house_pricing_analysis_ofallon.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import sys\n",
    "import pandas as pd\n",
    "pd.set_option('display.max_rows', 100)\n",
    "pd.set_option('display.max_columns', 100)\n",
    "pd.set_option('expand_frame_repr', False)\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import seaborn as sns\n",
    "import os\n",
    "import datetime as dt\n",
    "import sklearn\n",
    "import re"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Raw dataset shape: (32250, 32)\n",
      "Raw dataset shape with duplicates removed: (10659, 22)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>MAILING_ADDR</th>\n",
       "      <th>MAILING_ZIP</th>\n",
       "      <th>SUBDIVISION</th>\n",
       "      <th>SCHOOL_DISTRICT</th>\n",
       "      <th>LOT_SIZE</th>\n",
       "      <th>YEAR_BUILT</th>\n",
       "      <th>QUALITY_CODE</th>\n",
       "      <th>BEDROOMS</th>\n",
       "      <th>BATHROOMS</th>\n",
       "      <th>HALF_BATHROOMS</th>\n",
       "      <th>TOTAL_ROOMS</th>\n",
       "      <th>ARCHITECTURAL_STYLE</th>\n",
       "      <th>TOTAL_AREA</th>\n",
       "      <th>BASE_AREA</th>\n",
       "      <th>BASEMENT_AREA</th>\n",
       "      <th>BASEMENT_FINISHED_AREA</th>\n",
       "      <th>PARKING_AREA</th>\n",
       "      <th>LAND_VALUE</th>\n",
       "      <th>LAST_SALE_PRICE</th>\n",
       "      <th>LAST_SALE_DATE</th>\n",
       "      <th>MARKET_VALUE</th>\n",
       "      <th>LAST_SALE_UPSF</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>414 CORDES AVE</td>\n",
       "      <td>63366</td>\n",
       "      <td>WOODLAWN ADDN-OFALLON</td>\n",
       "      <td>Fort Zumwalt</td>\n",
       "      <td>75X120</td>\n",
       "      <td>2001</td>\n",
       "      <td>R - 30 - Average</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>1STY - 1 Story Building</td>\n",
       "      <td>1,496</td>\n",
       "      <td>1,496</td>\n",
       "      <td>1,496</td>\n",
       "      <td>0</td>\n",
       "      <td>660</td>\n",
       "      <td>$35,000.00</td>\n",
       "      <td>$9,000.00</td>\n",
       "      <td>10/02/2000</td>\n",
       "      <td>$210,354.00</td>\n",
       "      <td>$6.02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8564 HWY N</td>\n",
       "      <td>63367-4049</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Wentzville</td>\n",
       "      <td>3.2000 AC</td>\n",
       "      <td>2001</td>\n",
       "      <td>R - 35 - Average to Good</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>1STY - 1 Story Building</td>\n",
       "      <td>1,842</td>\n",
       "      <td>1,842</td>\n",
       "      <td>1,812</td>\n",
       "      <td>0</td>\n",
       "      <td>624</td>\n",
       "      <td>$80,000.00</td>\n",
       "      <td>$0.00</td>\n",
       "      <td>08/21/2000</td>\n",
       "      <td>$335,872.00</td>\n",
       "      <td>$0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>18 OLD GLORY CT</td>\n",
       "      <td>63366</td>\n",
       "      <td>PATRIOTS LANDING #1</td>\n",
       "      <td>Fort Zumwalt</td>\n",
       "      <td>0.2500 AC</td>\n",
       "      <td>2000</td>\n",
       "      <td>R - 30 - Average</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>1STY - 1 Story Building</td>\n",
       "      <td>1,675</td>\n",
       "      <td>1,675</td>\n",
       "      <td>1,647</td>\n",
       "      <td>0</td>\n",
       "      <td>644</td>\n",
       "      <td>$45,000.00</td>\n",
       "      <td>$141,900.00</td>\n",
       "      <td>02/15/2000</td>\n",
       "      <td>$247,055.00</td>\n",
       "      <td>$84.72</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1449 AMBER WAVE DR</td>\n",
       "      <td>63366</td>\n",
       "      <td>PATRIOTS LANDING #1</td>\n",
       "      <td>Fort Zumwalt</td>\n",
       "      <td>0.2000 AC</td>\n",
       "      <td>2000</td>\n",
       "      <td>R - 30 - Average</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>1STY - 1 Story Building</td>\n",
       "      <td>1,474</td>\n",
       "      <td>1,474</td>\n",
       "      <td>1,450</td>\n",
       "      <td>0</td>\n",
       "      <td>399</td>\n",
       "      <td>$45,000.00</td>\n",
       "      <td>$0.00</td>\n",
       "      <td>03/21/2001</td>\n",
       "      <td>$216,745.00</td>\n",
       "      <td>$0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>9 FOLK ART CT</td>\n",
       "      <td>63366</td>\n",
       "      <td>PATRIOTS LANDING #1</td>\n",
       "      <td>Fort Zumwalt</td>\n",
       "      <td>0.1700 AC</td>\n",
       "      <td>2000</td>\n",
       "      <td>R - 30 - Average</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>1STY - 1 Story Building</td>\n",
       "      <td>1,421</td>\n",
       "      <td>1,421</td>\n",
       "      <td>1,361</td>\n",
       "      <td>0</td>\n",
       "      <td>400</td>\n",
       "      <td>$45,000.00</td>\n",
       "      <td>$128,450.00</td>\n",
       "      <td>07/10/2000</td>\n",
       "      <td>$212,616.00</td>\n",
       "      <td>$90.39</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         MAILING_ADDR MAILING_ZIP            SUBDIVISION SCHOOL_DISTRICT   LOT_SIZE  YEAR_BUILT              QUALITY_CODE  BEDROOMS  BATHROOMS  HALF_BATHROOMS  TOTAL_ROOMS      ARCHITECTURAL_STYLE TOTAL_AREA BASE_AREA BASEMENT_AREA BASEMENT_FINISHED_AREA PARKING_AREA  LAND_VALUE LAST_SALE_PRICE LAST_SALE_DATE MARKET_VALUE LAST_SALE_UPSF\n",
       "0      414 CORDES AVE       63366  WOODLAWN ADDN-OFALLON    Fort Zumwalt     75X120        2001          R - 30 - Average         3          2               0            5  1STY - 1 Story Building      1,496     1,496         1,496                      0          660  $35,000.00       $9,000.00     10/02/2000  $210,354.00          $6.02\n",
       "1          8564 HWY N  63367-4049                    NaN      Wentzville  3.2000 AC        2001  R - 35 - Average to Good         3          2               0            6  1STY - 1 Story Building      1,842     1,842         1,812                      0          624  $80,000.00           $0.00     08/21/2000  $335,872.00          $0.00\n",
       "2     18 OLD GLORY CT       63366    PATRIOTS LANDING #1    Fort Zumwalt  0.2500 AC        2000          R - 30 - Average         3          2               0            5  1STY - 1 Story Building      1,675     1,675         1,647                      0          644  $45,000.00     $141,900.00     02/15/2000  $247,055.00         $84.72\n",
       "3  1449 AMBER WAVE DR       63366    PATRIOTS LANDING #1    Fort Zumwalt  0.2000 AC        2000          R - 30 - Average         3          2               0            5  1STY - 1 Story Building      1,474     1,474         1,450                      0          399  $45,000.00           $0.00     03/21/2001  $216,745.00          $0.00\n",
       "4       9 FOLK ART CT       63366    PATRIOTS LANDING #1    Fort Zumwalt  0.1700 AC        2000          R - 30 - Average         3          2               0            5  1STY - 1 Story Building      1,421     1,421         1,361                      0          400  $45,000.00     $128,450.00     07/10/2000  $212,616.00         $90.39"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "if os.getcwd()[9:14] == 'ssoma':\n",
    "    file_path = \"C:\\\\Users\\\\ssoma\\\\OneDrive - Bayer\\\\Personal Data-R90PQ30R\\\\Analytics\\\\house-pricing\\\\input-data\\\\ofallon\\\\\"\n",
    "else:\n",
    "    file_path = \"C:\\\\Users\\\\sivac\\\\Documents\\\\Analytics\\\\house-pricing\\\\input-data\\\\ofallon\\\\\"\n",
    "\n",
    "\n",
    "file_list = os.listdir(file_path)\n",
    "for i in file_list:\n",
    "    i = file_path + i\n",
    "    file = pd.read_csv(i)\n",
    "    \n",
    "    if 'df' in globals():\n",
    "        df = pd.concat([df, file])\n",
    "    else:\n",
    "        df = file\n",
    "\n",
    "print(f'Raw dataset shape: {df.shape}')\n",
    "\n",
    "df_new = df[['MAILING_ADDR', 'MAILING_ZIP', 'SUBDIVISION', 'SCHOOL_DISTRICT', 'LOT_SIZE', 'YEAR_BUILT', 'QUALITY_CODE', 'BEDROOMS', 'BATHROOMS',\n",
    "            'HALF_BATHROOMS', 'TOTAL_ROOMS', 'ARCHITECTURAL_STYLE', 'TOTAL_AREA', 'BASE_AREA', 'BASEMENT_AREA','BASEMENT_FINISHED_AREA',\n",
    "            'PARKING_AREA', 'LAND_VALUE', 'LAST_SALE_PRICE', 'LAST_SALE_DATE', 'MARKET_VALUE', 'LAST_SALE_UPSF']]\n",
    "\n",
    "df_new = df_new.drop_duplicates().reset_index(drop=True)\n",
    "print(f'Raw dataset shape with duplicates removed: {df_new.shape}')\n",
    "df_new.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Reading unseen dataset\n",
    "unseen_data = file_path.replace(\"input-data\\\\ofallon\\\\\",\"\")\n",
    "unseen_data = pd.read_csv(unseen_data + \"unseen_data.csv\")\n",
    "\n",
    "# Filtering source dataset as per unseen data\n",
    "\n",
    "subdivision = unseen_data['SUBDIVISION'].values\n",
    "address = unseen_data['MAILING_ADDR'].values\n",
    "\n",
    "#df_new = df_new[df_new['SUBDIVISION'].isin(subdivision)].copy()\n",
    "df_new = df_new[~df_new['MAILING_ADDR'].isin(address)].copy().reset_index(drop=True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>MAILING_ADDR</th>\n",
       "      <th>MAILING_ZIP</th>\n",
       "      <th>SUBDIVISION</th>\n",
       "      <th>SCHOOL_DISTRICT</th>\n",
       "      <th>LOT_SIZE</th>\n",
       "      <th>YEAR_BUILT</th>\n",
       "      <th>QUALITY_CODE</th>\n",
       "      <th>BEDROOMS</th>\n",
       "      <th>BATHROOMS</th>\n",
       "      <th>HALF_BATHROOMS</th>\n",
       "      <th>TOTAL_ROOMS</th>\n",
       "      <th>ARCHITECTURAL_STYLE</th>\n",
       "      <th>TOTAL_AREA</th>\n",
       "      <th>BASE_AREA</th>\n",
       "      <th>BASEMENT_AREA</th>\n",
       "      <th>BASEMENT_FINISHED_AREA</th>\n",
       "      <th>PARKING_AREA</th>\n",
       "      <th>LAND_VALUE</th>\n",
       "      <th>LAST_SALE_PRICE</th>\n",
       "      <th>LAST_SALE_DATE</th>\n",
       "      <th>MARKET_VALUE</th>\n",
       "      <th>LAST_SALE_UPSF</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>414 CORDES AVE</td>\n",
       "      <td>63366</td>\n",
       "      <td>WOODLAWN ADDN-OFALLON</td>\n",
       "      <td>Fort Zumwalt</td>\n",
       "      <td>75X120</td>\n",
       "      <td>2001</td>\n",
       "      <td>R - 30 - Average</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>1STY - 1 Story Building</td>\n",
       "      <td>1,496</td>\n",
       "      <td>1,496</td>\n",
       "      <td>1,496</td>\n",
       "      <td>0</td>\n",
       "      <td>660</td>\n",
       "      <td>$35,000.00</td>\n",
       "      <td>$9,000.00</td>\n",
       "      <td>10/02/2000</td>\n",
       "      <td>$210,354.00</td>\n",
       "      <td>$6.02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8564 HWY N</td>\n",
       "      <td>63367-4049</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Wentzville</td>\n",
       "      <td>3.2000 AC</td>\n",
       "      <td>2001</td>\n",
       "      <td>R - 35 - Average to Good</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>1STY - 1 Story Building</td>\n",
       "      <td>1,842</td>\n",
       "      <td>1,842</td>\n",
       "      <td>1,812</td>\n",
       "      <td>0</td>\n",
       "      <td>624</td>\n",
       "      <td>$80,000.00</td>\n",
       "      <td>$0.00</td>\n",
       "      <td>08/21/2000</td>\n",
       "      <td>$335,872.00</td>\n",
       "      <td>$0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>18 OLD GLORY CT</td>\n",
       "      <td>63366</td>\n",
       "      <td>PATRIOTS LANDING #1</td>\n",
       "      <td>Fort Zumwalt</td>\n",
       "      <td>0.2500 AC</td>\n",
       "      <td>2000</td>\n",
       "      <td>R - 30 - Average</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>1STY - 1 Story Building</td>\n",
       "      <td>1,675</td>\n",
       "      <td>1,675</td>\n",
       "      <td>1,647</td>\n",
       "      <td>0</td>\n",
       "      <td>644</td>\n",
       "      <td>$45,000.00</td>\n",
       "      <td>$141,900.00</td>\n",
       "      <td>02/15/2000</td>\n",
       "      <td>$247,055.00</td>\n",
       "      <td>$84.72</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1449 AMBER WAVE DR</td>\n",
       "      <td>63366</td>\n",
       "      <td>PATRIOTS LANDING #1</td>\n",
       "      <td>Fort Zumwalt</td>\n",
       "      <td>0.2000 AC</td>\n",
       "      <td>2000</td>\n",
       "      <td>R - 30 - Average</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>1STY - 1 Story Building</td>\n",
       "      <td>1,474</td>\n",
       "      <td>1,474</td>\n",
       "      <td>1,450</td>\n",
       "      <td>0</td>\n",
       "      <td>399</td>\n",
       "      <td>$45,000.00</td>\n",
       "      <td>$0.00</td>\n",
       "      <td>03/21/2001</td>\n",
       "      <td>$216,745.00</td>\n",
       "      <td>$0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>9 FOLK ART CT</td>\n",
       "      <td>63366</td>\n",
       "      <td>PATRIOTS LANDING #1</td>\n",
       "      <td>Fort Zumwalt</td>\n",
       "      <td>0.1700 AC</td>\n",
       "      <td>2000</td>\n",
       "      <td>R - 30 - Average</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>1STY - 1 Story Building</td>\n",
       "      <td>1,421</td>\n",
       "      <td>1,421</td>\n",
       "      <td>1,361</td>\n",
       "      <td>0</td>\n",
       "      <td>400</td>\n",
       "      <td>$45,000.00</td>\n",
       "      <td>$128,450.00</td>\n",
       "      <td>07/10/2000</td>\n",
       "      <td>$212,616.00</td>\n",
       "      <td>$90.39</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         MAILING_ADDR MAILING_ZIP            SUBDIVISION SCHOOL_DISTRICT   LOT_SIZE  YEAR_BUILT              QUALITY_CODE  BEDROOMS  BATHROOMS  HALF_BATHROOMS  TOTAL_ROOMS      ARCHITECTURAL_STYLE TOTAL_AREA BASE_AREA BASEMENT_AREA BASEMENT_FINISHED_AREA PARKING_AREA  LAND_VALUE LAST_SALE_PRICE LAST_SALE_DATE MARKET_VALUE LAST_SALE_UPSF\n",
       "0      414 CORDES AVE       63366  WOODLAWN ADDN-OFALLON    Fort Zumwalt     75X120        2001          R - 30 - Average         3          2               0            5  1STY - 1 Story Building      1,496     1,496         1,496                      0          660  $35,000.00       $9,000.00     10/02/2000  $210,354.00          $6.02\n",
       "1          8564 HWY N  63367-4049                    NaN      Wentzville  3.2000 AC        2001  R - 35 - Average to Good         3          2               0            6  1STY - 1 Story Building      1,842     1,842         1,812                      0          624  $80,000.00           $0.00     08/21/2000  $335,872.00          $0.00\n",
       "2     18 OLD GLORY CT       63366    PATRIOTS LANDING #1    Fort Zumwalt  0.2500 AC        2000          R - 30 - Average         3          2               0            5  1STY - 1 Story Building      1,675     1,675         1,647                      0          644  $45,000.00     $141,900.00     02/15/2000  $247,055.00         $84.72\n",
       "3  1449 AMBER WAVE DR       63366    PATRIOTS LANDING #1    Fort Zumwalt  0.2000 AC        2000          R - 30 - Average         3          2               0            5  1STY - 1 Story Building      1,474     1,474         1,450                      0          399  $45,000.00           $0.00     03/21/2001  $216,745.00          $0.00\n",
       "4       9 FOLK ART CT       63366    PATRIOTS LANDING #1    Fort Zumwalt  0.1700 AC        2000          R - 30 - Average         3          2               0            5  1STY - 1 Story Building      1,421     1,421         1,361                      0          400  $45,000.00     $128,450.00     07/10/2000  $212,616.00         $90.39"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_new = df_new[~((df_new.QUALITY_CODE == 'C - Not used') | (df_new.LAST_SALE_PRICE.isnull()))].copy().reset_index(drop=True) # Removing any non residential data\n",
    "df_backup = df_new.copy()\n",
    "df_new.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def convert_to_float(value):\n",
    "  try:\n",
    "    return float(value)\n",
    "  except:\n",
    "    return None"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\sivac\\miniconda3\\envs\\myDefaultEnv\\lib\\site-packages\\ipykernel_launcher.py:10: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  # Remove the CWD from sys.path while we load stuff.\n",
      "C:\\Users\\sivac\\miniconda3\\envs\\myDefaultEnv\\lib\\site-packages\\ipykernel_launcher.py:11: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  # This is added back by InteractiveShellApp.init_path()\n"
     ]
    }
   ],
   "source": [
    "df_new = df_backup.copy()\n",
    "\n",
    "# Removing any rows without sale year\n",
    "\n",
    "df_new = df_new[df_new.LAST_SALE_DATE.notnull()].copy().reset_index(drop=True)\n",
    "\n",
    "\n",
    "# Fixing the lot size\n",
    "df_new['start'] = np.where(df_new.LOT_SIZE.str.find('AC') == -1, df_new.LOT_SIZE.str.split('X') , df_new.LOT_SIZE.str.replace(pat=\" AC\", repl=\"\"))\n",
    "df_new['start'][df_new.LOT_SIZE.str.find('AC') == -1] = df_new['start'][df_new.LOT_SIZE.str.find('AC') == -1].apply(lambda x: list(map(convert_to_float, x)))\n",
    "df_new['start'][df_new.LOT_SIZE.str.find('AC') == -1]  = df_new['start'][df_new.LOT_SIZE.str.find('AC') == -1].apply(lambda x: np.prod(x[0:2])/43560)\n",
    "df_new['LOTS_SIZE'] = pd.to_numeric(df_new['start'])\n",
    "\n",
    "# Fixing all numeric data types\n",
    "df_new['TOTAL_AREA'] = df_new['TOTAL_AREA'].str.replace('[,$]','', regex=True).astype('float32')\n",
    "df_new['BASE_AREA'] = df_new['BASE_AREA'].str.replace('[,$]','', regex=True).astype('float32')\n",
    "df_new['BASEMENT_AREA'] = df_new['BASEMENT_AREA'].str.replace('[,$]','', regex=True).astype('float32')\n",
    "df_new['BASEMENT_FINISHED_AREA'] = df_new['BASEMENT_FINISHED_AREA'].str.replace('[,$]','', regex=True).astype('float32')\n",
    "df_new['PARKING_AREA'] = df_new['PARKING_AREA'].str.replace('[,$]','', regex=True).astype('float32')\n",
    "df_new['LAND_VALUE'] = df_new['LAND_VALUE'].str.replace('[,$]','', regex=True).astype('float64')\n",
    "df_new['LAST_SALE_PRICE'] = df_new['LAST_SALE_PRICE'].str.replace('[,$]','', regex=True).astype('float64')\n",
    "df_new = df_new[df_new['LAST_SALE_PRICE'] != 0].copy().reset_index(drop=True)\n",
    "df_new['MARKET_VALUE'] = df_new['MARKET_VALUE'].str.replace('[,$]','', regex=True).astype('float64')\n",
    "\n",
    "# Extracting features from date values\n",
    "df_new['LAST_SALE_DATE'] = pd.to_datetime(df_new['LAST_SALE_DATE'])\n",
    "\n",
    "# Generating useful information from date values\n",
    "\n",
    "df_new['SALE_DATE_YR'] = df_new['LAST_SALE_DATE'].dt.year\n",
    "df_new['AGE_OF_HOUSE_AT_SALE'] = df_new['SALE_DATE_YR'] - df_new['YEAR_BUILT']\n",
    "df_new['SALE_DATE_MNTH'] = df_new['LAST_SALE_DATE'].dt.month\n",
    "df_new['SALE_DATE_DT'] = df_new['LAST_SALE_DATE'].dt.day\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "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>BEDROOMS</th>\n",
       "      <th>BATHROOMS</th>\n",
       "      <th>HALF_BATHROOMS</th>\n",
       "      <th>TOTAL_ROOMS</th>\n",
       "      <th>TOTAL_AREA</th>\n",
       "      <th>BASE_AREA</th>\n",
       "      <th>BASEMENT_FINISHED_AREA</th>\n",
       "      <th>PARKING_AREA</th>\n",
       "      <th>LAND_VALUE</th>\n",
       "      <th>LAST_SALE_PRICE</th>\n",
       "      <th>MARKET_VALUE</th>\n",
       "      <th>AGE_OF_HOUSE_AT_SALE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1496.0</td>\n",
       "      <td>1496.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>660.0</td>\n",
       "      <td>35000.0</td>\n",
       "      <td>9000.0</td>\n",
       "      <td>210354.0</td>\n",
       "      <td>-1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1675.0</td>\n",
       "      <td>1675.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>644.0</td>\n",
       "      <td>45000.0</td>\n",
       "      <td>141900.0</td>\n",
       "      <td>247055.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1421.0</td>\n",
       "      <td>1421.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>400.0</td>\n",
       "      <td>45000.0</td>\n",
       "      <td>128450.0</td>\n",
       "      <td>212616.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1340.0</td>\n",
       "      <td>1340.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>526.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>123433.0</td>\n",
       "      <td>210286.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1505.0</td>\n",
       "      <td>1505.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>420.0</td>\n",
       "      <td>45000.0</td>\n",
       "      <td>124950.0</td>\n",
       "      <td>228551.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   BEDROOMS  BATHROOMS  HALF_BATHROOMS  TOTAL_ROOMS  TOTAL_AREA  BASE_AREA  BASEMENT_FINISHED_AREA  PARKING_AREA  LAND_VALUE  LAST_SALE_PRICE  MARKET_VALUE  AGE_OF_HOUSE_AT_SALE\n",
       "0       3.0        2.0             0.0          5.0      1496.0     1496.0                     0.0         660.0     35000.0           9000.0      210354.0                  -1.0\n",
       "1       3.0        2.0             0.0          5.0      1675.0     1675.0                     0.0         644.0     45000.0         141900.0      247055.0                   0.0\n",
       "2       3.0        2.0             0.0          5.0      1421.0     1421.0                     0.0         400.0     45000.0         128450.0      212616.0                   0.0\n",
       "3       3.0        2.0             0.0          6.0      1340.0     1340.0                     0.0         526.0         0.0         123433.0      210286.0                   0.0\n",
       "4       3.0        2.0             0.0          5.0      1505.0     1505.0                     0.0         420.0     45000.0         124950.0      228551.0                   0.0"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Removed Items - 'QUALITY_CODE', 'ARCHITECTURAL_STYLE', 'SALE_DATE_MNTH', 'LOTS_SIZE', 'SALE_DATE_YR', 'BASEMENT_AREA',\n",
    "\n",
    "df_new = df_new[[ 'BEDROOMS', 'BATHROOMS','HALF_BATHROOMS', 'TOTAL_ROOMS', 'TOTAL_AREA',\n",
    "          'BASE_AREA',  'BASEMENT_FINISHED_AREA', 'PARKING_AREA','LAND_VALUE', 'LAST_SALE_PRICE','MARKET_VALUE',\n",
    "           'AGE_OF_HOUSE_AT_SALE']].copy().drop_duplicates().reset_index(drop=True)\n",
    "\n",
    "df_new = pd.get_dummies(df_new)\n",
    "df_new.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Transformation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df_new['LAST_SALE_DATE'] = pd.to_datetime(df_new['LAST_SALE_DATE'], format=\"%b %d %Y %I:%M:%S:%f%p\")\n",
    "\n",
    "# # Generating useful information from date values\n",
    "\n",
    "# df_new['SALE_DATE_YR'] = df_new['LAST_SALE_DATE'].dt.year\n",
    "# df_new['AGE_OF_HOUSE_AT_SALE'] = df_new['SALE_DATE_YR'] - df_new['YRBLT']\n",
    "# df_new['SALE_DATE_MNTH'] = df_new['LAST_SALE_DATE'].dt.month\n",
    "# df_new['SALE_DATE_DT'] = df_new['LAST_SALE_DATE'].dt.day\n",
    "# df_new.drop(['LAST_SALE_DATE', 'YRBLT', 'USPF', 'PHYS'], axis=1, inplace=True)\n",
    "# df_new = pd.get_dummies(df_new)\n",
    "# df_new.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(8586, 12)\n",
      "(8586, 12)\n"
     ]
    }
   ],
   "source": [
    "print(df_new.shape)\n",
    "df_new.dropna(inplace=True, axis=0)\n",
    "print(df_new.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Training dataset shape: (6439, 11) | Testing dataset shape: (2147, 11)\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "df_y = df_new['LAST_SALE_PRICE']\n",
    "df_x = df_new.drop('LAST_SALE_PRICE', axis=1)\n",
    "\n",
    "x_train, x_test, y_train, y_test = train_test_split(df_x, df_y, test_size = 0.25, random_state = 2)\n",
    "print(f' Training dataset shape: {x_train.shape} | Testing dataset shape: {x_test.shape}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "        Actual      Predicted\n",
      "2926  287000.0  328999.210491\n",
      "6597  270000.0  238279.511353\n",
      "663   180275.0  279672.756820\n",
      "7309  397000.0  375383.134181\n",
      "6534  279000.0  254796.939242 \n",
      "\n",
      "Mean Absolute Error: 40851.04866531676\n",
      "Mean Squared Error: 3806182209.8847055\n",
      "Root Mean Squared Error: 61694.263994999616\n",
      "R-squared: 0.6838536738243048\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn import metrics\n",
    "\n",
    "model = LinearRegression()\n",
    "model.fit(x_train, y_train)\n",
    "\n",
    "# print(model.intercept_)\n",
    "# print(model.coef_)\n",
    "y_pred = model.predict(x_test)\n",
    "\n",
    "df = pd.DataFrame({'Actual': y_test, 'Predicted': y_pred})\n",
    "print(f'{df.head()} \\n')\n",
    "\n",
    "print('Mean Absolute Error:', metrics.mean_absolute_error(y_test, y_pred))\n",
    "print('Mean Squared Error:', metrics.mean_squared_error(y_test, y_pred))\n",
    "print('Root Mean Squared Error:', np.sqrt(metrics.mean_squared_error(y_test, y_pred)))\n",
    "print('R-squared:', metrics.r2_score(y_test, y_pred, multioutput='uniform_average'))\n",
    "\n",
    "# https://www.analyticsvidhya.com/blog/2021/05/all-you-need-to-know-about-your-first-machine-learning-model-linear-regression/\n",
    "# https://www.analyticsvidhya.com/blog/2021/07/an-introduction-to-linear-regression/\n",
    "# https://www.analyticsvidhya.com/blog/2021/06/getting-started-with-machine-learning%e2%80%8a-%e2%80%8aimplementing-linear-regression-from-scratch/\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Predicting unknown dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\sivac\\miniconda3\\envs\\myDefaultEnv\\lib\\site-packages\\ipykernel_launcher.py:3: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  This is separate from the ipykernel package so we can avoid doing imports until\n",
      "C:\\Users\\sivac\\miniconda3\\envs\\myDefaultEnv\\lib\\site-packages\\ipykernel_launcher.py:4: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  after removing the cwd from sys.path.\n"
     ]
    }
   ],
   "source": [
    "# Fixing the lot size\n",
    "unseen_data['start'] = np.where(unseen_data.LOT_SIZE.str.find('AC') == -1, unseen_data.LOT_SIZE.str.split('X') , unseen_data.LOT_SIZE.str.replace(pat=\" AC\", repl=\"\"))\n",
    "unseen_data['start'][unseen_data.LOT_SIZE.str.find('AC') == -1] = unseen_data['start'][unseen_data.LOT_SIZE.str.find('AC') == -1].apply(lambda x: list(map(convert_to_float, x)))\n",
    "unseen_data['start'][unseen_data.LOT_SIZE.str.find('AC') == -1]  = unseen_data['start'][unseen_data.LOT_SIZE.str.find('AC') == -1].apply(lambda x: np.prod(x[0:2])/43560)\n",
    "unseen_data['LOTS_SIZE'] = pd.to_numeric(unseen_data['start'])\n",
    "\n",
    "# Fixing all numeric data types\n",
    "unseen_data['TOTAL_AREA'] = unseen_data['TOTAL_AREA'].str.replace('[,$]','', regex=True).astype('float32')\n",
    "unseen_data['BASE_AREA'] = unseen_data['BASE_AREA'].str.replace('[,$]','', regex=True).astype('float32')\n",
    "unseen_data['BASEMENT_AREA'] = unseen_data['BASEMENT_AREA'].str.replace('[,$]','', regex=True).astype('float32')\n",
    "#unseen_data['BASEMENT_FINISHED_AREA'] = unseen_data['BASEMENT_FINISHED_AREA'].str.replace('[,$]','', regex=True).astype('float32')\n",
    "# unseen_data['PARKING_AREA'] = unseen_data['PARKING_AREA'].str.replace('[,$]','', regex=True).astype('float32')\n",
    "unseen_data['LAND_VALUE'] = unseen_data['LAND_VALUE'].str.replace('[,$]','', regex=True).astype('float64')\n",
    "unseen_data['LAST_SALE_PRICE'] = unseen_data['LAST_SALE_PRICE'].str.replace('[,$]','', regex=True).astype('float64')\n",
    "#unseen_data = unseen_data[unseen_data['LAST_SALE_PRICE'] != 0].copy().reset_index(drop=True)\n",
    "unseen_data['MARKET_VALUE'] = unseen_data['MARKET_VALUE'].str.replace('[,$]','', regex=True).astype('float64')\n",
    "\n",
    "# Extracting features from date values\n",
    "unseen_data['LAST_SALE_DATE'] = pd.to_datetime(unseen_data['LAST_SALE_DATE'])\n",
    "\n",
    "# Generating useful information from date values\n",
    "\n",
    "unseen_data['SALE_DATE_YR'] = dt.date.today().year\n",
    "unseen_data['AGE_OF_HOUSE_AT_SALE'] = (dt.date.today().year)  - (unseen_data['YEAR_BUILT'])\n",
    "unseen_data['SALE_DATE_MNTH'] = dt.date.today().month\n",
    "unseen_data['SALE_DATE_DT'] = dt.date.today().day\n",
    "\n",
    "\n",
    "unseen_data = unseen_data[[ 'BEDROOMS', 'BATHROOMS','HALF_BATHROOMS', 'TOTAL_ROOMS', 'TOTAL_AREA',\n",
    "          'BASE_AREA',  'BASEMENT_FINISHED_AREA', 'PARKING_AREA','LAND_VALUE', 'MARKET_VALUE',\n",
    "           'AGE_OF_HOUSE_AT_SALE']].copy().drop_duplicates().reset_index(drop=True)\n",
    "\n",
    "unseen_data = pd.get_dummies(unseen_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "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>BEDROOMS</th>\n",
       "      <th>BATHROOMS</th>\n",
       "      <th>HALF_BATHROOMS</th>\n",
       "      <th>TOTAL_ROOMS</th>\n",
       "      <th>TOTAL_AREA</th>\n",
       "      <th>BASE_AREA</th>\n",
       "      <th>BASEMENT_FINISHED_AREA</th>\n",
       "      <th>PARKING_AREA</th>\n",
       "      <th>LAND_VALUE</th>\n",
       "      <th>MARKET_VALUE</th>\n",
       "      <th>AGE_OF_HOUSE_AT_SALE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>1904.0</td>\n",
       "      <td>1904.0</td>\n",
       "      <td>0</td>\n",
       "      <td>400</td>\n",
       "      <td>40002.0</td>\n",
       "      <td>300645.0</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>1684.0</td>\n",
       "      <td>1684.0</td>\n",
       "      <td>0</td>\n",
       "      <td>400</td>\n",
       "      <td>40002.0</td>\n",
       "      <td>267114.0</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   BEDROOMS  BATHROOMS  HALF_BATHROOMS  TOTAL_ROOMS  TOTAL_AREA  BASE_AREA  BASEMENT_FINISHED_AREA  PARKING_AREA  LAND_VALUE  MARKET_VALUE  AGE_OF_HOUSE_AT_SALE\n",
       "0         4          2               0            6      1904.0     1904.0                       0           400     40002.0      300645.0                    17\n",
       "1         3          2               0            6      1684.0     1684.0                       0           400     40002.0      267114.0                    17"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#unseen_data = unseen_data_backup.copy()\n",
    "unseen_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [],
   "source": [
    "unseen_data_backup = unseen_data.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "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>Address</th>\n",
       "      <th>Predicted</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1132 SPRING ORCHARD DR</td>\n",
       "      <td>287388.515082</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1124 SPRING ORCHARD DR</td>\n",
       "      <td>262274.193667</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  Address      Predicted\n",
       "0  1132 SPRING ORCHARD DR  287388.515082\n",
       "1  1124 SPRING ORCHARD DR  262274.193667"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predicted_price = model.predict(unseen_data)\n",
    "df = pd.DataFrame({'Address': address, 'Predicted': predicted_price})\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Visualizations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1c338221f48>"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x648 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "numericCols = df_new.select_dtypes(include=['float64','int64'])\n",
    "\n",
    "# Calculate Correlation\n",
    "corr = numericCols.corr()\n",
    "# Generate a mask for the upper triangle\n",
    "mask = np.zeros_like(corr, dtype=np.bool)\n",
    "mask[np.triu_indices_from(mask)] = True\n",
    "\n",
    "# Set up the matplotlib figure\n",
    "f, ax = plt.subplots(figsize=(11, 9))\n",
    "\n",
    "# Generate a custom diverging colormap\n",
    "cmap = sns.diverging_palette(220, 10, as_cmap=True)\n",
    "\n",
    "# Draw the heatmap with the mask and correct aspect ratio\n",
    "sns.heatmap(corr, mask=mask, cmap=cmap, vmax=.3, center=0,\n",
    "            square=True, linewidths=.5, cbar_kws={\"shrink\": .5})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}