysis_ND = analysis_ND.loc[(analysis_ND[\"days_b_screening_arrest\"] > -30) & (analysis_ND[\"days_b_screening_arrest\"] <30)]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4bb64f28",
   "metadata": {},
   "source": [
    "3. Since my purpose is evalute Compass, I will remove all the cases that has the crime after the compass screening "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "0209e130",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Remove cases where the date of crime is after copmas screening\n",
    "analysis_ND = analysis_ND[analysis_ND[\"c_offense_date\"] < analysis_ND[\"compas_screening_date\"]]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e3a2601d",
   "metadata": {},
   "source": [
    "4. Now, I will create 3 groups: Age, Race and Score using CategoricalDtype"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "306550d8",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Group 1: Age\n",
    "grp_age = pd.CategoricalDtype(categories=[\"Less than 25\",\"25 - 45\",\"Greater than 45\"],ordered=True)\n",
    "analysis_ND[\"age_cat\"] = analysis_ND[\"age_cat\"].astype(grp_age)\n",
    "\n",
    "# Group 2: Race\n",
    "grp_race = pd.CategoricalDtype(categories=['African-American','Caucasian','Hispanic',\"Other\",'Asian',\n",
    "'Native American'],ordered=True)\n",
    "analysis_ND[\"race\"] = analysis_ND[\"race\"].astype(grp_race)\n",
    "\n",
    "# Group 3: Score\n",
    "grp_score = pd.CategoricalDtype(categories=[\"Low\",\"Medium\",\"High\"],ordered=True)\n",
    "analysis_ND[\"score_text\"] = analysis_ND[\"score_text\"].astype(grp_score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "45ee1bcd",
   "metadata": {},
   "outputs": [],
   "source": [
    "#convert the attributes between \"Sex\" and \"c_charge_degree\" to categories with the function astype\n",
    "\n",
    "for att in [\"sex\",\"c_charge_degree\"]:\n",
    "    analysis_ND[att] = analysis_ND[att].astype(\"category\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b24e929e",
   "metadata": {},
   "source": [
    "5. Remove all the rows that has empty value for the attribute score_text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "0081b1da",
   "metadata": {},
   "outputs": [],
   "source": [
    "#remove rows with score text equals NaN \n",
    "analysis_ND = analysis_ND[analysis_ND[\"score_text\"] != \"NaN\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "7ccf17ec",
   "metadata": {},
   "outputs": [],
   "source": [
    "#I will not consider traffic tickets & munipal violations as repeated offense (charge degree equals O)\n",
    "#analysis_ND = analysis_ND[analysis_ND[\"c_charge_degree\"] != \"O\"]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "528b8306",
   "metadata": {},
   "source": [
    "6. I want to know since when the Compass screening happened for both ordinary and violent offenses (I call both variables 2y_r and 2y_v), to do  that I first substract\n",
    "the columns of time, then I assign 3 different values for every case:  \n",
    "    - 0: the offender didnot commit any crime since 1st arrest \n",
    "    - 1: the offender committed a new crime in less than 2 years\n",
    "    - 2: the offender committed a new crime after 2 years    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "e21a2d32",
   "metadata": {},
   "outputs": [],
   "source": [
    "# I will call my function nbr_offenses\n",
    "def nbr_offenses(att,recid):\n",
    "   \n",
    "    # Subtract the columns of time\n",
    "    analysis_ND[\"days\"] = analysis_ND[att] - analysis_ND[\"compas_screening_date\"]\n",
    "    \n",
    "    # Convert the output to an integer by using the .days parameter  \n",
    "    analysis_ND[\"days\"] = analysis_ND[\"days\"].apply(lambda x:x.days)\n",
    "    \n",
    "    # Assign the values 0, 1 and 2 \n",
    "    analysis_ND[\"offence\"] = np.where(analysis_ND[recid]==0,0,\n",
    "                np.where((analysis_ND[recid]==1) & (analysis_ND[\"days\"] < 730),1,2))\n",
    "    \n",
    "    return analysis_ND[\"offence\"]\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "2288f4b5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# calcualte both 2y_r and 2y_v\n",
    "analysis_ND[\"2y_r\"] = nbr_offenses(\"r_offense_date\",\"is_recid\")\n",
    "analysis_ND[\"2y_v\"] = nbr_offenses(\"vr_offense_date\",\"is_violent_recid\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cc26be59",
   "metadata": {},
   "source": [
    "7. I want to foucs only on offenders who didnot or did repeat crime in less then 2 years, so I will remove Offenders who has 2y_r and 2y_v equals 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "bf8d1d05",
   "metadata": {},
   "outputs": [],
   "source": [
    "# remove offenders who committed new ordinary crimes after 2 years\n",
    "analysis_nd_r = analysis_ND[analysis_ND[\"2y_r\"] !=2].copy()\n",
    "\n",
    "# remove offenders who committed new violent crimes after 2 years\n",
    "analysis_nd_v = analysis_ND[analysis_ND[\"2y_v\"] != 2].copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "ff67ba61",
   "metadata": {},
   "outputs": [],
   "source": [
    "# reset the index to  make it easier for me to work with the dataset \n",
    "analysis_nd_r.reset_index(drop=True,inplace=True)\n",
    "analysis_nd_v.reset_index(drop=True,inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e99a7af4",
   "metadata": {},
   "source": [
    "8. In order to make data extraction easier, I add another 2 new attributes:\n",
    "    - <b>Attribute 1 :</b> devide cases in 2 categories based on the score using binary values,\n",
    "    so if the score >5, the case category will be 1, if the score is lower than 5 the case category will be 0\n",
    "    - <b>Attribute 2 :</b> This attribute will help me know if the prediction of Compass was correct or wrong,\n",
    "        so if it is \"True\", the prediction of Compass was correct, if \"False\" the prediction was wrong"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "8123af03",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Attribute 1: grouping cases in 2 categories (high score (>=5 and low score <5))\n",
    "analysis_nd_r[\"binary_score\"] = np.where(analysis_nd_r[\"decile_score\"] >=5,1,0)\n",
    "analysis_nd_v[\"binary_v_score\"] = np.where(analysis_nd_v[\"v_decile_score\"] >=5,1,0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "e3a2569c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Attribute 2: if the prediction is correct return True,else return false\n",
    "\n",
    "# For ordinary offense\n",
    "analysis_nd_r[\"prediction_recid\"] = analysis_nd_r['is_recid'] == analysis_nd_r[\"binary_score\"]\n",
    "\n",
    "# For violent offense\n",
    "analysis_nd_v[\"prediction_vrecid\"] = analysis_nd_v['is_violent_recid'] == analysis_nd_v[\"binary_v_score\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "5a87016e",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Index reset \n",
    "analysis_nd_r.reset_index(drop=True,inplace=True)\n",
    "analysis_nd_v.reset_index(drop=True,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "317f3bcf",
   "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>id</th>\n",
       "      <th>compas_screening_date</th>\n",
       "      <th>c_offense_date</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>age_cat</th>\n",
       "      <th>race</th>\n",
       "      <th>c_charge_degree</th>\n",
       "      <th>c_charge_desc</th>\n",
       "      <th>days_b_screening_arrest</th>\n",
       "      <th>...</th>\n",
       "      <th>v_decile_score</th>\n",
       "      <th>is_violent_recid</th>\n",
       "      <th>vr_offense_date</th>\n",
       "      <th>score_text</th>\n",
       "      <th>days</th>\n",
       "      <th>offence</th>\n",
       "      <th>2y_r</th>\n",
       "      <th>2y_v</th>\n",
       "      <th>binary_score</th>\n",
       "      <th>prediction_recid</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2.0</td>\n",
       "      <td>2013-08-14</td>\n",
       "      <td>2013-08-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>69</td>\n",
       "      <td>Greater than 45</td>\n",
       "      <td>Other</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Aggravated Assault w/Firearm</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4.0</td>\n",
       "      <td>2013-01-27</td>\n",
       "      <td>2013-01-26</td>\n",
       "      <td>Male</td>\n",
       "      <td>34</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Felony Battery w/Prior Convict</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2013-05-07</td>\n",
       "      <td>Low</td>\n",
       "      <td>100.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9.0</td>\n",
       "      <td>2013-04-14</td>\n",
       "      <td>2013-04-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>24</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Cocaine</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 24 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    id compas_screening_date c_offense_date   sex  age          age_cat  \\\n",
       "0  2.0            2013-08-14     2013-08-13  Male   69  Greater than 45   \n",
       "1  4.0            2013-01-27     2013-01-26  Male   34          25 - 45   \n",
       "2  9.0            2013-04-14     2013-04-13  Male   24     Less than 25   \n",
       "\n",
       "               race c_charge_degree                   c_charge_desc  \\\n",
       "0             Other            (F3)    Aggravated Assault w/Firearm   \n",
       "1  African-American            (F3)  Felony Battery w/Prior Convict   \n",
       "2  African-American            (F3)           Possession of Cocaine   \n",
       "\n",
       "   days_b_screening_arrest  ...  v_decile_score  is_violent_recid  \\\n",
       "0                     -1.0  ...               1                 0   \n",
       "1                     -1.0  ...               1                 1   \n",
       "2                     -1.0  ...               3                 0   \n",
       "\n",
       "  vr_offense_date score_text   days  offence 2y_r 2y_v  binary_score  \\\n",
       "0             NaT        Low    NaN        0    0    0             0   \n",
       "1      2013-05-07        Low  100.0        1    1    1             0   \n",
       "2             NaT        Low    NaN        0    1    0             0   \n",
       "\n",
       "   prediction_recid  \n",
       "0              True  \n",
       "1             False  \n",
       "2             False  \n",
       "\n",
       "[3 rows x 24 columns]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Checking the new values\n",
    "analysis_nd_r.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "4880252e",
   "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>id</th>\n",
       "      <th>compas_screening_date</th>\n",
       "      <th>c_offense_date</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>age_cat</th>\n",
       "      <th>race</th>\n",
       "      <th>c_charge_degree</th>\n",
       "      <th>c_charge_desc</th>\n",
       "      <th>days_b_screening_arrest</th>\n",
       "      <th>...</th>\n",
       "      <th>v_decile_score</th>\n",
       "      <th>is_violent_recid</th>\n",
       "      <th>vr_offense_date</th>\n",
       "      <th>score_text</th>\n",
       "      <th>days</th>\n",
       "      <th>offence</th>\n",
       "      <th>2y_r</th>\n",
       "      <th>2y_v</th>\n",
       "      <th>binary_v_score</th>\n",
       "      <th>prediction_vrecid</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2.0</td>\n",
       "      <td>2013-08-14</td>\n",
       "      <td>2013-08-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>69</td>\n",
       "      <td>Greater than 45</td>\n",
       "      <td>Other</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Aggravated Assault w/Firearm</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4.0</td>\n",
       "      <td>2013-01-27</td>\n",
       "      <td>2013-01-26</td>\n",
       "      <td>Male</td>\n",
       "      <td>34</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Felony Battery w/Prior Convict</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2013-05-07</td>\n",
       "      <td>Low</td>\n",
       "      <td>100.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9.0</td>\n",
       "      <td>2013-04-14</td>\n",
       "      <td>2013-04-13</td>\n",
       "      <td>Male</td>\n",
       "      <td>24</td>\n",
       "      <td>Less than 25</td>\n",
       "      <td>African-American</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>Possession of Cocaine</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>NaT</td>\n",
       "      <td>Low</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 24 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    id compas_screening_date c_offense_date   sex  age          age_cat  \\\n",
       "0  2.0            2013-08-14     2013-08-13  Male   69  Greater than 45   \n",
       "1  4.0            2013-01-27     2013-01-26  Male   34          25 - 45   \n",
       "2  9.0            2013-04-14     2013-04-13  Male   24     Less than 25   \n",
       "\n",
       "               race c_charge_degree                   c_charge_desc  \\\n",
       "0             Other            (F3)    Aggravated Assault w/Firearm   \n",
       "1  African-American            (F3)  Felony Battery w/Prior Convict   \n",
       "2  African-American            (F3)           Possession of Cocaine   \n",
       "\n",
       "   days_b_screening_arrest  ...  v_decile_score  is_violent_recid  \\\n",
       "0                     -1.0  ...               1                 0   \n",
       "1                     -1.0  ...               1                 1   \n",
       "2                     -1.0  ...               3                 0   \n",
       "\n",
       "  vr_offense_date score_text   days  offence 2y_r 2y_v  binary_v_score  \\\n",
       "0             NaT        Low    NaN        0    0    0               0   \n",
       "1      2013-05-07        Low  100.0        1    1    1               0   \n",
       "2             NaT        Low    NaN        0    1    0               0   \n",
       "\n",
       "   prediction_vrecid  \n",
       "0               True  \n",
       "1              False  \n",
       "2               True  \n",
       "\n",
       "[3 rows x 24 columns]"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "analysis_nd_v.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e91f80ad",
   "metadata": {},
   "source": [
    "#### Now my data is ready for the exporatory analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a248011b",
   "metadata": {},
   "source": [
    "# 3. Data Exploration\n",
    "In this step, I will be exploring my data to understand more about: the relationship between the crime, race, gender, age and the frequent commited crimes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "5919a0bf",
   "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>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>id</th>\n",
       "      <td>3527.0</td>\n",
       "      <td>5511.070031</td>\n",
       "      <td>3171.661078</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2703.0</td>\n",
       "      <td>5531.0</td>\n",
       "      <td>8284.0</td>\n",
       "      <td>10998.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>age</th>\n",
       "      <td>3527.0</td>\n",
       "      <td>34.908704</td>\n",
       "      <td>12.090702</td>\n",
       "      <td>18.0</td>\n",
       "      <td>25.0</td>\n",
       "      <td>32.0</td>\n",
       "      <td>42.0</td>\n",
       "      <td>96.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>days_b_screening_arrest</th>\n",
       "      <td>3527.0</td>\n",
       "      <td>-1.666005</td>\n",
       "      <td>4.413039</td>\n",
       "      <td>-29.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>29.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>decile_score</th>\n",
       "      <td>3527.0</td>\n",
       "      <td>4.199036</td>\n",
       "      <td>2.843142</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>10.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>is_recid</th>\n",
       "      <td>3527.0</td>\n",
       "      <td>0.322937</td>\n",
       "      <td>0.467665</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>v_decile_score</th>\n",
       "      <td>3527.0</td>\n",
       "      <td>3.442586</td>\n",
       "      <td>2.444397</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>10.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>is_violent_recid</th>\n",
       "      <td>3527.0</td>\n",
       "      <td>0.074568</td>\n",
       "      <td>0.262730</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>days</th>\n",
       "      <td>263.0</td>\n",
       "      <td>314.019011</td>\n",
       "      <td>250.267909</td>\n",
       "      <td>-218.0</td>\n",
       "      <td>123.5</td>\n",
       "      <td>295.0</td>\n",
       "      <td>489.0</td>\n",
       "      <td>1046.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>offence</th>\n",
       "      <td>3527.0</td>\n",
       "      <td>0.078537</td>\n",
       "      <td>0.283426</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2y_r</th>\n",
       "      <td>3527.0</td>\n",
       "      <td>0.322937</td>\n",
       "      <td>0.467665</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2y_v</th>\n",
       "      <td>3527.0</td>\n",
       "      <td>0.078537</td>\n",
       "      <td>0.283426</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>binary_score</th>\n",
       "      <td>3527.0</td>\n",
       "      <td>0.410547</td>\n",
       "      <td>0.492003</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          count         mean          std    min     25%  \\\n",
       "id                       3527.0  5511.070031  3171.661078    2.0  2703.0   \n",
       "age                      3527.0    34.908704    12.090702   18.0    25.0   \n",
       "days_b_screening_arrest  3527.0    -1.666005     4.413039  -29.0    -1.0   \n",
       "decile_score             3527.0     4.199036     2.843142   -1.0     2.0   \n",
       "is_recid                 3527.0     0.322937     0.467665    0.0     0.0   \n",
       "v_decile_score           3527.0     3.442586     2.444397   -1.0     1.0   \n",
       "is_violent_recid         3527.0     0.074568     0.262730    0.0     0.0   \n",
       "days                      263.0   314.019011   250.267909 -218.0   123.5   \n",
       "offence                  3527.0     0.078537     0.283426    0.0     0.0   \n",
       "2y_r                     3527.0     0.322937     0.467665    0.0     0.0   \n",
       "2y_v                     3527.0     0.078537     0.283426    0.0     0.0   \n",
       "binary_score             3527.0     0.410547     0.492003    0.0     0.0   \n",
       "\n",
       "                            50%     75%      max  \n",
       "id                       5531.0  8284.0  10998.0  \n",
       "age                        32.0    42.0     96.0  \n",
       "days_b_screening_arrest    -1.0    -1.0     29.0  \n",
       "decile_score                4.0     6.0     10.0  \n",
       "is_recid                    0.0     1.0      1.0  \n",
       "v_decile_score              3.0     5.0     10.0  \n",
       "is_violent_recid            0.0     0.0      1.0  \n",
       "days                      295.0   489.0   1046.0  \n",
       "offence                     0.0     0.0      2.0  \n",
       "2y_r                        0.0     1.0      1.0  \n",
       "2y_v                        0.0     0.0      2.0  \n",
       "binary_score                0.0     1.0      1.0  "
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check data\n",
    "analysis_nd_r.describe().T"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "10eca583",
   "metadata": {},
   "source": [
    "As shown in  the table above, from the ordinary 3527 cases: \n",
    "   - 75% of the cases assessed by compass got score 6\n",
    "   - 75% of the offenders are above 42 years old where 32% are recidivists and 7 % are violent recidivist"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "8f024dde",
   "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>count</th>\n",
       "      <th>unique</th>\n",
       "      <th>top</th>\n",
       "      <th>freq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>sex</th>\n",
       "      <td>3527</td>\n",
       "      <td>2</td>\n",
       "      <td>Male</td>\n",
       "      <td>2791</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>age_cat</th>\n",
       "      <td>3527</td>\n",
       "      <td>3</td>\n",
       "      <td>25 - 45</td>\n",
       "      <td>1978</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>race</th>\n",
       "      <td>3527</td>\n",
       "      <td>6</td>\n",
       "      <td>African-American</td>\n",
       "      <td>1725</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c_charge_degree</th>\n",
       "      <td>3527</td>\n",
       "      <td>12</td>\n",
       "      <td>(F3)</td>\n",
       "      <td>1907</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c_charge_desc</th>\n",
       "      <td>3526</td>\n",
       "      <td>321</td>\n",
       "      <td>Battery</td>\n",
       "      <td>737</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c_case_number</th>\n",
       "      <td>3527</td>\n",
       "      <td>3527</td>\n",
       "      <td>13011352CF10A</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>score_text</th>\n",
       "      <td>3523</td>\n",
       "      <td>3</td>\n",
       "      <td>Low</td>\n",
       "      <td>2075</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                count unique               top  freq\n",
       "sex              3527      2              Male  2791\n",
       "age_cat          3527      3           25 - 45  1978\n",
       "race             3527      6  African-American  1725\n",
       "c_charge_degree  3527     12              (F3)  1907\n",
       "c_charge_desc    3526    321           Battery   737\n",
       "c_case_number    3527   3527     13011352CF10A     1\n",
       "score_text       3523      3               Low  2075"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "analysis_nd_r.describe(include=[\"object\",\"category\"]).T"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d3ea2ece",
   "metadata": {},
   "source": [
    "From the table above, we notice:\n",
    " - One of the most common cases is Battery with charge degree F3 with the main race African Americans      \n",
    " - TMost reported cases are from males aging between 25 and 45 "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "54ba8518",
   "metadata": {},
   "source": [
    "<b> After cleaning data, and preparing it, now I will apply SVM to classify it"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "33eb3f2c",
   "metadata": {},
   "source": [
    "## 3.1 Explore Race"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fca10f5f",
   "metadata": {},
   "source": [
    " {change me}\n",
    "  - th race, when looking at the statistics, we must always keep in mind that there is very little data on Asians and Native Americans (a good idea would be to group them together with Other in a single category to make it more representative).\n",
    "\n",
    " - In general, we realize that almost 50% of the cases belong to African Americans and they are also the ones with the highest percentage of recidivism (almost 40% of the cases are African American).\n",
    "\n",
    " - This is important because the dataset is not balanced and can be very biased due to the predominance of this ethnic group."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "e54fb50b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "race_g = analysis_nd_r[\"race\"].value_counts(normalize=True,ascending=False).reset_index()\n",
    "\n",
    "plt.figure(figsize=(8,4))\n",
    "sns.barplot(x=race_g[\"index\"],y=race_g[\"race\"],color='#42b7bd')\n",
    "plt.xticks(rotation=45)\n",
    "plt.xlabel(\"\")\n",
    "plt.ylabel(\"% of cases\")\n",
    "plt.title(\"% of cases by race\")\n",
    "plt.show(block=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "42563b4e",
   "metadata": {},
   "source": [
    " - 50% of the cases belong to African-Americans, 35% to Caucasian, and the rest is below 1%   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "74f7b5b1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_b272e_row0_col1 {\n",
       "  background-color: #67000d;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_b272e_row1_col1 {\n",
       "  background-color: #d21f20;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_b272e_row2_col1 {\n",
       "  background-color: #f14432;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_b272e_row3_col1 {\n",
       "  background-color: #ef3c2c;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_b272e_row4_col1 {\n",
       "  background-color: #fb7a5a;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_b272e_row5_col1 {\n",
       "  background-color: #fff5f0;\n",
       "  color: #000000;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_b272e_\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th class=\"col_heading level0 col0\" >race</th>\n",
       "      <th class=\"col_heading level0 col1\" >2y_r</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_b272e_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_b272e_row0_col0\" class=\"data row0 col0\" >African-American</td>\n",
       "      <td id=\"T_b272e_row0_col1\" class=\"data row0 col1\" >0.385507</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b272e_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_b272e_row1_col0\" class=\"data row1 col0\" >Caucasian</td>\n",
       "      <td id=\"T_b272e_row1_col1\" class=\"data row1 col1\" >0.279376</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b272e_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_b272e_row2_col0\" class=\"data row2 col0\" >Hispanic</td>\n",
       "      <td id=\"T_b272e_row2_col1\" class=\"data row2 col1\" >0.231270</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b272e_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_b272e_row3_col0\" class=\"data row3 col0\" >Other</td>\n",
       "      <td id=\"T_b272e_row3_col1\" class=\"data row3 col1\" >0.240816</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b272e_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_b272e_row4_col0\" class=\"data row4 col0\" >Asian</td>\n",
       "      <td id=\"T_b272e_row4_col1\" class=\"data row4 col1\" >0.173913</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b272e_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_b272e_row5_col0\" class=\"data row5 col0\" >Native American</td>\n",
       "      <td id=\"T_b272e_row5_col1\" class=\"data row5 col1\" >0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7f91a8e4a730>"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "analysis_nd_r.groupby(\"race\",as_index=False)[\"2y_r\"].mean().style.background_gradient(cmap='Reds',axis=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1e1e3a1c",
   "metadata": {},
   "source": [
    " - For Recidivism rate, 39 % of African Americans are recidivist and for Caucasian 28% "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "8bdeadda",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_0972a_row0_col0 {\n",
       "  background-color: #fdce9e;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_0972a_row0_col1, #T_0972a_row1_col1, #T_0972a_row2_col1, #T_0972a_row3_col1, #T_0972a_row4_col1, #T_0972a_row5_col1 {\n",
       "  background-color: #7f2704;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_0972a_row0_col2, #T_0972a_row1_col0, #T_0972a_row2_col0, #T_0972a_row3_col2, #T_0972a_row4_col0, #T_0972a_row4_col2, #T_0972a_row5_col0 {\n",
       "  background-color: #fff5eb;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_0972a_row1_col2 {\n",
       "  background-color: #fdb475;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_0972a_row2_col2 {\n",
       "  background-color: #fedcbb;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_0972a_row3_col0 {\n",
       "  background-color: #fff3e6;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_0972a_row5_col2 {\n",
       "  background-color: #fda762;\n",
       "  color: #000000;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_0972a_\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >age_cat</th>\n",
       "      <th class=\"col_heading level0 col0\" >Less than 25</th>\n",
       "      <th class=\"col_heading level0 col1\" >25 - 45</th>\n",
       "      <th class=\"col_heading level0 col2\" >Greater than 45</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >race</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "      <th class=\"blank col2\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_0972a_level0_row0\" class=\"row_heading level0 row0\" >African-American</th>\n",
       "      <td id=\"T_0972a_row0_col0\" class=\"data row0 col0\" >0.267826</td>\n",
       "      <td id=\"T_0972a_row0_col1\" class=\"data row0 col1\" >0.572174</td>\n",
       "      <td id=\"T_0972a_row0_col2\" class=\"data row0 col2\" >0.160000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_0972a_level0_row1\" class=\"row_heading level0 row1\" >Caucasian</th>\n",
       "      <td id=\"T_0972a_row1_col0\" class=\"data row1 col0\" >0.156943</td>\n",
       "      <td id=\"T_0972a_row1_col1\" class=\"data row1 col1\" >0.548069</td>\n",
       "      <td id=\"T_0972a_row1_col2\" class=\"data row1 col2\" >0.294988</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_0972a_level0_row2\" class=\"row_heading level0 row2\" >Hispanic</th>\n",
       "      <td id=\"T_0972a_row2_col0\" class=\"data row2 col0\" >0.198697</td>\n",
       "      <td id=\"T_0972a_row2_col1\" class=\"data row2 col1\" >0.540717</td>\n",
       "      <td id=\"T_0972a_row2_col2\" class=\"data row2 col2\" >0.260586</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_0972a_level0_row3\" class=\"row_heading level0 row3\" >Other</th>\n",
       "      <td id=\"T_0972a_row3_col0\" class=\"data row3 col0\" >0.220408</td>\n",
       "      <td id=\"T_0972a_row3_col1\" class=\"data row3 col1\" >0.567347</td>\n",
       "      <td id=\"T_0972a_row3_col2\" class=\"data row3 col2\" >0.212245</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_0972a_level0_row4\" class=\"row_heading level0 row4\" >Asian</th>\n",
       "      <td id=\"T_0972a_row4_col0\" class=\"data row4 col0\" >0.217391</td>\n",
       "      <td id=\"T_0972a_row4_col1\" class=\"data row4 col1\" >0.565217</td>\n",
       "      <td id=\"T_0972a_row4_col2\" class=\"data row4 col2\" >0.217391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_0972a_level0_row5\" class=\"row_heading level0 row5\" >Native American</th>\n",
       "      <td id=\"T_0972a_row5_col0\" class=\"data row5 col0\" >0.100000</td>\n",
       "      <td id=\"T_0972a_row5_col1\" class=\"data row5 col1\" >0.600000</td>\n",
       "      <td id=\"T_0972a_row5_col2\" class=\"data row5 col2\" >0.300000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7f91a8cc8a90>"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "##Obtain percentages by rows with normalize\n",
    "pd.crosstab(analysis_nd_r[\"race\"],analysis_nd_r[\"age_cat\"],normalize=0).style.background_gradient(cmap='Oranges',axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "254ba817",
   "metadata": {},
   "source": [
    "- From the table above, we notice that the predominant age for the different races is between 25 and 45"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ceabe1ce",
   "metadata": {},
   "source": [
    "## 3.2 Explore link between sex and age"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "f383e940",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(8,5))\n",
    "sns.countplot(data=analysis_nd_r,x=\"sex\",hue=\"age_cat\",palette=\"Greens_r\")\n",
    "plt.title(\"Cases by age & sex\")\n",
    "plt.xlabel(\"\")\n",
    "plt.show(block=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cc28f3c5",
   "metadata": {},
   "source": [
    "Females are secondary compared to males"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "7e9f6c42",
   "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": [
    "decile_mean_age = analysis_nd_r.groupby(\"age_cat\")[\"decile_score\"].mean().reset_index()\n",
    "sns.barplot(data=decile_mean_age,x=\"age_cat\",y=\"decile_score\",palette=\"Greys_r\")\n",
    "plt.ylabel(\"Decile score mean\")\n",
    "plt.xlabel(\"\")\n",
    "plt.title(\"Average decile_score by different age ranges\")\n",
    "plt.show(block=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e3333c49",
   "metadata": {},
   "source": [
    "- There is a correlation between age and decile score, the younger the person, the more penalized is."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5c78a46b",
   "metadata": {},
   "source": [
    "## 3.3 Type of crime per sex (gender)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c3368ce5",
   "metadata": {},
   "source": [
    "I focus, because of their great variety, only on the 10 most typical.\n",
    "Many could be grouped by type of crime (drugs, vehicle, robbery and assault...) to facilitate the study of the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "fa3ee631",
   "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": [
    "#Top 5 crimes for women\n",
    "female_crime = analysis_nd_r[analysis_nd_r[\"sex\"]==\"Female\"][\"c_charge_desc\"].value_counts(normalize=True,ascending=False)[:5].reset_index()\n",
    "sns.barplot(data=female_crime,x=\"c_charge_desc\",y=\"index\",palette=\"Reds_r\")\n",
    "plt.title(\"Top 5 crimes for females\")\n",
    "plt.ylabel(\"\")\n",
    "plt.xlabel(\"Cases percentage\")\n",
    "plt.show(block=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "b3b9120d",
   "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": [
    "#Top 5 crimes for men\n",
    "male_crime= analysis_nd_r[analysis_nd_r[\"sex\"]==\"Male\"][\"c_charge_desc\"].value_counts(normalize=True,ascending=False)[:5].reset_index()\n",
    "\n",
    "sns.barplot(data=male_crime,x=\"c_charge_desc\",y=\"index\",palette=\"Greens_r\")\n",
    "plt.title(\"Top 5 crimes for females\")\n",
    "plt.ylabel(\"\")\n",
    "plt.xlabel(\"percentage of cases\")\n",
    "# so only the graphic appears without any text referring to the object type.\n",
    "plt.show(block=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b1ceafa2",
   "metadata": {},
   "source": [
    " Men and women commit the same top 3 offenses"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ab47bff2",
   "metadata": {},
   "source": [
    "## 3.4 Top offenses by both race and sex"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "0ab7a53d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Here I only consider the top 5 cases by counting their valuses then filtering with the function isin\n",
    "top_5_cases = analysis_nd_r[\"c_charge_desc\"].value_counts()[:5].index.tolist()\n",
    "top_cases = analysis_nd_r[analysis_nd_r[\"c_charge_desc\"].isin(top_5_cases)]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "283d5298",
   "metadata": {},
   "source": [
    "Lets explore in more depth top crimes by race"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "973a5f1c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_dc9d5_row0_col0, #T_dc9d5_row0_col1, #T_dc9d5_row0_col2, #T_dc9d5_row0_col3, #T_dc9d5_row0_col4, #T_dc9d5_row0_col5 {\n",
       "  background-color: #67000d;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_dc9d5_row1_col0 {\n",
       "  background-color: #fee1d4;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_dc9d5_row1_col1, #T_dc9d5_row1_col2, #T_dc9d5_row1_col3, #T_dc9d5_row1_col5, #T_dc9d5_row2_col0, #T_dc9d5_row2_col4, #T_dc9d5_row3_col4, #T_dc9d5_row3_col5, #T_dc9d5_row4_col5 {\n",
       "  background-color: #fff5f0;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_dc9d5_row1_col4 {\n",
       "  background-color: #fee3d6;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_dc9d5_row2_col1 {\n",
       "  background-color: #fff4ef;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_dc9d5_row2_col2, #T_dc9d5_row4_col2 {\n",
       "  background-color: #fee9df;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_dc9d5_row2_col3 {\n",
       "  background-color: #ffeee7;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_dc9d5_row2_col5 {\n",
       "  background-color: #fb694a;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_dc9d5_row3_col0 {\n",
       "  background-color: #f96044;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_dc9d5_row3_col1 {\n",
       "  background-color: #fcbea5;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_dc9d5_row3_col2 {\n",
       "  background-color: #fca486;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_dc9d5_row3_col3 {\n",
       "  background-color: #fcbfa7;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_dc9d5_row4_col0 {\n",
       "  background-color: #fc9474;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_dc9d5_row4_col1 {\n",
       "  background-color: #fcb095;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_dc9d5_row4_col3 {\n",
       "  background-color: #fff3ed;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_dc9d5_row4_col4 {\n",
       "  background-color: #fcc4ad;\n",
       "  color: #000000;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_dc9d5_\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >race</th>\n",
       "      <th class=\"col_heading level0 col0\" >African-American</th>\n",
       "      <th class=\"col_heading level0 col1\" >Caucasian</th>\n",
       "      <th class=\"col_heading level0 col2\" >Hispanic</th>\n",
       "      <th class=\"col_heading level0 col3\" >Other</th>\n",
       "      <th class=\"col_heading level0 col4\" >Asian</th>\n",
       "      <th class=\"col_heading level0 col5\" >Native American</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >c_charge_desc</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "      <th class=\"blank col2\" >&nbsp;</th>\n",
       "      <th class=\"blank col3\" >&nbsp;</th>\n",
       "      <th class=\"blank col4\" >&nbsp;</th>\n",
       "      <th class=\"blank col5\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_dc9d5_level0_row0\" class=\"row_heading level0 row0\" >Battery</th>\n",
       "      <td id=\"T_dc9d5_row0_col0\" class=\"data row0 col0\" >0.422006</td>\n",
       "      <td id=\"T_dc9d5_row0_col1\" class=\"data row0 col1\" >0.562500</td>\n",
       "      <td id=\"T_dc9d5_row0_col2\" class=\"data row0 col2\" >0.538462</td>\n",
       "      <td id=\"T_dc9d5_row0_col3\" class=\"data row0 col3\" >0.658537</td>\n",
       "      <td id=\"T_dc9d5_row0_col4\" class=\"data row0 col4\" >0.750000</td>\n",
       "      <td id=\"T_dc9d5_row0_col5\" class=\"data row0 col5\" >0.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_dc9d5_level0_row1\" class=\"row_heading level0 row1\" >Driving While License Revoked</th>\n",
       "      <td id=\"T_dc9d5_row1_col0\" class=\"data row1 col0\" >0.094708</td>\n",
       "      <td id=\"T_dc9d5_row1_col1\" class=\"data row1 col1\" >0.040323</td>\n",
       "      <td id=\"T_dc9d5_row1_col2\" class=\"data row1 col2\" >0.059829</td>\n",
       "      <td id=\"T_dc9d5_row1_col3\" class=\"data row1 col3\" >0.040650</td>\n",
       "      <td id=\"T_dc9d5_row1_col4\" class=\"data row1 col4\" >0.083333</td>\n",
       "      <td id=\"T_dc9d5_row1_col5\" class=\"data row1 col5\" >0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_dc9d5_level0_row2\" class=\"row_heading level0 row2\" >Felony Battery (Dom Strang)</th>\n",
       "      <td id=\"T_dc9d5_row2_col0\" class=\"data row2 col0\" >0.050139</td>\n",
       "      <td id=\"T_dc9d5_row2_col1\" class=\"data row2 col1\" >0.044355</td>\n",
       "      <td id=\"T_dc9d5_row2_col2\" class=\"data row2 col2\" >0.094017</td>\n",
       "      <td id=\"T_dc9d5_row2_col3\" class=\"data row2 col3\" >0.065041</td>\n",
       "      <td id=\"T_dc9d5_row2_col4\" class=\"data row2 col4\" >0.000000</td>\n",
       "      <td id=\"T_dc9d5_row2_col5\" class=\"data row2 col5\" >0.333333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_dc9d5_level0_row3\" class=\"row_heading level0 row3\" >Grand Theft in the 3rd Degree</th>\n",
       "      <td id=\"T_dc9d5_row3_col0\" class=\"data row3 col0\" >0.245125</td>\n",
       "      <td id=\"T_dc9d5_row3_col1\" class=\"data row3 col1\" >0.165323</td>\n",
       "      <td id=\"T_dc9d5_row3_col2\" class=\"data row3 col2\" >0.213675</td>\n",
       "      <td id=\"T_dc9d5_row3_col3\" class=\"data row3 col3\" >0.186992</td>\n",
       "      <td id=\"T_dc9d5_row3_col4\" class=\"data row3 col4\" >0.000000</td>\n",
       "      <td id=\"T_dc9d5_row3_col5\" class=\"data row3 col5\" >0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_dc9d5_level0_row4\" class=\"row_heading level0 row4\" >Possession of Cocaine</th>\n",
       "      <td id=\"T_dc9d5_row4_col0\" class=\"data row4 col0\" >0.188022</td>\n",
       "      <td id=\"T_dc9d5_row4_col1\" class=\"data row4 col1\" >0.187500</td>\n",
       "      <td id=\"T_dc9d5_row4_col2\" class=\"data row4 col2\" >0.094017</td>\n",
       "      <td id=\"T_dc9d5_row4_col3\" class=\"data row4 col3\" >0.048780</td>\n",
       "      <td id=\"T_dc9d5_row4_col4\" class=\"data row4 col4\" >0.166667</td>\n",
       "      <td id=\"T_dc9d5_row4_col5\" class=\"data row4 col5\" >0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7f91ab035c10>"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.crosstab(index=top_cases[\"c_charge_desc\"],columns=top_cases[\"race\"],normalize=1)\\\n",
    "    .style.background_gradient(cmap='Reds',axis=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f6aff3f0",
   "metadata": {},
   "source": [
    "- We already saw from the previous sections, Battery is top committed crime, but what we know more now is that its is taking the highest % for the different races. \n",
    "- For the top cases, we still have the sames result. \n",
    "- The exception we see that both Asian and Native American don't have any crimes on the category Driving while Licesne revoked and also no Grand Theft.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "bf542c8b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_de142_row0_col0, #T_de142_row1_col1, #T_de142_row2_col2, #T_de142_row3_col1, #T_de142_row4_col3, #T_de142_row4_col4 {\n",
       "  background-color: #7f2704;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_de142_row0_col1 {\n",
       "  background-color: #d14501;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_de142_row0_col2 {\n",
       "  background-color: #f87e2b;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_de142_row0_col3 {\n",
       "  background-color: #fc8b3a;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_de142_row0_col4 {\n",
       "  background-color: #fdb576;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_de142_row0_col5, #T_de142_row1_col4, #T_de142_row1_col5, #T_de142_row2_col4, #T_de142_row2_col5, #T_de142_row3_col4, #T_de142_row3_col5, #T_de142_row4_col5 {\n",
       "  background-color: #fff5eb;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_de142_row1_col0 {\n",
       "  background-color: #882a04;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_de142_row1_col2 {\n",
       "  background-color: #f67824;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_de142_row1_col3 {\n",
       "  background-color: #fda762;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_de142_row2_col0 {\n",
       "  background-color: #f26d17;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_de142_row2_col1 {\n",
       "  background-color: #d94801;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_de142_row2_col3 {\n",
       "  background-color: #fdb678;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_de142_row3_col0 {\n",
       "  background-color: #942f03;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_de142_row3_col2 {\n",
       "  background-color: #f36e19;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_de142_row3_col3 {\n",
       "  background-color: #ea5f0e;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_de142_row4_col0 {\n",
       "  background-color: #852904;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_de142_row4_col1 {\n",
       "  background-color: #d84801;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_de142_row4_col2 {\n",
       "  background-color: #fdb170;\n",
       "  color: #000000;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_de142_\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >race</th>\n",
       "      <th class=\"col_heading level0 col0\" >African-American</th>\n",
       "      <th class=\"col_heading level0 col1\" >Caucasian</th>\n",
       "      <th class=\"col_heading level0 col2\" >Hispanic</th>\n",
       "      <th class=\"col_heading level0 col3\" >Other</th>\n",
       "      <th class=\"col_heading level0 col4\" >Asian</th>\n",
       "      <th class=\"col_heading level0 col5\" >Native American</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >c_charge_desc</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "      <th class=\"blank col2\" >&nbsp;</th>\n",
       "      <th class=\"blank col3\" >&nbsp;</th>\n",
       "      <th class=\"blank col4\" >&nbsp;</th>\n",
       "      <th class=\"blank col5\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_de142_level0_row0\" class=\"row_heading level0 row0\" >Battery</th>\n",
       "      <td id=\"T_de142_row0_col0\" class=\"data row0 col0\" >0.316832</td>\n",
       "      <td id=\"T_de142_row0_col1\" class=\"data row0 col1\" >0.243728</td>\n",
       "      <td id=\"T_de142_row0_col2\" class=\"data row0 col2\" >0.174603</td>\n",
       "      <td id=\"T_de142_row0_col3\" class=\"data row0 col3\" >0.160494</td>\n",
       "      <td id=\"T_de142_row0_col4\" class=\"data row0 col4\" >0.111111</td>\n",
       "      <td id=\"T_de142_row0_col5\" class=\"data row0 col5\" >0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_de142_level0_row1\" class=\"row_heading level0 row1\" >Driving While License Revoked</th>\n",
       "      <td id=\"T_de142_row1_col0\" class=\"data row1 col0\" >0.485294</td>\n",
       "      <td id=\"T_de142_row1_col1\" class=\"data row1 col1\" >0.500000</td>\n",
       "      <td id=\"T_de142_row1_col2\" class=\"data row1 col2\" >0.285714</td>\n",
       "      <td id=\"T_de142_row1_col3\" class=\"data row1 col3\" >0.200000</td>\n",
       "      <td id=\"T_de142_row1_col4\" class=\"data row1 col4\" >0.000000</td>\n",
       "      <td id=\"T_de142_row1_col5\" class=\"data row1 col5\" >0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_de142_level0_row2\" class=\"row_heading level0 row2\" >Felony Battery (Dom Strang)</th>\n",
       "      <td id=\"T_de142_row2_col0\" class=\"data row2 col0\" >0.222222</td>\n",
       "      <td id=\"T_de142_row2_col1\" class=\"data row2 col1\" >0.272727</td>\n",
       "      <td id=\"T_de142_row2_col2\" class=\"data row2 col2\" >0.363636</td>\n",
       "      <td id=\"T_de142_row2_col3\" class=\"data row2 col3\" >0.125000</td>\n",
       "      <td id=\"T_de142_row2_col4\" class=\"data row2 col4\" >0.000000</td>\n",
       "      <td id=\"T_de142_row2_col5\" class=\"data row2 col5\" >0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_de142_level0_row3\" class=\"row_heading level0 row3\" >Grand Theft in the 3rd Degree</th>\n",
       "      <td id=\"T_de142_row3_col0\" class=\"data row3 col0\" >0.306818</td>\n",
       "      <td id=\"T_de142_row3_col1\" class=\"data row3 col1\" >0.329268</td>\n",
       "      <td id=\"T_de142_row3_col2\" class=\"data row3 col2\" >0.200000</td>\n",
       "      <td id=\"T_de142_row3_col3\" class=\"data row3 col3\" >0.217391</td>\n",
       "      <td id=\"T_de142_row3_col4\" class=\"data row3 col4\" >0.000000</td>\n",
       "      <td id=\"T_de142_row3_col5\" class=\"data row3 col5\" >0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_de142_level0_row4\" class=\"row_heading level0 row4\" >Possession of Cocaine</th>\n",
       "      <td id=\"T_de142_row4_col0\" class=\"data row4 col0\" >0.488889</td>\n",
       "      <td id=\"T_de142_row4_col1\" class=\"data row4 col1\" >0.376344</td>\n",
       "      <td id=\"T_de142_row4_col2\" class=\"data row4 col2\" >0.181818</td>\n",
       "      <td id=\"T_de142_row4_col3\" class=\"data row4 col3\" >0.500000</td>\n",
       "      <td id=\"T_de142_row4_col4\" class=\"data row4 col4\" >0.500000</td>\n",
       "      <td id=\"T_de142_row4_col5\" class=\"data row4 col5\" >0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7f91ab042d30>"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Recidivism % between race and the type of the case \n",
    "pd.pivot_table(data=top_cases,values=\"2y_r\",index=\"c_charge_desc\",columns=\"race\",fill_value=0)\\\n",
    "    .style.background_gradient(cmap='Oranges',axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "cfd73be6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# % of recidivism by crime type\n",
    "percentage_recid_cases= top_cases.groupby(\"c_charge_desc\")[\"2y_r\"].mean().reset_index().sort_values(by=\"2y_r\",ascending=False)\n",
    "\n",
    "plt.figure(figsize=(8,4))\n",
    "sns.barplot(data=percentage_recid_cases,y=\"c_charge_desc\",x=\"2y_r\", color=\"#7FFFD4\")\n",
    "plt.title(\" Recidivism by crime type\\n\",fontsize=15)\n",
    "plt.xlabel(\"\")\n",
    "plt.ylabel(\"\")\n",
    "plt.axvline(x=0.34,color=\"#4169E1\")\n",
    "plt.show(block=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d362b2a7",
   "metadata": {},
   "source": [
    "## 4. Check the fairness of the model & how to improve it\n",
    " - I will compare both ordinary & violent offences and see wich one has better accuracy using logisitc linear \n",
    " - I will also see the result by both race and sex"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "9e439f03",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.78      0.68      0.73      2388\n",
      "           1       0.48      0.61      0.53      1139\n",
      "\n",
      "    accuracy                           0.66      3527\n",
      "   macro avg       0.63      0.65      0.63      3527\n",
      "weighted avg       0.69      0.66      0.67      3527\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# first I will calcualte TPR, FPR, FNR, TNR for normal offences\n",
    "from sklearn.metrics import classification_report\n",
    "print(classification_report(analysis_nd_r['2y_r'],analysis_nd_r[\"binary_score\"]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "6ec5aae4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_72618_row0_col0, #T_72618_row1_col1 {\n",
       "  background-color: #67000d;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_72618_row0_col1, #T_72618_row1_col0 {\n",
       "  background-color: #fff5f0;\n",
       "  color: #000000;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_72618_\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >2y_r</th>\n",
       "      <th class=\"col_heading level0 col0\" >0</th>\n",
       "      <th class=\"col_heading level0 col1\" >1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >binary_score</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_72618_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_72618_row0_col0\" class=\"data row0 col0\" >0.683417</td>\n",
       "      <td id=\"T_72618_row0_col1\" class=\"data row0 col1\" >0.392450</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_72618_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_72618_row1_col0\" class=\"data row1 col0\" >0.316583</td>\n",
       "      <td id=\"T_72618_row1_col1\" class=\"data row1 col1\" >0.607550</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7f91ab1626d0>"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# calcualte TPR, FPR, FNR, TNR for normal offences\n",
    "cross_tab = pd.crosstab(analysis_nd_r[\"binary_score\"],analysis_nd_r['2y_r'],normalize=\"columns\").style.background_gradient(cmap='Reds',axis=1)\n",
    "cross_tab"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "58274f32",
   "metadata": {},
   "source": [
    "- This model has an accuray of 66%\n",
    "- Of the actual positive cases it predicted 61% where it is correct in 48% of the cases => False Positive Rate : 1-0.68 = 0.32 \n",
    "- Of the actual negative cases, the model predicted 68% of the cases where it was correct for the 78% cases \n",
    "  => False Negative Rate 1-0.61=0.39"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "35dcc7c8",
   "metadata": {},
   "source": [
    "Now I check the prediction for violent crimes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "fd433f62",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_baf0d_row0_col0, #T_baf0d_row1_col1 {\n",
       "  background-color: #00441b;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_baf0d_row0_col1, #T_baf0d_row1_col0 {\n",
       "  background-color: #f7fcf5;\n",
       "  color: #000000;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_baf0d_\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >2y_v</th>\n",
       "      <th class=\"col_heading level0 col0\" >0</th>\n",
       "      <th class=\"col_heading level0 col1\" >1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >binary_v_score</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_baf0d_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_baf0d_row0_col0\" class=\"data row0 col0\" >0.717326</td>\n",
       "      <td id=\"T_baf0d_row0_col1\" class=\"data row0 col1\" >0.469880</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_baf0d_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_baf0d_row1_col0\" class=\"data row1 col0\" >0.282674</td>\n",
       "      <td id=\"T_baf0d_row1_col1\" class=\"data row1 col1\" >0.530120</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7f91ab0db850>"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# analysis_nd_v only cases with violent crimes\n",
    "pd.crosstab(analysis_nd_v[\"binary_v_score\"],analysis_nd_v['2y_v'],normalize=\"columns\").style.background_gradient(cmap='Greens',axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "c1408607",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.95      0.72      0.82      3336\n",
      "           1       0.12      0.53      0.20       249\n",
      "\n",
      "    accuracy                           0.70      3585\n",
      "   macro avg       0.54      0.62      0.51      3585\n",
      "weighted avg       0.90      0.70      0.78      3585\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(analysis_nd_v['2y_v'],analysis_nd_v[\"binary_v_score\"]))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f8858d33",
   "metadata": {},
   "source": [
    "The prediction here is worse than the previous one even though the accuracy shows a higher score,\n",
    "as we see, the prediction for positive cases is higher compared to the previous one (18% more) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "c142cedd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    0.930544\n",
       "1    0.069456\n",
       "Name: 2y_v, dtype: float64"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "analysis_nd_v['2y_v'].value_counts(normalize=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "b92d3eae",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    0.677063\n",
       "1    0.322937\n",
       "Name: 2y_r, dtype: float64"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "analysis_nd_r[\"2y_r\"].value_counts(normalize=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5aed6044",
   "metadata": {},
   "source": [
    "- Despite the accuracy being higher (0.70 > 0.66) \n",
    "- In case of violent crimes, the model is less accurate for positive cases (diffirence 18% higher) \n",
    "- False Positive Rate : 0.46\n",
    "- False Negative Rate:  0.28\n",
    "\n",
    "Conclusion, it shows more accuracy with less accurate predicition for positive cases "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "daf33b23",
   "metadata": {},
   "source": [
    "## 4.1 Check Descrimination of COMPASS for both the race and sex"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "8bc63f1f",
   "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": [
    "# Comparaison between Revidivism &  Compas forcast\n",
    "reality = analysis_nd_r.groupby(\"race\",as_index=False)[\"2y_r\"].mean()\n",
    "compas_prediction = analysis_nd_r.groupby(\"race\")[\"binary_score\"].mean().reset_index()\n",
    "\n",
    "sns.barplot(data=compas_prediction,x=\"race\",y=\"binary_score\",color=\"powderblue\",label=\"Prediction\")\n",
    "\n",
    "ax= sns.barplot(data=reality,x=\"race\",y=\"2y_r\",color=\"blue\",label=\"Real\")\n",
    "\n",
    "plt.title(\"Recidivism rate in reality vs. Compas prediction\")\n",
    "plt.xticks(rotation=45)\n",
    "plt.xlabel(\"\")\n",
    "plt.ylabel(\"\")\n",
    "plt.legend()\n",
    "plt.show(block=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "767ceddf",
   "metadata": {},
   "source": [
    "Comparing the reality with the system COMPASS, shows hi less accuray for false positive prediction where the highest % goes for Afriacan american followed by Asians and Native Americans "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "7c705a2f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_9a611_row0_col0, #T_9a611_row1_col0, #T_9a611_row2_col0, #T_9a611_row3_col1, #T_9a611_row4_col0, #T_9a611_row5_col0 {\n",
       "  background-color: #f7fcf5;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_9a611_row0_col1, #T_9a611_row1_col1, #T_9a611_row2_col1, #T_9a611_row3_col0, #T_9a611_row4_col1, #T_9a611_row5_col1 {\n",
       "  background-color: #00441b;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_9a611_\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th class=\"col_heading level0 col0\" >2y_r</th>\n",
       "      <th class=\"col_heading level0 col1\" >binary_score</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >race</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_9a611_level0_row0\" class=\"row_heading level0 row0\" >African-American</th>\n",
       "      <td id=\"T_9a611_row0_col0\" class=\"data row0 col0\" >0.385507</td>\n",
       "      <td id=\"T_9a611_row0_col1\" class=\"data row0 col1\" >0.548986</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9a611_level0_row1\" class=\"row_heading level0 row1\" >Caucasian</th>\n",
       "      <td id=\"T_9a611_row1_col0\" class=\"data row1 col0\" >0.279376</td>\n",
       "      <td id=\"T_9a611_row1_col1\" class=\"data row1 col1\" >0.299918</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9a611_level0_row2\" class=\"row_heading level0 row2\" >Hispanic</th>\n",
       "      <td id=\"T_9a611_row2_col0\" class=\"data row2 col0\" >0.231270</td>\n",
       "      <td id=\"T_9a611_row2_col1\" class=\"data row2 col1\" >0.244300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9a611_level0_row3\" class=\"row_heading level0 row3\" >Other</th>\n",
       "      <td id=\"T_9a611_row3_col0\" class=\"data row3 col0\" >0.240816</td>\n",
       "      <td id=\"T_9a611_row3_col1\" class=\"data row3 col1\" >0.216327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9a611_level0_row4\" class=\"row_heading level0 row4\" >Asian</th>\n",
       "      <td id=\"T_9a611_row4_col0\" class=\"data row4 col0\" >0.173913</td>\n",
       "      <td id=\"T_9a611_row4_col1\" class=\"data row4 col1\" >0.304348</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9a611_level0_row5\" class=\"row_heading level0 row5\" >Native American</th>\n",
       "      <td id=\"T_9a611_row5_col0\" class=\"data row5 col0\" >0.000000</td>\n",
       "      <td id=\"T_9a611_row5_col1\" class=\"data row5 col1\" >0.100000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7f91ab03b9a0>"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "analysis_nd_r.groupby(\"race\").agg({\"2y_r\":\"mean\",\n",
    "                                    \"binary_score\":\"mean\"}).style.background_gradient(cmap='Greens',axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "48b899fd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXkAAAFCCAYAAAAUvq1TAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8rg+JYAAAACXBIWXMAAAsTAAALEwEAmpwYAAAypklEQVR4nO3dd5hcZdnH8e8vCaE3SQiEBELvCiQQeu9BQEGaIEEkIC8ggvjSRVAQaSIiUkQUqaJCkCgvICBFIEERKUYCUhKKCUSkk3K/f9zPksmyIbvJZmfm5Pe5rrl2Zs7ZOfe0e57zVEUEZmZWTd3qHYCZmc05TvJmZhXmJG9mVmFO8mZmFeYkb2ZWYU7yZmYV5iRvNheSFJJWqnccNuc5yTcJST0lnSbpGUnvSHpe0pWSBtTss4ukR8r21yVdI6lfzfah5ct9QavH3q3cf1W5PaDcfrtcnpd0fKv/uUrSZElLt7p/sRLXq5LekvTP2v8tx3pM0n8lTZD0R0nLd+6rVX1O0tZeTvLN4yZgV2A/YFHgM8CjwDYAkvYErgV+APQC1gQ+AO6XtHjN4zwL7CWpR819BwL/bOOYi0XEQsC+wKmSdizHWhDYA3gT2L/V/1wALASsXuLcFRhT/m8l4BfAsWXb8sDFwJQOvRJdpNVrVPnjdoZmjr2yIsKXBr8A2wLvAf1nsF3AC8A3W93fDXgCOL3cHgrcD/wBGFLu+xTwKnAOcFW5bwAQQI+axxoJfKNc/xLwEvA14IlWx3wC2H0Gce4JPNaB5z0E+Cvw33K801pt3xR4EPhP2T603D8/cF55Td4sz3l+YEtgbKvHeB7Ytlw/jfwx/WU55leADYA/l2O8AvwI6Fnz/2sCdwBvAK8BJwJLAe8CS9Tstx4wHpinjefZoeMCfyrvzzvA28De5f5dgMfK/zwIfPoTXtsAjgKeAyaU978b0LM8l7Vr9l2yPJ/ebTzOUOAB8sf9deA7wIrAH8vtCcA1ZIGh5X/6A78pr8frwI9qtn0ZeBqYCNwOLFfv71+zX+oegC/teJPge8C9n7B9tfKlXb6Nbd8G/lyuDy0Jbz/ghnLf4cCl5ct5VblvQHm8HuQPyCblS75N2X4X8H2gDzAZGFhzvCuAJ4GDgJVbxbIC8H5JCFsBC83keW8JrF2Sz6fJJLp72bYc8BZ5ljEPsASwTtl2MXAPsAzQHdgYmJf2JflJwO7lmPMDA4ENy2sxoCSgo8v+C5MJ+FhgvnJ7cNk2AvhqzXEuAC6awfPs0HHL/wSwUs3tdYF/A4PLcz6wPLd5Z3DMAO4mf+SXJc/kvlK2/Rg4u2bfrwG3zuBxhpbPwJEl1vmBlYDtymvem/xR+kHZvzvwt/J6LFhet03Ltt3Is77Vy2OdDDxY7+9fs1/qHoAv7XiT4HLg+k/Yvmn50s7XxrbDgGfK9aFMK9W+RlaZPEQm8baS/H/IEtXTwFFl27LAVKYl1NuBC2uONz9Zmn20JK4xwE412zcEbiRLce8DVzGTZF/zvz8ALijXTwB+28Y+3cizns+0sW1LZp7k/zSTGI5uOS75A/PXGey3N/BAud6dPFvaYAb7dui45XbrJH8JcEar/xkNbDGDxwtgx5rbhwN3leuDgRcBldujgL1m8DhDgRdnEvvuLa8TsFF573u0sd/vgYNbvZfv4tL8bF1cJ98cXgeW/oTtE8rftvZZumY7ABHxHnAbWVJaIiIemMHj9oqIxSNi9Yj4YbnvAODpiHis3L4G2E/SPC2PHRFnRsRAsnR9I/ArSZ8q2x+KiL0iojewGbA5cFJbB5c0WNLdksZLepP8wepVNvcn2xc+FjNZOmxrW3u81CqGVST9rjQk/xc4sx0xANwCrFEalbcD3oyIRzrpuG1ZDjhW0n9aLiW+vu085gst+0bEw2Ry3VLSamTJfHgHYu8j6XpJ40rsv2T61+yFiJg8g+dwYU38b5Bnkst8wrFtJpzkm8OdwAa1PWVaGQ2MBb5Qe6ekbmQD6V1t/E9LA+gvOxjLl4AVSvJ5FTif/ALv3HrHiGhJTguSjaytt48k62bXmsGxriWTS/+IWBT4Cfmlh0wsK7bxPxPIM4S2tr0DLNByQ1J3sjphurBa3b4E+AdZ9bQIeZZSG8MKbQUeEe+TP3D7kz+MV7e13ywety0vAd+NiMVqLgtExHWf8D/9a64vC7xcc/vnNbHfVJ5Pe2M/s9y3dol9f6Z/zZadQQPtS8ChrZ7D/BHx4Ccc22bCSb4JRMSdZOPebyUNlNRD0sKSDpP05chz228AJ0vaT9J8kpYi68cXIes/W7uXLGFe1N44JG1EJs8NgHXKZS0yGX+p7HOKpPVLl8/5yPrc/wCjJW0q6RBJS5Z9VyN73zw0g0MuDLwREe9L2oBsS2hxDbCtpL3K67GEpHUiYipwJXC+pL6SukvaSNK8ZL3zfJKGlDOPk8l640+yMNkY+naJ96s1234HLC3paEnzlvdkcM32X5DVGbsy8yTfkeNCVrfV/sBcDhxWzn4kacHyPBf+hGMcJ2lxSf3J9+mGmm2/BD5HJuhfzELsbwNvSloGOK5m2yNkO8b3SozzSdqkbPsJcIKkNQEkLSppuoKLdZyTfPPYk2zMu4HsMfIEMIgs5RMRN5Clrq+T1TtPkfXjm0TE660fLNJdEfFGB2I4ELglIv4eEa+2XIALgV1KlUwAPyNL1C+TPyRDIuJtMtnvCvxd0ttkL5/fko24bTkcOF3SW8CpZMm4Jf4XybOHY8nT+sfIbqWQP3h/J3sEvQGcDXSLiDfLY14BjCNL9mNn8py/Qf64vEUm0o8SYUS8VZ7fZ8k692fIBuWW7Q+Q7Rd/iYgXZnKcdh+3OA34eana2CsiRgGHkL1wJpJtIUNncoxbyLaTx8jqu5/WxP4S8Bfy/byvg7F/m+xN9GZ53N/UPO4U8vVaiaz3H0u2XxARvyXfq+tLNc8TwE4dPLa10tKwYmZzgKQ/AtdGxBX1jqWjJF0JvBwRJ9c7Fpt1TvJmc4ik9clqtv6l1N80lCOpHwPWjYh/1Tcamx2urjGbAyT9nKxKO7oJE/wZZFXJOU7wzc8leTOzCnNJ3syswuo2mVCvXr1iwIAB9Tq8mVlTevTRRyeUwYTtUrckP2DAAEaNGlWvw5uZNSVJHeqO6+oaM7MKc5I3M6swJ3kzswpzkjczqzAneTOzCnOSNzOrMCd5M7MKc5I3M6swJ3kzswqr24hXM7Ou9M7oO+sdQocsuOq2nfI4LsmbmVWYk7yZWYU5yZuZVZiTvJlZJ7jjvpGsu+NBfHr7Aznvsus/tv3Fca8xZOhxDN51GDsecCzjXh3fJXE5yZuZzaYpU6ZwzOkX8ZvLz2TU767gV7fdzdNjpp8R+MTvX8p+u23Hw8Mv4/j/2Z9vnf/TLonNSd7MbDaNenw0Kyzbl+X7L03PnvOw585bcttdD063zz+efZEtNlwHgC0Gr8Ntd/25S2Jzkjczm00vvzaBfktPW6xpmaV68fJrE6bbZ+1VV+CWO+4HYPgd9/PWO+/y+sT/zvHYnOTNzLrAmd8cxv0jH2fjzx3G/SMfp2+fXnTvPudTsAdDmZnNpr59ejH2lWkNqeNenUDfPr2m22fpPr247qLTAHj7nfe45f/uZ7FFFprjsbkkb2Y2mwauvSrPvjCO58e+wocfTuKmEfew89YbTbfPhIlvMnXqVADOvew6Dthjhy6JzSV5M7PZ1KNHd8475Qh2P/gEpkydygF77MAaKw/gjB9exXprrcKQrTfmvof/xmkX/BQhNll/bc4/9cguiU0R0SUHam3QoEExatSouhzbzOY+VZm7RtKjETGovY/j6hozswpzkjczqzAneTOzCnOSNzOrMCd5M7MKc5I3M6swJ3kzswprV5KXtKOk0ZLGSDq+je3LSrpb0l8lPS5p584P1czMOmqmSV5Sd+BiYCdgDWBfSWu02u1k4MaIWBfYB/hxZwdqZmYd156S/AbAmIh4LiI+BK4Hdmu1TwCLlOuLAi93XohmZjar2pPklwFeqrk9ttxX6zRgf0ljgRFAm5MySBomaZSkUePHd83SV2Zmc7POanjdF7gqIvoBOwNXS/rYY0fEZRExKCIG9e7d+2MPYmZmnas9SX4c0L/mdr9yX62DgRsBIuLPwHxAL8zMrK7ak+RHAitLWl5ST7JhdXirfV4EtgGQtDqZ5F0fY2ZWZzOdTz4iJks6Argd6A5cGRFPSjodGBURw4FjgcslfZ1shB0a9ZrD2MxmSVWm4rXptWvRkIgYQTao1t53as31p4BNOjc0MzObXR7xamZWYU7yZmYV5iRvZlZhXsjbzLrEHfeN5Jvf/TFTpk7lwD134thh+0y3/X/PuoQ/PfwYAO+99wHj3/gP40beDMDJ51zOH+59mKlTp7L1xgM556TDuzj65uUkb2Zz3JQpUzjm9IsYfuXZLNOnF5t/4Qh23nojVl9puY/2OfuEr350/ZKrb+bxp8cA8NBfnuShvzzBw7dcCsB2+32d+x55nJ1W265rn0STcnWNmc1xox4fzQrL9mX5/kvTs+c87Lnzltx214Mz3P+m2+7mC0O2AkAS738wiQ8nTeaDDycxafJkluy1WBdF3vxckjdrp2brRw6N05f85dcm0G/paVOZLLNUL0b+7R9t7vviuNd4ftyrbLHhOgAMXncNNh/8GVbabG8igmFf3I3VVlyuzf+1j3NJ3swayk0j7mb37Teje/fuADz7wjhGP/cio++5jn/eez1/eugxHhj19zpH2Tyc5M1sjuvbpxdjX5k208m4VyfQt0/b01vdNOKej6pqAG698wHW/8zqLLTg/Cy04Pxst/n6PPLYU3M85qpwkjezOW7g2qvy7AvjeH7sK3z44SRuGnEPO2+90cf2G/3ci/znzbcZvO60dYn6Lb0k9498nMmTpzBp0mTuH/k4q66wbFeG39RcJ2/WIGani+Eia+zAmqsMAKD/0kty4yVndGHkM9ejR3fOO+UIdj/4BKZMncoBe+zAGisP4IwfXsV6a63CkK03BuCm2+5hzyFbIumj//3cDptx70OPscGuhyCJ7TZdv80fCGub6jWP2KBBg2LUqFF1ObbZrJiTDa9TpkxhnR0Pmq6L4c/OO3G6Loa1WroYXnLmNwDos95nee0vt35sv440vDZbw3JHG5Wr8vwkPRoRg9r7OK6uMWsAs9PF0OyTOMmbNYC2uhi+/NqENvdt3cUQ4P0PPmSzPQ5nq72P5NY7H5jT4VoTcZ28WZNp3cUQ4Ok/XkPfPr3410uvMOTA41hzleVZYdm+dYzSGoWTvHWaZqvzhMYZLNTRLobnn3Lkx/4fYPn+S7PZBp/mb0+NcZI3wEm+SzkJ2ozUdjHsu2QvbhpxD1eee8LH9muri+HEN99igfnnZd6ePZkw8U0e+uuTHP2VvbsyfGtgTvJmDWB2uhiOfvZFjvrWD+jWrRtTp07lmEP2mWGvHJv7OMmbNYgdthjMDlsMnu6+U44aOt3tk4780sf+b8P11uSRWy+fk6FZE3OSt6YxO4OFdv/KCYz829NstN5a3HTpd7o4crP6cZK3pjA785EDfO3gL/Deex9w5Q23dWncZvXmfvIVcsd9I1l3x4P49PYHct5l139s+/+edQkb7X4oG+1+KOvsMJRl1t+964OcRbM7WGirjdZjoQUX6IpQzRqKS/IVMbsl3UY3O/ORm83NXJKvCA+Ln6atwUJmcysn+YqY3WHxjW525iM3m5s5yc+FmrGkOzvzkZvNzZzkK6LqJd3awUIDhxzM53fa/KPBQrf9cVq1VFuDhQC2++LXOeDoM7jnob+yyhb7cud9I7v6KZjVhRteK2J2hsU3i1kdLARwxzUXzKmwzBqak3xFzM6weDOrLif5Cpmdkq6ZVZPr5M3MKsxJ3syswpzkzcwqzEnezKzCnOTNzCrMSd7MrMKc5M3MKqxdSV7SjpJGSxoj6fgZ7LOXpKckPSnp2s4N08zMZsVMB0NJ6g5cDGwHjAVGShoeEU/V7LMycAKwSURMlLTknArYzMzarz0jXjcAxkTEcwCSrgd2A56q2ecQ4OKImAgQEf+e1YDeGX3nrP5rXSy46rb1DsHMbIbaU12zDPBSze2x5b5aqwCrSHpA0kOSdmzrgSQNkzRK0qjx48e3tYuZmXWizmp47QGsDGwJ7AtcLmmx1jtFxGURMSgiBvXu3bv1ZjMz62TtSfLjgP41t/uV+2qNBYZHxKSI+BfwTzLpm5lZHbUnyY8EVpa0vKSewD7A8Fb73EyW4pHUi6y+ea7zwjQzs1kx0yQfEZOBI4DbgaeBGyPiSUmnS9q17HY78Lqkp4C7geMi4vU5FfTsuOO+kay740F8evsDOe+y69vc59e/v5eBQw5m0C5f4aBjz+ziCM3MOk+75pOPiBHAiFb3nVpzPYBjyqVhTZkyhWNOv4jhV57NMn16sfkXjmDnrTdi9ZWW+2ifMc+P5bzLruPOa3/A4osuzL9fn1jHiM3MZs9cNeJ11OOjWWHZvizff2l69pyHPXfektvuenC6fa761e8Ztt+uLL7owgAsucTi9QjVzKxTzFVJ/uXXJtBv6Wm9epZZqhcvvzZhun3GPD+WMc+PY9t9v8ZWex/JHV7w2cyamJf/a2Xy5CmMeWEcv//FeYx7bTw77H8sDw+/jMUWWajeoZmZddhcVZLv26cXY1+ZNghr3KsT6Nun1/T7LNWLIVttyDzz9GBAv6VZacAyPPtC6x6jZmbNYa5K8gPXXpVnXxjH82Nf4cMPJ3HTiHvYeeuNptvns9tuwn2PPA7AhIlvMub5cQzot3Q9wjUzm21zVXVNjx7dOe+UI9j94BOYMnUqB+yxA2usPIAzfngV6621CkO23phtNx3EXfc/ysAhB9O9Wze+c9whLLH4IvUO3cxslsxVSR5ghy0Gs8MWg6e775Sjhn50XRLfO+EwvsdhXRyZmVnnm6uqa8zM5jZO8mZmFeYkb2ZWYU7yZmYV5iRvZlZhTvJmZhXmJG9mVmFO8mZmFeYkb2ZWYU7yZmYV5iRvZlZhTvJmZhXmJG9mVmFO8mZmFeYkb2ZWYU7yZmYV5iRvZlZhTvJmZhXmJG9mVmFO8mZmFeYkb2ZWYU7yZmYV5iRvZlZhTvJmZhXmJG9mVmFO8mZmFeYkb2ZWYU7yZmYV5iRvZlZhTvJmZhXWriQvaUdJoyWNkXT8J+y3h6SQNKjzQjQzs1k10yQvqTtwMbATsAawr6Q12thvYeBrwMOdHaSZmc2a9pTkNwDGRMRzEfEhcD2wWxv7nQGcDbzfifGZmdlsaE+SXwZ4qeb22HLfRyStB/SPiNs+6YEkDZM0StKo8ePHdzhYMzPrmNlueJXUDTgfOHZm+0bEZRExKCIG9e7de3YPbWZmM9GeJD8O6F9zu1+5r8XCwFrAPZKeBzYEhrvx1cys/tqT5EcCK0taXlJPYB9geMvGiHgzInpFxICIGAA8BOwaEaPmSMRmZtZuM03yETEZOAK4HXgauDEinpR0uqRd53SAZmY263q0Z6eIGAGMaHXfqTPYd8vZD8vMzDqDR7yamVWYk7yZWYU5yZuZVZiTvJlZhTnJm5lVmJO8mVmFOcmbmVWYk7yZWYU5yZuZVZiTvJlZhTnJm5lVmJO8mVmFOcmbmVWYk7yZWYU5yZuZVZiTvJlZhTnJm5lVmJO8mVmFOcmbmVWYk7yZWYU5yZuZVZiTvJlZhTnJm5lVmJO8mVmFOcmbmVWYk7yZWYU5yZuZVZiTvJlZhTnJm5lVmJO8mVmFOcmbmVWYk7yZWYU5yZuZVZiTvJlZhTnJm5lVmJO8mVmFOcmbmVVYu5K8pB0ljZY0RtLxbWw/RtJTkh6XdJek5To/VDMz66iZJnlJ3YGLgZ2ANYB9Ja3Rare/AoMi4tPATcD3OztQMzPruPaU5DcAxkTEcxHxIXA9sFvtDhFxd0S8W24+BPTr3DDNzGxWtCfJLwO8VHN7bLlvRg4Gft/WBknDJI2SNGr8+PHtj9LMzGZJpza8StofGASc09b2iLgsIgZFxKDevXt35qHNzKwNPdqxzzigf83tfuW+6UjaFjgJ2CIiPuic8MzMbHa0pyQ/ElhZ0vKSegL7AMNrd5C0LnApsGtE/LvzwzQzs1kx0yQfEZOBI4DbgaeBGyPiSUmnS9q17HYOsBDwK0mPSRo+g4czM7Mu1J7qGiJiBDCi1X2n1lzftpPjMjOzTuARr2ZmFeYkb2ZWYU7yZmYV5iRvZlZhTvJmZhXmJG9mVmFO8mZmFeYkb2ZWYU7yZmYV5iRvZlZhTvJmZhXmJG9mVmFO8mZmFeYkb2ZWYU7yZmYV5iRvZlZhTvJmZhXmJG9mVmFO8mZmFeYkb2ZWYU7yZmYV5iRvZlZhTvJmZhXmJG9mVmFO8mZmFeYkb2ZWYU7yZmYV5iRvZlZhTvJmZhXmJG9mVmFO8mZmFeYkb2ZWYU7yZmYV5iRvZlZhTvJmZhXmJG9mVmFO8mZmFdauJC9pR0mjJY2RdHwb2+eVdEPZ/rCkAZ0eqZmZddhMk7yk7sDFwE7AGsC+ktZotdvBwMSIWAm4ADi7swM1M7OOa09JfgNgTEQ8FxEfAtcDu7XaZzfg5+X6TcA2ktR5YZqZ2azo0Y59lgFeqrk9Fhg8o30iYrKkN4ElgAm1O0kaBgwrN9+WNHpWgp5FvVrHUzF+fs2rys8N/Pw623Id2bk9Sb7TRMRlwGVdecwWkkZFxKB6HLsr+Pk1ryo/N/Dzq7f2VNeMA/rX3O5X7mtzH0k9gEWB1zsjQDMzm3XtSfIjgZUlLS+pJ7APMLzVPsOBA8v1PYE/RkR0XphmZjYrZlpdU+rYjwBuB7oDV0bEk5JOB0ZFxHDgp8DVksYAb5A/BI2mLtVEXcjPr3lV+bmBn19dyQVuM7Pq8ohXM7MKc5I3M6swJ3mzBuTBhNZZnOTNGlBL7zRJfSX5ezqX6owfe394OkHtGyFpvnrG0h5VTRot74OkFeodS2eQdCBwFjBPvWPpLFX97M0JklTzY7+5pG1n5XH8gs+mVm/EIcAVkg6VtECdQ5uhiJgKIGkHSetLWrHcbtoqgpb3QdKOwAhJqzX58/kysD5wZkR8UO94OkvNZ29rSatKWqbeMTWqmrzyVeBHwAWSfiRppY48jpP8bKp5I/YC9gd+B3wZOELS0vWMrbVWZxwHAVeSsV4kacOSJJsqMZZZUimxfxo4FxgWEf9opgF5NWchLa//psDhNdu7dAqSztbGZ+8XwMnAKZIG1i2wBidpa2CXiPg0sA6wIHBYS8GsPZzkO4GkQcD/AGdHxPXk1MsbAQc0Ukml5gdpN2AFcqK5o4FbgDObLdFL6gvsKqmlOqMncFdE/EnSPC0/AI2u9mwQWBYgIr4M/Bi4VVLPMiixaRN9zWdvX2B5YF3gNGA0cJSk9eoXXWOS1BvYAVhL0noRMQU4lpz88ThJy7fncZzkZ0EbSXABYCJwsKSlI+IJ4CTyDdqrUZKNpG4llhOAPQBKVcC1wA3AjyWt30Ql4OWBfwILSVoSeAf4vKRNI2JSREyRtF2pRmtYNQnwCOBiSRdIOjwijgAeAB6RNG9ETK5roLOh5jtzAjA0IsZHxLPACOBvwMmS1qlXfI2gdV6JiPHAD4BfAwdKWisi3gC+AQT5eZ/54zbP97kxtKqDXwMYHxHjJa0GHAJMAn4QEa9KWhV4JyLGNki8C0TEu5LmJUvvL0XEIWXbIuS8Q3dFxAv1irejStw/AR4GrgB2J8+kriCnfz0HOCkiflevGNtD0q7kl3cIcDPwbEQMK9uuA5aPiA1blfobWqvPXs+yHgWSHiWf317l9urANsBvIuLlugVcR61eq/9h2qSQZwELAYcBCwNXRcRjkrq1tG/M9LGb5PPScCQdCewLjALmi4hhktYiJ2qbHzg9Iv5d5xhbNwqvCzwF/Iz8MbodeKYmmTRFAmkdp6SdyfmS/gT8EVgLOAJ4DbghIn7X6M9N0heBKeRnZ1+yHvZDSQMi4vlyhvhKfaNsv1afva8AA4BXIuLict9IMtHvU27PExGT6hVvo5B0NLArcBxZXfc2sDc5z9hxwGTgFGBSuz/PEeFLBy9kQnkAWIxc7vBF4NaybV3gu8CS9Y6zJt5DgfvJVb5eKh+eFck67L8CP6x3jLPwnLYpr/MW5QuwGXANeTY1T9lHtX8b8QJ0K383A8YAD9RsOxK4COhR7zhn4/kdXj57g4H/kg3jC5RtY8iSad3jrOPro5rrSwAXkiX2b5LVND8E7gM+BSwF9O7oMVySb4c2So4bA88CnyeXPvwCcAfwekQMqT01rYfSEPweWWrvA5wBHE/+OO1T7l+ALBGMA/pGE1TR1HSTHARcCvyF/KF6mvwyrEuWdv4A/CwasOth6XH1WkRMrWmEvBN4FPge2U52F9AbOAo4MLKNpylI2pRcLvTV0ungh2QhY29gL7Ie+RXg8Ij4QNLyEfGv+kVcP63OdtaMnN33U2QB7AfkD/+ywN3AM8CO0c4qmlpueJ2JVm/ElyVtHBEPklMqDwK+HRFvAfcC80nqV88EX3yGXLRl0Yh4lexB0w/YPSI2I0vA25FdPmmGBA8fdZNclyztHBHZnnA9ufzaUcBjwHnAQw2a4PuTDY97ShoKnAjMR67HsDU5ZfcTZP3rFmQDZdMk+GI9oEdp/xlHdtFdBdgzIrYgP4v7AUeW79ZcmeBhugb3w4GfSuod2bA6GXicPEPdnPxcHDwrCR6c5Geq5o04ivzyvVU2dSdfv40kfZNMrPtEfRtZuwFExE/JxprflW6R75R4+0paCFiTbKj8aTRfPehkMmkMBYiI35NnUf2BY8jqjsfqFdxMTAD+BQwEdgb2johTyWqZc4EVI+Jn5BnisIj4e90i7aCaz94PyeqGJ0t7wpuUPKMcDb4ScCvZVjLXVyNI2oWsYtwnsjcNwHiyDeNy8uzu1xHxUtuP0A71rpNq1Au5hGFLdVZvMpEsVW631KNuRfbeuAVYp87x1tbt9S9/TyF7amwACDibTO5/B9as92vckecFLEKemUAOChkJnFKz307A6vWOdwbPYTFKGw1ZGPg6eeZ3KjBvuX8PYCzZ4Fr3mGflPSrXe5W/55bP2nJk1eCFZLXDU8Bq9Y65jq/VYGCzmtv7kb2/AOYFupfrC5NjWfrN7jFdJ9+GMprsKOC4yB4OfcgP6G4R8YykHpGDU5aIiNdbbtc36iTpGDIJfi0iJko6kRw9eSpZnbEaMDHyVLqh1dTB70Y24HUDfhERVytHt14M/CkiTqproDMhaStgY2Bx8ot8GpnUVyN7Z/0qIiaVbpRPRMRz9Yp1dkg6FlgDODoi3pJ0NtlAvjvZ02k94NVokurBOaH0BHsMmBIRr0naBvg+cFBEPF72OQyYEBE3dcpB6/3L1qgXsvQ1GNik3D4NOJ/yy0pWF/yRrBZpiN4bZPfN+4DFy+35yt/DyB4O69U7xll4TtuTjZJ9ydLhO8D/lG3rkiX6lRrlPWgVe7/yOdoC+C3ZyL1T2bYw2c3zfLJff9P2oCnP5xCyC2vLGUtLD6cTyU4Ky9Y7xjq/PrVnO6uWz+3nyBqDrwNXl9sHkD/8nXZW2rTDpOeElhFnke9EH7JHwNqlNPw7cqDKXZJ+C3wW2Csi3q5XvG1YHrgNWEHS9sAWksZGxFeUE6aN/+R/b0iLAl8lG7nXJ78El0maPyLOlbR1ZMN3QylnHycAL5Ml+M3JBu+tJP07Ih6VdDHwv2QPigXILoZNQVKfiHit5q4Vyflo+knaH9hG0iMR8e0y+K4hRn3XQ+3AJeVkYy+RvcO+CrxL5pbXyR/7iWSp/ulOO375ZbEakj5HNnztJOl4YEPgTLJP+XbAVOCfUcfT6la9frpFdsnbnOxfuwhwFfA8WS1wSmSrfcOrqaKZN0oPmfIDdS1wQUTcK+mnwJZk3WbDjZAs1TOXkoOanidL7b8ku3z+i6xrPQtYkmxg+1OzvD8AytHdT5H17P+IiEvLYK6dyILGz4EPyB/lY6MBezrVQ6mO+yzZI2+spC8BXwLOj4gRpfFakXPUdBqX5FspI0N3Bb4FEBHfk3QSedp5bkSMqGd88LEEfyjQW9JbEXGhpFHA1Ih4X9Lnyb62TTMfeUnwu5DdDOcjexf8k2ws3lI5jcFCZG+EhkvwxcbkALNHJc0X2W6zH/ArYGmy0fVaMsFv2UwJvngbeBB4FdijdGu9G/huSwlU0u7k67AoUNeR3/WmnC9qMeBG4MGS4LtFxC8kTQG+I2lqRPxhThx/ru9C2VJFU2MiWS3z0ax4EfFdMskcqQZYFKQmwR9JVl8MB86WdAE5XQElqZwJ7NvqtLqhKaedPZ2so3yZnN1zR3ISq8XIpH9NRIysV4wzUvNZ6kf23Qf4QFL3iHiRPB1flBzsdCKwdUQ83+WBzqbIbsKPkN+Rncm6+H2ASyV9ujT+nwgcEHWe2qNeWuWVbhHxOtmGtJ6ko1qqbyLiGvIz3WnVMx9T7waJel6YvjFkPaBPub4j8D6wfav9l2igeJcmhz23jIy8neyydhVZ0h0MDKj3a9zB57cSWa97Sc19w8gRrAuX2y3vUcM1tNbEvA3Z5XZgud2NPJvqB9zU8lya8cK0Kt6e5EC0pciqs+fIieKGkw3la9c71ka4kIPBLiEb2ZcDVifr34/sqhjm6uqaaPnUZsljZ+CZUhNyuKQDgOskDY2IW8v+r9cr1lZVNP1iWp3eIOALEbFZ6fr5DFlfek7L/o2spg5+YbLh8Q3gM5K2jIh7IuKy0kYyiKwS+DdMe+8a1EPk3EZ7l+c3CpiqnA5jCZq4mrS8Vy2l1GfIEcYDgWMi4mblHOdvRvNVQXU6ScPIUeWnktMULB4RZ0jaAnhc0qSI+MmcjqNpP2ydpXzxdo6IbSVdk3epR0T8qjSEXCzpj8C79UwsNQn+aGB7SftHxBuSPgBeVY5kXZucYvemBk+CwMeW7DuUHBhyEXAQMEQ598njZJezCdDwyR2AiHhH0uXAV4DzJP0Z+JCcynnfiJhY1wBnU3kPPpT0S7J94eKIuLlsm2unKahVqnX7kfNa7Uj2bDurdCh4QtKaZAeOOR9LE3xnOlVtibjc3oocufoGucjH7pETJ20SEQ9IWigapJukpL3Joft7RhnmrJyz/giyG94a5A/WM/WLcuZKHfWUcn1bclbMQyLi3nJfL3Ju9V3ImQovjog71IE5tBuBpPnJM5AdyB+p30fE6PpG1bmUc/AMAL4fEe/WN5r6aZ1Xyn0nkyX5FyJih3Lf4cAbkSvIdYm5qiTfqspj4cj+1U+SDX0LRMTAsu1wYCflLIHtWn2liywEXBkRL7X8+ETEaEnnk32xJzZ6SUo5enh7STdGdq3bhOwrPqr0rx5Gno2cTc6vDtOqaJomwQNExHvk4LT76h3LHPQQOdfOXK0mr2wDfCoifkW2k61HDpps6QzxVfKMrsvMdSV5+GhAwtbkjH8/IPuqrkwOTHiRHL13QDTIDICS1ibrPw8EvhwRg2u27Q28GBF/rld8HVGeyySyZBvklAtXkiMA7yWHvx9FjkdYlJzOYCLZffW9esRsn0xlxbF6x1EPrQqOX2XaJIYvkoPf1iGra/qSC8IM6+q8MleU5DX9iLMNyGR5HNl1aSrwG3Io8Z5kw9gXI+KpOoU7HUkrk0lvVOSgk8GS7iS7Fm5OLuy7az1jbA/losRDgZ9EzmtyMblS0HckbUmehYxVLs59CDB/RIyRdAnwHyf4xuUE/1Ed/HzkAL3/SrqC/I6eD1xH9oJ7P3JWzq6Nc24qyZdTqfnJ2SSvUC7g8BOyce/CiJhQ1wDbIGkesg/yILKv/i/ImS8XI/tiH9coP0ifRNJm5BnTWHK05ybkXB0vAtdFxLhSPXYycHJE/Latek6zRtAqwX+DnIStN3nGeXm5/zLyO/r1qOOkbJUuybd6Iw4iT58eAdaS9PeIeFi5/uQNZZ9vNUq9r6R9yEEU1yoXcp5ETnT1TkR8rewzX0S8X884O+AhsjrsS+TCEeeTQ9/3AfaSdC25YtCxEfEHJ3hrZDV5ZWOyyvFYchqHz0p6IyJ+Hbnu84WUAYr1UukkX/NG7EyWhDchS5JfBU6S9N2S6PciZ2ysW4JvI6lNBb4t6YOI+LWkX5MDnI5WzuVyJZkkG1bpM/1GRLwZOZXu38ih7v8hz6jOIuvlv0xWoZ0b00YCOsFbQ9K0uaI2BH4E3FXyyNNkL70vla6S17YUyOqpkkm+pv91DzJZHkD2te4XEf8q/eEBvi/p2MjBKnXTutcPMDkibpQ0CfhW6XJ4o6THyOR4a5MkwRWBv0pavMR7E9nT5DqyBH8cWfXUnayTb4izKLO2lPaxN6NM1RARD0n6AzBQ0uoR8bSkW8i6+d0k3Qq8Xe/vauXq5Mvp045kv923y33zkn2xPyQnURpbGgL3AG6L2VlaqxOVur1BZEv8sRExUtJnyVn9hpNnIkMi4p91DLNDykCnH5O9gx6KiG+V+7chB4q8Ss7KV60PolWKcubNlhHXx5CLevytbDuXnH3z1MjFuBcg1wdoiKmjK5PkpY+GWt9MTud5G/B/5Eo7d5dE/xMy0Z8ZES/Ue3CNcjKu7mTD74Hkavbbkx+m5cjuVrdLWotcl3VURDxbr3hnVUnot5MLSdQOi98aeDk6ce5sszmltI2tQU7I1gf4e0ScUbadSH6ej4iIf9Qvyo+rTJJvIWkVshrgWbLBYz9yDdara6pq/k32Sqnbkn2lhHsGOYz/AfLsYziZ6AcD9wDfAYZGxPA6hdlpSrvIhcBGjdiLyWxGWkZoS9qB7Pd+Ndke9gS5lN+95Nnq4cAvI2fpbBhVrJN/g3wDHo+ciP89chKltZXzvBxD9lqpZ4Lfgkzu+0WZMlfSj8mpCYYAu0bE22UE6FGS7mj2fuLlvZgCPClptWjy+Vts7hHTFvF4llyP9R5yMZhJ5Ayp65EFye0joq49adpSuSQfEROUE0KdIGkJckDCgeQSW1cAPRugDn4g8KNS594jIiaXaozxZO+fzysXGngaOL3ZE3yLUvX0ZeAz5BfFrCmUzhFjJH2L7Nm2BHBoRNxStvdpxAQPFUvyLb1UIuIaSesDF5OjV28tu+xXx/Bqe9EsD7SMfKtd6msyuTjGZsBGwN71HEQxJ0TEbdD2hE5mjarms/pncqWymyPilpqxKg27OErVVoYSgHKJuJHAn1sSvHLa4Lqq+aD8FthQ0sBSgu9W6v0+JE8BLwY2j4gn6xbsHOYEb42oplNA6/u7A0SusvZX4PDSceP9cn/Dfp7rnvhmRe0bUfrCt5QMp5aeHD8iE2lvSadCw81g+DBwP7moxMCImFoadvYlq5ZeDy+6YNalWo1XOaBcDoasl6/JO2eTU4wsWadQO6Tpete0eiOGAhsCvwduJV/064DLI6cD2Bh4qQHq4D9GuSDGweRScaOA98gJ0vaMBpn90mxupFyYZzeysPhd4KyI+HnZ1pLo52uWtrKmS/ItlHO7HEnOO3MQuY7iA+QKTv8q+zT0IhPKRSUGAtuS87bc3UwDncyqoGaEfDdyzYYfR8T+kk4gp+HYHZg3mnS2zaZM8pIGk12Z/rcMLd6SnCBoOLn0nbvnmVmHSFoxIp6VNJzsir0I2c35feUEh09FxMP1jbLjmqJOvuUUSQXZfWkqcJhyhaR7yKR/APC5RmhkNbPmIWk5YISkpcjR2TuT0xS8L+lL5Pial+sZ46xq+JJ8qzr4AWSj5FulNL8/uUDuuRHxbiPXwZtZ46mpqukOnEZ2iLgTOJ5sM7uNnDPqgGbt7dbwSb6FpCPJhsl/AOMj4mRJmwB7k/OUn96sdWZm1nUkLRllJklJG0TEI+X6wWT73laRU2NvSo5n+U8zFxybolqjnC7tTS4YPBn4sqQrIuIBsqtkd2CBOoZoZk1AucbB/pIWKCPiz5R0jXJZ0J+RI7G/WUr490fE35s5wUODJvk2BiS8Ri4Vtw8wgFyBZXBJ9HeTdWee9MrMZuZ14KdkHtkS2ImcaGw/4C5yMOKARh7c1FENN61Bqzr4/YFHgTvIxTI2A06MXA/0XmCQpKUi4tX6RWxmjU7SQuTA1P8qF4v/ArAUuXLZWWWfYeRgxLUknVxGtza9hkvyNQn+KHI90KFlJKvIqpr1JG1NLqyxi0vwZvZJyniUzYCFJK1K9sy7jKwZ2E3SIhFxS0RcJmkE8F5EvF7HkDtVwyR5SQtGxDvlej+ykXVIRLxWSvdvl/6rm5ELaBzjBG9mMxMR70l6n1y/YXFy4r9XJF1Nltw3k7RARFwXDTYXfGdoiDp5SasDp0iap9zVHViUaQtVt/wY/S4ijgR2iojHujZKM2smrdr2HgSeKn/XktSvFBIvA/4LrK5cX7lyGqYLZWnpHgBMiojHJf2QfPG/X+rRDiYbX/cEPqhSw4iZda5WbXurAS+SS39uCHyRHL16Uam+WQx4pqqTAta9uqZmfplFgSOAxZXrJd4A7ArcI+lmclm8vVum9jQzm5GaBH8E2XPmfjLJfwuYlxwZ/1tgXWCTqiZ4qGNJXtK8EfFBub438PmI2FvSOWSr99nAc2Tf+A+AxyLimboEa2ZNQdLCEfFWub4vue7qbsD5ZEIfBQwD+pEjWR+NiNF1CrdL1CXJKxd17h8Rl0o6lOyvek5E/KVsP59cDf0sT7trZu0haUXgUHKSwkck7QT8BdiDnEnyKOAnwEvAV1oKmVXX5Q2vZcTZpcAfy10fkKNZl2nZJyKOAd4CjpXUs6tjNLOmtCjZPfJzktaOiN+T7XobAodFxD+AMWTHjiXqF2bX6vKSfEna3yOrZFaOiPUlfRM4iawbe6Jm34/mmDAza4ukxSLiP+X6msC+QE/gxogYJen/gJvJcTb7APvMTXmly0rykvoAlHVMA9iFHEZMRHyf7MN6h6R1Wv5nbnojzKzjJG0LPCLpQknrk/PA/wh4hyzRLwecDGxEduQ4Zm7LK11Ski9dmJ4CLiLXRvwbsBawAjm67Myy36nk4ITVya6U7iZpZjNUCoUPkT1nTgS+RnbaWI2chnxJshv2a5J6lkLmXKWrSvJvk0vzvUzOGXEoMBF4HphP0nEAEXE6sH5EfOgEb2YzUwZFrke27f0X2J6slx9ItvV9HThe0jxzY4KHLqyTLz1mliEHIuwNfBZYCfg52cXp5xFxde0gBjOz9ihVNXcCX4uIq8oiIJ8hk/4tEfF0XQOsozme5GtWXukJ/AI4mjyVugq4lUz0a5MT9bsfvJnNkpLo/w84KSJ+XO94GsUcH/FaEnzLHBLPAOeRp1JHR8TNkvoD71Zp1jcz63oRMbI0xI6U9H5EXFnvmBpBl3ahLPNE3AtcHBFndNmBzWyuIWldsuBY6ZGs7VWPfvJDyYnIvu81Wc3M5qx6TDX8ENkabmZmc1i95q5ZwKV4M7M5r2Hmkzczs87XECtDmZnZnOEkb2ZWYU7yZmYV5iRvZlZhTvJmZhXmJG9mVmH/D4da9eVLiJZxAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "race_acc = analysis_nd_r.groupby(\"race\")[\"prediction_recid\"].mean().reset_index()\n",
    "ax= sns.barplot(data=race_acc,x=\"race\",y=\"prediction_recid\",color=\"#FED8B1\")\n",
    "plt.title(\"  COMPASS accuracy rate by race\")\n",
    "plt.xticks(rotation=45)\n",
    "plt.xlabel(\"\")\n",
    "plt.ylabel(\"\")\n",
    "for num,text in zip(range(6),round(race_acc[\"prediction_recid\"],2)):\n",
    "    ax.text(num,text-0.05,text)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "5a25723c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_d7fa2_row0_col1 {\n",
       "  background-color: #aedea7;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_d7fa2_row0_col2, #T_d7fa2_row1_col2 {\n",
       "  background-color: #f7fcf5;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_d7fa2_row0_col3, #T_d7fa2_row1_col3 {\n",
       "  background-color: #00441b;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "#T_d7fa2_row1_col1 {\n",
       "  background-color: #cdecc7;\n",
       "  color: #000000;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_d7fa2_\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th class=\"col_heading level0 col0\" >sex</th>\n",
       "      <th class=\"col_heading level0 col1\" >decile_score</th>\n",
       "      <th class=\"col_heading level0 col2\" >2y_r</th>\n",
       "      <th class=\"col_heading level0 col3\" >accuracy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_d7fa2_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_d7fa2_row0_col0\" class=\"data row0 col0\" >Female</td>\n",
       "      <td id=\"T_d7fa2_row0_col1\" class=\"data row0 col1\" >0.360000</td>\n",
       "      <td id=\"T_d7fa2_row0_col2\" class=\"data row0 col2\" >0.210000</td>\n",
       "      <td id=\"T_d7fa2_row0_col3\" class=\"data row0 col3\" >0.660000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_d7fa2_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_d7fa2_row1_col0\" class=\"data row1 col0\" >Male</td>\n",
       "      <td id=\"T_d7fa2_row1_col1\" class=\"data row1 col1\" >0.420000</td>\n",
       "      <td id=\"T_d7fa2_row1_col2\" class=\"data row1 col2\" >0.350000</td>\n",
       "      <td id=\"T_d7fa2_row1_col3\" class=\"data row1 col3\" >0.660000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7f91aaf4c3d0>"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Recidivism by gender\n",
    "prediction_sex = analysis_nd_r.groupby(\"sex\")[\"binary_score\"].mean()\n",
    "recidivism_sex = analysis_nd_r.groupby(\"sex\")[\"2y_r\"].mean()\n",
    "comparacion_sex = analysis_nd_r.groupby(\"sex\")[\"prediction_recid\"].mean()\n",
    "\n",
    "dt_comp_sex_recidivism = pd.concat([prediction_sex,recidivism_sex,comparacion_sex],axis=1).reset_index()\n",
    "dt_comp_sex_recidivism.columns = [\"sex\",\"decile_score\",\"2y_r\",\"accuracy\"]\n",
    "dt_comp_sex_recidivism.round(2).style.background_gradient(cmap='Greens',axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4f063866",
   "metadata": {},
   "source": [
    "<b>If we look into sex, in real world males are more recidivists than females and we see same in compass. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "349d6726",
   "metadata": {},
   "source": [
    "# 4.2 Percentage of false positives and negative positives"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "99235224",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create 2 attributes one for the name and one for the result\n",
    "def wrong_prediction(att):\n",
    "    # Build 2 lists one for the FPR and one for FNR\n",
    "    list_att = []\n",
    "    FPR = []\n",
    "    FNR = []\n",
    "    for x in analysis_nd_r[att].unique().tolist():    \n",
    "        #Filter by race or gender\n",
    "        data = analysis_nd_r[analysis_nd_r[att]==x] \n",
    "        # create sorting report (with output_dict we return a dictionary)   \n",
    "        classif_race = classification_report(data['2y_r'],data[\"binary_score\"],output_dict=True)\n",
    "        list_att.append(x)\n",
    "        # False Positive Rate is 1-exahusivity(recall)\n",
    "        false_positive = 1 - classif_race.get(\"0\")[\"recall\"]\n",
    "        FPR.append(false_positive)\n",
    "        # False Negative Rate 1-TPR\n",
    "        false_negative = 1 - classif_race.get(\"1\")[\"recall\"]\n",
    "        FNR.append(false_negative)\n",
    "    # creamos dataframe\n",
    "    df_fpr = pd.DataFrame({x:list_att,\"FPR\":FPR,\"FNR\":FNR})\n",
    "    return df_fpr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "744da005",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_03a1d_row0_col1, #T_03a1d_row1_col2, #T_03a1d_row2_col1, #T_03a1d_row3_col1, #T_03a1d_row4_col1, #T_03a1d_row5_col1 {\n",
       "  background-color: #fff5eb;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_03a1d_row0_col2, #T_03a1d_row1_col1, #T_03a1d_row2_col2, #T_03a1d_row3_col2, #T_03a1d_row4_col2, #T_03a1d_row5_col2 {\n",
       "  background-color: #7f2704;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_03a1d_\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th class=\"col_heading level0 col0\" >Native American</th>\n",
       "      <th class=\"col_heading level0 col1\" >FPR</th>\n",
       "      <th class=\"col_heading level0 col2\" >FNR</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_03a1d_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_03a1d_row0_col0\" class=\"data row0 col0\" >Other</td>\n",
       "      <td id=\"T_03a1d_row0_col1\" class=\"data row0 col1\" >0.150538</td>\n",
       "      <td id=\"T_03a1d_row0_col2\" class=\"data row0 col2\" >0.576271</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_03a1d_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_03a1d_row1_col0\" class=\"data row1 col0\" >African-American</td>\n",
       "      <td id=\"T_03a1d_row1_col1\" class=\"data row1 col1\" >0.455660</td>\n",
       "      <td id=\"T_03a1d_row1_col2\" class=\"data row1 col2\" >0.302256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_03a1d_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_03a1d_row2_col0\" class=\"data row2 col0\" >Caucasian</td>\n",
       "      <td id=\"T_03a1d_row2_col1\" class=\"data row2 col1\" >0.220068</td>\n",
       "      <td id=\"T_03a1d_row2_col2\" class=\"data row2 col2\" >0.494118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_03a1d_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_03a1d_row3_col0\" class=\"data row3 col0\" >Hispanic</td>\n",
       "      <td id=\"T_03a1d_row3_col1\" class=\"data row3 col1\" >0.199153</td>\n",
       "      <td id=\"T_03a1d_row3_col2\" class=\"data row3 col2\" >0.605634</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_03a1d_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_03a1d_row4_col0\" class=\"data row4 col0\" >Asian</td>\n",
       "      <td id=\"T_03a1d_row4_col1\" class=\"data row4 col1\" >0.210526</td>\n",
       "      <td id=\"T_03a1d_row4_col2\" class=\"data row4 col2\" >0.250000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_03a1d_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_03a1d_row5_col0\" class=\"data row5 col0\" >Native American</td>\n",
       "      <td id=\"T_03a1d_row5_col1\" class=\"data row5 col1\" >0.100000</td>\n",
       "      <td id=\"T_03a1d_row5_col2\" class=\"data row5 col2\" >1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7f91aaea0af0>"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "wrong_prediction(\"race\").style.background_gradient(cmap='Oranges',axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ff3e06d3",
   "metadata": {},
   "source": [
    "We can see clearly in the table above, that the African- American has the highest % of False positive rate compared to the rest of races. \n",
    "=> <b> This shows clearly the mistreat discrimination "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "1b3772f6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_133a7_row0_col1, #T_133a7_row1_col1 {\n",
       "  background-color: #f7fbff;\n",
       "  color: #000000;\n",
       "}\n",
       "#T_133a7_row0_col2, #T_133a7_row1_col2 {\n",
       "  background-color: #08306b;\n",
       "  color: #f1f1f1;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_133a7_\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th class=\"col_heading level0 col0\" >Female</th>\n",
       "      <th class=\"col_heading level0 col1\" >FPR</th>\n",
       "      <th class=\"col_heading level0 col2\" >FNR</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_133a7_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_133a7_row0_col0\" class=\"data row0 col0\" >Male</td>\n",
       "      <td id=\"T_133a7_row0_col1\" class=\"data row0 col1\" >0.317829</td>\n",
       "      <td id=\"T_133a7_row0_col2\" class=\"data row0 col2\" >0.383756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_133a7_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_133a7_row1_col0\" class=\"data row1 col0\" >Female</td>\n",
       "      <td id=\"T_133a7_row1_col1\" class=\"data row1 col1\" >0.312715</td>\n",
       "      <td id=\"T_133a7_row1_col2\" class=\"data row1 col2\" >0.448052</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7f91aae0ee50>"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "wrong_prediction(\"sex\").style.background_gradient(cmap='Blues',axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7eeb6687",
   "metadata": {},
   "source": [
    "For Gender, Recidivism is same for both gender.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "a9ed10e5",
   "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": [
    "d = analysis_nd_r.groupby([\"decile_score\",\"race\"]).agg({\"2y_r\":\"mean\"}).reset_index()\n",
    "d = d[d[\"race\"].isin([\"African-American\",\"Caucasian\"])]\n",
    "im = sns.scatterplot(data=d,x=\"decile_score\",y=\"2y_r\",hue=\"race\")\n",
    "im.set(ylim=(0,1))\n",
    "plt.show(block=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cede2631",
   "metadata": {},
   "source": [
    "# 4.SUMMARY\n",
    "1. Compas evaluates by race and gender; there are some differences in results, but we must not forget that one of the races has the highest prevalence (nearly 50%) and they are also the most likely to recidivate (nearly 40%).\n",
    "2. Adding the two_year_r feature based on compas report was for the purpose to see if adding this feature will help to narrow the prediction coming from COMPASS \n",
    "3. We saw that is_violent_racid got better compard to is_recid but no improvement on Positive cases prediction it became worse. \n",
    "My suggestion, is to remove the race from the full data and train the model in a way that it will be always independant from the race. In that way the discrimination will be avoided, maybe not fully but at least it would be better.   "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e5834621",
   "metadata": {},
   "source": [
    "# 9.References"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "55cba2a6",
   "metadata": {},
   "source": [
    "kaggle.com. (n.d.). Compas: thoroughly investigating the controversial. [online] Available at: https://www.kaggle.com/code/gonzalogarciafuste/compas-thoroughly-investigating-the-controversial/data [Accessed 25 OCt. 2022]."
   ]
  }
 ],
 "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": "3.8.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}