Notebook2-ML_model_predictions.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# **Response to mailings - Part 2 notebook**\n",
    "In this notebook we evaluate several machine learning models and identify the best feature to predict mailings response "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Table of contents"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. [Overview](#Overview)\n",
    "2. [Data Processing](#Data_Understanding)\n",
    "3. [Data Modelling](#Data_Modelling)\n",
    "3. [Model Evaluation](#Model_Evaluation)\n",
    "4. [Conclusion](#Conclusion)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1. Overview"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this notebook I will try and answer the following questions that could possible help the business to improve mailing campaigns:\n",
    "* Can the response to mailings be predicted?\n",
    "* Which features could be most important to the response?\n",
    "* Can we identify an ideal mailing target individual?<br>\n",
    "\n",
    "I will be using data modeling and feature engineering to prepare the data, where I will then evaluate different types of prediction models and pick the best one that performs will with the data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id='Data_Understanding'></a>\n",
    "# 2. Data processing: \n",
    "- features engineering\n",
    "- encoding categorical features\n",
    "- scaling numerical features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Importing Libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# base\n",
    "import os\n",
    "import sys\n",
    "import time\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "import pickle\n",
    "%matplotlib inline\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "# models\n",
    "import sklearn.ensemble\n",
    "from sklearn.dummy import DummyClassifier\n",
    "from sklearn.naive_bayes import BernoulliNB\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.linear_model import Perceptron\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.svm import SVC\n",
    "import xgboost as xgb\n",
    "\n",
    "\n",
    "# evaluation\n",
    "\n",
    "from sklearn import feature_selection\n",
    "from sklearn import model_selection\n",
    "\n",
    "from sklearn.metrics import classification_report, accuracy_score, f1_score, precision_score, recall_score\n",
    "from scipy.stats import pearsonr\n",
    "from sklearn.model_selection import learning_curve, ShuffleSplit, StratifiedKFold\n",
    "\n",
    "from sklearn.metrics import accuracy_score, f1_score, precision_score, recall_score, classification_report, confusion_matrix\n",
    "\n",
    "from sklearn.model_selection import train_test_split, cross_val_score, RandomizedSearchCV\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Loading Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Loading csv and pickle files into dataframe or list\n",
    "df_mailings = pd.read_csv('data/population.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# eliminate useless features\n",
    "df_mailings = df_mailings.drop(['Unnamed: 0', 'label'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# group zip codes by bins of 10000 \n",
    "df_mailings['zipcode_group'] = df_mailings['zip code']//10000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>name</th>\n",
       "      <th>age</th>\n",
       "      <th>lifestyle</th>\n",
       "      <th>zip code</th>\n",
       "      <th>family status</th>\n",
       "      <th>car</th>\n",
       "      <th>sports</th>\n",
       "      <th>earnings</th>\n",
       "      <th>living area</th>\n",
       "      <th>response</th>\n",
       "      <th>zipcode_group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>VnSEFOuL</td>\n",
       "      <td>62</td>\n",
       "      <td>cozily</td>\n",
       "      <td>50168</td>\n",
       "      <td>married</td>\n",
       "      <td>practical</td>\n",
       "      <td>athletics</td>\n",
       "      <td>102526</td>\n",
       "      <td>urban</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8Tv0hcce</td>\n",
       "      <td>34</td>\n",
       "      <td>active</td>\n",
       "      <td>66479</td>\n",
       "      <td>married</td>\n",
       "      <td>expensive</td>\n",
       "      <td>soccer</td>\n",
       "      <td>33006</td>\n",
       "      <td>urban</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Zny9ysbk</td>\n",
       "      <td>69</td>\n",
       "      <td>healthy</td>\n",
       "      <td>16592</td>\n",
       "      <td>single</td>\n",
       "      <td>expensive</td>\n",
       "      <td>badminton</td>\n",
       "      <td>118760</td>\n",
       "      <td>urban</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       name  age lifestyle  zip code family status        car     sports  \\\n",
       "0  VnSEFOuL   62    cozily     50168       married  practical  athletics   \n",
       "1  8Tv0hcce   34    active     66479       married  expensive     soccer   \n",
       "2  Zny9ysbk   69   healthy     16592        single  expensive  badminton   \n",
       "\n",
       "   earnings living area  response  zipcode_group  \n",
       "0    102526       urban         0              5  \n",
       "1     33006       urban         0              6  \n",
       "2    118760       urban         1              1  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Displaying all columns\n",
    "pd.set_option('display.max_columns', None)\n",
    "# show listings data sample\n",
    "df_mailings.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.1 Binary encoding:\n",
    "\n",
    "- lives in New York state?\n",
    "- married?\n",
    "- owns an expensive car?\n",
    "- urban citizen?\n",
    "\n",
    "Yes = 1\n",
    "No = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "binary_features = ['family status', 'car', 'living area']\n",
    "positive_values = ['married', 'expensive', 'urban']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_binary_feature(df, variables, positives):\n",
    "    \n",
    "    \"\"\"\n",
    "    Creates one binary encoding for the specified column\n",
    "    using the unique items specified\n",
    "    \n",
    "    Parameters:\n",
    "    \n",
    "    Dataframe, variables(list of strings), unique_items(list)\n",
    "    \n",
    "    Returns:\n",
    "    \n",
    "    Dataframe\n",
    "    \n",
    "    \"\"\"\n",
    "    for variable in variables: \n",
    "       \n",
    "        item_lists = df[variable]\n",
    "\n",
    "        # Loop through all the labels\n",
    "        for label in positives:\n",
    "\n",
    "            # Creating new column for each label\n",
    "            df[variable+'_n'] = item_lists.apply(lambda x: 1 if label in x else 0)\n",
    "\n",
    "        # Removing old column\n",
    "        del df[variable]\n",
    "            \n",
    "    # Return the results as a dataframe\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "make_binary_feature(df_mailings, binary_features, positive_values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_mailings.rename(columns = {'family status_n' : 'married', 'car_n' : 'expensive car owner', 'living area_n':'urban citizen'}, inplace = True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.2 One hot encoding:\n",
    "\n",
    "- sport\n",
    "- lyfestyle\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "categorical_features = ['sports', 'lifestyle','zipcode_group']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def onehot_df(df, variables):\n",
    "    \n",
    "    \"\"\"\n",
    "    Creates one hot encoding for the specified columns\n",
    "    using the unique items specified\n",
    "    \n",
    "    Parameters:\n",
    "    \n",
    "    Dataframe, variables(list of strings)\n",
    "    \n",
    "    Returns:\n",
    "    \n",
    "    Dataframe\n",
    "    \n",
    "    \"\"\"\n",
    "    \n",
    "    for variable in variables: \n",
    "        # use pd.concat to join the new columns with your original dataframe\n",
    "        df = pd.concat([df,pd.get_dummies(df[variable], prefix=variable)],axis=1)\n",
    "\n",
    "# now drop the original variable column (you don't need it anymore)\n",
    "        # Removing old column\n",
    "        del df[variable]        \n",
    "        # Return the results as a dataframe\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_encoded = onehot_df(df_mailings, categorical_features)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.3 Scaling numerical features:\n",
    "\n",
    "- age\n",
    "- earnings\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "min_max_scaler = MinMaxScaler()\n",
    "df_encoded[[\"age\", \"earnings\"]] = min_max_scaler.fit_transform(df_encoded[[\"age\", \"earnings\"]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Dropping unwanted columns\n",
    "df_encoded = df_encoded.drop(['name', 'zip code'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>age</th>\n",
       "      <th>earnings</th>\n",
       "      <th>response</th>\n",
       "      <th>married</th>\n",
       "      <th>expensive car owner</th>\n",
       "      <th>urban citizen</th>\n",
       "      <th>sports_athletics</th>\n",
       "      <th>sports_badminton</th>\n",
       "      <th>sports_soccer</th>\n",
       "      <th>sports_unknown</th>\n",
       "      <th>lifestyle_active</th>\n",
       "      <th>lifestyle_cozily</th>\n",
       "      <th>lifestyle_healthy</th>\n",
       "      <th>zipcode_group_1</th>\n",
       "      <th>zipcode_group_2</th>\n",
       "      <th>zipcode_group_3</th>\n",
       "      <th>zipcode_group_4</th>\n",
       "      <th>zipcode_group_5</th>\n",
       "      <th>zipcode_group_6</th>\n",
       "      <th>zipcode_group_7</th>\n",
       "      <th>zipcode_group_8</th>\n",
       "      <th>zipcode_group_9</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.870370</td>\n",
       "      <td>0.634921</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.351852</td>\n",
       "      <td>0.100024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.759827</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.857304</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.944444</td>\n",
       "      <td>0.584732</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.111111</td>\n",
       "      <td>0.487578</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.129630</td>\n",
       "      <td>0.229439</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.814815</td>\n",
       "      <td>0.162808</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.879702</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.296296</td>\n",
       "      <td>0.682386</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        age  earnings  response  married  expensive car owner  urban citizen  \\\n",
       "0  0.870370  0.634921         0        0                    0              1   \n",
       "1  0.351852  0.100024         0        0                    0              1   \n",
       "2  1.000000  0.759827         1        0                    0              1   \n",
       "3  0.777778  0.857304         1        0                    0              1   \n",
       "4  0.944444  0.584732         1        0                    0              1   \n",
       "5  0.111111  0.487578         0        0                    0              1   \n",
       "6  0.129630  0.229439         0        0                    0              0   \n",
       "7  0.814815  0.162808         1        0                    0              1   \n",
       "8  0.888889  0.879702         1        0                    0              1   \n",
       "9  0.296296  0.682386         0        0                    0              0   \n",
       "\n",
       "   sports_athletics  sports_badminton  sports_soccer  sports_unknown  \\\n",
       "0                 1                 0              0               0   \n",
       "1                 0                 0              1               0   \n",
       "2                 0                 1              0               0   \n",
       "3                 0                 0              1               0   \n",
       "4                 0                 1              0               0   \n",
       "5                 0                 1              0               0   \n",
       "6                 0                 1              0               0   \n",
       "7                 0                 0              1               0   \n",
       "8                 0                 1              0               0   \n",
       "9                 1                 0              0               0   \n",
       "\n",
       "   lifestyle_active  lifestyle_cozily  lifestyle_healthy  zipcode_group_1  \\\n",
       "0                 0                 1                  0                0   \n",
       "1                 1                 0                  0                0   \n",
       "2                 0                 0                  1                1   \n",
       "3                 0                 1                  0                0   \n",
       "4                 0                 1                  0                0   \n",
       "5                 0                 0                  1                0   \n",
       "6                 0                 0                  1                0   \n",
       "7                 0                 1                  0                0   \n",
       "8                 1                 0                  0                0   \n",
       "9                 1                 0                  0                0   \n",
       "\n",
       "   zipcode_group_2  zipcode_group_3  zipcode_group_4  zipcode_group_5  \\\n",
       "0                0                0                0                1   \n",
       "1                0                0                0                0   \n",
       "2                0                0                0                0   \n",
       "3                0                0                0                1   \n",
       "4                0                1                0                0   \n",
       "5                0                0                0                0   \n",
       "6                0                0                0                0   \n",
       "7                0                0                1                0   \n",
       "8                0                0                0                1   \n",
       "9                0                1                0                0   \n",
       "\n",
       "   zipcode_group_6  zipcode_group_7  zipcode_group_8  zipcode_group_9  \n",
       "0                0                0                0                0  \n",
       "1                1                0                0                0  \n",
       "2                0                0                0                0  \n",
       "3                0                0                0                0  \n",
       "4                0                0                0                0  \n",
       "5                1                0                0                0  \n",
       "6                1                0                0                0  \n",
       "7                0                0                0                0  \n",
       "8                0                0                0                0  \n",
       "9                0                0                0                0  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_encoded.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id='Data_Modelling'></a>\n",
    "## 3. Data Modeling\n",
    "\n",
    "Now we are ready to train a model and predict the required solution. There are 60+ predictive modelling algorithms to choose from. We must understand the type of problem and solution requirement to narrow down to a select few models which we can evaluate. Our problem is a classification and regression problem. We want to identify relationship between output (response or not) with other variables or features (income, age, lifestyle...). We are also perfoming a category of machine learning which is called supervised learning as we are training our model with a given dataset. With these two criteria - Supervised Learning plus Classification and Regression, we can narrow down our choice of models to a few. These include:\n",
    "\n",
    "    - Logistic Regression\n",
    "    - KNN or k-Nearest Neighbors\n",
    "    - Support Vector Machines\n",
    "    - Naive Bayes Bernoulli classifier\n",
    "    - Random Forest\n",
    "    - XG-Boost\n",
    "    - Perceptron\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.1 Identify target and predictors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Extract features and labels\n",
    "X = df_encoded.drop(\"response\", axis=1) # drop labels for training set\n",
    "y = df_encoded[\"response\"].copy()\n",
    "\n",
    "# List of features for later use\n",
    "X_columns = list(X.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Train / Test split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Applying the train / test split \n",
    "X_train, X_test, y_train, y_test = train_test_split(X,y,test_size=0.2, random_state=42)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(8000, 21)\n",
      "(2000, 21)\n",
      "(8000,)\n",
      "(2000,)\n"
     ]
    }
   ],
   "source": [
    "# Confirming the rows and columns are the \n",
    "# same for train and test data\n",
    "print(X_train.shape)\n",
    "print(X_test.shape)\n",
    "print(y_train.shape)\n",
    "print(y_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id='Model_Evaluation'></a>\n",
    "## 4. Model evaluation\n",
    "\n",
    "Compare results from several machine learning classification models"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "When it comes to data modeling, the beginner’s question is always, \"what is the best machine learning algorithm?\" To this the beginner must learn, the No Free Lunch Theorem (NFLT) of Machine Learning. In short, NFLT states, there is no super algorithm, that works best in all situations, for all datasets. So the best approach is to try multiple MLAs, tune them, and compare them for your specific scenario.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Machine Learning Algorithm (MLA) Selection and Initialization\n",
    "Dummy_frequent = DummyClassifier(random_state=0, strategy=\"most_frequent\")# Dummy Classifier (baseline)\n",
    "Logistic_Regression = LogisticRegression()\n",
    "Bernoulli_Naive_Bayes = BernoulliNB() # naive Bayes algorithm for multinomially distributed data,\n",
    "Support_Vector_Classifier = SVC()\n",
    "Perceptron_Classifier = Perceptron()\n",
    "Random_Forest = RandomForestClassifier()\n",
    "KNN = KNeighborsClassifier(n_neighbors = 3)\n",
    "XG_boost = xgb.XGBClassifier(objective='binary:logistic',random_state=42)\n",
    "\n",
    "MLA = [Dummy_frequent,Logistic_Regression, Support_Vector_Classifier, Bernoulli_Naive_Bayes, Perceptron_Classifier, Random_Forest, KNN, XG_boost]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "def compare_classifiers(models_list):    \n",
    "    MLA_names = []\n",
    "    fit_times =[]\n",
    "    acc_scores = []\n",
    "    f1_scores = []\n",
    "    precision_scores = []\n",
    "    recall_scores = []\n",
    "       \n",
    "    for alg in MLA:\n",
    "        MLA_name = alg.__class__.__name__\n",
    "                \n",
    "        #initiate time count\n",
    "        global _start_time \n",
    "        start_time = time.time()\n",
    "     \n",
    "        alg.fit(X_train, y_train) # MLA fit\n",
    "                \n",
    "        y_pred = alg.predict(X_test) #MLA_predict\n",
    "        # estimate time elapsed\n",
    "        time_taken = round(((time.time() - start_time) / 60),3)\n",
    "        #model scores  \n",
    "        score = accuracy_score(y_test, y_pred)\n",
    "\n",
    "        F1 = f1_score(y_test, y_pred)\n",
    "\n",
    "        precision = precision_score(y_test, y_pred)\n",
    "\n",
    "        recall = recall_score(y_test, y_pred)\n",
    "\n",
    "        # add results to lists for a final comparison\n",
    "        MLA_names.append(MLA_name)\n",
    "        fit_times.append(time_taken)\n",
    "        acc_scores.append(round(score, 2))\n",
    "        precision_scores.append(round(precision, 2))\n",
    "        recall_scores.append(round(recall, 2))\n",
    "        f1_scores.append(round(F1, 2))\n",
    "    \n",
    "    #create table to compare MLA metrics\n",
    "    MLA_columns = ['MLA Name', 'Fit_time', 'Accuracy', 'Precision', 'Recall', 'F1']\n",
    "    zippedList =  list(zip(MLA_names, fit_times, acc_scores, f1_scores, precision_scores, recall_scores))\n",
    "    df = pd.DataFrame(zippedList, columns = MLA_columns).sort_values(by=['Recall'],ascending=False)\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/marco/anaconda3/lib/python3.8/site-packages/sklearn/metrics/_classification.py:1221: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n"
     ]
    },
    {
     "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>MLA Name</th>\n",
       "      <th>Fit_time</th>\n",
       "      <th>Accuracy</th>\n",
       "      <th>Precision</th>\n",
       "      <th>Recall</th>\n",
       "      <th>F1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>RandomForestClassifier</td>\n",
       "      <td>0.008</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.84</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>XGBClassifier</td>\n",
       "      <td>0.005</td>\n",
       "      <td>0.88</td>\n",
       "      <td>0.83</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.83</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>SVC</td>\n",
       "      <td>0.025</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.76</td>\n",
       "      <td>0.82</td>\n",
       "      <td>0.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>LogisticRegression</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.81</td>\n",
       "      <td>0.70</td>\n",
       "      <td>0.78</td>\n",
       "      <td>0.64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>KNeighborsClassifier</td>\n",
       "      <td>0.003</td>\n",
       "      <td>0.79</td>\n",
       "      <td>0.69</td>\n",
       "      <td>0.74</td>\n",
       "      <td>0.65</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>BernoulliNB</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.29</td>\n",
       "      <td>0.61</td>\n",
       "      <td>0.19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Perceptron</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.63</td>\n",
       "      <td>0.61</td>\n",
       "      <td>0.49</td>\n",
       "      <td>0.80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>DummyClassifier</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.64</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": [
       "                 MLA Name  Fit_time  Accuracy  Precision  Recall    F1\n",
       "5  RandomForestClassifier     0.008      0.89       0.84    0.84  0.84\n",
       "7           XGBClassifier     0.005      0.88       0.83    0.84  0.83\n",
       "2                     SVC     0.025      0.84       0.76    0.82  0.70\n",
       "1      LogisticRegression     0.001      0.81       0.70    0.78  0.64\n",
       "6    KNeighborsClassifier     0.003      0.79       0.69    0.74  0.65\n",
       "3             BernoulliNB     0.000      0.67       0.29    0.61  0.19\n",
       "4              Perceptron     0.000      0.63       0.61    0.49  0.80\n",
       "0         DummyClassifier     0.000      0.64       0.00    0.00  0.00"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_models = compare_classifiers(MLA)\n",
    "df_models"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Metrics importance**\n",
    "We have thousands of customers reached by our mailings. We wish they would all respond, but it is imposible, so we are interested in those with good chances to be a buyer (respond to mailings). We don't care to email a false positive, i.e. a guy that is not going to buy (so precision is not important) but for us is very important that all of them with high temperature are always in my selection, so they don't go without responding and hopefully buying. That means that my model needs to have a high recall (as little false negatives as possible), no matter if the precision is conversely affected."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4.1 Fine tuning the best model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "# define functions to measure time, to estimate our search efficiency\n",
    "def process_time_starts():\n",
    "    global _start_time \n",
    "    _start_time = time.time()\n",
    "\n",
    "def time_elapsed():\n",
    "    t_sec = round(time.time() - _start_time)\n",
    "    (t_min, t_sec) = divmod(t_sec,60)\n",
    "    (t_hour,t_min) = divmod(t_min,60) \n",
    "    print('The process took: {}hour:{}min:{}sec'.format(t_hour,t_min,t_sec))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next we set up our stratified folds and grid search parameters. I am using AUC as a scoring function, since we saw earlier that recall is quite important in this case."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 10 candidates, totalling 30 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done  30 out of  30 | elapsed:    1.1s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The process took: 0hour:0min:1sec\n"
     ]
    }
   ],
   "source": [
    "# Applying RandomizedSearchCV on the Random Forest Classifier algorithm\n",
    "\n",
    " # create training and validation dataset\n",
    "n_estimators = [100, 50, 20]\n",
    "max_depth = [5, 8, 20]\n",
    "min_samples_split = [2, 5, 10, 15, 100]\n",
    "min_samples_leaf = [1, 2, 5, 10] \n",
    "\n",
    "hyperF = dict(n_estimators = n_estimators, max_depth = max_depth,  \n",
    "              min_samples_split = min_samples_split, \n",
    "             min_samples_leaf = min_samples_leaf)\n",
    "\n",
    "\n",
    "# Here we go\n",
    "process_time_starts() # timing starts from this point \n",
    "\n",
    "randF = RandomizedSearchCV(Random_Forest, hyperF, cv = 3, verbose = 1, n_jobs = -1, scoring='roc_auc')\n",
    "randF.fit(X_train, y_train)\n",
    "\n",
    "time_elapsed() # timing ends here for the parameters search\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Best score:\n",
      "0.9191699216061924\n",
      "\n",
      " Best hyperparameters:\n",
      "{'n_estimators': 100, 'min_samples_split': 100, 'min_samples_leaf': 1, 'max_depth': 8}\n"
     ]
    }
   ],
   "source": [
    "print('\\n Best score:')\n",
    "print(randF.best_score_)\n",
    "print('\\n Best hyperparameters:')\n",
    "print(randF.best_params_)\n",
    "results = pd.DataFrame(randF.cv_results_)\n",
    "results.to_csv('data/RF-grid-search-results-01.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "fitting the data with the optimized random forest model took 5.0 milliseconds vs. 9.0 prior to optimization\n"
     ]
    }
   ],
   "source": [
    "#  best model!!\n",
    "rf1 = RandomForestClassifier( n_estimators = 100, max_depth = 8) #randF.best_estimator_\n",
    "\n",
    "#initiate time count\n",
    "global _start_time \n",
    "start_time = time.time()\n",
    "rf1.fit(X_train,y_train)\n",
    "time_taken = round(((time.time() - start_time) / 60),3)\n",
    "print(\"fitting the data with the optimized random forest model took\",time_taken*1000,'milliseconds vs. 9.0 prior to optimization')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.2 How do the best two models compare regarding efficiency? "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "# \"https://scikit-learn.org/stable/auto_examples/model_selection/plot_learning_curve.html\"\n",
    "\n",
    "def plot_learning_curve(estimator, title, X, y, axes=None, ylim=None, cv=None,\n",
    "                        n_jobs=None, train_sizes=np.linspace(.1, 1.0, 5)):\n",
    "    \n",
    "    \"\"\"\n",
    "    Generate 3 plots: the test and training learning curve, the training\n",
    "    samples vs fit times curve, the fit times vs score curve.\n",
    "\n",
    "    Parameters\n",
    "    ----------\n",
    "    estimator : estimator instance\n",
    "        An estimator instance implementing `fit` and `predict` methods which\n",
    "        will be cloned for each validation.\n",
    "\n",
    "    title : str\n",
    "        Title for the chart.\n",
    "\n",
    "    X : array-like of shape (n_samples, n_features)\n",
    "        Training vector, where ``n_samples`` is the number of samples and\n",
    "        ``n_features`` is the number of features.\n",
    "\n",
    "    y : array-like of shape (n_samples) or (n_samples, n_features)\n",
    "        Target relative to ``X`` for classification or regression;\n",
    "        None for unsupervised learning.\n",
    "\n",
    "    axes : array-like of shape (3,), default=None\n",
    "        Axes to use for plotting the curves.\n",
    "\n",
    "    ylim : tuple of shape (2,), default=None\n",
    "        Defines minimum and maximum y-values plotted, e.g. (ymin, ymax).\n",
    "\n",
    "    cv : int, cross-validation generator or an iterable, default=None\n",
    "        Determines the cross-validation splitting strategy.\n",
    "        Possible inputs for cv are:\n",
    "\n",
    "          - None, to use the default 5-fold cross-validation,\n",
    "          - integer, to specify the number of folds.\n",
    "          - :term:`CV splitter`,\n",
    "          - An iterable yielding (train, test) splits as arrays of indices.\n",
    "\n",
    "        For integer/None inputs, if ``y`` is binary or multiclass,\n",
    "        :class:`StratifiedKFold` used. If the estimator is not a classifier\n",
    "        or if ``y`` is neither binary nor multiclass, :class:`KFold` is used.\n",
    "\n",
    "        Refer :ref:`User Guide <cross_validation>` for the various\n",
    "        cross-validators that can be used here.\n",
    "\n",
    "    n_jobs : int or None, default=None\n",
    "        Number of jobs to run in parallel.\n",
    "        ``None`` means 1 unless in a :obj:`joblib.parallel_backend` context.\n",
    "        ``-1`` means using all processors. See :term:`Glossary <n_jobs>`\n",
    "        for more details.\n",
    "\n",
    "    train_sizes : array-like of shape (n_ticks,)\n",
    "        Relative or absolute numbers of training examples that will be used to\n",
    "        generate the learning curve. If the ``dtype`` is float, it is regarded\n",
    "        as a fraction of the maximum size of the training set (that is\n",
    "        determined by the selected validation method), i.e. it has to be within\n",
    "        (0, 1]. Otherwise it is interpreted as absolute sizes of the training\n",
    "        sets. Note that for classification the number of samples usually have\n",
    "        to be big enough to contain at least one sample from each class.\n",
    "        (default: np.linspace(0.1, 1.0, 5))\n",
    "    \"\"\"\n",
    "    if axes is None:\n",
    "        _, axes = plt.subplots(1, 3, figsize=(20, 5))\n",
    "\n",
    "    axes[0].set_title(title)\n",
    "    if ylim is not None:\n",
    "        axes[0].set_ylim(*ylim)\n",
    "    axes[0].set_xlabel(\"Training examples\")\n",
    "    axes[0].set_ylabel(\"Score\")\n",
    "\n",
    "    train_sizes, train_scores, test_scores, fit_times, _ = \\\n",
    "        learning_curve(estimator, X, y, cv=cv, n_jobs=n_jobs,\n",
    "                       train_sizes=train_sizes,\n",
    "                       return_times=True)\n",
    "    train_scores_mean = np.mean(train_scores, axis=1)\n",
    "    train_scores_std = np.std(train_scores, axis=1)\n",
    "    test_scores_mean = np.mean(test_scores, axis=1)\n",
    "    test_scores_std = np.std(test_scores, axis=1)\n",
    "    fit_times_mean = np.mean(fit_times, axis=1)\n",
    "    fit_times_std = np.std(fit_times, axis=1)\n",
    "\n",
    "    # Plot learning curve\n",
    "    axes[0].grid()\n",
    "    axes[0].fill_between(train_sizes, train_scores_mean - train_scores_std,\n",
    "                         train_scores_mean + train_scores_std, alpha=0.1,\n",
    "                         color=\"r\")\n",
    "    axes[0].fill_between(train_sizes, test_scores_mean - test_scores_std,\n",
    "                         test_scores_mean + test_scores_std, alpha=0.1,\n",
    "                         color=\"g\")\n",
    "    axes[0].plot(train_sizes, train_scores_mean, 'o-', color=\"r\",\n",
    "                 label=\"Training score\")\n",
    "    axes[0].plot(train_sizes, test_scores_mean, 'o-', color=\"g\",\n",
    "                 label=\"Cross-validation score\")\n",
    "    axes[0].legend(loc=\"best\")\n",
    "\n",
    "    # Plot n_samples vs fit_times\n",
    "    axes[1].grid()\n",
    "    axes[1].plot(train_sizes, fit_times_mean, 'o-')\n",
    "    axes[1].fill_between(train_sizes, fit_times_mean - fit_times_std,\n",
    "                         fit_times_mean + fit_times_std, alpha=0.1)\n",
    "    axes[1].set_xlabel(\"Training examples\")\n",
    "    axes[1].set_ylabel(\"fit_times\")\n",
    "    axes[1].set_title(\"Scalability of the model\")\n",
    "\n",
    "    # Plot fit_time vs score\n",
    "    axes[2].grid()\n",
    "    axes[2].plot(fit_times_mean, test_scores_mean, 'o-')\n",
    "    axes[2].fill_between(fit_times_mean, test_scores_mean - test_scores_std,\n",
    "                         test_scores_mean + test_scores_std, alpha=0.1)\n",
    "    axes[2].set_xlabel(\"fit_times\")\n",
    "    axes[2].set_ylabel(\"Score\")\n",
    "    axes[2].set_title(\"Performance of the model\")\n",
    "\n",
    "    return plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now let's fit and obtain scores"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Visualising Train and Test scores by the number of training points"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Random Forest Classifier:\n",
    "title = r\"Learning Curves for Random Forest Classifier\"\n",
    "estimator = rf1#RandomForestClassifier()\n",
    "plot_learning_curve(estimator, title,  X_train, y_train,cv=5, n_jobs=5)\n",
    "\n",
    "# Xtreme Gradient Boosting Classifier:\n",
    "title = r\"Learning Curves for XGB Classifier\"\n",
    "estimator= xgb.XGBClassifier(objective= 'binary:logistic')\n",
    "plot_learning_curve(estimator, title,  X_train, y_train,cv=5, n_jobs=5)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4.3 Feature Importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "accuracy: 0.8935 precision: 0.8676470588235294 recall: 0.8274894810659187 F1: 0.8470926058865758\n"
     ]
    }
   ],
   "source": [
    "# Final model and parameters\n",
    "\n",
    "y_pred = rf1.predict(X_test)\n",
    "\n",
    "acc = accuracy_score(y_test, y_pred)\n",
    "\n",
    "F1 = f1_score(y_test, y_pred)\n",
    "\n",
    "precision = precision_score(y_test, y_pred)\n",
    "\n",
    "recall = recall_score(y_test, y_pred)\n",
    "\n",
    "print('accuracy:',acc, 'precision:',precision ,'recall:', recall, 'F1:', F1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "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>Feature</th>\n",
       "      <th>% weight</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>age</td>\n",
       "      <td>64.279738</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>earnings</td>\n",
       "      <td>25.449822</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>zipcode_group_1</td>\n",
       "      <td>5.710067</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>urban citizen</td>\n",
       "      <td>0.632585</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>lifestyle_active</td>\n",
       "      <td>0.432359</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>lifestyle_cozily</td>\n",
       "      <td>0.402619</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>lifestyle_healthy</td>\n",
       "      <td>0.394437</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>sports_soccer</td>\n",
       "      <td>0.371999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>sports_badminton</td>\n",
       "      <td>0.321236</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>sports_athletics</td>\n",
       "      <td>0.313571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>sports_unknown</td>\n",
       "      <td>0.255339</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>zipcode_group_4</td>\n",
       "      <td>0.250885</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>zipcode_group_9</td>\n",
       "      <td>0.208548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>zipcode_group_3</td>\n",
       "      <td>0.206766</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>zipcode_group_7</td>\n",
       "      <td>0.186144</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>zipcode_group_2</td>\n",
       "      <td>0.156282</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>zipcode_group_6</td>\n",
       "      <td>0.154727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>zipcode_group_5</td>\n",
       "      <td>0.137772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>zipcode_group_8</td>\n",
       "      <td>0.135105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>expensive car owner</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>married</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Feature   % weight\n",
       "0                   age  64.279738\n",
       "1              earnings  25.449822\n",
       "12      zipcode_group_1   5.710067\n",
       "4         urban citizen   0.632585\n",
       "9      lifestyle_active   0.432359\n",
       "10     lifestyle_cozily   0.402619\n",
       "11    lifestyle_healthy   0.394437\n",
       "7         sports_soccer   0.371999\n",
       "6      sports_badminton   0.321236\n",
       "5      sports_athletics   0.313571\n",
       "8        sports_unknown   0.255339\n",
       "15      zipcode_group_4   0.250885\n",
       "20      zipcode_group_9   0.208548\n",
       "14      zipcode_group_3   0.206766\n",
       "18      zipcode_group_7   0.186144\n",
       "13      zipcode_group_2   0.156282\n",
       "17      zipcode_group_6   0.154727\n",
       "16      zipcode_group_5   0.137772\n",
       "19      zipcode_group_8   0.135105\n",
       "3   expensive car owner   0.000000\n",
       "2               married   0.000000"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# feature importance scores\n",
    "list_pc_importance = (rf1.feature_importances_)*100\n",
    "feat = X.columns.tolist()\n",
    "df_features = pd.DataFrame({'Feature':feat, '% weight': list_pc_importance}).sort_values(by=['% weight'], ascending = False )\n",
    "\n",
    "#print(xgb2.feature_importances_)\n",
    "df_features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot graph of feature importances for better visualization\n",
    "plt.figure(figsize= (6,6))\n",
    "plt.title(\"Feature importances for mailings response\")\n",
    "\n",
    "plt.bar(df_features[\"Feature\"][:5], df_features[\"% weight\"][:5], label = 'importance (%)')\n",
    "\n",
    "plt.legend()\n",
    "plt.ylabel('% weight',fontsize = 16)\n",
    "plt.xlabel('Responder feature',fontsize = 16)\n",
    "plt.xticks(rotation=90)\n",
    "plt.savefig('outcome.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Are categorical features correlated or anticorrelated?**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_binary = df_encoded[['zipcode_group_1','lifestyle_healthy','urban citizen','response' ]] #select features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "zipcode_group_1      0.182073\n",
       "lifestyle_healthy    0.005076\n",
       "urban citizen       -0.002543\n",
       "response             1.000000\n",
       "Name: response, dtype: float64"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## find correlation to response\n",
    "df_binary[df_binary.columns[0:]].corr()['response'][:]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id='Conclusion'></a>\n",
    "# 5. Conclusions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "With my final investigation I wanted to see if I could predict how likely someone is going to respond to mailings based on personal informations.\n",
    "\n",
    "**The prediction model that performed the best, was the \"Random Forest\" algorithm. I could predict response with a 89% accuracy on test data, and most important with 85% recall. This means I can have only very few false negatives and target the most likely responders.**\n",
    "\n",
    "\n",
    "**When I investigated the features of the results, I found that the conclusions are very similar to those from simple EDA:**\n",
    "-  Main predictors are age, income and whether the target lives or not in the group 1 of zip codes (10000-19999) corresponding to NY state in US.\n",
    "\n",
    "-  Based on the model it is more likely to obtain response from people living in rural area and with a healthy lifestyle. These are minor predictors though (around 1% importance)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}