Project2_Classification.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### <font color='red'> Project 2\n",
    "\n",
    "Project Description:\n",
    "- Use same datasets as Project 1.\n",
    "- Preprocess data: Explore data and apply data scaling.\n",
    "\n",
    "Regression Task:\n",
    "- Apply any two models with bagging and any two models with pasting.\n",
    "- Apply any two models with adaboost boosting\n",
    "- Apply one model with gradient boosting\n",
    "- Apply PCA on data and then apply all the models in project 1 again on data you get from PCA. Compare your results with results in project 2. You don't need to apply all the models twice. Just copy the result table from project 1, prepare similar table for all the models after PCA and compare both tables. Does PCA help in getting better results?\n",
    "- Apply deep learning models covered in class\n",
    "\n",
    "Classification Task:\n",
    "- Apply two voting classifiers - one with hard voting and one with soft voting\n",
    "- Apply any two models with bagging and any two models with pasting.\n",
    "- Apply any two models with adaboost boosting\n",
    "- Apply one model with gradient boosting\n",
    "- Apply PCA on data and then apply all the models in project 1 again on data you get from PCA. Compare your results with results in project 1. You don't need to apply all the models twice. Just copy the result table from project 1, prepare similar table for all the models after PCA and compare both tables. Does PCA help in getting better results?\n",
    "- Apply deep learning models covered in class\n",
    "\n",
    "Deliverables:\n",
    "- Use markdown to provide inline comments for this project.\n",
    "- Your outputs should be clearly executed in the notebook i.e. we should not need to rerun the code to obtain the outputs.\n",
    "- Visualization encouraged.\n",
    "- If you are submitting two different files, then please only one group member submit both the files. If you submit two files separately from different accounts, it will be submitted as two different attempts.\n",
    "- If you are submitting two different files, then please follow below naming convetion:\n",
    "    Project2_Regression_GroupXX_Firstname1_Firstname2.ipynb\n",
    "    Project2_Classification_GroupXX_Firstname1_Firstname2.ipynb\n",
    "- If you are submitting single file, then please follow below naming convetion:\n",
    "    Project2_Both_GroupXX_Firstname1_Firstname2.ipynb\n",
    "\n",
    "Questions regarding the project:\n",
    "- We have created a discussion board under Projects folder on e-learning. Create threads over there and post your queries related to project there.\n",
    "- We will also answer queries there. We will not be answering any project related queries through the mail."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "Group = 'Group 3'\n",
    "Name = 'My Pham (Classification), Monica Kulkarnk (Regression)'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Sources of Datasets"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. Regression: Bike Sharing Data ('bikesharing.csv')\n",
    "\n",
    "Source: https://code.datasciencedojo.com/datasciencedojo/datasets/tree/master/Bike%20Sharing\n",
    "\n",
    "Introduction: \n",
    "\n",
    "This data set, consisting of 17,379 rows and 17 columns, contains the hourly and daily count of rental bikes between years 2011 and 2012 in Capital bike-share system with the corresponding weather and seasonal information. Bike-sharing rental process is highly correlated to the environmental and seasonal settings.\n",
    "\n",
    "2. Classification: Heart Disease Prediction ('framingham_heart_disease.csv')\n",
    "\n",
    "Source: https://www.kaggle.com/dileep070/heart-disease-prediction-using-logistic-regression\n",
    "\n",
    "Introduction: \n",
    "\n",
    "World Health Organization has estimated 12 million deaths occur worldwide, every year due to Heart diseases. Half the deaths in the United States and other developed countries are due to cardio vascular diseases. The early prognosis of cardiovascular diseases can aid in making decisions on lifestyle changes in high risk patients and in turn reduce the complications. This research intends to pinpoint the most relevant/risk factors of heart disease as well as predict the overall risk using logistic regression Data Preparation. The dataset is publically available on the Kaggle website, and it is from an ongoing cardiovascular study on residents of the town of Framingham, Massachusetts. The classification goal is to predict whether the patient has 10-year risk of future coronary heart disease (CHD). The dataset provides the patients’ information. It includes over 4,000 records and 15 attributes. Variables Each attribute is a potential risk factor. There are both demographic, behavioral and medical risk factors.\n",
    "\n",
    "Column description:\n",
    "- male: 0 = Female; 1 = Male (boolean)\n",
    "- age: age of patient at exam time (integer)\n",
    "- education: 1 = Some High School; 2 = High School or GED; 3 = Some College or Vocational School; 4 = college (ordinal)\n",
    "- currentSmoker: 0 = nonsmoker; 1 = smoker (boolean)\n",
    "- cigsPerDay: number of cigarettes smoked per day (estimated average) (float)\n",
    "- BPMeds: 0 = Not on Blood Pressure medications; 1 = Is on Blood Pressure medications (boolean)\n",
    "- prevalentStroke: 0 = previously had no stroke; 1 = previously had a stroke (boolean)\n",
    "- prevalentHyp: 0 = patient was not hypertensive; 1 = patient was hypertensive (boolean)\n",
    "- diabetes: 0 = No; 1 = Yes (boolean)\n",
    "- totChol: total cholesterol level (mg/dL) (float)\n",
    "- sysBP: systolic blood pressure (mmHg) (float)\n",
    "- diaBP: diastolic blood pressure (mmHg) (float)\n",
    "- BMI: Body Mass Index = Weight (kg)/Height(meter-squared) (float)\n",
    "- heartRate: Beats/Min (Ventricular) (float)\n",
    "- glucose: glucose level (mg/dL) (float)\n",
    "- TenYearCHD: 10 year risk of coronary heart disease CHD (0 = No; 1 = Yes) (boolean)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Part 2: Classification Task:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data Preprocess"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. Exploratory Data Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Surpress warnings and import libraries\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "import numpy as np \n",
    "import pandas as pd \n",
    "import matplotlib.pyplot as plt \n",
    "import seaborn as sns\n",
    "from sklearn.metrics import confusion_matrix, precision_score, recall_score, accuracy_score, f1_score\n",
    "from sklearn import model_selection\n",
    "import matplotlib.mlab as mlab\n",
    "%matplotlib inline "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "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>male</th>\n",
       "      <th>age</th>\n",
       "      <th>education</th>\n",
       "      <th>currentSmoker</th>\n",
       "      <th>cigsPerDay</th>\n",
       "      <th>BPMeds</th>\n",
       "      <th>prevalentStroke</th>\n",
       "      <th>prevalentHyp</th>\n",
       "      <th>diabetes</th>\n",
       "      <th>totChol</th>\n",
       "      <th>sysBP</th>\n",
       "      <th>diaBP</th>\n",
       "      <th>BMI</th>\n",
       "      <th>heartRate</th>\n",
       "      <th>glucose</th>\n",
       "      <th>TenYearCHD</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>39</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>195.0</td>\n",
       "      <td>106.0</td>\n",
       "      <td>70.0</td>\n",
       "      <td>26.97</td>\n",
       "      <td>80.0</td>\n",
       "      <td>77.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>46</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>250.0</td>\n",
       "      <td>121.0</td>\n",
       "      <td>81.0</td>\n",
       "      <td>28.73</td>\n",
       "      <td>95.0</td>\n",
       "      <td>76.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>48</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>20.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>245.0</td>\n",
       "      <td>127.5</td>\n",
       "      <td>80.0</td>\n",
       "      <td>25.34</td>\n",
       "      <td>75.0</td>\n",
       "      <td>70.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>61</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1</td>\n",
       "      <td>30.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>225.0</td>\n",
       "      <td>150.0</td>\n",
       "      <td>95.0</td>\n",
       "      <td>28.58</td>\n",
       "      <td>65.0</td>\n",
       "      <td>103.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>46</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1</td>\n",
       "      <td>23.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>285.0</td>\n",
       "      <td>130.0</td>\n",
       "      <td>84.0</td>\n",
       "      <td>23.10</td>\n",
       "      <td>85.0</td>\n",
       "      <td>85.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   male  age  education  currentSmoker  cigsPerDay  BPMeds  prevalentStroke  \\\n",
       "0     1   39        4.0              0         0.0     0.0                0   \n",
       "1     0   46        2.0              0         0.0     0.0                0   \n",
       "2     1   48        1.0              1        20.0     0.0                0   \n",
       "3     0   61        3.0              1        30.0     0.0                0   \n",
       "4     0   46        3.0              1        23.0     0.0                0   \n",
       "\n",
       "   prevalentHyp  diabetes  totChol  sysBP  diaBP    BMI  heartRate  glucose  \\\n",
       "0             0         0    195.0  106.0   70.0  26.97       80.0     77.0   \n",
       "1             0         0    250.0  121.0   81.0  28.73       95.0     76.0   \n",
       "2             0         0    245.0  127.5   80.0  25.34       75.0     70.0   \n",
       "3             1         0    225.0  150.0   95.0  28.58       65.0    103.0   \n",
       "4             0         0    285.0  130.0   84.0  23.10       85.0     85.0   \n",
       "\n",
       "   TenYearCHD  \n",
       "0           0  \n",
       "1           0  \n",
       "2           0  \n",
       "3           1  \n",
       "4           0  "
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Import dataset\n",
    "hd = pd.read_csv('framingham_heart_disease.csv')\n",
    "hd.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The size of the dataset is:  (4238, 16)\n"
     ]
    }
   ],
   "source": [
    "print(\"The size of the dataset is: \", hd.shape) # Number of rows and columns met the condition"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Column names are:  Index(['male', 'age', 'education', 'currentSmoker', 'cigsPerDay', 'BPMeds',\n",
      "       'prevalentStroke', 'prevalentHyp', 'diabetes', 'totChol', 'sysBP',\n",
      "       'diaBP', 'BMI', 'heartRate', 'glucose', 'TenYearCHD'],\n",
      "      dtype='object')\n"
     ]
    }
   ],
   "source": [
    "print(\"Column names are: \", hd.keys())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "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>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>education</th>\n",
       "      <th>currentSmoker</th>\n",
       "      <th>cigsPerDay</th>\n",
       "      <th>BPMeds</th>\n",
       "      <th>prevalentStroke</th>\n",
       "      <th>prevalentHyp</th>\n",
       "      <th>diabetes</th>\n",
       "      <th>totChol</th>\n",
       "      <th>sysBP</th>\n",
       "      <th>diaBP</th>\n",
       "      <th>BMI</th>\n",
       "      <th>heartRate</th>\n",
       "      <th>glucose</th>\n",
       "      <th>TenYearCHD</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>39</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>195.0</td>\n",
       "      <td>106.0</td>\n",
       "      <td>70.0</td>\n",
       "      <td>26.97</td>\n",
       "      <td>80.0</td>\n",
       "      <td>77.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>46</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>250.0</td>\n",
       "      <td>121.0</td>\n",
       "      <td>81.0</td>\n",
       "      <td>28.73</td>\n",
       "      <td>95.0</td>\n",
       "      <td>76.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>48</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>20.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>245.0</td>\n",
       "      <td>127.5</td>\n",
       "      <td>80.0</td>\n",
       "      <td>25.34</td>\n",
       "      <td>75.0</td>\n",
       "      <td>70.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>61</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1</td>\n",
       "      <td>30.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>225.0</td>\n",
       "      <td>150.0</td>\n",
       "      <td>95.0</td>\n",
       "      <td>28.58</td>\n",
       "      <td>65.0</td>\n",
       "      <td>103.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>46</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1</td>\n",
       "      <td>23.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>285.0</td>\n",
       "      <td>130.0</td>\n",
       "      <td>84.0</td>\n",
       "      <td>23.10</td>\n",
       "      <td>85.0</td>\n",
       "      <td>85.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   sex  age  education  currentSmoker  cigsPerDay  BPMeds  prevalentStroke  \\\n",
       "0    1   39        4.0              0         0.0     0.0                0   \n",
       "1    0   46        2.0              0         0.0     0.0                0   \n",
       "2    1   48        1.0              1        20.0     0.0                0   \n",
       "3    0   61        3.0              1        30.0     0.0                0   \n",
       "4    0   46        3.0              1        23.0     0.0                0   \n",
       "\n",
       "   prevalentHyp  diabetes  totChol  sysBP  diaBP    BMI  heartRate  glucose  \\\n",
       "0             0         0    195.0  106.0   70.0  26.97       80.0     77.0   \n",
       "1             0         0    250.0  121.0   81.0  28.73       95.0     76.0   \n",
       "2             0         0    245.0  127.5   80.0  25.34       75.0     70.0   \n",
       "3             1         0    225.0  150.0   95.0  28.58       65.0    103.0   \n",
       "4             0         0    285.0  130.0   84.0  23.10       85.0     85.0   \n",
       "\n",
       "   TenYearCHD  \n",
       "0           0  \n",
       "1           0  \n",
       "2           0  \n",
       "3           1  \n",
       "4           0  "
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Rename Male column to Sex\n",
    "hd.rename(columns={'male':'sex'}, inplace=True)\n",
    "hd.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Male column needs to be renamed to Sex so that Male = 1, Female = 0."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2. Missing data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "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>Total</th>\n",
       "      <th>Percent</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>glucose</th>\n",
       "      <td>388</td>\n",
       "      <td>9.155262</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>education</th>\n",
       "      <td>105</td>\n",
       "      <td>2.477584</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BPMeds</th>\n",
       "      <td>53</td>\n",
       "      <td>1.250590</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>totChol</th>\n",
       "      <td>50</td>\n",
       "      <td>1.179802</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cigsPerDay</th>\n",
       "      <td>29</td>\n",
       "      <td>0.684285</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BMI</th>\n",
       "      <td>19</td>\n",
       "      <td>0.448325</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>heartRate</th>\n",
       "      <td>1</td>\n",
       "      <td>0.023596</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TenYearCHD</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>diaBP</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sysBP</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>diabetes</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>prevalentHyp</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>prevalentStroke</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>currentSmoker</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>age</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sex</th>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Total   Percent\n",
       "glucose            388  9.155262\n",
       "education          105  2.477584\n",
       "BPMeds              53  1.250590\n",
       "totChol             50  1.179802\n",
       "cigsPerDay          29  0.684285\n",
       "BMI                 19  0.448325\n",
       "heartRate            1  0.023596\n",
       "TenYearCHD           0  0.000000\n",
       "diaBP                0  0.000000\n",
       "sysBP                0  0.000000\n",
       "diabetes             0  0.000000\n",
       "prevalentHyp         0  0.000000\n",
       "prevalentStroke      0  0.000000\n",
       "currentSmoker        0  0.000000\n",
       "age                  0  0.000000\n",
       "sex                  0  0.000000"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Count missing data in each column\n",
    "total = hd.isnull().sum().sort_values(ascending = False)\n",
    "percent = 100*total/len(hd)\n",
    "missingCol = pd.concat([total, percent], axis = 1, keys = ['Total', 'Percent'])\n",
    "missingCol"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of rows with missing data is:  582\n",
      "Percentage of rows with missing data compared to the entire dataset is:  14\n"
     ]
    }
   ],
   "source": [
    "# Count number of rows with missing data\n",
    "count=0\n",
    "for i in hd.isnull().sum(axis=1):\n",
    "    if i>0:\n",
    "        count=count+1\n",
    "print('Number of rows with missing data is: ', count)\n",
    "print('Percentage of rows with missing data compared to the entire dataset is: ',round((count/len(hd.index))*100))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "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>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>education</th>\n",
       "      <th>currentSmoker</th>\n",
       "      <th>cigsPerDay</th>\n",
       "      <th>BPMeds</th>\n",
       "      <th>prevalentStroke</th>\n",
       "      <th>prevalentHyp</th>\n",
       "      <th>diabetes</th>\n",
       "      <th>totChol</th>\n",
       "      <th>sysBP</th>\n",
       "      <th>diaBP</th>\n",
       "      <th>BMI</th>\n",
       "      <th>heartRate</th>\n",
       "      <th>glucose</th>\n",
       "      <th>TenYearCHD</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>39</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>195.0</td>\n",
       "      <td>106.0</td>\n",
       "      <td>70.0</td>\n",
       "      <td>26.97</td>\n",
       "      <td>80.0</td>\n",
       "      <td>77.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>46</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>250.0</td>\n",
       "      <td>121.0</td>\n",
       "      <td>81.0</td>\n",
       "      <td>28.73</td>\n",
       "      <td>95.0</td>\n",
       "      <td>76.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>48</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>20.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>245.0</td>\n",
       "      <td>127.5</td>\n",
       "      <td>80.0</td>\n",
       "      <td>25.34</td>\n",
       "      <td>75.0</td>\n",
       "      <td>70.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>61</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1</td>\n",
       "      <td>30.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>225.0</td>\n",
       "      <td>150.0</td>\n",
       "      <td>95.0</td>\n",
       "      <td>28.58</td>\n",
       "      <td>65.0</td>\n",
       "      <td>103.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>46</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1</td>\n",
       "      <td>23.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>285.0</td>\n",
       "      <td>130.0</td>\n",
       "      <td>84.0</td>\n",
       "      <td>23.10</td>\n",
       "      <td>85.0</td>\n",
       "      <td>85.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0</td>\n",
       "      <td>43</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>228.0</td>\n",
       "      <td>180.0</td>\n",
       "      <td>110.0</td>\n",
       "      <td>30.30</td>\n",
       "      <td>77.0</td>\n",
       "      <td>99.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0</td>\n",
       "      <td>63</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>205.0</td>\n",
       "      <td>138.0</td>\n",
       "      <td>71.0</td>\n",
       "      <td>33.11</td>\n",
       "      <td>60.0</td>\n",
       "      <td>85.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0</td>\n",
       "      <td>45</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>20.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>313.0</td>\n",
       "      <td>100.0</td>\n",
       "      <td>71.0</td>\n",
       "      <td>21.68</td>\n",
       "      <td>79.0</td>\n",
       "      <td>78.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1</td>\n",
       "      <td>52</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>260.0</td>\n",
       "      <td>141.5</td>\n",
       "      <td>89.0</td>\n",
       "      <td>26.36</td>\n",
       "      <td>76.0</td>\n",
       "      <td>79.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1</td>\n",
       "      <td>43</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>30.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>225.0</td>\n",
       "      <td>162.0</td>\n",
       "      <td>107.0</td>\n",
       "      <td>23.61</td>\n",
       "      <td>93.0</td>\n",
       "      <td>88.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   sex  age  education  currentSmoker  cigsPerDay  BPMeds  prevalentStroke  \\\n",
       "0    1   39        4.0              0         0.0     0.0                0   \n",
       "1    0   46        2.0              0         0.0     0.0                0   \n",
       "2    1   48        1.0              1        20.0     0.0                0   \n",
       "3    0   61        3.0              1        30.0     0.0                0   \n",
       "4    0   46        3.0              1        23.0     0.0                0   \n",
       "5    0   43        2.0              0         0.0     0.0                0   \n",
       "6    0   63        1.0              0         0.0     0.0                0   \n",
       "7    0   45        2.0              1        20.0     0.0                0   \n",
       "8    1   52        1.0              0         0.0     0.0                0   \n",
       "9    1   43        1.0              1        30.0     0.0                0   \n",
       "\n",
       "   prevalentHyp  diabetes  totChol  sysBP  diaBP    BMI  heartRate  glucose  \\\n",
       "0             0         0    195.0  106.0   70.0  26.97       80.0     77.0   \n",
       "1             0         0    250.0  121.0   81.0  28.73       95.0     76.0   \n",
       "2             0         0    245.0  127.5   80.0  25.34       75.0     70.0   \n",
       "3             1         0    225.0  150.0   95.0  28.58       65.0    103.0   \n",
       "4             0         0    285.0  130.0   84.0  23.10       85.0     85.0   \n",
       "5             1         0    228.0  180.0  110.0  30.30       77.0     99.0   \n",
       "6             0         0    205.0  138.0   71.0  33.11       60.0     85.0   \n",
       "7             0         0    313.0  100.0   71.0  21.68       79.0     78.0   \n",
       "8             1         0    260.0  141.5   89.0  26.36       76.0     79.0   \n",
       "9             1         0    225.0  162.0  107.0  23.61       93.0     88.0   \n",
       "\n",
       "   TenYearCHD  \n",
       "0           0  \n",
       "1           0  \n",
       "2           0  \n",
       "3           1  \n",
       "4           0  \n",
       "5           0  \n",
       "6           1  \n",
       "7           0  \n",
       "8           0  \n",
       "9           0  "
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Replace missing data with median of columns\n",
    "hd.fillna(hd.median(), inplace=True)\n",
    "hd.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Missing data may impact the results of the models so they can be replaced with the median of each column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "sex                0\n",
       "age                0\n",
       "education          0\n",
       "currentSmoker      0\n",
       "cigsPerDay         0\n",
       "BPMeds             0\n",
       "prevalentStroke    0\n",
       "prevalentHyp       0\n",
       "diabetes           0\n",
       "totChol            0\n",
       "sysBP              0\n",
       "diaBP              0\n",
       "BMI                0\n",
       "heartRate          0\n",
       "glucose            0\n",
       "TenYearCHD         0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check for missing data in the updated dataset\n",
    "hd.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "sex                  int64\n",
       "age                  int64\n",
       "education          float64\n",
       "currentSmoker        int64\n",
       "cigsPerDay         float64\n",
       "BPMeds             float64\n",
       "prevalentStroke      int64\n",
       "prevalentHyp         int64\n",
       "diabetes             int64\n",
       "totChol            float64\n",
       "sysBP              float64\n",
       "diaBP              float64\n",
       "BMI                float64\n",
       "heartRate          float64\n",
       "glucose            float64\n",
       "TenYearCHD           int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Data have been converted to numeric types\n",
    "hd.dtypes    # At least 2 categorical/ordinal columns condition is met"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "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>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>education</th>\n",
       "      <th>currentSmoker</th>\n",
       "      <th>cigsPerDay</th>\n",
       "      <th>BPMeds</th>\n",
       "      <th>prevalentStroke</th>\n",
       "      <th>prevalentHyp</th>\n",
       "      <th>diabetes</th>\n",
       "      <th>totChol</th>\n",
       "      <th>sysBP</th>\n",
       "      <th>diaBP</th>\n",
       "      <th>BMI</th>\n",
       "      <th>heartRate</th>\n",
       "      <th>glucose</th>\n",
       "      <th>TenYearCHD</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>4238.000000</td>\n",
       "      <td>4238.000000</td>\n",
       "      <td>4238.000000</td>\n",
       "      <td>4238.000000</td>\n",
       "      <td>4238.000000</td>\n",
       "      <td>4238.000000</td>\n",
       "      <td>4238.000000</td>\n",
       "      <td>4238.000000</td>\n",
       "      <td>4238.000000</td>\n",
       "      <td>4238.000000</td>\n",
       "      <td>4238.000000</td>\n",
       "      <td>4238.000000</td>\n",
       "      <td>4238.000000</td>\n",
       "      <td>4238.000000</td>\n",
       "      <td>4238.000000</td>\n",
       "      <td>4238.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.429212</td>\n",
       "      <td>49.584946</td>\n",
       "      <td>1.979471</td>\n",
       "      <td>0.494101</td>\n",
       "      <td>8.941482</td>\n",
       "      <td>0.029259</td>\n",
       "      <td>0.005899</td>\n",
       "      <td>0.310524</td>\n",
       "      <td>0.025720</td>\n",
       "      <td>236.689476</td>\n",
       "      <td>132.352407</td>\n",
       "      <td>82.893464</td>\n",
       "      <td>25.800205</td>\n",
       "      <td>75.878716</td>\n",
       "      <td>81.603587</td>\n",
       "      <td>0.151958</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.495022</td>\n",
       "      <td>8.572160</td>\n",
       "      <td>1.007081</td>\n",
       "      <td>0.500024</td>\n",
       "      <td>11.902399</td>\n",
       "      <td>0.168552</td>\n",
       "      <td>0.076587</td>\n",
       "      <td>0.462763</td>\n",
       "      <td>0.158316</td>\n",
       "      <td>44.327427</td>\n",
       "      <td>22.038097</td>\n",
       "      <td>11.910850</td>\n",
       "      <td>4.071041</td>\n",
       "      <td>12.025185</td>\n",
       "      <td>22.865246</td>\n",
       "      <td>0.359023</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>107.000000</td>\n",
       "      <td>83.500000</td>\n",
       "      <td>48.000000</td>\n",
       "      <td>15.540000</td>\n",
       "      <td>44.000000</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>42.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>206.000000</td>\n",
       "      <td>117.000000</td>\n",
       "      <td>75.000000</td>\n",
       "      <td>23.080000</td>\n",
       "      <td>68.000000</td>\n",
       "      <td>72.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>49.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>234.000000</td>\n",
       "      <td>128.000000</td>\n",
       "      <td>82.000000</td>\n",
       "      <td>25.400000</td>\n",
       "      <td>75.000000</td>\n",
       "      <td>78.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>56.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>262.000000</td>\n",
       "      <td>144.000000</td>\n",
       "      <td>89.875000</td>\n",
       "      <td>28.037500</td>\n",
       "      <td>83.000000</td>\n",
       "      <td>85.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>70.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>70.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>696.000000</td>\n",
       "      <td>295.000000</td>\n",
       "      <td>142.500000</td>\n",
       "      <td>56.800000</td>\n",
       "      <td>143.000000</td>\n",
       "      <td>394.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               sex          age    education  currentSmoker   cigsPerDay  \\\n",
       "count  4238.000000  4238.000000  4238.000000    4238.000000  4238.000000   \n",
       "mean      0.429212    49.584946     1.979471       0.494101     8.941482   \n",
       "std       0.495022     8.572160     1.007081       0.500024    11.902399   \n",
       "min       0.000000    32.000000     1.000000       0.000000     0.000000   \n",
       "25%       0.000000    42.000000     1.000000       0.000000     0.000000   \n",
       "50%       0.000000    49.000000     2.000000       0.000000     0.000000   \n",
       "75%       1.000000    56.000000     3.000000       1.000000    20.000000   \n",
       "max       1.000000    70.000000     4.000000       1.000000    70.000000   \n",
       "\n",
       "            BPMeds  prevalentStroke  prevalentHyp     diabetes      totChol  \\\n",
       "count  4238.000000      4238.000000   4238.000000  4238.000000  4238.000000   \n",
       "mean      0.029259         0.005899      0.310524     0.025720   236.689476   \n",
       "std       0.168552         0.076587      0.462763     0.158316    44.327427   \n",
       "min       0.000000         0.000000      0.000000     0.000000   107.000000   \n",
       "25%       0.000000         0.000000      0.000000     0.000000   206.000000   \n",
       "50%       0.000000         0.000000      0.000000     0.000000   234.000000   \n",
       "75%       0.000000         0.000000      1.000000     0.000000   262.000000   \n",
       "max       1.000000         1.000000      1.000000     1.000000   696.000000   \n",
       "\n",
       "             sysBP        diaBP          BMI    heartRate      glucose  \\\n",
       "count  4238.000000  4238.000000  4238.000000  4238.000000  4238.000000   \n",
       "mean    132.352407    82.893464    25.800205    75.878716    81.603587   \n",
       "std      22.038097    11.910850     4.071041    12.025185    22.865246   \n",
       "min      83.500000    48.000000    15.540000    44.000000    40.000000   \n",
       "25%     117.000000    75.000000    23.080000    68.000000    72.000000   \n",
       "50%     128.000000    82.000000    25.400000    75.000000    78.000000   \n",
       "75%     144.000000    89.875000    28.037500    83.000000    85.000000   \n",
       "max     295.000000   142.500000    56.800000   143.000000   394.000000   \n",
       "\n",
       "        TenYearCHD  \n",
       "count  4238.000000  \n",
       "mean      0.151958  \n",
       "std       0.359023  \n",
       "min       0.000000  \n",
       "25%       0.000000  \n",
       "50%       0.000000  \n",
       "75%       0.000000  \n",
       "max       1.000000  "
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Description of updated dataset\n",
    "hd.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 4238 entries, 0 to 4237\n",
      "Data columns (total 16 columns):\n",
      "sex                4238 non-null int64\n",
      "age                4238 non-null int64\n",
      "education          4238 non-null float64\n",
      "currentSmoker      4238 non-null int64\n",
      "cigsPerDay         4238 non-null float64\n",
      "BPMeds             4238 non-null float64\n",
      "prevalentStroke    4238 non-null int64\n",
      "prevalentHyp       4238 non-null int64\n",
      "diabetes           4238 non-null int64\n",
      "totChol            4238 non-null float64\n",
      "sysBP              4238 non-null float64\n",
      "diaBP              4238 non-null float64\n",
      "BMI                4238 non-null float64\n",
      "heartRate          4238 non-null float64\n",
      "glucose            4238 non-null float64\n",
      "TenYearCHD         4238 non-null int64\n",
      "dtypes: float64(9), int64(7)\n",
      "memory usage: 529.8 KB\n"
     ]
    }
   ],
   "source": [
    "hd.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "3. Data visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "    #T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col0 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col1 {\n",
       "            background-color:  #6a8bef;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col2 {\n",
       "            background-color:  #6b8df0;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col3 {\n",
       "            background-color:  #abc8fd;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col4 {\n",
       "            background-color:  #c9d7f0;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col5 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col6 {\n",
       "            background-color:  #4358cb;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col7 {\n",
       "            background-color:  #5977e3;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col8 {\n",
       "            background-color:  #4b64d5;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col9 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col10 {\n",
       "            background-color:  #5470de;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col11 {\n",
       "            background-color:  #6a8bef;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col12 {\n",
       "            background-color:  #80a3fa;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col13 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col14 {\n",
       "            background-color:  #4e68d8;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col15 {\n",
       "            background-color:  #6485ec;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col0 {\n",
       "            background-color:  #536edd;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col1 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col2 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col3 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col4 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col5 {\n",
       "            background-color:  #6e90f2;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col6 {\n",
       "            background-color:  #5572df;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col7 {\n",
       "            background-color:  #b7cff9;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col8 {\n",
       "            background-color:  #6687ed;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col9 {\n",
       "            background-color:  #a1c0ff;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col10 {\n",
       "            background-color:  #d3dbe7;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col11 {\n",
       "            background-color:  #98b9ff;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col12 {\n",
       "            background-color:  #90b2fe;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col13 {\n",
       "            background-color:  #5673e0;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col14 {\n",
       "            background-color:  #6f92f3;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col15 {\n",
       "            background-color:  #92b4fe;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col0 {\n",
       "            background-color:  #5f7fe8;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col1 {\n",
       "            background-color:  #465ecf;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col2 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col3 {\n",
       "            background-color:  #779af7;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col4 {\n",
       "            background-color:  #6f92f3;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col5 {\n",
       "            background-color:  #465ecf;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col6 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col7 {\n",
       "            background-color:  #4055c8;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col8 {\n",
       "            background-color:  #3c4ec2;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col9 {\n",
       "            background-color:  #485fd1;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col10 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col11 {\n",
       "            background-color:  #465ecf;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col12 {\n",
       "            background-color:  #4257c9;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col13 {\n",
       "            background-color:  #4b64d5;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col14 {\n",
       "            background-color:  #4055c8;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col15 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col0 {\n",
       "            background-color:  #98b9ff;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col1 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col2 {\n",
       "            background-color:  #6c8ff1;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col3 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col4 {\n",
       "            background-color:  #ee8468;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col5 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col6 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col7 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col8 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col9 {\n",
       "            background-color:  #4055c8;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col10 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col11 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col12 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col13 {\n",
       "            background-color:  #6e90f2;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col14 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col15 {\n",
       "            background-color:  #4f69d9;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col0 {\n",
       "            background-color:  #bcd2f7;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col1 {\n",
       "            background-color:  #3f53c6;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col2 {\n",
       "            background-color:  #688aef;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col3 {\n",
       "            background-color:  #ed8366;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col4 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col5 {\n",
       "            background-color:  #3c4ec2;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col6 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col7 {\n",
       "            background-color:  #445acc;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col8 {\n",
       "            background-color:  #3c4ec2;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col9 {\n",
       "            background-color:  #465ecf;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col10 {\n",
       "            background-color:  #455cce;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col11 {\n",
       "            background-color:  #485fd1;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col12 {\n",
       "            background-color:  #4e68d8;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col13 {\n",
       "            background-color:  #7093f3;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col14 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col15 {\n",
       "            background-color:  #5b7ae5;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col0 {\n",
       "            background-color:  #4b64d5;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col1 {\n",
       "            background-color:  #96b7ff;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col2 {\n",
       "            background-color:  #6384eb;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col3 {\n",
       "            background-color:  #6485ec;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col4 {\n",
       "            background-color:  #6180e9;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col5 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col6 {\n",
       "            background-color:  #688aef;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col7 {\n",
       "            background-color:  #a7c5fe;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col8 {\n",
       "            background-color:  #5673e0;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col9 {\n",
       "            background-color:  #6687ed;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col10 {\n",
       "            background-color:  #abc8fd;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col11 {\n",
       "            background-color:  #94b6ff;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col12 {\n",
       "            background-color:  #85a8fc;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col13 {\n",
       "            background-color:  #5f7fe8;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col14 {\n",
       "            background-color:  #5977e3;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col15 {\n",
       "            background-color:  #6384eb;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col0 {\n",
       "            background-color:  #5977e3;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col1 {\n",
       "            background-color:  #84a7fc;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col2 {\n",
       "            background-color:  #5d7ce6;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col3 {\n",
       "            background-color:  #6a8bef;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col4 {\n",
       "            background-color:  #6485ec;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col5 {\n",
       "            background-color:  #6c8ff1;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col6 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col7 {\n",
       "            background-color:  #6e90f2;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col8 {\n",
       "            background-color:  #4961d2;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col9 {\n",
       "            background-color:  #4e68d8;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col10 {\n",
       "            background-color:  #6f92f3;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col11 {\n",
       "            background-color:  #6687ed;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col12 {\n",
       "            background-color:  #6e90f2;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col13 {\n",
       "            background-color:  #5572df;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col14 {\n",
       "            background-color:  #506bda;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col15 {\n",
       "            background-color:  #5b7ae5;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col0 {\n",
       "            background-color:  #5b7ae5;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col1 {\n",
       "            background-color:  #c9d7f0;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col2 {\n",
       "            background-color:  #506bda;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col3 {\n",
       "            background-color:  #5673e0;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col4 {\n",
       "            background-color:  #5b7ae5;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col5 {\n",
       "            background-color:  #9dbdff;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col6 {\n",
       "            background-color:  #5b7ae5;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col7 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col8 {\n",
       "            background-color:  #5e7de7;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col9 {\n",
       "            background-color:  #81a4fb;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col10 {\n",
       "            background-color:  #f5a081;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col11 {\n",
       "            background-color:  #f7bca1;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col12 {\n",
       "            background-color:  #c0d4f5;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col13 {\n",
       "            background-color:  #88abfd;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col14 {\n",
       "            background-color:  #6485ec;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col15 {\n",
       "            background-color:  #82a6fb;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col0 {\n",
       "            background-color:  #5f7fe8;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col1 {\n",
       "            background-color:  #90b2fe;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col2 {\n",
       "            background-color:  #5b7ae5;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col3 {\n",
       "            background-color:  #6687ed;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col4 {\n",
       "            background-color:  #6384eb;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col5 {\n",
       "            background-color:  #5977e3;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col6 {\n",
       "            background-color:  #465ecf;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col7 {\n",
       "            background-color:  #6f92f3;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col8 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col9 {\n",
       "            background-color:  #5a78e4;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col10 {\n",
       "            background-color:  #80a3fa;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col11 {\n",
       "            background-color:  #6788ee;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col12 {\n",
       "            background-color:  #81a4fb;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col13 {\n",
       "            background-color:  #688aef;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col14 {\n",
       "            background-color:  #f5c4ac;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col15 {\n",
       "            background-color:  #6788ee;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col0 {\n",
       "            background-color:  #465ecf;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col1 {\n",
       "            background-color:  #bcd2f7;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col2 {\n",
       "            background-color:  #6180e9;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col3 {\n",
       "            background-color:  #6687ed;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col4 {\n",
       "            background-color:  #6687ed;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col5 {\n",
       "            background-color:  #6180e9;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col6 {\n",
       "            background-color:  #445acc;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col7 {\n",
       "            background-color:  #89acfd;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col8 {\n",
       "            background-color:  #536edd;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col9 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col10 {\n",
       "            background-color:  #9ebeff;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col11 {\n",
       "            background-color:  #8badfd;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col12 {\n",
       "            background-color:  #89acfd;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col13 {\n",
       "            background-color:  #7699f6;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col14 {\n",
       "            background-color:  #5875e1;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col15 {\n",
       "            background-color:  #6282ea;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col0 {\n",
       "            background-color:  #506bda;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col1 {\n",
       "            background-color:  #dddcdc;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col2 {\n",
       "            background-color:  #4358cb;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col3 {\n",
       "            background-color:  #4f69d9;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col4 {\n",
       "            background-color:  #5572df;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col5 {\n",
       "            background-color:  #9abbff;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col6 {\n",
       "            background-color:  #5572df;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col7 {\n",
       "            background-color:  #f6a385;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col8 {\n",
       "            background-color:  #6a8bef;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col9 {\n",
       "            background-color:  #90b2fe;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col10 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col11 {\n",
       "            background-color:  #ec8165;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col12 {\n",
       "            background-color:  #c6d6f1;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col13 {\n",
       "            background-color:  #93b5fe;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col14 {\n",
       "            background-color:  #7597f6;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col15 {\n",
       "            background-color:  #8fb1fe;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col0 {\n",
       "            background-color:  #6c8ff1;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col1 {\n",
       "            background-color:  #aec9fc;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col2 {\n",
       "            background-color:  #5572df;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col3 {\n",
       "            background-color:  #5572df;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col4 {\n",
       "            background-color:  #5e7de7;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col5 {\n",
       "            background-color:  #86a9fc;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col6 {\n",
       "            background-color:  #516ddb;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col7 {\n",
       "            background-color:  #f6bda2;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col8 {\n",
       "            background-color:  #5673e0;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col9 {\n",
       "            background-color:  #81a4fb;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col10 {\n",
       "            background-color:  #ec7f63;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col11 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col12 {\n",
       "            background-color:  #d4dbe6;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col13 {\n",
       "            background-color:  #93b5fe;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col14 {\n",
       "            background-color:  #5d7ce6;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col15 {\n",
       "            background-color:  #779af7;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col0 {\n",
       "            background-color:  #7396f5;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col1 {\n",
       "            background-color:  #9abbff;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col2 {\n",
       "            background-color:  #4257c9;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col3 {\n",
       "            background-color:  #455cce;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col4 {\n",
       "            background-color:  #5470de;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col5 {\n",
       "            background-color:  #6788ee;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col6 {\n",
       "            background-color:  #4b64d5;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col7 {\n",
       "            background-color:  #b5cdfa;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col8 {\n",
       "            background-color:  #6180e9;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col9 {\n",
       "            background-color:  #7295f4;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col10 {\n",
       "            background-color:  #c1d4f4;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col11 {\n",
       "            background-color:  #cbd8ee;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col12 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col13 {\n",
       "            background-color:  #6f92f3;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col14 {\n",
       "            background-color:  #6384eb;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col15 {\n",
       "            background-color:  #6180e9;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col0 {\n",
       "            background-color:  #3b4cc0;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col1 {\n",
       "            background-color:  #6f92f3;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col2 {\n",
       "            background-color:  #5875e1;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col3 {\n",
       "            background-color:  #85a8fc;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col4 {\n",
       "            background-color:  #84a7fc;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col5 {\n",
       "            background-color:  #4e68d8;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col6 {\n",
       "            background-color:  #3f53c6;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col7 {\n",
       "            background-color:  #85a8fc;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col8 {\n",
       "            background-color:  #5572df;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col9 {\n",
       "            background-color:  #6a8bef;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col10 {\n",
       "            background-color:  #96b7ff;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col11 {\n",
       "            background-color:  #90b2fe;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col12 {\n",
       "            background-color:  #7b9ff9;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col13 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col14 {\n",
       "            background-color:  #6485ec;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col15 {\n",
       "            background-color:  #506bda;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col0 {\n",
       "            background-color:  #5e7de7;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col1 {\n",
       "            background-color:  #94b6ff;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col2 {\n",
       "            background-color:  #5d7ce6;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col3 {\n",
       "            background-color:  #6384eb;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col4 {\n",
       "            background-color:  #5d7ce6;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col5 {\n",
       "            background-color:  #5875e1;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col6 {\n",
       "            background-color:  #4a63d3;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col7 {\n",
       "            background-color:  #7093f3;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col8 {\n",
       "            background-color:  #f4c5ad;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col9 {\n",
       "            background-color:  #5b7ae5;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col10 {\n",
       "            background-color:  #88abfd;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col11 {\n",
       "            background-color:  #6a8bef;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col12 {\n",
       "            background-color:  #80a3fa;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col13 {\n",
       "            background-color:  #7597f6;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col14 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col15 {\n",
       "            background-color:  #6f92f3;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col0 {\n",
       "            background-color:  #7699f6;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col1 {\n",
       "            background-color:  #b3cdfb;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col2 {\n",
       "            background-color:  #5875e1;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col3 {\n",
       "            background-color:  #799cf8;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col4 {\n",
       "            background-color:  #7ea1fa;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col5 {\n",
       "            background-color:  #6384eb;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col6 {\n",
       "            background-color:  #5673e0;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col7 {\n",
       "            background-color:  #8fb1fe;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col8 {\n",
       "            background-color:  #6485ec;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col9 {\n",
       "            background-color:  #6788ee;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col10 {\n",
       "            background-color:  #a1c0ff;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col11 {\n",
       "            background-color:  #85a8fc;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col12 {\n",
       "            background-color:  #7da0f9;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col13 {\n",
       "            background-color:  #6180e9;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col14 {\n",
       "            background-color:  #7093f3;\n",
       "            color:  #000000;\n",
       "        }    #T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col15 {\n",
       "            background-color:  #b40426;\n",
       "            color:  #f1f1f1;\n",
       "        }</style><table id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2\" ><thead>    <tr>        <th class=\"blank level0\" ></th>        <th class=\"col_heading level0 col0\" >sex</th>        <th class=\"col_heading level0 col1\" >age</th>        <th class=\"col_heading level0 col2\" >education</th>        <th class=\"col_heading level0 col3\" >currentSmoker</th>        <th class=\"col_heading level0 col4\" >cigsPerDay</th>        <th class=\"col_heading level0 col5\" >BPMeds</th>        <th class=\"col_heading level0 col6\" >prevalentStroke</th>        <th class=\"col_heading level0 col7\" >prevalentHyp</th>        <th class=\"col_heading level0 col8\" >diabetes</th>        <th class=\"col_heading level0 col9\" >totChol</th>        <th class=\"col_heading level0 col10\" >sysBP</th>        <th class=\"col_heading level0 col11\" >diaBP</th>        <th class=\"col_heading level0 col12\" >BMI</th>        <th class=\"col_heading level0 col13\" >heartRate</th>        <th class=\"col_heading level0 col14\" >glucose</th>        <th class=\"col_heading level0 col15\" >TenYearCHD</th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                        <th id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2level0_row0\" class=\"row_heading level0 row0\" >sex</th>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col0\" class=\"data row0 col0\" >1</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col1\" class=\"data row0 col1\" >-0.02898</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col2\" class=\"data row0 col2\" >0.01721</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col3\" class=\"data row0 col3\" >0.1976</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col4\" class=\"data row0 col4\" >0.3156</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col5\" class=\"data row0 col5\" >-0.05154</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col6\" class=\"data row0 col6\" >-0.004546</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col7\" class=\"data row0 col7\" >0.005313</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col8\" class=\"data row0 col8\" >0.01571</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col9\" class=\"data row0 col9\" >-0.06955</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col10\" class=\"data row0 col10\" >-0.03599</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col11\" class=\"data row0 col11\" >0.05793</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col12\" class=\"data row0 col12\" >0.08165</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col13\" class=\"data row0 col13\" >-0.1166</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col14\" class=\"data row0 col14\" >0.01007</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row0_col15\" class=\"data row0 col15\" >0.08843</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2level0_row1\" class=\"row_heading level0 row1\" >age</th>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col0\" class=\"data row1 col0\" >-0.02898</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col1\" class=\"data row1 col1\" >1</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col2\" class=\"data row1 col2\" >-0.1635</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col3\" class=\"data row1 col3\" >-0.2137</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col4\" class=\"data row1 col4\" >-0.1918</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col5\" class=\"data row1 col5\" >0.121</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col6\" class=\"data row1 col6\" >0.05765</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col7\" class=\"data row1 col7\" >0.3072</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col8\" class=\"data row1 col8\" >0.1013</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col9\" class=\"data row1 col9\" >0.2601</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col10\" class=\"data row1 col10\" >0.3943</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col11\" class=\"data row1 col11\" >0.2061</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col12\" class=\"data row1 col12\" >0.1352</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col13\" class=\"data row1 col13\" >-0.01285</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col14\" class=\"data row1 col14\" >0.1178</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row1_col15\" class=\"data row1 col15\" >0.2253</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2level0_row2\" class=\"row_heading level0 row2\" >education</th>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col0\" class=\"data row2 col0\" >0.01721</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col1\" class=\"data row2 col1\" >-0.1635</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col2\" class=\"data row2 col2\" >1</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col3\" class=\"data row2 col3\" >0.01827</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col4\" class=\"data row2 col4\" >0.007618</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col5\" class=\"data row2 col5\" >-0.01036</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col6\" class=\"data row2 col6\" >-0.03515</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col7\" class=\"data row2 col7\" >-0.08102</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col8\" class=\"data row2 col8\" >-0.03814</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col9\" class=\"data row2 col9\" >-0.02236</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col10\" class=\"data row2 col10\" >-0.1283</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col11\" class=\"data row2 col11\" >-0.06172</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col12\" class=\"data row2 col12\" >-0.1355</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col13\" class=\"data row2 col13\" >-0.05368</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col14\" class=\"data row2 col14\" >-0.03398</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row2_col15\" class=\"data row2 col15\" >-0.05338</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2level0_row3\" class=\"row_heading level0 row3\" >currentSmoker</th>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col0\" class=\"data row3 col0\" >0.1976</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col1\" class=\"data row3 col1\" >-0.2137</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col2\" class=\"data row3 col2\" >0.01827</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col3\" class=\"data row3 col3\" >1</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col4\" class=\"data row3 col4\" >0.7602</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col5\" class=\"data row3 col5\" >-0.04836</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col6\" class=\"data row3 col6\" >-0.03299</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col7\" class=\"data row3 col7\" >-0.1033</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col8\" class=\"data row3 col8\" >-0.0443</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col9\" class=\"data row3 col9\" >-0.04647</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col10\" class=\"data row3 col10\" >-0.1302</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col11\" class=\"data row3 col11\" >-0.1077</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col12\" class=\"data row3 col12\" >-0.1672</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col13\" class=\"data row3 col13\" >0.06233</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col14\" class=\"data row3 col14\" >-0.05518</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row3_col15\" class=\"data row3 col15\" >0.01946</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2level0_row4\" class=\"row_heading level0 row4\" >cigsPerDay</th>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col0\" class=\"data row4 col0\" >0.3156</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col1\" class=\"data row4 col1\" >-0.1918</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col2\" class=\"data row4 col2\" >0.007618</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col3\" class=\"data row4 col3\" >0.7602</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col4\" class=\"data row4 col4\" >1</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col5\" class=\"data row4 col5\" >-0.04467</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col6\" class=\"data row4 col6\" >-0.03224</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col7\" class=\"data row4 col7\" >-0.06505</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col8\" class=\"data row4 col8\" >-0.03615</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col9\" class=\"data row4 col9\" >-0.02698</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col10\" class=\"data row4 col10\" >-0.08838</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col11\" class=\"data row4 col11\" >-0.05669</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col12\" class=\"data row4 col12\" >-0.09221</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col13\" class=\"data row4 col13\" >0.07385</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col14\" class=\"data row4 col14\" >-0.05686</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row4_col15\" class=\"data row4 col15\" >0.05886</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2level0_row5\" class=\"row_heading level0 row5\" >BPMeds</th>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col0\" class=\"data row5 col0\" >-0.05154</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col1\" class=\"data row5 col1\" >0.121</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col2\" class=\"data row5 col2\" >-0.01036</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col3\" class=\"data row5 col3\" >-0.04836</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col4\" class=\"data row5 col4\" >-0.04467</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col5\" class=\"data row5 col5\" >1</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col6\" class=\"data row5 col6\" >0.1146</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col7\" class=\"data row5 col7\" >0.2587</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col8\" class=\"data row5 col8\" >0.05139</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col9\" class=\"data row5 col9\" >0.07858</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col10\" class=\"data row5 col10\" >0.2515</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col11\" class=\"data row5 col11\" >0.1924</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col12\" class=\"data row5 col12\" >0.09958</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col13\" class=\"data row5 col13\" >0.01514</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col14\" class=\"data row5 col14\" >0.04912</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row5_col15\" class=\"data row5 col15\" >0.08642</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2level0_row6\" class=\"row_heading level0 row6\" >prevalentStroke</th>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col0\" class=\"data row6 col0\" >-0.004546</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col1\" class=\"data row6 col1\" >0.05765</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col2\" class=\"data row6 col2\" >-0.03515</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col3\" class=\"data row6 col3\" >-0.03299</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col4\" class=\"data row6 col4\" >-0.03224</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col5\" class=\"data row6 col5\" >0.1146</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col6\" class=\"data row6 col6\" >1</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col7\" class=\"data row6 col7\" >0.07483</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col8\" class=\"data row6 col8\" >0.006949</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col9\" class=\"data row6 col9\" >0.0001226</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col10\" class=\"data row6 col10\" >0.05701</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col11\" class=\"data row6 col11\" >0.04519</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col12\" class=\"data row6 col12\" >0.02426</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col13\" class=\"data row6 col13\" >-0.01767</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col14\" class=\"data row6 col14\" >0.01872</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row6_col15\" class=\"data row6 col15\" >0.06181</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2level0_row7\" class=\"row_heading level0 row7\" >prevalentHyp</th>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col0\" class=\"data row7 col0\" >0.005313</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col1\" class=\"data row7 col1\" >0.3072</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col2\" class=\"data row7 col2\" >-0.08102</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col3\" class=\"data row7 col3\" >-0.1033</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col4\" class=\"data row7 col4\" >-0.06505</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col5\" class=\"data row7 col5\" >0.2587</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col6\" class=\"data row7 col6\" >0.07483</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col7\" class=\"data row7 col7\" >1</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col8\" class=\"data row7 col8\" >0.07781</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col9\" class=\"data row7 col9\" >0.1631</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col10\" class=\"data row7 col10\" >0.6968</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col11\" class=\"data row7 col11\" >0.6158</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col12\" class=\"data row7 col12\" >0.3005</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col13\" class=\"data row7 col13\" >0.1472</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col14\" class=\"data row7 col14\" >0.08357</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row7_col15\" class=\"data row7 col15\" >0.1776</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2level0_row8\" class=\"row_heading level0 row8\" >diabetes</th>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col0\" class=\"data row8 col0\" >0.01571</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col1\" class=\"data row8 col1\" >0.1013</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col2\" class=\"data row8 col2\" >-0.03814</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col3\" class=\"data row8 col3\" >-0.0443</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col4\" class=\"data row8 col4\" >-0.03615</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col5\" class=\"data row8 col5\" >0.05139</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col6\" class=\"data row8 col6\" >0.006949</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col7\" class=\"data row8 col7\" >0.07781</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col8\" class=\"data row8 col8\" >1</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col9\" class=\"data row8 col9\" >0.04012</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col10\" class=\"data row8 col10\" >0.1113</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col11\" class=\"data row8 col11\" >0.05033</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col12\" class=\"data row8 col12\" >0.08603</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col13\" class=\"data row8 col13\" >0.049</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col14\" class=\"data row8 col14\" >0.6065</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row8_col15\" class=\"data row8 col15\" >0.09732</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2level0_row9\" class=\"row_heading level0 row9\" >totChol</th>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col0\" class=\"data row9 col0\" >-0.06955</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col1\" class=\"data row9 col1\" >0.2601</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col2\" class=\"data row9 col2\" >-0.02236</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col3\" class=\"data row9 col3\" >-0.04647</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col4\" class=\"data row9 col4\" >-0.02698</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col5\" class=\"data row9 col5\" >0.07858</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col6\" class=\"data row9 col6\" >0.0001226</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col7\" class=\"data row9 col7\" >0.1631</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col8\" class=\"data row9 col8\" >0.04012</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col9\" class=\"data row9 col9\" >1</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col10\" class=\"data row9 col10\" >0.2075</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col11\" class=\"data row9 col11\" >0.1639</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col12\" class=\"data row9 col12\" >0.1148</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col13\" class=\"data row9 col13\" >0.09054</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col14\" class=\"data row9 col14\" >0.04556</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row9_col15\" class=\"data row9 col15\" >0.08157</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2level0_row10\" class=\"row_heading level0 row10\" >sysBP</th>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col0\" class=\"data row10 col0\" >-0.03599</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col1\" class=\"data row10 col1\" >0.3943</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col2\" class=\"data row10 col2\" >-0.1283</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col3\" class=\"data row10 col3\" >-0.1302</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col4\" class=\"data row10 col4\" >-0.08838</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col5\" class=\"data row10 col5\" >0.2515</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col6\" class=\"data row10 col6\" >0.05701</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col7\" class=\"data row10 col7\" >0.6968</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col8\" class=\"data row10 col8\" >0.1113</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col9\" class=\"data row10 col9\" >0.2075</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col10\" class=\"data row10 col10\" >1</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col11\" class=\"data row10 col11\" >0.784</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col12\" class=\"data row10 col12\" >0.3252</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col13\" class=\"data row10 col13\" >0.1821</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col14\" class=\"data row10 col14\" >0.1351</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row10_col15\" class=\"data row10 col15\" >0.2164</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2level0_row11\" class=\"row_heading level0 row11\" >diaBP</th>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col0\" class=\"data row11 col0\" >0.05793</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col1\" class=\"data row11 col1\" >0.2061</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col2\" class=\"data row11 col2\" >-0.06172</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col3\" class=\"data row11 col3\" >-0.1077</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col4\" class=\"data row11 col4\" >-0.05669</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col5\" class=\"data row11 col5\" >0.1924</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col6\" class=\"data row11 col6\" >0.04519</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col7\" class=\"data row11 col7\" >0.6158</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col8\" class=\"data row11 col8\" >0.05033</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col9\" class=\"data row11 col9\" >0.1639</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col10\" class=\"data row11 col10\" >0.784</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col11\" class=\"data row11 col11\" >1</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col12\" class=\"data row11 col12\" >0.3766</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col13\" class=\"data row11 col13\" >0.1813</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col14\" class=\"data row11 col14\" >0.05904</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row11_col15\" class=\"data row11 col15\" >0.1453</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2level0_row12\" class=\"row_heading level0 row12\" >BMI</th>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col0\" class=\"data row12 col0\" >0.08165</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col1\" class=\"data row12 col1\" >0.1352</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col2\" class=\"data row12 col2\" >-0.1355</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col3\" class=\"data row12 col3\" >-0.1672</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col4\" class=\"data row12 col4\" >-0.09221</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col5\" class=\"data row12 col5\" >0.09958</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col6\" class=\"data row12 col6\" >0.02426</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col7\" class=\"data row12 col7\" >0.3005</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col8\" class=\"data row12 col8\" >0.08603</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col9\" class=\"data row12 col9\" >0.1148</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col10\" class=\"data row12 col10\" >0.3252</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col11\" class=\"data row12 col11\" >0.3766</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col12\" class=\"data row12 col12\" >1</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col13\" class=\"data row12 col13\" >0.06752</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col14\" class=\"data row12 col14\" >0.08222</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row12_col15\" class=\"data row12 col15\" >0.07422</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2level0_row13\" class=\"row_heading level0 row13\" >heartRate</th>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col0\" class=\"data row13 col0\" >-0.1166</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col1\" class=\"data row13 col1\" >-0.01285</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col2\" class=\"data row13 col2\" >-0.05368</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col3\" class=\"data row13 col3\" >0.06233</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col4\" class=\"data row13 col4\" >0.07385</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col5\" class=\"data row13 col5\" >0.01514</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col6\" class=\"data row13 col6\" >-0.01767</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col7\" class=\"data row13 col7\" >0.1472</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col8\" class=\"data row13 col8\" >0.049</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col9\" class=\"data row13 col9\" >0.09054</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col10\" class=\"data row13 col10\" >0.1821</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col11\" class=\"data row13 col11\" >0.1813</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col12\" class=\"data row13 col12\" >0.06752</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col13\" class=\"data row13 col13\" >1</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col14\" class=\"data row13 col14\" >0.08732</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row13_col15\" class=\"data row13 col15\" >0.02286</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2level0_row14\" class=\"row_heading level0 row14\" >glucose</th>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col0\" class=\"data row14 col0\" >0.01007</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col1\" class=\"data row14 col1\" >0.1178</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col2\" class=\"data row14 col2\" >-0.03398</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col3\" class=\"data row14 col3\" >-0.05518</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col4\" class=\"data row14 col4\" >-0.05686</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col5\" class=\"data row14 col5\" >0.04912</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col6\" class=\"data row14 col6\" >0.01872</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col7\" class=\"data row14 col7\" >0.08357</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col8\" class=\"data row14 col8\" >0.6065</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col9\" class=\"data row14 col9\" >0.04556</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col10\" class=\"data row14 col10\" >0.1351</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col11\" class=\"data row14 col11\" >0.05904</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col12\" class=\"data row14 col12\" >0.08222</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col13\" class=\"data row14 col13\" >0.08732</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col14\" class=\"data row14 col14\" >1</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row14_col15\" class=\"data row14 col15\" >0.1213</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2level0_row15\" class=\"row_heading level0 row15\" >TenYearCHD</th>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col0\" class=\"data row15 col0\" >0.08843</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col1\" class=\"data row15 col1\" >0.2253</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col2\" class=\"data row15 col2\" >-0.05338</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col3\" class=\"data row15 col3\" >0.01946</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col4\" class=\"data row15 col4\" >0.05886</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col5\" class=\"data row15 col5\" >0.08642</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col6\" class=\"data row15 col6\" >0.06181</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col7\" class=\"data row15 col7\" >0.1776</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col8\" class=\"data row15 col8\" >0.09732</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col9\" class=\"data row15 col9\" >0.08157</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col10\" class=\"data row15 col10\" >0.2164</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col11\" class=\"data row15 col11\" >0.1453</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col12\" class=\"data row15 col12\" >0.07422</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col13\" class=\"data row15 col13\" >0.02286</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col14\" class=\"data row15 col14\" >0.1213</td>\n",
       "                        <td id=\"T_4a871e12_16f5_11ea_84e3_448500d700c2row15_col15\" class=\"data row15 col15\" >1</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x16ec47ceb70>"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Correlation matrix\n",
    "corr = hd.corr()\n",
    "corr.style.background_gradient(cmap='coolwarm').set_precision(4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['sex', 'age', 'education', 'currentSmoker', 'cigsPerDay', 'BPMeds', 'prevalentStroke', 'prevalentHyp', 'diabetes', 'totChol', 'sysBP', 'diaBP', 'BMI', 'heartRate', 'glucose']\n"
     ]
    }
   ],
   "source": [
    "# Get list of features\n",
    "features = hd.drop('TenYearCHD', 1).columns\n",
    "print(features.tolist())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1440 with 16 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Histograms for features\n",
    "def plot_hists(hd, features, rows, cols):\n",
    "    fig=plt.figure(figsize=(20,20))\n",
    "    for i, feature in enumerate(features):\n",
    "        ax=fig.add_subplot(rows,cols,i+1)\n",
    "        hd[feature].hist(bins=20,ax=ax,facecolor='blue')\n",
    "        ax.set_title('Distribution: '+ feature) \n",
    "    fig.tight_layout() \n",
    "    plt.show()\n",
    "\n",
    "plot_hists(hd,hd.columns,5,4)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As each feature has a different range of values, data scaling will be necessary."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "5. Data scaling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Separate dataset into X (input features) and y (output label) and convert dataset to a NumPy array\n",
    "hd.to_numpy()\n",
    "X = hd.drop('TenYearCHD',1).to_numpy()\n",
    "y = hd['TenYearCHD'].to_numpy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Min-Max Scaling is useful in this case as the features have hard boundaries and the technique produces values of range [0,1]."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Scale data\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.model_selection import train_test_split, GridSearchCV, cross_val_score\n",
    "\n",
    "X_train_org, X_test_org, y_train, y_test = train_test_split(X, y, random_state = 0)\n",
    "\n",
    "scaler = MinMaxScaler()\n",
    "X_train = scaler.fit_transform(X_train_org)\n",
    "X_test = scaler.transform(X_test_org)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Classification Models"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1. Voting Classifiers"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "    a. Logistic Regression with Hard Voting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Logistic Regression with Hard Voting:\n",
      "Train score: 0.8597\n",
      "Test score: 0.8434\n",
      "Accuracy score: 0.8434\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import VotingClassifier\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "log_clf = LogisticRegression(C=1, penalty='l1')\n",
    "log_clf.fit(X_train, y_train)\n",
    "\n",
    "voting_clf = VotingClassifier(estimators=[('lr', log_clf)], voting='hard')\n",
    "lr_voting = voting_clf.fit(X_train, y_train)\n",
    "y_pred_lr = lr_voting.predict(X_test)\n",
    "\n",
    "print('Logistic Regression with Hard Voting:')\n",
    "print('Train score: {:.4f}'.format(voting_clf.score(X_train, y_train)))\n",
    "print('Test score: {:.4f}'.format(voting_clf.score(X_test, y_test)))\n",
    "print('Accuracy score: {:.4f}'.format(accuracy_score(y_test, y_pred_lr)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "    b. KNN with Soft Voting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "KNN with Soft Voting:\n",
      "Train score: 0.8521\n",
      "Test score: 0.8396\n",
      "Accuracy score: 0.8396\n"
     ]
    }
   ],
   "source": [
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "\n",
    "knn_clf = KNeighborsClassifier(20)\n",
    "knn_clf.fit(X_train, y_train)\n",
    "\n",
    "voting_clf = VotingClassifier(estimators=[('knn', knn_clf)], voting='soft')\n",
    "knn_voting = voting_clf.fit(X_train, y_train)\n",
    "y_pred_knn = knn_voting.predict(X_test)\n",
    "\n",
    "print('KNN with Soft Voting:')\n",
    "print('Train score: {:.4f}'.format(voting_clf.score(X_train, y_train)))\n",
    "print('Test score: {:.4f}'.format(voting_clf.score(X_test, y_test)))\n",
    "print('Accuracy score: {:.4f}'.format(accuracy_score(y_test, y_pred_knn)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 2. Bagging"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "    a. Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best parameters: {'max_samples': 100, 'n_estimators': 200}\n",
      "Best cross-validation score: 0.8508\n",
      "Average cross-validation score: 0.8509\n"
     ]
    }
   ],
   "source": [
    "# GridSearch\n",
    "\n",
    "from sklearn.ensemble import BaggingClassifier\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "\n",
    "dt_clf = DecisionTreeClassifier(random_state=0, max_depth=1, max_leaf_nodes=2, min_samples_split=2)\n",
    "bag_dt = BaggingClassifier(dt_clf, bootstrap=True, random_state=0, oob_score=True)\n",
    "\n",
    "param = {'n_estimators':[200,500], 'max_samples':[100,200]}\n",
    "bag_dt_grid_search = GridSearchCV(BaggingClassifier(dt_clf), param, cv=5)\n",
    "bag_dt_grid_search.fit(X_train,y_train)\n",
    "print(\"Best parameters: {}\".format(bag_dt_grid_search.best_params_))\n",
    "print(\"Best cross-validation score: {:.4f}\".format(bag_dt_grid_search.best_score_))\n",
    "\n",
    "# Cross-validation\n",
    "bag_dt_scores = cross_val_score(bag_dt, X_train, y_train, cv=5)\n",
    "print(\"Average cross-validation score: {:.4f}\".format(bag_dt_scores.mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Bagging with Decision Tree:\n",
      "Train score: 0.8508\n",
      "Test score: 0.8396\n",
      "Accuracy score: 0.8396\n"
     ]
    }
   ],
   "source": [
    "# Best parameters\n",
    "\n",
    "bag_clf = BaggingClassifier(dt_clf, n_estimators=200, max_samples=100, bootstrap=True, random_state=0, oob_score=True)\n",
    "bag_clf.fit(X_train, y_train)\n",
    "y_pred_bag_dt = bag_clf.predict(X_test)\n",
    "\n",
    "print('Bagging with Decision Tree:')\n",
    "print('Train score: {:.4f}'.format(bag_clf.score(X_train, y_train)))\n",
    "print('Test score: {:.4f}'.format(bag_clf.score(X_test, y_test)))\n",
    "print('Accuracy score: {:.4f}'.format(accuracy_score(y_test, y_pred_bag_dt)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "    b. LinearSVC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# GridSearch\n",
    "\n",
    "from sklearn.svm import LinearSVC\n",
    "\n",
    "svc_clf = LinearSVC(C=10)\n",
    "bag_svc = BaggingClassifier(svc_clf, bootstrap=True, random_state=0, oob_score=True)\n",
    "\n",
    "param = {'n_estimators':[200,500], 'max_samples':[100,200]}\n",
    "bag_svc_grid_search = GridSearchCV(BaggingClassifier(svc_clf), param, cv=5)\n",
    "bag_svc_grid_search.fit(X_train,y_train)\n",
    "print(\"Best parameters: {}\".format(bag_svc_grid_search.best_params_))\n",
    "print(\"Best cross-validation score: {:.4f}\".format(bag_svc_grid_search.best_score_))\n",
    "\n",
    "# Cross-validation\n",
    "bag_svc_scores = cross_val_score(bag_svc, X_train, y_train, cv=5)\n",
    "print(\"Average cross-validation score: {:.4f}\".format(bag_svc_scores.mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Best parameters\n",
    "\n",
    "bag_clf = BaggingClassifier(svc_clf, n_estimators=200, max_samples=100, bootstrap=True, random_state=0, oob_score=True)\n",
    "bag_clf.fit(X_train, y_train)\n",
    "y_pred_bag_svc = bag_clf.predict(X_test)\n",
    "\n",
    "print('Bagging with LinearSVC:')\n",
    "print('Train score: {:.4f}'.format(bag_clf.score(X_train, y_train)))\n",
    "print('Test score: {:.4f}'.format(bag_clf.score(X_test, y_test)))\n",
    "print('Accuracy score: {:.4f}'.format(accuracy_score(y_test, y_pred_bag_svc)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3. Pasting"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "    a. Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# GridSearch\n",
    "\n",
    "from sklearn.ensemble import BaggingClassifier\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "\n",
    "dt_clf = DecisionTreeClassifier(random_state=0, max_depth=1, max_leaf_nodes=2, min_samples_split=2)\n",
    "bag_dt = BaggingClassifier(dt_clf, bootstrap=False, random_state=0)\n",
    "\n",
    "param = {'n_estimators':[200,500], 'max_samples':[100,200]}\n",
    "bag_dt_grid_search = GridSearchCV(BaggingClassifier(dt_clf), param, cv=5)\n",
    "bag_dt_grid_search.fit(X_train,y_train)\n",
    "print(\"Best parameters: {}\".format(bag_dt_grid_search.best_params_))\n",
    "print(\"Best cross-validation score: {:.4f}\".format(bag_dt_grid_search.best_score_))\n",
    "\n",
    "# Cross-validation\n",
    "bag_dt_scores = cross_val_score(bag_dt, X_train, y_train, cv=5)\n",
    "print(\"Average cross-validation score: {:.4f}\".format(bag_dt_scores.mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Best parameters\n",
    "\n",
    "bag_clf = BaggingClassifier(dt_clf, n_estimators=200, max_samples=100, bootstrap=False, random_state=0)\n",
    "bag_clf.fit(X_train, y_train)\n",
    "y_pred_bag_dt = bag_clf.predict(X_test)\n",
    "\n",
    "print('Pasting with Decision Tree:')\n",
    "print('Train score: {:.4f}'.format(bag_clf.score(X_train, y_train)))\n",
    "print('Test score: {:.4f}'.format(bag_clf.score(X_test, y_test)))\n",
    "print('Accuracy score: {:.4f}'.format(accuracy_score(y_test, y_pred_bag_dt)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "    b. LinearSVC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# GridSearch\n",
    "\n",
    "from sklearn.svm import LinearSVC\n",
    "\n",
    "svc_clf = LinearSVC(C=10)\n",
    "bag_svc = BaggingClassifier(svc_clf, bootstrap=False, random_state=0)\n",
    "\n",
    "param = {'n_estimators':[200,500], 'max_samples':[100,200]}\n",
    "bag_svc_grid_search = GridSearchCV(BaggingClassifier(svc_clf), param, cv=5)\n",
    "bag_svc_grid_search.fit(X_train,y_train)\n",
    "print(\"Best parameters: {}\".format(bag_svc_grid_search.best_params_))\n",
    "print(\"Best cross-validation score: {:.4f}\".format(bag_svc_grid_search.best_score_))\n",
    "\n",
    "# Cross-validation\n",
    "bag_svc_scores = cross_val_score(bag_svc, X_train, y_train, cv=5)\n",
    "print(\"Average cross-validation score: {:.4f}\".format(bag_svc_scores.mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Best parameters\n",
    "\n",
    "bag_clf = BaggingClassifier(svc_clf, n_estimators=200, max_samples=100, bootstrap=False, random_state=0)\n",
    "bag_clf.fit(X_train, y_train)\n",
    "y_pred_bag_svc = bag_clf.predict(X_test)\n",
    "\n",
    "print('Pasting with LinearSVC:')\n",
    "print('Train score: {:.4f}'.format(bag_clf.score(X_train, y_train)))\n",
    "print('Test score: {:.4f}'.format(bag_clf.score(X_test, y_test)))\n",
    "print('Accuracy score: {:.4f}'.format(accuracy_score(y_test, y_pred_bag_svc)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 4. AdaBoost"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "    a. Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# GridSearch\n",
    "\n",
    "from sklearn.ensemble import AdaBoostClassifier\n",
    "\n",
    "dt_clf = DecisionTreeClassifier(random_state=0, max_depth=1, max_leaf_nodes=2, min_samples_split=2)\n",
    "ada_dt = AdaBoostClassifier(dt_clf)\n",
    "\n",
    "param = {'base_estimator__criterion':[\"gini\", \"entropy\"],'base_estimator__splitter':[\"best\",\"random\"], 'n_estimators':[100,200], 'learning_rate':[0.8, 1.0]}\n",
    "ada_dt_grid_search = GridSearchCV(AdaBoostClassifier(dt_clf), param)\n",
    "ada_dt_grid_search.fit(X_train,y_train)\n",
    "print(\"Best parameters: {}\".format(ada_dt_grid_search.best_params_))\n",
    "print(\"Best cross-validation score: {:.4f}\".format(ada_dt_grid_search.best_score_))\n",
    "\n",
    "# Cross-validation\n",
    "ada_dt_scores = cross_val_score(ada_dt, X_train, y_train, cv=5)\n",
    "print(\"Average cross-validation score: {:.4f}\".format(ada_dt_scores.mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Best parameters\n",
    "\n",
    "ada_clf = AdaBoostClassifier(dt_clf, n_estimators=100, algorithm=\"SAMME.R\", learning_rate=1.0, random_state=0)\n",
    "ada_clf.fit(X_train, y_train)\n",
    "y_pred_ada_dt = ada_clf.predict(X_test)\n",
    "\n",
    "print('AdaBoost with Decision Tree:')\n",
    "print('Train score: {:.4f}'.format(ada_clf.score(X_train, y_train)))\n",
    "print('Test score: {:.4f}'.format(ada_clf.score(X_test, y_test)))\n",
    "print('Accuracy score: {:.4f}'.format(accuracy_score(y_test, y_pred_ada_dt)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "    b. LinearSVC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# GridSearch\n",
    "\n",
    "svc_clf = LinearSVC(C=10)\n",
    "ada_svc = AdaBoostClassifier(svc_clf, algorithm='SAMME')\n",
    "\n",
    "param = {'n_estimators':[100,200], 'learning_rate':[0.8, 1.0]}\n",
    "ada_svc_grid_search = GridSearchCV(AdaBoostClassifier(svc_clf, algorithm='SAMME'), param)\n",
    "ada_svc_grid_search.fit(X_train,y_train)\n",
    "print(\"Best parameters: {}\".format(ada_svc_grid_search.best_params_))\n",
    "print(\"Best cross-validation score: {:.4f}\".format(ada_svc_grid_search.best_score_))\n",
    "\n",
    "# Cross-validation\n",
    "ada_svc_scores = cross_val_score(ada_svc, X_train, y_train, cv=5)\n",
    "print(\"Average cross-validation score: {:.4f}\".format(ada_svc_scores.mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Best parameters\n",
    "\n",
    "ada_clf = AdaBoostClassifier(svc_clf, n_estimators=100, algorithm='SAMME', learning_rate=0.8, random_state=0)\n",
    "ada_clf.fit(X_train, y_train)\n",
    "y_pred_ada_svc = ada_clf.predict(X_test)\n",
    "\n",
    "print('AdaBoost with LinearSVC:')\n",
    "print('Train score: {:.4f}'.format(ada_clf.score(X_train, y_train)))\n",
    "print('Test score: {:.4f}'.format(ada_clf.score(X_test, y_test)))\n",
    "print('Accuracy score: {:.4f}'.format(accuracy_score(y_test, y_pred_ada_svc)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 5. Gradient Boosting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# GridSearch\n",
    "\n",
    "from  sklearn.ensemble import GradientBoostingClassifier\n",
    "\n",
    "gb = GradientBoostingClassifier(random_state=42)\n",
    "\n",
    "param = {'max_depth':[1,2,3], 'n_estimators':[100,200], 'learning_rate':[0.8, 1.0]}\n",
    "gb_grid_search = GridSearchCV(gb, param, cv=5)\n",
    "gb_grid_search.fit(X_train,y_train)\n",
    "print(\"Best parameters: {}\".format(gb_grid_search.best_params_))\n",
    "print(\"Best cross-validation score: {:.4f}\".format(gb_grid_search.best_score_))\n",
    "\n",
    "# Cross-validation\n",
    "gb_scores = cross_val_score(gb, X_train, y_train, cv=5)\n",
    "print(\"Average cross-validation score: {:.4f}\".format(gb_scores.mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Best parameters\n",
    "\n",
    "gb = GradientBoostingClassifier(max_depth=1, n_estimators=100, learning_rate=1.0, random_state=42)\n",
    "gb.fit(X_train, y_train)\n",
    "y_pred_gb = gb.predict(X_test)\n",
    "\n",
    "print('Gradient Boosting:')\n",
    "print('Train score: {:.4f}'.format(gb.score(X_train, y_train)))\n",
    "print('Test score: {:.4f}'.format(gb.score(X_test, y_test)))\n",
    "print('Accuracy score: {:.4f}'.format(accuracy_score(y_test, y_pred_gb)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 6. PCA Dimensionality Reduction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Use PCA to reduce the size of the feature space while retaining 95% of explained variance\n",
    "\n",
    "from sklearn.decomposition import PCA\n",
    "\n",
    "pca = PCA()\n",
    "pca.fit(X_train)\n",
    "\n",
    "cumsum = np.cumsum(pca.explained_variance_ratio_)\n",
    "d = np.argmax(cumsum >= 0.95) + 1\n",
    "d"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pca = PCA(n_components=0.95)\n",
    "X_reduced = pca.fit_transform(X_train)\n",
    "X_reduced_test = pca.transform(X_test)\n",
    "X_reduced.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pca.n_components_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.sum(pca.explained_variance_ratio_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 7. Classification Models using Reduced Dataset from PCA"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "    a. KNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_score_array = []\n",
    "test_score_array = []\n",
    "\n",
    "n = [2, 3, 10, 15, 20, 50]\n",
    "for k in n:\n",
    "    knn = KNeighborsClassifier(k)\n",
    "    knn.fit(X_reduced, y_train)\n",
    "    train_score_array.append(knn.score(X_reduced, y_train))\n",
    "    test_score_array.append(knn.score(X_reduced_test, y_test))\n",
    "\n",
    "x_axis = [2, 3, 10, 15, 20, 50]\n",
    "%matplotlib inline\n",
    "plt.plot(x_axis, train_score_array, label = 'Train Score', c = 'g')\n",
    "plt.plot(x_axis, test_score_array, label = 'Test Score', c='b')\n",
    "plt.xlabel('k')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Gridsearch\n",
    "knn_param_grid = {'n_neighbors': [2, 3, 5, 10, 15, 20, 50], 'weights': ['uniform','distance']}\n",
    "knn_grid_search = GridSearchCV(KNeighborsClassifier(), knn_param_grid, cv=5, return_train_score=True)\n",
    "knn_grid_search.fit(X_reduced, y_train)\n",
    "print(\"Best parameters: {}\".format(knn_grid_search.best_params_))\n",
    "print(\"Best cross-validation score: {:.4f}\".format(knn_grid_search.best_score_))\n",
    "\n",
    "# Cross-validation\n",
    "knn_scores = cross_val_score(KNeighborsClassifier(), X_reduced, y_train, cv=5)\n",
    "print(\"Average cross-validation score: {:.4f}\".format(knn_scores.mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Best parameters\n",
    "knn = KNeighborsClassifier(10, weights='uniform')\n",
    "knn.fit(X_reduced, y_train)\n",
    "y_pred_knn = knn.predict(X_reduced_test)\n",
    "\n",
    "print('KNN Classifier:')\n",
    "print('Train score: {:.4f}'.format(knn.score(X_reduced, y_train)))\n",
    "print('Test score: {:.4f}'.format(knn.score(X_reduced_test, y_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from mlxtend.plotting import plot_decision_regions\n",
    "\n",
    "Xt = X_reduced[:, [1,7]]\n",
    "yt = y_train\n",
    "\n",
    "knn.fit(Xt, yt) \n",
    "plot_decision_regions(Xt, yt, clf = knn)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "    b. Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "c_range = [0.1, 1, 10]\n",
    "train_score_l1 = []\n",
    "train_score_l2 = []\n",
    "test_score_l1 = []\n",
    "test_score_l2 = []\n",
    "\n",
    "for c in c_range:\n",
    "    log_l1 = LogisticRegression(penalty = 'l1', C = c, random_state=0,solver='liblinear')\n",
    "    log_l2 = LogisticRegression(penalty = 'l2', C = c, random_state=0,solver='liblinear')\n",
    "    log_l1.fit(X_reduced, y_train)\n",
    "    log_l2.fit(X_reduced, y_train)\n",
    "    train_score_l1.append(log_l1.score(X_reduced, y_train))\n",
    "    train_score_l2.append(log_l2.score(X_reduced, y_train))\n",
    "    test_score_l1.append(log_l1.score(X_reduced_test, y_test))\n",
    "    test_score_l2.append(log_l2.score(X_reduced_test, y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.plot(c_range, train_score_l1, label = 'Train score, penalty = l1')\n",
    "plt.plot(c_range, test_score_l1, label = 'Test score, penalty = l1')\n",
    "plt.plot(c_range, train_score_l2, label = 'Train score, penalty = l2')\n",
    "plt.plot(c_range, test_score_l2, label = 'Test score, penalty = l2')\n",
    "plt.legend()\n",
    "plt.xlabel('Regularization parameter: C')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.xscale('log')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Gridsearch\n",
    "lreg = LogisticRegression()\n",
    "hyperparameters = dict(C=c_range, penalty=['l1','l2'])\n",
    "lreg_grid_search = GridSearchCV(LogisticRegression(), hyperparameters, cv=5, return_train_score=True)\n",
    "lreg_grid_search.fit(X_reduced, y_train)\n",
    "print(\"Best parameters: {}\".format(lreg_grid_search.best_params_))\n",
    "print(\"Best cross-validation score: {:.4f}\".format(lreg_grid_search.best_score_))\n",
    "\n",
    "# Cross-validation\n",
    "lreg_scores = cross_val_score(lreg, X_reduced, y_train, cv=5)\n",
    "print(\"Average cross-validation score: {:.4f}\".format(lreg_scores.mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Best parameters\n",
    "log_clf = LogisticRegression(C=1, penalty='l2')\n",
    "log_clf.fit(X_reduced, y_train)\n",
    "y_pred_lr = log_clf.predict(X_reduced_test)\n",
    "\n",
    "print('Logistic Regression:')\n",
    "print('Train score: {:.4f}'.format(log_clf.score(X_reduced, y_train)))\n",
    "print('Test score: {:.4f}'.format(log_clf.score(X_reduced_test, y_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "log_clf.fit(Xt, yt) \n",
    "plot_decision_regions(Xt, yt, clf = log_clf)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "    c. Linear Support Vector Machine"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Gridsearch\n",
    "Linearsvc_param_grid = {'C': [0.001, 0.01, 0.1, 1, 10, 100]}\n",
    "Linearsvc_grid_search = GridSearchCV(LinearSVC(), Linearsvc_param_grid, cv=5, return_train_score=True)\n",
    "Linearsvc_grid_search.fit(X_reduced, y_train)\n",
    "print(\"Best parameters: {}\".format(Linearsvc_grid_search.best_params_))\n",
    "print(\"Best cross-validation score: {:.4f}\".format(Linearsvc_grid_search.best_score_))\n",
    "\n",
    "# Cross-validation\n",
    "Linearsvc_scores = cross_val_score(LinearSVC(), X_reduced, y_train, cv=5)\n",
    "print(\"Average cross-validation score: {:.4f}\".format(Linearsvc_scores.mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Best parameter\n",
    "svc = LinearSVC(C=1)\n",
    "svc.fit(X_reduced, y_train)\n",
    "y_pred_svc = svc.predict(X_reduced_test)\n",
    "\n",
    "print('LinearSVC:')\n",
    "print('Train score: {:.4f}'.format(svc.score(X_reduced, y_train)))\n",
    "print('Test score: {:.4f}'.format(svc.score(X_reduced_test, y_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "svc.fit(Xt, yt)\n",
    "plot_decision_regions(Xt, yt, clf=svc)\n",
    "plt.xlabel(\"Feature 1\")\n",
    "plt.ylabel(\"Feature 2\")\n",
    "plt.legend(['CHD', 'NoCHD'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "    d. Kernelized Support Vector Machine"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "rbf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Gridsearch\n",
    "from sklearn.svm import SVC\n",
    "\n",
    "rbf_svc_param_grid = {'C': [0.001, 0.01, 0.1, 1, 10, 100], 'gamma': [0.001, 0.01, 0.1, 1, 10, 100]}\n",
    "rbf_svc_grid_search = GridSearchCV(SVC(kernel='rbf'), rbf_svc_param_grid, cv=5, return_train_score=True)\n",
    "rbf_svc_grid_search.fit(X_reduced, y_train)\n",
    "print(\"Best parameters: {}\".format(rbf_svc_grid_search.best_params_))\n",
    "print(\"Best cross-validation score: {:.4f}\".format(rbf_svc_grid_search.best_score_))\n",
    "\n",
    "# Cross-validation\n",
    "rbf_svc_scores = cross_val_score(SVC(kernel='rbf'), X_reduced, y_train, cv=5)\n",
    "print(\"Average cross-validation score: {:.4f}\".format(rbf_svc_scores.mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Best parameters\n",
    "rbf = SVC(kernel='rbf', C=0.001, gamma=0.001)\n",
    "rbf.fit(X_reduced, y_train)\n",
    "y_pred_rbf = rbf.predict(X_reduced_test)\n",
    "\n",
    "print('SVC with kernel = rbf:')\n",
    "print('Train score: {:.4f}'.format(rbf.score(X_reduced, y_train)))\n",
    "print('Test score: {:.4f}'.format(rbf.score(X_reduced_test, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "poly"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Gridsearch\n",
    "poly_svc_param_grid = {'C': [0.001, 0.01, 0.1, 1, 10, 100], 'degree': range(1,10)}\n",
    "poly_svc_grid_search = GridSearchCV(SVC(kernel='poly'), poly_svc_param_grid, cv=5, return_train_score=True)\n",
    "poly_svc_grid_search.fit(X_reduced, y_train)\n",
    "print(\"Best parameters: {}\".format(poly_svc_grid_search.best_params_))\n",
    "print(\"Best cross-validation score: {:.4f}\".format(poly_svc_grid_search.best_score_))\n",
    "\n",
    "# Cross-validation\n",
    "poly_svc_scores = cross_val_score(SVC(kernel='poly'), X_reduced, y_train, cv=5)\n",
    "print(\"Average cross-validation score: {:.4f}\".format(poly_svc_scores.mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Best parameters\n",
    "poly = SVC(kernel='poly', C=0.001, degree=3)\n",
    "poly.fit(X_reduced, y_train)\n",
    "y_pred_poly = poly.predict(X_reduced_test)\n",
    "\n",
    "print('SVC with kernel = poly:')\n",
    "print('Train score: {:.4f}'.format(poly.score(X_reduced, y_train)))\n",
    "print('Test score: {:.4f}'.format(poly.score(X_reduced_test, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "linear"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Gridsearch\n",
    "linear_svc_param_grid = {'C': [0.001, 0.01, 0.1, 1, 10, 100]}\n",
    "linear_svc_grid_search = GridSearchCV(SVC(kernel='linear'), linear_svc_param_grid, cv=5, return_train_score=True)\n",
    "linear_svc_grid_search.fit(X_reduced, y_train)\n",
    "print(\"Best parameters: {}\".format(linear_svc_grid_search.best_params_))\n",
    "print(\"Best cross-validation score: {:.4f}\".format(linear_svc_grid_search.best_score_))\n",
    "\n",
    "# Cross-validation\n",
    "linear_svc_scores = cross_val_score(SVC(kernel='linear'), X_reduced, y_train, cv=5)\n",
    "print(\"Average cross-validation score: {:.4f}\".format(linear_svc_scores.mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Best parameters\n",
    "linear = SVC(kernel='linear', C=0.001)\n",
    "linear.fit(X_reduced, y_train)\n",
    "y_pred_linear = linear.predict(X_reduced_test)\n",
    "\n",
    "print('SVC with kernel = linear:')\n",
    "print('Train score: {:.4f}'.format(linear.score(X_reduced, y_train)))\n",
    "print('Test score: {:.4f}'.format(linear.score(X_reduced_test, y_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# SVC Best Parameters\n",
    "\n",
    "import matplotlib.gridspec as gridspec\n",
    "import itertools\n",
    "\n",
    "clf1 = LinearSVC(C=1)\n",
    "clf2 = SVC(kernel='linear', C=0.001)\n",
    "clf3 = SVC(kernel='rbf', gamma=0.001, C=0.001)\n",
    "clf4 = SVC(kernel='poly', degree=1, C=0.001)\n",
    "\n",
    "gs = gridspec.GridSpec(2, 2)\n",
    "fig = plt.figure(figsize=(10, 8))\n",
    "\n",
    "labels = ['LinearSVC','SVC with linear (C=0.001) kernel','SVC with RBF (gamma=0.001, C=0.001) kernel','SVC with polynomial (degree=3, C=100) kernel']\n",
    "\n",
    "for clf, lab, grd in zip([clf1, clf2, clf3, clf4],labels,itertools.product([0, 1],repeat=2)):\n",
    "    clf.fit(Xt, yt)\n",
    "    ax = plt.subplot(gs[grd[0], grd[1]])\n",
    "    fig = plot_decision_regions(X=Xt, y=yt,clf=clf, legend=2)\n",
    "    plt.title(lab)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "    d. Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Gridsearch\n",
    "dtree = DecisionTreeClassifier(random_state=0)\n",
    "param_DT = {\"max_depth\": range(1,10),\"min_samples_split\": range(2,50,1),\"max_leaf_nodes\": range(2,5)}\n",
    "dtree_grid_search = GridSearchCV(dtree,param_DT,cv=5)\n",
    "dtree_grid_search.fit(X_reduced, y_train)\n",
    "print(\"Best parameters: {}\".format(dtree_grid_search.best_params_))\n",
    "print(\"Best cross-validation score: {:.4f}\".format(dtree_grid_search.best_score_))\n",
    "\n",
    "# Cross-validation\n",
    "dtree_scores = cross_val_score(dtree, X_reduced, y_train, cv=5)\n",
    "print(\"Average cross-validation score: {:.4f}\".format(dtree_scores.mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Best parameters\n",
    "dtree = DecisionTreeClassifier(max_depth=1, max_leaf_nodes=2, min_samples_split=2)\n",
    "dtree.fit(X_reduced, y_train)\n",
    "y_pred_dtree = dtree.predict(X_reduced_test)\n",
    "\n",
    "print('Decision Tree:')\n",
    "print('Train score: {:.4f}'.format(dtree.score(X_reduced, y_train)))\n",
    "print('Test score: {:.4f}'.format(dtree.score(X_reduced_test, y_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Feature importances\n",
    "dtree.fit(X_reduced, y_train)\n",
    "\n",
    "plt.figure(figsize=(20,10))\n",
    "def plot_feature_importances_hd(model):\n",
    "    n_features = X_reduced.shape[1]\n",
    "    plt.barh(range(n_features), model.feature_importances_, align='center')\n",
    "    plt.yticks(np.arange(n_features))\n",
    "    plt.xlabel(\"Feature importance\")\n",
    "    plt.ylabel(\"Feature\")\n",
    "    plt.ylim(-1, n_features)\n",
    "\n",
    "plot_feature_importances_hd(dtree)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Feature 1 of the reduced dataset is the most important."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dtree.fit(Xt, yt)\n",
    "plot_decision_regions(Xt, yt, clf=dtree)\n",
    "plt.xlabel(\"Feature 1\")\n",
    "plt.ylabel(\"Feature 2\")\n",
    "plt.legend(['CHD', 'NoCHD'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 8. Model Evaluation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Confusion Matrices\n",
    "\n",
    "seed = 15\n",
    "\n",
    "reports = []\n",
    "reports.append(('KNN', y_pred_knn))\n",
    "reports.append(('Logistic Regression', y_pred_lr))\n",
    "reports.append(('LinearSVC', y_pred_svc))\n",
    "reports.append(('SVC with kernel = rbf', y_pred_rbf))\n",
    "reports.append(('SVC with kernel = poly', y_pred_poly))\n",
    "reports.append(('SVC with kernel = linear', y_pred_linear))\n",
    "reports.append(('Decision Tree', y_pred_dtree))\n",
    "\n",
    "from sklearn.metrics import confusion_matrix\n",
    "\n",
    "for clas, report in reports:\n",
    "    print(clas)\n",
    "    print(confusion_matrix(y_test, report))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Cross-validation scores\n",
    "\n",
    "seed = 10\n",
    "\n",
    "results = []\n",
    "results.append(('KNN', knn_grid_search))\n",
    "results.append(('Logistic Regression', lreg_grid_search))\n",
    "results.append(('LinearSVC', Linearsvc_grid_search))\n",
    "results.append(('SVC with kernel = rbf', rbf_svc_grid_search))\n",
    "results.append(('SVC with kernel = poly', poly_svc_grid_search))\n",
    "results.append(('SVC with kernel = linear', linear_svc_grid_search))\n",
    "results.append(('Decision Tree', dtree_grid_search))\n",
    "params = []\n",
    "cv_scores = []\n",
    "\n",
    "for clas, result in results:\n",
    "    param = result.best_params_\n",
    "    cv_score = result.best_score_.round(4)\n",
    "    params.append(param)\n",
    "    cv_scores.append(cv_score)\n",
    "    \n",
    "table = [params, cv_scores]\n",
    "reshaped = np.array(table).reshape(2,7)\n",
    "table = pd.DataFrame(reshaped, index=['Best parameters', 'Best cross-validation score'], columns=['KNN', 'Logistic Regression', 'LinearSVC', 'SVC with kernel = rbf', 'SVC with kernel = poly', 'SVC with kernel = linear', 'Decision Tree'])\n",
    "print(\"Cross-validation scores with PCA: \")\n",
    "table = table.T\n",
    "table"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# AUC and ROC Curve\n",
    "\n",
    "seed = 12\n",
    "\n",
    "auc_scores = []\n",
    "classes = []\n",
    "\n",
    "for clas, result in results:\n",
    "    auc = roc_auc_score(y_test, model.predict_proba(X_reduced_test)[:,1]).round(4)\n",
    "    auc_scores.append(auc)\n",
    "    classes.append(clas)\n",
    "\n",
    "auc_results = [classes, auc_scores]\n",
    "reshaped2 = np.array(auc_results).reshape(2,7)\n",
    "auc_results = pd.DataFrame(reshaped2, index=['Name', 'AUC'])\n",
    "auc_results.rename(columns=auc_results.iloc[0], inplace=True)\n",
    "auc_results.drop(auc_results.index[0], inplace=True)\n",
    "auc_results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "for clas, result in results:\n",
    "    y_pred_proba = result.predict_proba(X_test)[::,1]\n",
    "    fpr, tpr, threshold = roc_curve(y_test, y_pred_proba)\n",
    "    roc_auc = roc_auc_score(y_test, y_pred_proba)\n",
    "    print(clas)\n",
    "    plt.plot(fpr,tpr,label=\"auc=\"+str(roc_auc))\n",
    "    plt.legend(loc=4)\n",
    "    plt.ylabel('Recall')\n",
    "    plt.xlabel('1-specificity')\n",
    "    plt.title('ROC Curve')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Average cross-validation scores\n",
    "\n",
    "seed = 20\n",
    "\n",
    "models = []\n",
    "models.append(('KNN', knn_scores))\n",
    "models.append(('Logistic Regression', lreg_scores))\n",
    "models.append(('LinearSVC', Linearsvc_scores))\n",
    "models.append(('SVC with kernel = rbf', rbf_svc_scores))\n",
    "models.append(('SVC with kernel = poly', poly_svc_scores))\n",
    "models.append(('SVC with kernel = linear', linear_svc_scores))\n",
    "models.append(('Decision Tree', dtree_scores))\n",
    "\n",
    "avg_scores = []\n",
    "\n",
    "for clas, model in models:\n",
    "    avg_score = model.mean().round(4)\n",
    "    avg_scores.append(avg_score)\n",
    "    \n",
    "table1 = [avg_scores]\n",
    "reshaped2 = np.array(table1).reshape(1,7)\n",
    "table1 = pd.DataFrame(reshaped2, index=['Average cross-validation score'], columns=['KNN', 'Logistic Regression', 'LinearSVC', 'SVC with kernel = rbf', 'SVC with kernel = poly', 'SVC with kernel = linear', 'Decision Tree'])\n",
    "print(\"Average cross-validation scores with PCA: \")\n",
    "table1 = table1.T\n",
    "table1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Classification Reports\n",
    "    \n",
    "precision_scores = []\n",
    "recall_scores = []\n",
    "accuracy_scores = []\n",
    "f1_scores = []\n",
    "classes = []\n",
    "\n",
    "for clas, report in reports:\n",
    "    precision = precision_score(y_test, report).round(4)\n",
    "    recall = recall_score(y_test, report).round(4)\n",
    "    accuracy = accuracy_score(y_test, report).round(4)\n",
    "    f1 = f1_score(y_test, report).round(4)\n",
    "    precision_scores.append(precision)\n",
    "    recall_scores.append(recall)\n",
    "    accuracy_scores.append(accuracy)\n",
    "    f1_scores.append(f1)\n",
    "    classes.append(clas)\n",
    "\n",
    "classification = [precision_scores, recall_scores, accuracy_scores, f1_scores]\n",
    "reshaped1 = np.array(classification).reshape(4,7)\n",
    "classification = pd.DataFrame(reshaped1, index=['Precision score', 'Recall score', 'Accuracy score', 'F1 score'], columns=['KNN', 'Logistic Regression', 'LinearSVC', 'SVC with kernel = rbf', 'SVC with kernel = poly', 'SVC with kernel = linear', 'Decision Tree'])\n",
    "print(\"Classification Report with PCA: \")\n",
    "classification = classification.T\n",
    "classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Combine all results\n",
    "result = [table, table1, classification]\n",
    "result = pd.concat(result, axis=1)\n",
    "print('Results with PCA: ')\n",
    "result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Results without PCA\n",
    "\n",
    "data = {'Model Selection':[\"KNN\", \"Logistic Regression\", \"LinearSVC\", \"SVC kernel rbf\", \"SVC kernel poly\", \"SVC kernel linear\", \"Decision Tree\"],\n",
    "        'Best parameters':[{'n_neighbors': 20, 'weights': 'uniform'},{'C': 1, 'penalty': 'l1'},{'C': 10},{'C': 0.001, 'gamma': 0.001},{'C': 100, 'degree': 3},{'C': 0.001},{'max_depth': 1, 'max_leaf_nodes': 2, 'min_samples_split': 2}],\n",
    "       'Best cross-validation score':[0.8512, 0.8581, 0.8537, 0.8508, 0.8515, 0.8508, 0.8508],\n",
    "       'Average cross-validation score':[0.8364, 0.8581, 0.8534, 0.8509, 0.8509, 0.8483, 0.7555],\n",
    "       'Precision score':[0.0000, 0.6250, 0.6250, 0.0000, 0.7500, 0.0000, 0.0000],\n",
    "       'Recall score':[0.0000, 0.0588, 0.0294, 0.0000, 0.0176, 0.0000, 0.0000],\n",
    "        'Accuracy score':[0.8396, 0.8434, 0.8415, 0.8396, 0.8415, 0.8396, 0.8396],\n",
    "       'F1 score':[0.0000, 0.1075, 0.0562, 0.0000, 0.0345, 0.0000, 0.0000],} \n",
    "\n",
    "df = pd.DataFrame(data)\n",
    "print(\"Results without PCA: \")\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From the two result tables above, it is clear that all the scores are almost the same as those without using PCA but are still lower. In fact, the accuracy scores of all models decreased with PCA. Without using PCA, the best model with the highest accuracy score (0.8434) is Logistic Regression. After using PCA, the best model with the highest accuracy score (0.8406) is LinearSVC while all the other models have the same accuracy score as 0.8396. Thus, PCA didn't help in getting better results in this dataset.\n",
    "\n",
    "Since we have a small number of features and an intermediate number of training instances, SVMs are not very efficient computationally and work better for image analysis tasks. As we would like to have the best accuracy in our prediction, Logistic Regression generated the best results (highest best cross-validation score with a high accuracy score). Besides, it is also incredibly easy to implement and very efficient to train. In this situation, Logistic Regression is the best model and should be selected."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 9. Deep Learning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense, Activation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Get number of input dim\n",
    "X_train.shape    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define the keras model\n",
    "model = Sequential()\n",
    "\n",
    "# Hidden layers\n",
    "model.add(Dense(12, input_dim=15, activation='relu'))\n",
    "model.add(Dense(8, activation='relu'))\n",
    "\n",
    "# Output\n",
    "model.add(Dense(1, activation='sigmoid'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Compile the model\n",
    "model.compile(loss='binary_crossentropy' , optimizer='adam' , metrics = ['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Train the model: fit epochs and batch_size\n",
    "model.fit(X_train, y_train, epochs = 10, batch_size = 150)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Evaluation\n",
    "model.evaluate(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Evaluation for test data\n",
    "loss_and_metrics = model.evaluate(X_test, y_test)\n",
    "print(\"Test Loss:\", loss_and_metrics[0])\n",
    "print(\"Test Accuracy:\", loss_and_metrics[1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The model generated a good result of accuracy."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Evaluate both correctly and incorrectly classified examples\n",
    "predicted_classes = model.predict_classes(X_test)\n",
    "\n",
    "# See which we predicted correctly and incorrectly\n",
    "correct_indices = np.nonzero(predicted_classes == y_test)[0]\n",
    "incorrect_indices = np.nonzero(predicted_classes != y_test)[0]\n",
    "print()\n",
    "print(len(correct_indices),\" classified correctly\")\n",
    "print(len(incorrect_indices),\" classified incorrectly\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Summarize the first 5 cases\n",
    "for i in range(5):\n",
    "\tprint('%s => %d (expected %d)' % (X_test[i].tolist(), predicted_classes[i], y_test[i]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The neural network model actually provided an accuracy score that is as good as our previous models. The training score of this model is higher than that of every model we performed earlier. The test score, though not the highest, is still higher than most of the test scores of all the other models."
   ]
  }
 ],
 "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
}