iv_and_woe_in_python.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Information Value (IV) and Weight of Evidence (WOE) in Python\n",
    "\n",
    "Information value (IV) and weight of evidence (WOE) are simple and powerful techniques of conducting attribute relevance analysis. They provide a great framework for exploratory analysis and have been used extensively in the credit risk world for several decades.\n",
    "\n",
    "This simple module enables quick attribute relevance analysis: calculation of IV and WOE, creation of readable visualizations and evaluation of statistical significance and effect size.\n",
    "\n",
    "- In order to see how it can be used in practice, check [**Telco Customer Churn Analysis**](https://github.com/klaudia-nazarko/iv-and-woe-python/blob/master/telco_customer_churn_analysis.ipynb).\n",
    "- What about statistical significance? Check [**Information Value vs Statistical Significance Analysis**](https://github.com/klaudia-nazarko/iv-and-woe-python/blob/master/statistical_significance_of_information_value.ipynb)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from pandas.api.types import is_numeric_dtype\n",
    "\n",
    "from Features import *\n",
    "from AttributeRelevance import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.options.display.float_format = '{:,.2f}'.format\n",
    "pd.set_option('display.max_columns', None)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('data/telco_churn.csv', na_values=[' '])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.columns = [c[0].lower() + c[1:] for c in df.columns]\n",
    "df['label'] = df['churn'].map({'Yes': 0, 'No': 1})\n",
    "df['seniorCitizen'] = df['seniorCitizen'].map({1: 'Yes', 0: 'No'})\n",
    "df.drop(['customerID', 'churn'], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>gender</th>\n",
       "      <th>seniorCitizen</th>\n",
       "      <th>partner</th>\n",
       "      <th>dependents</th>\n",
       "      <th>tenure</th>\n",
       "      <th>phoneService</th>\n",
       "      <th>multipleLines</th>\n",
       "      <th>internetService</th>\n",
       "      <th>onlineSecurity</th>\n",
       "      <th>onlineBackup</th>\n",
       "      <th>deviceProtection</th>\n",
       "      <th>techSupport</th>\n",
       "      <th>streamingTV</th>\n",
       "      <th>streamingMovies</th>\n",
       "      <th>contract</th>\n",
       "      <th>paperlessBilling</th>\n",
       "      <th>paymentMethod</th>\n",
       "      <th>monthlyCharges</th>\n",
       "      <th>totalCharges</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>94</th>\n",
       "      <td>Male</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>72</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Fiber optic</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Two year</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Bank transfer (automatic)</td>\n",
       "      <td>100.90</td>\n",
       "      <td>7,459.05</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>Male</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>25</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Fiber optic</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Month-to-month</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Electronic check</td>\n",
       "      <td>105.50</td>\n",
       "      <td>2,686.05</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6381</th>\n",
       "      <td>Male</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>1</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>No internet service</td>\n",
       "      <td>Month-to-month</td>\n",
       "      <td>No</td>\n",
       "      <td>Mailed check</td>\n",
       "      <td>19.55</td>\n",
       "      <td>19.55</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6869</th>\n",
       "      <td>Female</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>45</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>DSL</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>One year</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Credit card (automatic)</td>\n",
       "      <td>54.15</td>\n",
       "      <td>2,319.80</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4217</th>\n",
       "      <td>Male</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>14</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Fiber optic</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Month-to-month</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Electronic check</td>\n",
       "      <td>75.10</td>\n",
       "      <td>1,013.35</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      gender seniorCitizen partner dependents  tenure phoneService  \\\n",
       "94      Male            No      No         No      72          Yes   \n",
       "14      Male            No      No         No      25          Yes   \n",
       "6381    Male            No      No        Yes       1          Yes   \n",
       "6869  Female            No     Yes        Yes      45          Yes   \n",
       "4217    Male           Yes     Yes         No      14          Yes   \n",
       "\n",
       "     multipleLines internetService       onlineSecurity         onlineBackup  \\\n",
       "94             Yes     Fiber optic                   No                  Yes   \n",
       "14              No     Fiber optic                  Yes                   No   \n",
       "6381            No              No  No internet service  No internet service   \n",
       "6869            No             DSL                   No                  Yes   \n",
       "4217           Yes     Fiber optic                   No                   No   \n",
       "\n",
       "         deviceProtection          techSupport          streamingTV  \\\n",
       "94                    Yes                   No                  Yes   \n",
       "14                    Yes                  Yes                  Yes   \n",
       "6381  No internet service  No internet service  No internet service   \n",
       "6869                   No                  Yes                   No   \n",
       "4217                   No                   No                   No   \n",
       "\n",
       "          streamingMovies        contract paperlessBilling  \\\n",
       "94                    Yes        Two year              Yes   \n",
       "14                    Yes  Month-to-month              Yes   \n",
       "6381  No internet service  Month-to-month               No   \n",
       "6869                   No        One year              Yes   \n",
       "4217                   No  Month-to-month              Yes   \n",
       "\n",
       "                  paymentMethod  monthlyCharges  totalCharges  label  \n",
       "94    Bank transfer (automatic)          100.90      7,459.05      1  \n",
       "14             Electronic check          105.50      2,686.05      1  \n",
       "6381               Mailed check           19.55         19.55      1  \n",
       "6869    Credit card (automatic)           54.15      2,319.80      0  \n",
       "4217           Electronic check           75.10      1,013.35      1  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sample(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create Feature Objects\n",
    "\n",
    "Basic object in IV & WOE analysis is a single feature. There are two types of features: categorical and continuous.\n",
    "\n",
    "Create feature objects by instantiating CategoricalFeature and ContinuousFeature class.\n",
    "\n",
    "**CategoricalFeature(df, colname)**\n",
    "\n",
    "Creates a feature object.\n",
    "* df - DataFrame with data\n",
    "* colname - name of the column containing a feature\n",
    "\n",
    "Attributes and methods:\n",
    "* feature - returns name of a column\n",
    "* df - returns a dataframe\n",
    "* df_lite - returns a new dataframe limited to bin and label\n",
    "\n",
    "**ContinuousFeature(df, colname)**\n",
    "\n",
    "Creates a feature object. Groups continuous values into bins.\n",
    "* df - DataFrame with data\n",
    "* colname - name of the column containing a feature\n",
    "\n",
    "Attributes and methods:\n",
    "* feature - returns name of a column\n",
    "* df - returns a dataframe\n",
    "* df_lite - returns a new dataframe limited to bin and label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "feat_gender = CategoricalFeature(df, 'gender')\n",
    "feat_charges = ContinuousFeature(df, 'totalCharges')\n",
    "feat_contract = CategoricalFeature(df, 'contract')\n",
    "feat_tenure = ContinuousFeature(df, 'tenure')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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>bin</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Female</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Male</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Male</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Male</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Female</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      bin  label\n",
       "0  Female      1\n",
       "1    Male      1\n",
       "2    Male      0\n",
       "3    Male      1\n",
       "4  Female      0"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feat_gender.df_lite.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create a list of features that will be analyzed. It can be done manually or dynamically, as below:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "feats_manual = [feat_gender, feat_charges, feat_contract, feat_tenure]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "feats_dict = {}\n",
    "\n",
    "for col in [c for c in df.columns if c != 'label']:\n",
    "    if is_numeric_dtype(df[col]):\n",
    "        feats_dict[col] = ContinuousFeature(df, col)\n",
    "    else:\n",
    "        feats_dict[col] = CategoricalFeature(df, col)\n",
    "\n",
    "feats = list(feats_dict.values())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Information Value Class\n",
    "\n",
    "**IV()**\n",
    "\n",
    "Attributes and methods:\n",
    "- group_by_feature(feature) - returns a dataframe with number of observations in each bin and count of good and bad\n",
    "- calculate_iv(feature) - returns a dataframe (with calculation of WOE and IV) and IV for a single feature\n",
    "- draw_woe(feature) - draws a chart with visualization of WOE values for a single feature\n",
    "- interpretation(iv_value) - returns the predictive power based on input IV\n",
    "- interpret_iv(feature) - returns the predictive power based on IV of a single feature\n",
    "- print_iv(feature) - prints interpretation of IV of a single feature"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "iv = IV()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>totalCharges</th>\n",
       "      <th>count</th>\n",
       "      <th>good</th>\n",
       "      <th>bad</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>18.80</td>\n",
       "      <td>1172</td>\n",
       "      <td>599</td>\n",
       "      <td>573</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>197.95</td>\n",
       "      <td>1172</td>\n",
       "      <td>819</td>\n",
       "      <td>353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>678.37</td>\n",
       "      <td>1172</td>\n",
       "      <td>889</td>\n",
       "      <td>283</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1,397.47</td>\n",
       "      <td>1172</td>\n",
       "      <td>899</td>\n",
       "      <td>273</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2,745.37</td>\n",
       "      <td>1172</td>\n",
       "      <td>944</td>\n",
       "      <td>228</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>4,919.84</td>\n",
       "      <td>1172</td>\n",
       "      <td>1013</td>\n",
       "      <td>159</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>MISSING</td>\n",
       "      <td>11</td>\n",
       "      <td>11</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  totalCharges  count  good  bad\n",
       "0        18.80   1172   599  573\n",
       "1       197.95   1172   819  353\n",
       "2       678.37   1172   889  283\n",
       "3     1,397.47   1172   899  273\n",
       "4     2,745.37   1172   944  228\n",
       "5     4,919.84   1172  1013  159\n",
       "6      MISSING     11    11    0"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "iv.group_by_feature(feat_charges)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>totalCharges</th>\n",
       "      <th>count</th>\n",
       "      <th>good</th>\n",
       "      <th>bad</th>\n",
       "      <th>perc_good</th>\n",
       "      <th>perc_bad</th>\n",
       "      <th>perc_diff</th>\n",
       "      <th>woe</th>\n",
       "      <th>iv</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>18.80</td>\n",
       "      <td>1172</td>\n",
       "      <td>599</td>\n",
       "      <td>573</td>\n",
       "      <td>0.12</td>\n",
       "      <td>0.31</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>-0.97</td>\n",
       "      <td>0.19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>197.95</td>\n",
       "      <td>1172</td>\n",
       "      <td>819</td>\n",
       "      <td>353</td>\n",
       "      <td>0.16</td>\n",
       "      <td>0.19</td>\n",
       "      <td>-0.03</td>\n",
       "      <td>-0.18</td>\n",
       "      <td>0.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>678.37</td>\n",
       "      <td>1172</td>\n",
       "      <td>889</td>\n",
       "      <td>283</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0.15</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.13</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1,397.47</td>\n",
       "      <td>1172</td>\n",
       "      <td>899</td>\n",
       "      <td>273</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0.15</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2,745.37</td>\n",
       "      <td>1172</td>\n",
       "      <td>944</td>\n",
       "      <td>228</td>\n",
       "      <td>0.18</td>\n",
       "      <td>0.12</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.40</td>\n",
       "      <td>0.02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>4,919.84</td>\n",
       "      <td>1172</td>\n",
       "      <td>1013</td>\n",
       "      <td>159</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.09</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.83</td>\n",
       "      <td>0.09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>MISSING</td>\n",
       "      <td>11</td>\n",
       "      <td>11</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  totalCharges  count  good  bad  perc_good  perc_bad  perc_diff   woe   iv\n",
       "0        18.80   1172   599  573       0.12      0.31      -0.19 -0.97 0.19\n",
       "1       197.95   1172   819  353       0.16      0.19      -0.03 -0.18 0.01\n",
       "2       678.37   1172   889  283       0.17      0.15       0.02  0.13 0.00\n",
       "3     1,397.47   1172   899  273       0.17      0.15       0.03  0.17 0.00\n",
       "4     2,745.37   1172   944  228       0.18      0.12       0.06  0.40 0.02\n",
       "5     4,919.84   1172  1013  159       0.20      0.09       0.11  0.83 0.09\n",
       "6      MISSING     11    11    0       0.00      0.00       0.00  0.00 0.00"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Information value:  0.3152298256031499\n"
     ]
    }
   ],
   "source": [
    "iv_df, iv_value = iv.calculate_iv(feat_charges)\n",
    "display(iv_df)\n",
    "print('Information value: ', iv_value)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "iv.draw_woe(feat_charges)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Totalcharges is a strong predictor.\n",
      "A feature with IV = 0.2 is a medium predictor.\n",
      "\n",
      "Information value: 0.32\n",
      "Totalcharges is a strong predictor\n"
     ]
    }
   ],
   "source": [
    "print('%s is a %s predictor.' % (feat_charges.feature.capitalize(), iv.interpret_iv(feat_charges)))\n",
    "print('A feature with IV = 0.2 is a %s predictor.' % iv.interpretation(0.2))\n",
    "print()\n",
    "iv.print_iv(feat_charges)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Statistical Significance Class\n",
    "\n",
    "**StatsSignificance()**\n",
    "\n",
    "Attributes and methods:\n",
    "- group_by_feature(feature) - returns a dataframe with number of observations in each bin and count of good and bad\n",
    "- calculate_chi(feature) - returns p-value and effect size (Cramer's V) for a single feature\n",
    "- interpretation(effect_size) - returns the predictive power based on input effect size\n",
    "- interpret_iv(feature) - returns the predictive power based on effect size of a single feature\n",
    "- print_iv(feature) - prints interpretation of p-value and effect size of a single feature"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "s = StatsSignificance()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "p-value: 0.00\n",
      "effect size: 0.25\n"
     ]
    }
   ],
   "source": [
    "p_value, effect_size = s.calculate_chi(feat_charges)\n",
    "print('p-value: %0.2f' % p_value)\n",
    "print('effect size: %0.2f' % effect_size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Totalcharges is a medium predictor.\n",
      "A feature with effect size = 0.2 is a medium predictor.\n",
      "\n",
      "P-value: 0.00\n",
      "Effect size: 0.25\n",
      "Totalcharges is a medium predictor\n"
     ]
    }
   ],
   "source": [
    "print('%s is a %s predictor.' % (feat_charges.feature.capitalize(), s.interpret_chi(feat_charges)))\n",
    "print('A feature with effect size = 0.2 is a %s predictor.' % s.interpretation(0.2))\n",
    "print()\n",
    "s.print_chi(feat_charges)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Attribute Relevance Class\n",
    "\n",
    "**AttributeRelevance()**\n",
    "\n",
    "Attributes and methods:\n",
    "- bulk_iv(features_list, iv, woe_extremes=False) - returns IV for features from the list; optionally: returns min and max WOE\n",
    "- bulk_stats(features_list, s) - returns p-value and effect size for features from the list\n",
    "- analyze(features_list, iv, s=None, interpretation=False) - returns IV (and p-value and effect size) for features from the list; optionally: returns interpretation of predictive power based on IV and effect size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "ar = AttributeRelevance()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "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>iv</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>gender</th>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>seniorCitizen</th>\n",
       "      <td>0.11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>partner</th>\n",
       "      <td>0.12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dependents</th>\n",
       "      <td>0.16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>tenure</th>\n",
       "      <td>0.78</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                iv\n",
       "gender        0.00\n",
       "seniorCitizen 0.11\n",
       "partner       0.12\n",
       "dependents    0.16\n",
       "tenure        0.78"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ar.bulk_iv(feats, iv).head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "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>iv</th>\n",
       "      <th>woe_min</th>\n",
       "      <th>woe_max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>gender</th>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>0.02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>seniorCitizen</th>\n",
       "      <td>0.11</td>\n",
       "      <td>-0.68</td>\n",
       "      <td>0.16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>partner</th>\n",
       "      <td>0.12</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>0.39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dependents</th>\n",
       "      <td>0.16</td>\n",
       "      <td>-0.23</td>\n",
       "      <td>0.68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>tenure</th>\n",
       "      <td>0.78</td>\n",
       "      <td>-1.27</td>\n",
       "      <td>1.87</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                iv  woe_min  woe_max\n",
       "gender        0.00    -0.02     0.02\n",
       "seniorCitizen 0.11    -0.68     0.16\n",
       "partner       0.12    -0.31     0.39\n",
       "dependents    0.16    -0.23     0.68\n",
       "tenure        0.78    -1.27     1.87"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ar.bulk_iv(feats, iv, woe_extremes=True).head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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>p-value</th>\n",
       "      <th>effect_size</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>gender</th>\n",
       "      <td>0.49</td>\n",
       "      <td>0.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>seniorCitizen</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>partner</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dependents</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>tenure</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.37</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               p-value  effect_size\n",
       "gender            0.49         0.01\n",
       "seniorCitizen     0.00         0.15\n",
       "partner           0.00         0.15\n",
       "dependents        0.00         0.16\n",
       "tenure            0.00         0.37"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ar.bulk_stats(feats, s).head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "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>iv</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>contract</th>\n",
       "      <td>1.24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>tenure</th>\n",
       "      <td>0.78</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>onlineSecurity</th>\n",
       "      <td>0.72</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>techSupport</th>\n",
       "      <td>0.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>internetService</th>\n",
       "      <td>0.62</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  iv\n",
       "contract        1.24\n",
       "tenure          0.78\n",
       "onlineSecurity  0.72\n",
       "techSupport     0.70\n",
       "internetService 0.62"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ar.analyze(feats, iv).head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "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>iv</th>\n",
       "      <th>p-value</th>\n",
       "      <th>effect_size</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>contract</th>\n",
       "      <td>1.24</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>tenure</th>\n",
       "      <td>0.78</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>onlineSecurity</th>\n",
       "      <td>0.72</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>techSupport</th>\n",
       "      <td>0.70</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>internetService</th>\n",
       "      <td>0.62</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.32</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  iv  p-value  effect_size\n",
       "contract        1.24     0.00         0.41\n",
       "tenure          0.78     0.00         0.37\n",
       "onlineSecurity  0.72     0.00         0.35\n",
       "techSupport     0.70     0.00         0.34\n",
       "internetService 0.62     0.00         0.32"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ar.analyze(feats, iv, s).head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "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>iv</th>\n",
       "      <th>p-value</th>\n",
       "      <th>effect_size</th>\n",
       "      <th>iv_interpretation</th>\n",
       "      <th>es_interpretation</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>contract</th>\n",
       "      <td>1.24</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.41</td>\n",
       "      <td>suspicious</td>\n",
       "      <td>strong</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>tenure</th>\n",
       "      <td>0.78</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.37</td>\n",
       "      <td>suspicious</td>\n",
       "      <td>medium</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>onlineSecurity</th>\n",
       "      <td>0.72</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.35</td>\n",
       "      <td>suspicious</td>\n",
       "      <td>medium</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>techSupport</th>\n",
       "      <td>0.70</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.34</td>\n",
       "      <td>suspicious</td>\n",
       "      <td>medium</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>internetService</th>\n",
       "      <td>0.62</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.32</td>\n",
       "      <td>suspicious</td>\n",
       "      <td>medium</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  iv  p-value  effect_size iv_interpretation es_interpretation\n",
       "contract        1.24     0.00         0.41        suspicious            strong\n",
       "tenure          0.78     0.00         0.37        suspicious            medium\n",
       "onlineSecurity  0.72     0.00         0.35        suspicious            medium\n",
       "techSupport     0.70     0.00         0.34        suspicious            medium\n",
       "internetService 0.62     0.00         0.32        suspicious            medium"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ar.analyze(feats, iv, s, interpretation=True).head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Visualizations of Information Value\n",
    "\n",
    "**IV()**\n",
    "\n",
    "Attributes and methods:\n",
    "- draw_iv(features_list, iv) - draws a chart with visualization of IV of features from the list\n",
    "- draw extremes(features_list, iv) - draws a chart with visualization of min and max WOE of features from the list\n",
    "- draw multiplot(features_list, iv) - draws a multi-chart with visualization of WOE of features from the list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ar.draw_iv(feats, iv)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ar.draw_woe_extremes(feats, iv)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ar.draw_woe_multiplot(feats[:6], iv)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Grouping continuous values into bins\n",
    "\n",
    "Values in continuous features are grouped according to the following rules:\n",
    "- Each bin should have at least 5% of the observations.\n",
    "- Each bin should be non-zero for both non-events and events.\n",
    "- The WOE should be monotonic, i.e. either growing or decreasing with the groupings.\n",
    "- Missing values are binned separately."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Reference\n",
    "\n",
    "1. https://stats.stackexchange.com/questions/93170/why-do-we-calculate-information-value\n",
    "2. https://www.listendata.com/2015/03/weight-of-evidence-woe-and-information.html#What-is-Information-Value-IV-\n",
    "3. https://towardsdatascience.com/attribute-relevance-analysis-in-python-iv-and-woe-b5651443fc04\n",
    "4. https://www.kaggle.com/pavansanagapati/weight-of-evidence-woe-information-value-iv\n",
    "5. https://medium.com/@sundarstyles89/weight-of-evidence-and-information-value-using-python-6f05072e83eb\n",
    "6. https://multithreaded.stitchfix.com/blog/2015/08/13/weight-of-evidence/"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "master",
   "language": "python",
   "name": "master"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}