house-prices-advanced-regression-techniques.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# House Price Prediction - Regression Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import csv\n",
    "\n",
    "import scipy\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from xgboost import XGBRegressor\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn import metrics\n",
    "import matplotlib.pyplot as plt\n",
    "from IPython.display import FileLink, FileLinks, display, HTML\n",
    "from sklearn.model_selection import GridSearchCV , KFold, cross_val_score, train_test_split\n",
    "from scipy import stats\n",
    "from scipy.stats import norm, skew #for some statistics\n",
    "from scipy.special import boxcox1p\n",
    "import seaborn as sns\n",
    "color = sns.color_palette()\n",
    "sns.set_style('darkgrid')\n",
    "\n",
    "from sklearn.linear_model import ElasticNet, Lasso,  BayesianRidge, LassoLarsIC\n",
    "from sklearn.ensemble import RandomForestRegressor,  GradientBoostingRegressor\n",
    "from sklearn.kernel_ridge import KernelRidge\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.preprocessing import RobustScaler\n",
    "from sklearn.base import BaseEstimator, TransformerMixin, RegressorMixin, clone\n",
    "from sklearn.model_selection import KFold, cross_val_score, train_test_split\n",
    "from sklearn.metrics import mean_squared_error\n",
    "import lightgbm as lgb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set these options to display all the columns\n",
    "pd.set_option('display.max_rows', 500)\n",
    "pd.set_option('display.max_columns', 500)\n",
    "pd.set_option('display.width', 1000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def write_csv(output_fn, ids, test_y):\n",
    "    with open(output_fn, \"w\") as fp:\n",
    "        writer = csv.writer(fp)\n",
    "        writer.writerow(['Id', 'SalePrice'])\n",
    "        for _id, y in zip(ids, test_y):\n",
    "            writer.writerow([_id, y])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "base_path = \"./data\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def train_model(train_X, train_y, **kwargs):\n",
    "    model = XGBRegressor(**kwargs)\n",
    "    model.fit(train_X, train_y, verbose=True)\n",
    "    return model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "=Baseline model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "testdf = pd.read_csv(f\"{base_path}/test.csv\")\n",
    "traindf = pd.read_csv(f\"{base_path}/train.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Check for null values and delete them before we proceed. We don't have any in our case"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "traindf['SalePrice'].isnull().values.any()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Understand some basics about the data. For convenience i split them into two parts "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Id</th>\n",
       "      <th>MSSubClass</th>\n",
       "      <th>LotFrontage</th>\n",
       "      <th>LotArea</th>\n",
       "      <th>OverallQual</th>\n",
       "      <th>OverallCond</th>\n",
       "      <th>YearBuilt</th>\n",
       "      <th>YearRemodAdd</th>\n",
       "      <th>MasVnrArea</th>\n",
       "      <th>BsmtFinSF1</th>\n",
       "      <th>BsmtFinSF2</th>\n",
       "      <th>BsmtUnfSF</th>\n",
       "      <th>TotalBsmtSF</th>\n",
       "      <th>1stFlrSF</th>\n",
       "      <th>2ndFlrSF</th>\n",
       "      <th>LowQualFinSF</th>\n",
       "      <th>GrLivArea</th>\n",
       "      <th>BsmtFullBath</th>\n",
       "      <th>BsmtHalfBath</th>\n",
       "      <th>FullBath</th>\n",
       "      <th>HalfBath</th>\n",
       "      <th>BedroomAbvGr</th>\n",
       "      <th>KitchenAbvGr</th>\n",
       "      <th>TotRmsAbvGrd</th>\n",
       "      <th>Fireplaces</th>\n",
       "      <th>GarageYrBlt</th>\n",
       "      <th>GarageCars</th>\n",
       "      <th>GarageArea</th>\n",
       "      <th>WoodDeckSF</th>\n",
       "      <th>OpenPorchSF</th>\n",
       "      <th>EnclosedPorch</th>\n",
       "      <th>3SsnPorch</th>\n",
       "      <th>ScreenPorch</th>\n",
       "      <th>PoolArea</th>\n",
       "      <th>MiscVal</th>\n",
       "      <th>MoSold</th>\n",
       "      <th>YrSold</th>\n",
       "      <th>SalePrice</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1201.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1452.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1379.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "      <td>1460.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>730.500000</td>\n",
       "      <td>56.897260</td>\n",
       "      <td>70.049958</td>\n",
       "      <td>10516.828082</td>\n",
       "      <td>6.099315</td>\n",
       "      <td>5.575342</td>\n",
       "      <td>1971.267808</td>\n",
       "      <td>1984.865753</td>\n",
       "      <td>103.685262</td>\n",
       "      <td>443.639726</td>\n",
       "      <td>46.549315</td>\n",
       "      <td>567.240411</td>\n",
       "      <td>1057.429452</td>\n",
       "      <td>1162.626712</td>\n",
       "      <td>346.992466</td>\n",
       "      <td>5.844521</td>\n",
       "      <td>1515.463699</td>\n",
       "      <td>0.425342</td>\n",
       "      <td>0.057534</td>\n",
       "      <td>1.565068</td>\n",
       "      <td>0.382877</td>\n",
       "      <td>2.866438</td>\n",
       "      <td>1.046575</td>\n",
       "      <td>6.517808</td>\n",
       "      <td>0.613014</td>\n",
       "      <td>1978.506164</td>\n",
       "      <td>1.767123</td>\n",
       "      <td>472.980137</td>\n",
       "      <td>94.244521</td>\n",
       "      <td>46.660274</td>\n",
       "      <td>21.954110</td>\n",
       "      <td>3.409589</td>\n",
       "      <td>15.060959</td>\n",
       "      <td>2.758904</td>\n",
       "      <td>43.489041</td>\n",
       "      <td>6.321918</td>\n",
       "      <td>2007.815753</td>\n",
       "      <td>180921.195890</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>421.610009</td>\n",
       "      <td>42.300571</td>\n",
       "      <td>24.284752</td>\n",
       "      <td>9981.264932</td>\n",
       "      <td>1.382997</td>\n",
       "      <td>1.112799</td>\n",
       "      <td>30.202904</td>\n",
       "      <td>20.645407</td>\n",
       "      <td>181.066207</td>\n",
       "      <td>456.098091</td>\n",
       "      <td>161.319273</td>\n",
       "      <td>441.866955</td>\n",
       "      <td>438.705324</td>\n",
       "      <td>386.587738</td>\n",
       "      <td>436.528436</td>\n",
       "      <td>48.623081</td>\n",
       "      <td>525.480383</td>\n",
       "      <td>0.518911</td>\n",
       "      <td>0.238753</td>\n",
       "      <td>0.550916</td>\n",
       "      <td>0.502885</td>\n",
       "      <td>0.815778</td>\n",
       "      <td>0.220338</td>\n",
       "      <td>1.625393</td>\n",
       "      <td>0.644666</td>\n",
       "      <td>24.689725</td>\n",
       "      <td>0.747315</td>\n",
       "      <td>213.804841</td>\n",
       "      <td>125.338794</td>\n",
       "      <td>66.256028</td>\n",
       "      <td>61.119149</td>\n",
       "      <td>29.317331</td>\n",
       "      <td>55.757415</td>\n",
       "      <td>40.177307</td>\n",
       "      <td>496.123024</td>\n",
       "      <td>2.703626</td>\n",
       "      <td>1.328095</td>\n",
       "      <td>79442.502883</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>21.000000</td>\n",
       "      <td>1300.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1872.000000</td>\n",
       "      <td>1950.000000</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>334.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>334.000000</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>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1900.000000</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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2006.000000</td>\n",
       "      <td>34900.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>365.750000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>59.000000</td>\n",
       "      <td>7553.500000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>1954.000000</td>\n",
       "      <td>1967.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>223.000000</td>\n",
       "      <td>795.750000</td>\n",
       "      <td>882.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1129.500000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1961.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>334.500000</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>0.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>2007.000000</td>\n",
       "      <td>129975.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>730.500000</td>\n",
       "      <td>50.000000</td>\n",
       "      <td>69.000000</td>\n",
       "      <td>9478.500000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>1973.000000</td>\n",
       "      <td>1994.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>383.500000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>477.500000</td>\n",
       "      <td>991.500000</td>\n",
       "      <td>1087.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1464.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1980.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>480.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>25.000000</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>6.000000</td>\n",
       "      <td>2008.000000</td>\n",
       "      <td>163000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1095.250000</td>\n",
       "      <td>70.000000</td>\n",
       "      <td>80.000000</td>\n",
       "      <td>11601.500000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>2004.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>712.250000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>808.000000</td>\n",
       "      <td>1298.250000</td>\n",
       "      <td>1391.250000</td>\n",
       "      <td>728.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1776.750000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2002.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>576.000000</td>\n",
       "      <td>168.000000</td>\n",
       "      <td>68.000000</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>8.000000</td>\n",
       "      <td>2009.000000</td>\n",
       "      <td>214000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1460.000000</td>\n",
       "      <td>190.000000</td>\n",
       "      <td>313.000000</td>\n",
       "      <td>215245.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>2010.000000</td>\n",
       "      <td>2010.000000</td>\n",
       "      <td>1600.000000</td>\n",
       "      <td>5644.000000</td>\n",
       "      <td>1474.000000</td>\n",
       "      <td>2336.000000</td>\n",
       "      <td>6110.000000</td>\n",
       "      <td>4692.000000</td>\n",
       "      <td>2065.000000</td>\n",
       "      <td>572.000000</td>\n",
       "      <td>5642.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2010.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>1418.000000</td>\n",
       "      <td>857.000000</td>\n",
       "      <td>547.000000</td>\n",
       "      <td>552.000000</td>\n",
       "      <td>508.000000</td>\n",
       "      <td>480.000000</td>\n",
       "      <td>738.000000</td>\n",
       "      <td>15500.000000</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>2010.000000</td>\n",
       "      <td>755000.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Id   MSSubClass  LotFrontage        LotArea  OverallQual  OverallCond    YearBuilt  YearRemodAdd   MasVnrArea   BsmtFinSF1   BsmtFinSF2    BsmtUnfSF  TotalBsmtSF     1stFlrSF     2ndFlrSF  LowQualFinSF    GrLivArea  BsmtFullBath  BsmtHalfBath     FullBath     HalfBath  BedroomAbvGr  KitchenAbvGr  TotRmsAbvGrd   Fireplaces  GarageYrBlt   GarageCars   GarageArea   WoodDeckSF  OpenPorchSF  EnclosedPorch    3SsnPorch  ScreenPorch     PoolArea       MiscVal       MoSold       YrSold      SalePrice\n",
       "count  1460.000000  1460.000000  1201.000000    1460.000000  1460.000000  1460.000000  1460.000000   1460.000000  1452.000000  1460.000000  1460.000000  1460.000000  1460.000000  1460.000000  1460.000000   1460.000000  1460.000000   1460.000000   1460.000000  1460.000000  1460.000000   1460.000000   1460.000000   1460.000000  1460.000000  1379.000000  1460.000000  1460.000000  1460.000000  1460.000000    1460.000000  1460.000000  1460.000000  1460.000000   1460.000000  1460.000000  1460.000000    1460.000000\n",
       "mean    730.500000    56.897260    70.049958   10516.828082     6.099315     5.575342  1971.267808   1984.865753   103.685262   443.639726    46.549315   567.240411  1057.429452  1162.626712   346.992466      5.844521  1515.463699      0.425342      0.057534     1.565068     0.382877      2.866438      1.046575      6.517808     0.613014  1978.506164     1.767123   472.980137    94.244521    46.660274      21.954110     3.409589    15.060959     2.758904     43.489041     6.321918  2007.815753  180921.195890\n",
       "std     421.610009    42.300571    24.284752    9981.264932     1.382997     1.112799    30.202904     20.645407   181.066207   456.098091   161.319273   441.866955   438.705324   386.587738   436.528436     48.623081   525.480383      0.518911      0.238753     0.550916     0.502885      0.815778      0.220338      1.625393     0.644666    24.689725     0.747315   213.804841   125.338794    66.256028      61.119149    29.317331    55.757415    40.177307    496.123024     2.703626     1.328095   79442.502883\n",
       "min       1.000000    20.000000    21.000000    1300.000000     1.000000     1.000000  1872.000000   1950.000000     0.000000     0.000000     0.000000     0.000000     0.000000   334.000000     0.000000      0.000000   334.000000      0.000000      0.000000     0.000000     0.000000      0.000000      0.000000      2.000000     0.000000  1900.000000     0.000000     0.000000     0.000000     0.000000       0.000000     0.000000     0.000000     0.000000      0.000000     1.000000  2006.000000   34900.000000\n",
       "25%     365.750000    20.000000    59.000000    7553.500000     5.000000     5.000000  1954.000000   1967.000000     0.000000     0.000000     0.000000   223.000000   795.750000   882.000000     0.000000      0.000000  1129.500000      0.000000      0.000000     1.000000     0.000000      2.000000      1.000000      5.000000     0.000000  1961.000000     1.000000   334.500000     0.000000     0.000000       0.000000     0.000000     0.000000     0.000000      0.000000     5.000000  2007.000000  129975.000000\n",
       "50%     730.500000    50.000000    69.000000    9478.500000     6.000000     5.000000  1973.000000   1994.000000     0.000000   383.500000     0.000000   477.500000   991.500000  1087.000000     0.000000      0.000000  1464.000000      0.000000      0.000000     2.000000     0.000000      3.000000      1.000000      6.000000     1.000000  1980.000000     2.000000   480.000000     0.000000    25.000000       0.000000     0.000000     0.000000     0.000000      0.000000     6.000000  2008.000000  163000.000000\n",
       "75%    1095.250000    70.000000    80.000000   11601.500000     7.000000     6.000000  2000.000000   2004.000000   166.000000   712.250000     0.000000   808.000000  1298.250000  1391.250000   728.000000      0.000000  1776.750000      1.000000      0.000000     2.000000     1.000000      3.000000      1.000000      7.000000     1.000000  2002.000000     2.000000   576.000000   168.000000    68.000000       0.000000     0.000000     0.000000     0.000000      0.000000     8.000000  2009.000000  214000.000000\n",
       "max    1460.000000   190.000000   313.000000  215245.000000    10.000000     9.000000  2010.000000   2010.000000  1600.000000  5644.000000  1474.000000  2336.000000  6110.000000  4692.000000  2065.000000    572.000000  5642.000000      3.000000      2.000000     3.000000     2.000000      8.000000      3.000000     14.000000     3.000000  2010.000000     4.000000  1418.000000   857.000000   547.000000     552.000000   508.000000   480.000000   738.000000  15500.000000    12.000000  2010.000000  755000.000000"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "traindf.select_dtypes(exclude=['object']).describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "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>MSZoning</th>\n",
       "      <th>Street</th>\n",
       "      <th>Alley</th>\n",
       "      <th>LotShape</th>\n",
       "      <th>LandContour</th>\n",
       "      <th>Utilities</th>\n",
       "      <th>LotConfig</th>\n",
       "      <th>LandSlope</th>\n",
       "      <th>Neighborhood</th>\n",
       "      <th>Condition1</th>\n",
       "      <th>Condition2</th>\n",
       "      <th>BldgType</th>\n",
       "      <th>HouseStyle</th>\n",
       "      <th>RoofStyle</th>\n",
       "      <th>RoofMatl</th>\n",
       "      <th>Exterior1st</th>\n",
       "      <th>Exterior2nd</th>\n",
       "      <th>MasVnrType</th>\n",
       "      <th>ExterQual</th>\n",
       "      <th>ExterCond</th>\n",
       "      <th>Foundation</th>\n",
       "      <th>BsmtQual</th>\n",
       "      <th>BsmtCond</th>\n",
       "      <th>BsmtExposure</th>\n",
       "      <th>BsmtFinType1</th>\n",
       "      <th>BsmtFinType2</th>\n",
       "      <th>Heating</th>\n",
       "      <th>HeatingQC</th>\n",
       "      <th>CentralAir</th>\n",
       "      <th>Electrical</th>\n",
       "      <th>KitchenQual</th>\n",
       "      <th>Functional</th>\n",
       "      <th>FireplaceQu</th>\n",
       "      <th>GarageType</th>\n",
       "      <th>GarageFinish</th>\n",
       "      <th>GarageQual</th>\n",
       "      <th>GarageCond</th>\n",
       "      <th>PavedDrive</th>\n",
       "      <th>PoolQC</th>\n",
       "      <th>Fence</th>\n",
       "      <th>MiscFeature</th>\n",
       "      <th>SaleType</th>\n",
       "      <th>SaleCondition</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1460</td>\n",
       "      <td>1460</td>\n",
       "      <td>91</td>\n",
       "      <td>1460</td>\n",
       "      <td>1460</td>\n",
       "      <td>1460</td>\n",
       "      <td>1460</td>\n",
       "      <td>1460</td>\n",
       "      <td>1460</td>\n",
       "      <td>1460</td>\n",
       "      <td>1460</td>\n",
       "      <td>1460</td>\n",
       "      <td>1460</td>\n",
       "      <td>1460</td>\n",
       "      <td>1460</td>\n",
       "      <td>1460</td>\n",
       "      <td>1460</td>\n",
       "      <td>1452</td>\n",
       "      <td>1460</td>\n",
       "      <td>1460</td>\n",
       "      <td>1460</td>\n",
       "      <td>1423</td>\n",
       "      <td>1423</td>\n",
       "      <td>1422</td>\n",
       "      <td>1423</td>\n",
       "      <td>1422</td>\n",
       "      <td>1460</td>\n",
       "      <td>1460</td>\n",
       "      <td>1460</td>\n",
       "      <td>1459</td>\n",
       "      <td>1460</td>\n",
       "      <td>1460</td>\n",
       "      <td>770</td>\n",
       "      <td>1379</td>\n",
       "      <td>1379</td>\n",
       "      <td>1379</td>\n",
       "      <td>1379</td>\n",
       "      <td>1460</td>\n",
       "      <td>7</td>\n",
       "      <td>281</td>\n",
       "      <td>54</td>\n",
       "      <td>1460</td>\n",
       "      <td>1460</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>25</td>\n",
       "      <td>9</td>\n",
       "      <td>8</td>\n",
       "      <td>5</td>\n",
       "      <td>8</td>\n",
       "      <td>6</td>\n",
       "      <td>8</td>\n",
       "      <td>15</td>\n",
       "      <td>16</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>9</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>RL</td>\n",
       "      <td>Pave</td>\n",
       "      <td>Grvl</td>\n",
       "      <td>Reg</td>\n",
       "      <td>Lvl</td>\n",
       "      <td>AllPub</td>\n",
       "      <td>Inside</td>\n",
       "      <td>Gtl</td>\n",
       "      <td>NAmes</td>\n",
       "      <td>Norm</td>\n",
       "      <td>Norm</td>\n",
       "      <td>1Fam</td>\n",
       "      <td>1Story</td>\n",
       "      <td>Gable</td>\n",
       "      <td>CompShg</td>\n",
       "      <td>VinylSd</td>\n",
       "      <td>VinylSd</td>\n",
       "      <td>None</td>\n",
       "      <td>TA</td>\n",
       "      <td>TA</td>\n",
       "      <td>PConc</td>\n",
       "      <td>TA</td>\n",
       "      <td>TA</td>\n",
       "      <td>No</td>\n",
       "      <td>Unf</td>\n",
       "      <td>Unf</td>\n",
       "      <td>GasA</td>\n",
       "      <td>Ex</td>\n",
       "      <td>Y</td>\n",
       "      <td>SBrkr</td>\n",
       "      <td>TA</td>\n",
       "      <td>Typ</td>\n",
       "      <td>Gd</td>\n",
       "      <td>Attchd</td>\n",
       "      <td>Unf</td>\n",
       "      <td>TA</td>\n",
       "      <td>TA</td>\n",
       "      <td>Y</td>\n",
       "      <td>Gd</td>\n",
       "      <td>MnPrv</td>\n",
       "      <td>Shed</td>\n",
       "      <td>WD</td>\n",
       "      <td>Normal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>1151</td>\n",
       "      <td>1454</td>\n",
       "      <td>50</td>\n",
       "      <td>925</td>\n",
       "      <td>1311</td>\n",
       "      <td>1459</td>\n",
       "      <td>1052</td>\n",
       "      <td>1382</td>\n",
       "      <td>225</td>\n",
       "      <td>1260</td>\n",
       "      <td>1445</td>\n",
       "      <td>1220</td>\n",
       "      <td>726</td>\n",
       "      <td>1141</td>\n",
       "      <td>1434</td>\n",
       "      <td>515</td>\n",
       "      <td>504</td>\n",
       "      <td>864</td>\n",
       "      <td>906</td>\n",
       "      <td>1282</td>\n",
       "      <td>647</td>\n",
       "      <td>649</td>\n",
       "      <td>1311</td>\n",
       "      <td>953</td>\n",
       "      <td>430</td>\n",
       "      <td>1256</td>\n",
       "      <td>1428</td>\n",
       "      <td>741</td>\n",
       "      <td>1365</td>\n",
       "      <td>1334</td>\n",
       "      <td>735</td>\n",
       "      <td>1360</td>\n",
       "      <td>380</td>\n",
       "      <td>870</td>\n",
       "      <td>605</td>\n",
       "      <td>1311</td>\n",
       "      <td>1326</td>\n",
       "      <td>1340</td>\n",
       "      <td>3</td>\n",
       "      <td>157</td>\n",
       "      <td>49</td>\n",
       "      <td>1267</td>\n",
       "      <td>1198</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       MSZoning Street Alley LotShape LandContour Utilities LotConfig LandSlope Neighborhood Condition1 Condition2 BldgType HouseStyle RoofStyle RoofMatl Exterior1st Exterior2nd MasVnrType ExterQual ExterCond Foundation BsmtQual BsmtCond BsmtExposure BsmtFinType1 BsmtFinType2 Heating HeatingQC CentralAir Electrical KitchenQual Functional FireplaceQu GarageType GarageFinish GarageQual GarageCond PavedDrive PoolQC  Fence MiscFeature SaleType SaleCondition\n",
       "count      1460   1460    91     1460        1460      1460      1460      1460         1460       1460       1460     1460       1460      1460     1460        1460        1460       1452      1460      1460       1460     1423     1423         1422         1423         1422    1460      1460       1460       1459        1460       1460         770       1379         1379       1379       1379       1460      7    281          54     1460          1460\n",
       "unique        5      2     2        4           4         2         5         3           25          9          8        5          8         6        8          15          16          4         4         5          6        4        4            4            6            6       6         5          2          5           4          7           5          6            3          5          5          3      3      4           4        9             6\n",
       "top          RL   Pave  Grvl      Reg         Lvl    AllPub    Inside       Gtl        NAmes       Norm       Norm     1Fam     1Story     Gable  CompShg     VinylSd     VinylSd       None        TA        TA      PConc       TA       TA           No          Unf          Unf    GasA        Ex          Y      SBrkr          TA        Typ          Gd     Attchd          Unf         TA         TA          Y     Gd  MnPrv        Shed       WD        Normal\n",
       "freq       1151   1454    50      925        1311      1459      1052      1382          225       1260       1445     1220        726      1141     1434         515         504        864       906      1282        647      649     1311          953          430         1256    1428       741       1365       1334         735       1360         380        870          605       1311       1326       1340      3    157          49     1267          1198"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.select_dtypes(include=['object']).describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Convert the string data into categorical. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "traindf[traindf.select_dtypes(['object']).columns] = traindf.select_dtypes(['object']).apply(lambda x: x.astype('category'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = traindf.SalePrice\n",
    "X = traindf.drop(['SalePrice'], axis=1).select_dtypes(exclude=['category'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = train_model(X, y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "On our first try lets remove all the categorical data and try. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_X = testdf.select_dtypes(exclude=['object'])\n",
    "test_y = model.predict(test_X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<a href='house_price_predict_1.csv' target='_blank'>house_price_predict_1.csv</a><br>"
      ],
      "text/plain": [
       "/media/disk6tb/myfiles/fastai_v1/notebooks/house_price_predict_1.csv"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "write_csv(\"house_price_predict_1.csv\", test_X['Id'], test_y)\n",
    "FileLink(\"house_price_predict_1.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After submission, we got about 0.14179 score with our test data. We are approximately at 50 percentile ! position. You might get a different score because of randomization. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def show_missing_data(df):\n",
    "    total = df.isnull().sum().sort_values(ascending=False)\n",
    "    percent = (df.isnull().sum()/df.isnull().count()).sort_values(ascending=False)\n",
    "    missing_data = pd.concat([total, percent], axis=1)\n",
    "    display(missing_data.head(30))\n",
    "\n",
    "#show_missing_data(traindf)\n",
    "#show_missing_data(testdf)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## More serious model :)\n",
    "You can see that data is missing in some columns. These columns are different for test vs train. We need to come up with some strategy to replace the null fields. For convenience we will merge the train and test sets together and apply the same transformations for both the datasets."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "DataFrame Shape: (2917, 80)\n"
     ]
    }
   ],
   "source": [
    "testdf = pd.read_csv(f\"{base_path}/test.csv\")\n",
    "traindf = pd.read_csv(f\"{base_path}/train.csv\")\n",
    "\n",
    "# Remove outliers. \n",
    "traindf = traindf.drop(traindf[traindf['Id'] == 1299].index)\n",
    "traindf = traindf.drop(traindf[traindf['Id'] == 524].index)\n",
    "\n",
    "# The assumption here is, it is easier to model linear data. If you plot our sales price, the distribution is skewed\n",
    "# We do log transformation for making it linear\n",
    "traindf[\"SalePrice\"] = np.log1p(traindf[\"SalePrice\"])\n",
    "\n",
    "y = traindf.SalePrice\n",
    "df = pd.concat([traindf.drop(['SalePrice'], axis=1), testdf], axis=0)\n",
    "\n",
    "print(f\"DataFrame Shape: {df.shape}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x13136ba58>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x648 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Correlation matrix. Higher the numberm, the more they are correlated. You can also find some colinear columns. \n",
    "# For example check, GarageCars and GarageArea. Also TotalBsmtSF and 1stFlrSF\n",
    "corrmat = traindf.corr()\n",
    "plt.subplots(figsize=(12,9))\n",
    "sns.heatmap(corrmat, vmax=0.9, square=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "DataFrame Shape Before modifying: (2917, 80)\n",
      "DataFrame Shape After modifying: (2917, 80)\n",
      "DataFrame Shape after get_dummies: (2917, 218)\n"
     ]
    }
   ],
   "source": [
    "rating = {\n",
    "    \"Ex\": 5,\n",
    "    \"Gd\": 4,\n",
    "    \"TA\": 3,\n",
    "    \"Fa\": 2,\n",
    "    \"Po\": 1,\n",
    "    \"None\": 0,\n",
    "}\n",
    "\n",
    "basement_type = {\n",
    "    \"GLQ\": 6,\n",
    "    \"ALQ\": 5,\n",
    "    \"BLQ\": 4,\n",
    "    \"Rec\": 3,\n",
    "    \"LwQ\": 2,\n",
    "    \"Unf\": 1,\n",
    "    \"None\": 0,\n",
    "}\n",
    "\n",
    "fence_type = {\n",
    "    \"GdPrv\": 4,\n",
    "    \"MnPrv\": 1,\n",
    "    \"GdWo\": 3,\n",
    "    \"MnWw\": 2,\n",
    "    \"NA\": 0,\n",
    "    \"None\": 0,   \n",
    "}\n",
    "\n",
    "# This is the single most important transformation in the whole excercise. We need to provide sensible values to the \n",
    "# ordinal transformation. For example regular shaped plots are more prefered than irregular shaped plots. Go \n",
    "# through the data description multiple times and provide sensible values.\n",
    "MAPPING = {\n",
    "    \"PoolQC\":  rating,\n",
    "    \"FireplaceQu\": rating,\n",
    "    \"GarageCond\": rating,\n",
    "    \"GarageQual\": rating,\n",
    "    \"BsmtQual\": rating,\n",
    "    \"BsmtCond\": rating,\n",
    "    \"BsmtExposure\": {\n",
    "        \"Gd\": 4,\n",
    "        \"Av\": 3,\n",
    "        \"Mn\": 2,\n",
    "        \"No\": 1,\n",
    "        \"None\": 0,\n",
    "    },\n",
    "    \"KitchenQual\": rating,\n",
    "    \"ExterQual\": rating,\n",
    "    \"ExterCond\": rating,\n",
    "    \"LotShape\": {\n",
    "        \"Reg\": 3,\n",
    "        \"IR1\": 2,\n",
    "        \"IR2\": 1,\n",
    "        \"IR3\": 0,\n",
    "    },\n",
    "    \"BsmtFinType1\": basement_type,\n",
    "    \"BsmtFinType2\": basement_type,\n",
    "    \"MasVnrType\": {\n",
    "        \"Stone\": 2,\n",
    "        \"BrkFace\": 1,\n",
    "        \"BrkCmn\": 0,\n",
    "        \"None\": 0,\n",
    "    },\n",
    "    \"Functional\": {\n",
    "        np.NaN: 7,\n",
    "        \"None\": 7,\n",
    "        \"Sal\": 0,\n",
    "        \"Sev\": 1,\n",
    "        \"Maj2\": 2,\n",
    "        \"Maj1\": 3,\n",
    "        \"Mod\": 4,\n",
    "        \"Min2\": 5,\n",
    "        \"Min1\": 6,\n",
    "        \"Typ\": 7,    \n",
    "    },\n",
    "    \"HeatingQC\": rating,\n",
    "    \"Fence\": fence_type,\n",
    "    \"GarageFinish\": {\n",
    "        \"Fin\": 3,\n",
    "        \"RFn\": 2,\n",
    "        \"Unf\": 1,\n",
    "        \"NA\": 0,\n",
    "        \"None\": 0\n",
    "    },\n",
    "    \"LandSlope\": {\n",
    "        \"Gtl\": 3,\n",
    "        \"Mod\": 2,\n",
    "        \"Sev\": 1,\n",
    "    },\n",
    "    \"PavedDrive\": {\n",
    "        \"Y\": 3,\n",
    "        \"P\": 2,\n",
    "        \"N\": 1\n",
    "    },\n",
    "    \"Street\": {\n",
    "        \"Grvl\": 1,\n",
    "        \"Pave\": 2\n",
    "    },\n",
    "    \"Alley\": {\n",
    "        \"Grvl\": 1,\n",
    "        \"Pave\": 2,\n",
    "    },\n",
    "    \"CentralAir\": {\n",
    "        \"Y\": 2,\n",
    "        \"N\": 1,\n",
    "    },\n",
    "    \"MSSubClass\": {\n",
    "        20: 1,\n",
    "        30: 2,\n",
    "        40: 3,\n",
    "        45: 4,\n",
    "        50: 5,\n",
    "        60: 6,\n",
    "        70: 7,\n",
    "        75: 8,\n",
    "        80: 9,\n",
    "        85: 10,\n",
    "        90: 11,\n",
    "        120: 12,\n",
    "        150: 13,\n",
    "        160: 14,\n",
    "        180: 15,\n",
    "        190: 16,\n",
    "    },\n",
    "    \"YrSold\": {\n",
    "        2006: 1,\n",
    "        2007: 2,\n",
    "        2008: 3,\n",
    "        2009: 4,\n",
    "        2010: 5\n",
    "    }\n",
    "}\n",
    "\n",
    "def encode_data(df, columns, mapping):\n",
    "    for column in columns:\n",
    "        df[column] = df[column].map(mapping[column])\n",
    "\n",
    "cols_ordinal = list(MAPPING.keys())\n",
    "cols_to_drop = ['Utilities']\n",
    "cols_numeric = ['OverallCond', 'MoSold', 'BsmtFullBath', 'BsmtHalfBath', 'BsmtFinSF1', 'BsmtFinSF2', 'BsmtUnfSF', 'MasVnrArea', \"1stFlrSF\", \"2ndFlrSF\", \"Id\", \"YearBuilt\", \"YearRemodAdd\",  \"PoolArea\", \"LotArea\", \"GarageArea\", \"GarageYrBlt\", \"OverallQual\", \"WoodDeckSF\", \"TotRmsAbvGrd\", \"3SsnPorch\", \"BedroomAbvGr\", \"EnclosedPorch\", \"Fireplaces\", \"FullBath\", \"GarageCars\", \"HalfBath\", \"KitchenAbvGr\", \"LotFrontage\", \"LowQualFinSF\", \"MiscVal\",  \"OpenPorchSF\", \"ScreenPorch\", \"TotalBsmtSF\", \"GrLivArea\"]\n",
    "cols_categorical1 =  ['MiscFeature', 'LotConfig', 'GarageType', 'MSZoning', 'Exterior1st', 'Exterior2nd', 'Electrical', 'SaleType', 'SaleCondition']\n",
    "cols_categorical2 = [\"BldgType\", \"Condition1\", \"Condition2\", \"Foundation\", \"Heating\", \"HouseStyle\", \"LandContour\",  \"Neighborhood\", \"RoofMatl\", \"RoofStyle\", \"SaleCondition\"]\n",
    "cols_category = cols_categorical1 + cols_categorical2\n",
    "skewed_cols = ['MiscVal', 'PoolArea', 'LotArea', 'LowQualFinSF', '3SsnPorch', 'KitchenAbvGr', 'BsmtFinSF2', 'EnclosedPorch', 'ScreenPorch', 'BsmtHalfBath', 'MasVnrArea', 'OpenPorchSF', 'WoodDeckSF', '1stFlrSF', 'GrLivArea', 'LotFrontage', 'BsmtFinSF1', 'BsmtUnfSF', '2ndFlrSF']\n",
    "lam = 0.15\n",
    "\n",
    "def transform(df):\n",
    "    neighborhood_lot = dict(df.groupby('Neighborhood')['LotFrontage'].mean())\n",
    "    \n",
    "    # Replace the empty LotFrontage's with neighborhood average\n",
    "    df['LotFrontage'] = df.apply(lambda row: row['LotFrontage'] if pd.notnull(row['LotFrontage']) else neighborhood_lot[row['Neighborhood']], axis=1)\n",
    "    \n",
    "    # Replace null years for GarageYrBlt with the year built\n",
    "    df['GarageYrBlt'] = df.apply(lambda row: row['GarageYrBlt'] if pd.notnull(row['GarageYrBlt']) else row['YearBuilt'], axis=1)\n",
    "    \n",
    "    # Remove unnecessary columns. They have values only for couple of examples\n",
    "    df = df.drop(cols_to_drop, axis=1)\n",
    "\n",
    "    # Ordinal encoding. \n",
    "    encode_data(df, cols_ordinal, MAPPING)\n",
    "\n",
    "    # Missing values: According to data description, the missing value means the most common one ! \n",
    "    df['MSZoning'] = df['MSZoning'].fillna(df['MSZoning'].mode()[0])\n",
    "    df['Electrical'] = df['Electrical'].fillna(df['Electrical'].mode()[0])\n",
    "    df['KitchenQual'] = df['KitchenQual'].fillna(df['KitchenQual'].mode()[0])\n",
    "    df['Exterior1st'] = df['Exterior1st'].fillna(df['Exterior1st'].mode()[0])\n",
    "    df['Exterior2nd'] = df['Exterior2nd'].fillna(df['Exterior2nd'].mode()[0])\n",
    "    df['SaleType'] = df['SaleType'].fillna(df['SaleType'].mode()[0])\n",
    "    \n",
    "    # Replace null values\n",
    "    df[cols_category] = df[cols_category].fillna('NA')\n",
    "    df[cols_numeric] = df[cols_numeric].fillna(0)\n",
    "    df[cols_ordinal] = df[cols_ordinal].fillna(0)\n",
    "\n",
    "    # Convert some to categorical. This helps because of limited data. \n",
    "    #df[cols_convert_category] = df[cols_convert_category].astype(str)\n",
    "    \n",
    "    # Convert others to corresponding type\n",
    "    df[cols_ordinal] = df[cols_ordinal].apply(lambda x: pd.to_numeric(x))\n",
    "    df[cols_category] = df[cols_category].apply(lambda x: x.astype('category'))\n",
    "\n",
    "    for feat in skewed_cols:\n",
    "        df[feat] = boxcox1p(df[feat], lam)\n",
    "        \n",
    "    # Add a new feature\n",
    "    df['TotalSF'] = df['TotalBsmtSF'] + df['1stFlrSF'] + df['2ndFlrSF']\n",
    "\n",
    "    return df\n",
    "\n",
    "print(f\"DataFrame Shape Before modifying: {df.shape}\")\n",
    "df_trans = transform(df)\n",
    "print(f\"DataFrame Shape After modifying: {df_trans.shape}\")\n",
    "\n",
    "df_trans = pd.get_dummies(df_trans)\n",
    "print(f\"DataFrame Shape after get_dummies: {df_trans.shape}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train: (1458, 218), y: 1458, X_test: (1459, 218)\n"
     ]
    }
   ],
   "source": [
    "# Create train and test\n",
    "no_train = len(y)\n",
    "#df_trans = df_trans.drop(['Id'], axis=1)\n",
    "X_train = df_trans[:no_train]\n",
    "X_test = df_trans[no_train:]\n",
    "print(f\"X_train: {X_train.shape}, y: {no_train}, X_test: {X_test.shape}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<a href='house_price_after_tune.csv' target='_blank'>house_price_after_tune.csv</a><br>"
      ],
      "text/plain": [
       "/Users/pulikunt/myfiles/notebooks/house_price_after_tune.csv"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Generate model and dump the test data\n",
    "model = train_model(X_train, y, colsample_bytree=0.4603, \n",
    "                    n_estimators=2200, max_depth=3, learning_rate=0.05, min_child_weight=1.7817,\n",
    "                   reg_alpha=0.4640, reg_lambda=0.8571,subsample=0.5213, nthread = -1)\n",
    "\n",
    "y_pred = model.predict(X_test)\n",
    "y_pred = np.expm1(y_pred)\n",
    "\n",
    "test_data_id = df[no_train:]['Id']\n",
    "write_csv(\"house_price_after_tune.csv\", test_data_id, y_pred)\n",
    "FileLink(\"house_price_after_tune.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "def train_publish_model(model, X_train, y, X_test):\n",
    "    model.fit(X_train, y)\n",
    "    \n",
    "    y_pred = model.predict(X_test)\n",
    "    y_pred = np.expm1(y_pred)\n",
    "        \n",
    "    test_data_id = df[no_train:]['Id']\n",
    "    write_csv(\"house_price_after_tune.csv\", test_data_id, y_pred)\n",
    "    FileLink(\"house_price_after_tune.csv\") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>SaleCondition_Partial</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LandContour_Low</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Neighborhood_ClearCr</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Neighborhood_BrkSide</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Neighborhood_BrDale</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Neighborhood_Blueste</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Neighborhood_Blmngtn</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LotConfig_Inside</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LotConfig_FR3</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LotConfig_FR2</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LotConfig_CulDSac</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LotConfig_Corner</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LandContour_Lvl</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LandContour_HLS</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PoolArea</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LandContour_Bnk</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSZoning_RM</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSZoning_RL</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSZoning_RH</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSZoning_FV</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MSZoning_C (all)</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TotalSF</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>YrSold</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MoSold</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MiscVal</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fence</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Neighborhood_CollgCr</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Neighborhood_Crawfor</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Neighborhood_Edwards</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Neighborhood_Gilbert</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       0    1\n",
       "SaleCondition_Partial  0  0.0\n",
       "LandContour_Low        0  0.0\n",
       "Neighborhood_ClearCr   0  0.0\n",
       "Neighborhood_BrkSide   0  0.0\n",
       "Neighborhood_BrDale    0  0.0\n",
       "Neighborhood_Blueste   0  0.0\n",
       "Neighborhood_Blmngtn   0  0.0\n",
       "LotConfig_Inside       0  0.0\n",
       "LotConfig_FR3          0  0.0\n",
       "LotConfig_FR2          0  0.0\n",
       "LotConfig_CulDSac      0  0.0\n",
       "LotConfig_Corner       0  0.0\n",
       "LandContour_Lvl        0  0.0\n",
       "LandContour_HLS        0  0.0\n",
       "PoolArea               0  0.0\n",
       "LandContour_Bnk        0  0.0\n",
       "MSZoning_RM            0  0.0\n",
       "MSZoning_RL            0  0.0\n",
       "MSZoning_RH            0  0.0\n",
       "MSZoning_FV            0  0.0\n",
       "MSZoning_C (all)       0  0.0\n",
       "TotalSF                0  0.0\n",
       "YrSold                 0  0.0\n",
       "MoSold                 0  0.0\n",
       "MiscVal                0  0.0\n",
       "Fence                  0  0.0\n",
       "Neighborhood_CollgCr   0  0.0\n",
       "Neighborhood_Crawfor   0  0.0\n",
       "Neighborhood_Edwards   0  0.0\n",
       "Neighborhood_Gilbert   0  0.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Make sure there are no missing values. \n",
    "show_missing_data(df_trans)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "#We use cross validation to find out the predicted score for different techniques. \n",
    "n_folds = 5\n",
    "\n",
    "def rmsle_cv(model, X, y):\n",
    "    kf = KFold(n_folds, shuffle=True, random_state=42).get_n_splits(X)\n",
    "    rmse= np.sqrt(-cross_val_score(model, X, y, scoring=\"neg_mean_squared_error\", cv = kf))\n",
    "    return(rmse)\n",
    "\n",
    "def rmsle(y, y_pred):\n",
    "    return np.sqrt(mean_squared_error(y, y_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Xgboost score: 0.1158 (0.0063)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# XGB Model\n",
    "xgb_model = XGBRegressor(colsample_bytree=0.4603, \n",
    "                    n_estimators=2200, max_depth=3, learning_rate=0.05, min_child_weight=1.7817,\n",
    "                   reg_alpha=0.4640, reg_lambda=0.8571,subsample=0.5213, nthread = -1)\n",
    "cv_scores = rmsle_cv(xgb_model, X_train, y)\n",
    "print(\"Xgboost score: {:.4f} ({:.4f})\\n\".format(cv_scores.mean(), cv_scores.std()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Lasso score: 0.1104 (0.0055)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# LASSO\n",
    "lasso = make_pipeline(RobustScaler(), Lasso(alpha =0.0005, random_state=1))\n",
    "score = rmsle_cv(lasso, X_train, y)\n",
    "print(\"\\nLasso score: {:.4f} ({:.4f})\\n\".format(score.mean(), score.std()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "ENet score: 0.1104 (0.0055)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "ENet = make_pipeline(RobustScaler(), ElasticNet(alpha=0.0005, l1_ratio=.9, random_state=3, max_iter=3000))\n",
    "score = rmsle_cv(ENet, X_train, y)\n",
    "print(\"\\nENet score: {:.4f} ({:.4f})\\n\".format(score.mean(), score.std()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "KernelRidge score: 0.1130 (0.0047)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "KRR = KernelRidge(alpha=0.1, kernel='polynomial', degree=1, coef0=2.5)\n",
    "score = rmsle_cv(KRR, X_train, y)\n",
    "print(\"\\nKernelRidge score: {:.4f} ({:.4f})\\n\".format(score.mean(), score.std()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "GBoost score: 0.1166 (0.0069)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "GBoost = GradientBoostingRegressor(n_estimators=3000, learning_rate=0.05,\n",
    "                                   max_depth=4, max_features='sqrt',\n",
    "                                   min_samples_leaf=15, min_samples_split=10, \n",
    "                                   loss='huber', random_state =5)\n",
    "score = rmsle_cv(GBoost, X_train, y)\n",
    "print(\"\\nGBoost score: {:.4f} ({:.4f})\\n\".format(score.mean(), score.std()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "LGB score: 0.1165 (0.0060)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "model_lgb = lgb.LGBMRegressor(objective='regression',num_leaves=5,\n",
    "                              learning_rate=0.05, n_estimators=720,\n",
    "                              max_bin = 55, bagging_fraction = 0.8,\n",
    "                              bagging_freq = 5, feature_fraction = 0.2319,\n",
    "                              feature_fraction_seed=9, bagging_seed=9,\n",
    "                              min_data_in_leaf =6, min_sum_hessian_in_leaf = 11)\n",
    "score = rmsle_cv(model_lgb, X_train, y)\n",
    "print(\"\\nLGB score: {:.4f} ({:.4f})\\n\".format(score.mean(), score.std()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "class AveragingModels(BaseEstimator, RegressorMixin, TransformerMixin):\n",
    "    def __init__(self, models):\n",
    "        self.models = models\n",
    "        \n",
    "    # we define clones of the original models to fit the data in\n",
    "    def fit(self, X, y):\n",
    "        self.models_ = [clone(x) for x in self.models]\n",
    "        \n",
    "        # Train cloned base models\n",
    "        for model in self.models_:\n",
    "            model.fit(X, y)\n",
    "\n",
    "        return self\n",
    "    \n",
    "    #Now we do the predictions for cloned models and average them\n",
    "    def predict(self, X):\n",
    "        predictions = np.column_stack([\n",
    "            model.predict(X) for model in self.models_\n",
    "        ])\n",
    "        return np.mean(predictions, axis=1)   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Averaged base models score: 0.1084 (0.0052)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "averaged_models = AveragingModels(models = (ENet, GBoost, KRR, lasso))\n",
    "\n",
    "score = rmsle_cv(averaged_models, X_train, y)\n",
    "print(\" Averaged base models score: {:.4f} ({:.4f})\\n\".format(score.mean(), score.std()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_publish_model(averaged_models, X_train, y, X_test)\n",
    "FileLink(\"house_price_after_tune.csv\") \n",
    "# Got around .11840 in public leaderboard"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Averaged1 base models score: 0.1084 (0.0052)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "averaged_models1 = AveragingModels(models = (lasso, xgb_model, model_lgb))\n",
    "score = rmsle_cv(averaged_models, X_train, y)\n",
    "print(\" Averaged1 base models score: {:.4f} ({:.4f})\\n\".format(score.mean(), score.std()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.08355134474984019\n"
     ]
    }
   ],
   "source": [
    "averaged_models.fit(X_train, y)\n",
    "average_train_pred = averaged_models.predict(X_train)\n",
    "average_pred = np.expm1(averaged_models.predict(X_test))\n",
    "print(rmsle(y, average_train_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<a href='house_price_after_tune.csv' target='_blank'>house_price_after_tune.csv</a><br>"
      ],
      "text/plain": [
       "/Users/pulikunt/myfiles/notebooks/house_price_after_tune.csv"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_publish_model(averaged_models1, X_train, y, X_test)\n",
    "FileLink(\"house_price_after_tune.csv\") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [],
   "source": [
    "class StackingAveragedModels(BaseEstimator, RegressorMixin, TransformerMixin):\n",
    "    def __init__(self, base_models, meta_model, n_folds=5):\n",
    "        self.base_models = base_models\n",
    "        self.meta_model = meta_model\n",
    "        self.n_folds = n_folds\n",
    "   \n",
    "    # We again fit the data on clones of the original models\n",
    "    def fit(self, X, y):\n",
    "        self.base_models_ = [list() for x in self.base_models]\n",
    "        self.meta_model_ = clone(self.meta_model)\n",
    "        kfold = KFold(n_splits=self.n_folds, shuffle=True, random_state=156)\n",
    "        \n",
    "        # Train cloned base models then create out-of-fold predictions\n",
    "        # that are needed to train the cloned meta-model\n",
    "        out_of_fold_predictions = np.zeros((X.shape[0], len(self.base_models)))\n",
    "        for i, model in enumerate(self.base_models):\n",
    "            for train_index, holdout_index in kfold.split(X, y):\n",
    "                instance = clone(model)\n",
    "                self.base_models_[i].append(instance)\n",
    "                instance.fit(X[train_index], y[train_index])\n",
    "                y_pred = instance.predict(X[holdout_index])\n",
    "                out_of_fold_predictions[holdout_index, i] = y_pred\n",
    "                \n",
    "        # Now train the cloned  meta-model using the out-of-fold predictions as new feature\n",
    "        self.meta_model_.fit(out_of_fold_predictions, y)\n",
    "        return self\n",
    "   \n",
    "    #Do the predictions of all base models on the test data and use the averaged predictions as \n",
    "    #meta-features for the final prediction which is done by the meta-model\n",
    "    def predict(self, X):\n",
    "        meta_features = np.column_stack([\n",
    "            np.column_stack([model.predict(X) for model in base_models]).mean(axis=1)\n",
    "            for base_models in self.base_models_ ])\n",
    "        return self.meta_model_.predict(meta_features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Stacking Averaged models score: 0.1074 (0.0056)\n"
     ]
    }
   ],
   "source": [
    "stacked_averaged_models = StackingAveragedModels(base_models = (ENet, GBoost, KRR),\n",
    "                                                 meta_model = lasso)\n",
    "score = rmsle_cv(stacked_averaged_models, X_train.values, y.values)\n",
    "print(\"Stacking Averaged models score: {:.4f} ({:.4f})\".format(score.mean(), score.std()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<a href='house_price_after_tune.csv' target='_blank'>house_price_after_tune.csv</a><br>"
      ],
      "text/plain": [
       "/Users/pulikunt/myfiles/notebooks/house_price_after_tune.csv"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_publish_model(stacked_averaged_models, X_train.values, y.values, X_test.values)\n",
    "FileLink(\"house_price_after_tune.csv\") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.08185286388069687\n"
     ]
    }
   ],
   "source": [
    "stacked_averaged_models.fit(X_train.values, y.values)\n",
    "stacked_train_pred = stacked_averaged_models.predict(X_train.values)\n",
    "stacked_pred = np.expm1(stacked_averaged_models.predict(X_test.values))\n",
    "print(rmsle(y, stacked_train_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.041794705773656106\n"
     ]
    }
   ],
   "source": [
    "xgb_model.fit(X_train, y)\n",
    "xgb_train_pred = xgb_model.predict(X_train)\n",
    "xgb_pred = np.expm1(xgb_model.predict(X_test))\n",
    "print(rmsle(y, xgb_train_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.07250378640167945\n"
     ]
    }
   ],
   "source": [
    "model_lgb.fit(X_train, y)\n",
    "lgb_train_pred = model_lgb.predict(X_train)\n",
    "lgb_pred = np.expm1(model_lgb.predict(X_test))\n",
    "print(rmsle(y, lgb_train_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSLE score on train data:\n",
      "0.07240731250426967\n"
     ]
    }
   ],
   "source": [
    "'''RMSE on the entire Train data when averaging'''\n",
    "\n",
    "print('RMSLE score on train data:')\n",
    "print(rmsle(y,stacked_train_pred*0.70 +\n",
    "               xgb_train_pred*0.15 + lgb_train_pred*0.15 ))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSLE score on train data:\n",
      "0.06365876508149622\n"
     ]
    }
   ],
   "source": [
    "# This did not improve. Some how xgb overfitted here. \n",
    "print('RMSLE score on train data:')\n",
    "print(rmsle(y,stacked_train_pred*0.50 +\n",
    "               xgb_train_pred*0.35 + lgb_train_pred*0.15 ))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSLE score on train data:\n",
      "0.07387623566809344\n"
     ]
    }
   ],
   "source": [
    "print('RMSLE score on train data:')\n",
    "print(rmsle(y,stacked_train_pred*0.40 + lgb_train_pred*0.60 ))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [],
   "source": [
    "ensemble = stacked_pred*0.70 + xgb_pred*0.15 + lgb_pred*0.15"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [],
   "source": [
    "ensemble = stacked_pred*0.40 + lgb_pred*0.60\n",
    "#"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<a href='submission.csv' target='_blank'>submission.csv</a><br>"
      ],
      "text/plain": [
       "/Users/pulikunt/myfiles/notebooks/submission.csv"
      ]
     },
     "execution_count": 139,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sub = pd.DataFrame()\n",
    "sub['Id'] = df[no_train:]['Id']\n",
    "sub['SalePrice'] = ensemble\n",
    "sub.to_csv('submission.csv', index=False)\n",
    "FileLink(\"submission.csv\") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.9230487755090074,\n",
       " {'max_depth': 3, 'min_child_weight': 4, 'n_estimators': 975},\n",
       " 0.9230487755090074)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## MODEL TUNING\n",
    "\n",
    "param_test1 = {\n",
    " 'max_depth':range(3, 5, 1),\n",
    " 'min_child_weight':range(2, 6, 1),\n",
    " 'n_estimators': range(500, 1000, 25)   \n",
    "}\n",
    "\n",
    "gsearch1 = GridSearchCV(estimator = XGBRegressor(learning_rate =0.05, n_estimators=140, max_depth=5,\n",
    " min_child_weight=1, scale_pos_weight=1, seed=100, base_score=y.mean()), \n",
    " param_grid = param_test1, n_jobs=4,iid=False, cv=5)\n",
    "\n",
    "gsearch1.fit(X_train, y)\n",
    "gsearch1.best_score_, gsearch1.best_params_, gsearch1.best_score_\n",
    "#gsearch1.grid_scores_, gsearch1.best_params_, gsearch1.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def model_report():\n",
    "    #Print model report:\n",
    "    print \"\\nModel Report\"\n",
    "    print \"Accuracy : %.4g\" % metrics.accuracy_score(dtrain['Disbursed'].values, dtrain_predictions)\n",
    "    print \"AUC Score (Train): %f\" % metrics.roc_auc_score(dtrain['Disbursed'], dtrain_predprob)\n",
    "\n",
    "    feat_imp = pd.Series(alg.booster().get_fscore()).sort_values(ascending=False)\n",
    "    feat_imp.plot(kind='bar', title='Feature Importances')\n",
    "    plt.ylabel('Feature Importance Score')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After above split i got 0.13139 \n",
    "\n",
    "After removing the outliers it is improved to 0.12993\n",
    "\n",
    "After setting every numberic to 0 and every non numeric to 'NA' 0.12987\n",
    "\n",
    "After log transform it increased to 0.13376. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Label encoding\n",
    "??OrdinalEncoder"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. http://fastml.com/how-to-use-pd-dot-get-dummies-with-the-test-set/\n",
    "1. https://stackoverflow.com/questions/41335718/keep-same-dummy-variable-in-training-and-testing-data\n",
    "1. https://stackoverflow.com/questions/32011359/convert-categorical-data-in-pandas-dataframe\n",
    "1. https://stackoverflow.com/questions/24458645/label-encoding-across-multiple-columns-in-scikit-learn\n",
    "1. https://www.kaggle.com/pmarcelino/comprehensive-data-exploration-with-python \n",
    "1. https://www.kaggle.com/erikbruin/house-prices-lasso-xgboost-and-a-detailed-eda\n",
    "1. https://www.kaggle.com/serigne/stacked-regressions-top-4-on-leaderboard"
   ]
  }
 ],
 "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": "3.7.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}