classification_code.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 514,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.impute import SimpleImputer\n",
    "from sklearn.preprocessing import OneHotEncoder, OrdinalEncoder\n",
    "\n",
    "from sklearn.preprocessing import StandardScaler, MinMaxScaler\n",
    "\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.compose import ColumnTransformer\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "\n",
    "from sklearn.metrics import accuracy_score, classification_report, confusion_matrix, ConfusionMatrixDisplay\n",
    "\n",
    "from sklearn.base import BaseEstimator, TransformerMixin\n",
    "\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "from sklearn import set_config\n",
    "import warnings \n",
    "warnings.filterwarnings(\"ignore\")\n",
    "# Set the config so that we can view our preprocessor, and to transform output from numpy arrays to pandas dataframes\n",
    "set_config(display=\"diagram\")\n",
    "set_config(transform_output=\"pandas\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 515,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 1460 entries, 0 to 1459\n",
      "Data columns (total 81 columns):\n",
      " #   Column         Non-Null Count  Dtype  \n",
      "---  ------         --------------  -----  \n",
      " 0   LotArea        1460 non-null   int64  \n",
      " 1   LotFrontage    1201 non-null   float64\n",
      " 2   TotalBsmtSF    1460 non-null   int64  \n",
      " 3   BedroomAbvGr   1460 non-null   int64  \n",
      " 4   Fireplaces     1460 non-null   int64  \n",
      " 5   PoolArea       1460 non-null   int64  \n",
      " 6   GarageCars     1460 non-null   int64  \n",
      " 7   WoodDeckSF     1460 non-null   int64  \n",
      " 8   ScreenPorch    1460 non-null   int64  \n",
      " 9   Expensive      1460 non-null   int64  \n",
      " 10  MSZoning       1460 non-null   object \n",
      " 11  Condition1     1460 non-null   object \n",
      " 12  Heating        1460 non-null   object \n",
      " 13  Street         1460 non-null   object \n",
      " 14  CentralAir     1460 non-null   object \n",
      " 15  Foundation     1460 non-null   object \n",
      " 16  ExterQual      1460 non-null   object \n",
      " 17  ExterCond      1460 non-null   object \n",
      " 18  BsmtQual       1423 non-null   object \n",
      " 19  BsmtCond       1423 non-null   object \n",
      " 20  BsmtExposure   1422 non-null   object \n",
      " 21  BsmtFinType1   1423 non-null   object \n",
      " 22  KitchenQual    1460 non-null   object \n",
      " 23  FireplaceQu    770 non-null    object \n",
      " 24  MSSubClass     1460 non-null   int64  \n",
      " 25  OverallQual    1460 non-null   int64  \n",
      " 26  OverallCond    1460 non-null   int64  \n",
      " 27  YearBuilt      1460 non-null   int64  \n",
      " 28  YearRemodAdd   1460 non-null   int64  \n",
      " 29  MasVnrArea     1452 non-null   float64\n",
      " 30  BsmtFinSF1     1460 non-null   int64  \n",
      " 31  BsmtFinSF2     1460 non-null   int64  \n",
      " 32  BsmtUnfSF      1460 non-null   int64  \n",
      " 33  1stFlrSF       1460 non-null   int64  \n",
      " 34  2ndFlrSF       1460 non-null   int64  \n",
      " 35  LowQualFinSF   1460 non-null   int64  \n",
      " 36  GrLivArea      1460 non-null   int64  \n",
      " 37  BsmtFullBath   1460 non-null   int64  \n",
      " 38  BsmtHalfBath   1460 non-null   int64  \n",
      " 39  FullBath       1460 non-null   int64  \n",
      " 40  HalfBath       1460 non-null   int64  \n",
      " 41  KitchenAbvGr   1460 non-null   int64  \n",
      " 42  TotRmsAbvGrd   1460 non-null   int64  \n",
      " 43  GarageYrBlt    1379 non-null   float64\n",
      " 44  GarageArea     1460 non-null   int64  \n",
      " 45  OpenPorchSF    1460 non-null   int64  \n",
      " 46  EnclosedPorch  1460 non-null   int64  \n",
      " 47  3SsnPorch      1460 non-null   int64  \n",
      " 48  MiscVal        1460 non-null   int64  \n",
      " 49  MoSold         1460 non-null   int64  \n",
      " 50  YrSold         1460 non-null   int64  \n",
      " 51  Id             1460 non-null   int64  \n",
      " 52  Alley          91 non-null     object \n",
      " 53  LotShape       1460 non-null   object \n",
      " 54  LandContour    1460 non-null   object \n",
      " 55  Utilities      1460 non-null   object \n",
      " 56  LotConfig      1460 non-null   object \n",
      " 57  LandSlope      1460 non-null   object \n",
      " 58  Neighborhood   1460 non-null   object \n",
      " 59  Condition2     1460 non-null   object \n",
      " 60  BldgType       1460 non-null   object \n",
      " 61  HouseStyle     1460 non-null   object \n",
      " 62  RoofStyle      1460 non-null   object \n",
      " 63  RoofMatl       1460 non-null   object \n",
      " 64  Exterior1st    1460 non-null   object \n",
      " 65  Exterior2nd    1460 non-null   object \n",
      " 66  MasVnrType     1452 non-null   object \n",
      " 67  BsmtFinType2   1422 non-null   object \n",
      " 68  HeatingQC      1460 non-null   object \n",
      " 69  Electrical     1459 non-null   object \n",
      " 70  Functional     1460 non-null   object \n",
      " 71  GarageType     1379 non-null   object \n",
      " 72  GarageFinish   1379 non-null   object \n",
      " 73  GarageQual     1379 non-null   object \n",
      " 74  GarageCond     1379 non-null   object \n",
      " 75  PavedDrive     1460 non-null   object \n",
      " 76  PoolQC         7 non-null      object \n",
      " 77  Fence          281 non-null    object \n",
      " 78  MiscFeature    54 non-null     object \n",
      " 79  SaleType       1460 non-null   object \n",
      " 80  SaleCondition  1460 non-null   object \n",
      "dtypes: float64(3), int64(35), object(43)\n",
      "memory usage: 924.0+ KB\n"
     ]
    }
   ],
   "source": [
    "df = pd.read_csv(\"data\\housing_iteration_5_classification.csv\")\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 516,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "# Assuming df is your DataFrame\n",
    "# Check for missing values in each column\n",
    "missing_values = df.isnull()\n",
    "\n",
    "# Create a heatmap to visualize missing values\n",
    "plt.figure(figsize=(10, 6))\n",
    "sns.heatmap(missing_values, cbar=False, cmap='viridis')\n",
    "\n",
    "plt.title('Missing Values in DataFrame')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 517,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1460, 80)"
      ]
     },
     "execution_count": 517,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = df.set_index(\"Id\")\n",
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 518,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "           Column  Missing Values\n",
      "0          PoolQC            1453\n",
      "1     MiscFeature            1406\n",
      "2           Alley            1369\n",
      "3           Fence            1179\n",
      "4     FireplaceQu             690\n",
      "..            ...             ...\n",
      "75   YearRemodAdd               0\n",
      "76      YearBuilt               0\n",
      "77    OverallCond               0\n",
      "78    OverallQual               0\n",
      "79  SaleCondition               0\n",
      "\n",
      "[80 rows x 2 columns]\n"
     ]
    }
   ],
   "source": [
    "# Create a DataFrame with missing values count for each column\n",
    "missing_values_count = df.isnull().sum().sort_values(ascending=False)\n",
    "missing_df = pd.DataFrame({'Column': missing_values_count.index, 'Missing Values': missing_values_count.values})\n",
    "\n",
    "# Display the DataFrame\n",
    "print(missing_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 519,
   "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>Column</th>\n",
       "      <th>Missing Values</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>GarageCars</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Column  Missing Values\n",
       "56  GarageCars               0"
      ]
     },
     "execution_count": 519,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "missing_df.loc[missing_df[\"Column\"]==\"GarageCars\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 520,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.drop([\"PoolQC\", \"MiscFeature\", \"Alley\", \"Fence\"], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 521,
   "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>LotArea</th>\n",
       "      <th>LotFrontage</th>\n",
       "      <th>TotalBsmtSF</th>\n",
       "      <th>BedroomAbvGr</th>\n",
       "      <th>Fireplaces</th>\n",
       "      <th>PoolArea</th>\n",
       "      <th>GarageCars</th>\n",
       "      <th>WoodDeckSF</th>\n",
       "      <th>ScreenPorch</th>\n",
       "      <th>Expensive</th>\n",
       "      <th>...</th>\n",
       "      <th>HeatingQC</th>\n",
       "      <th>Electrical</th>\n",
       "      <th>Functional</th>\n",
       "      <th>GarageType</th>\n",
       "      <th>GarageFinish</th>\n",
       "      <th>GarageQual</th>\n",
       "      <th>GarageCond</th>\n",
       "      <th>PavedDrive</th>\n",
       "      <th>SaleType</th>\n",
       "      <th>SaleCondition</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8450</td>\n",
       "      <td>65.0</td>\n",
       "      <td>856</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>Ex</td>\n",
       "      <td>SBrkr</td>\n",
       "      <td>Typ</td>\n",
       "      <td>Attchd</td>\n",
       "      <td>RFn</td>\n",
       "      <td>TA</td>\n",
       "      <td>TA</td>\n",
       "      <td>Y</td>\n",
       "      <td>WD</td>\n",
       "      <td>Normal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9600</td>\n",
       "      <td>80.0</td>\n",
       "      <td>1262</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>298</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>Ex</td>\n",
       "      <td>SBrkr</td>\n",
       "      <td>Typ</td>\n",
       "      <td>Attchd</td>\n",
       "      <td>RFn</td>\n",
       "      <td>TA</td>\n",
       "      <td>TA</td>\n",
       "      <td>Y</td>\n",
       "      <td>WD</td>\n",
       "      <td>Normal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>11250</td>\n",
       "      <td>68.0</td>\n",
       "      <td>920</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>Ex</td>\n",
       "      <td>SBrkr</td>\n",
       "      <td>Typ</td>\n",
       "      <td>Attchd</td>\n",
       "      <td>RFn</td>\n",
       "      <td>TA</td>\n",
       "      <td>TA</td>\n",
       "      <td>Y</td>\n",
       "      <td>WD</td>\n",
       "      <td>Normal</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 76 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    LotArea  LotFrontage  TotalBsmtSF  BedroomAbvGr  Fireplaces  PoolArea  \\\n",
       "Id                                                                          \n",
       "1      8450         65.0          856             3           0         0   \n",
       "2      9600         80.0         1262             3           1         0   \n",
       "3     11250         68.0          920             3           1         0   \n",
       "\n",
       "    GarageCars  WoodDeckSF  ScreenPorch  Expensive  ... HeatingQC Electrical  \\\n",
       "Id                                                  ...                        \n",
       "1            2           0            0          0  ...        Ex      SBrkr   \n",
       "2            2         298            0          0  ...        Ex      SBrkr   \n",
       "3            2           0            0          0  ...        Ex      SBrkr   \n",
       "\n",
       "   Functional GarageType GarageFinish GarageQual GarageCond PavedDrive  \\\n",
       "Id                                                                       \n",
       "1         Typ     Attchd          RFn         TA         TA          Y   \n",
       "2         Typ     Attchd          RFn         TA         TA          Y   \n",
       "3         Typ     Attchd          RFn         TA         TA          Y   \n",
       "\n",
       "   SaleType SaleCondition  \n",
       "Id                         \n",
       "1        WD        Normal  \n",
       "2        WD        Normal  \n",
       "3        WD        Normal  \n",
       "\n",
       "[3 rows x 76 columns]"
      ]
     },
     "execution_count": 521,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 522,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['LotArea', 'LotFrontage', 'TotalBsmtSF', 'BedroomAbvGr', 'Fireplaces',\n",
       "       'PoolArea', 'GarageCars', 'WoodDeckSF', 'ScreenPorch', 'Expensive',\n",
       "       'MSZoning', 'Condition1', 'Heating', 'Street', 'CentralAir',\n",
       "       'Foundation', 'ExterQual', 'ExterCond', 'BsmtQual', 'BsmtCond',\n",
       "       'BsmtExposure', 'BsmtFinType1', 'KitchenQual', 'FireplaceQu',\n",
       "       'MSSubClass', 'OverallQual', 'OverallCond', 'YearBuilt', 'YearRemodAdd',\n",
       "       'MasVnrArea', 'BsmtFinSF1', 'BsmtFinSF2', 'BsmtUnfSF', '1stFlrSF',\n",
       "       '2ndFlrSF', 'LowQualFinSF', 'GrLivArea', 'BsmtFullBath', 'BsmtHalfBath',\n",
       "       'FullBath', 'HalfBath', 'KitchenAbvGr', 'TotRmsAbvGrd', 'GarageYrBlt',\n",
       "       'GarageArea', 'OpenPorchSF', 'EnclosedPorch', '3SsnPorch', 'MiscVal',\n",
       "       'MoSold', 'YrSold', 'LotShape', 'LandContour', 'Utilities', 'LotConfig',\n",
       "       'LandSlope', 'Neighborhood', 'Condition2', 'BldgType', 'HouseStyle',\n",
       "       'RoofStyle', 'RoofMatl', 'Exterior1st', 'Exterior2nd', 'MasVnrType',\n",
       "       'BsmtFinType2', 'HeatingQC', 'Electrical', 'Functional', 'GarageType',\n",
       "       'GarageFinish', 'GarageQual', 'GarageCond', 'PavedDrive', 'SaleType',\n",
       "       'SaleCondition'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 522,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 523,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Id\n",
       "1       SoldBetween05-15\n",
       "2       SoldBetween05-15\n",
       "3       SoldBetween05-15\n",
       "4       SoldBetween05-15\n",
       "5       SoldBetween05-15\n",
       "              ...       \n",
       "1456    SoldBetween05-15\n",
       "1457    SoldBetween05-15\n",
       "1458    SoldBetween05-15\n",
       "1459    SoldBetween05-15\n",
       "1460    SoldBetween05-15\n",
       "Name: YrSold, Length: 1460, dtype: object"
      ]
     },
     "execution_count": 523,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Assuming df is your DataFrame\n",
    "labels = ['SoldBefore05', 'SoldBetween05-15', 'SoldAfter15']\n",
    "\n",
    "# Define a custom function to assign categories based on the 'YrSold' values\n",
    "def categorize_year_sold(year_sold):\n",
    "    if year_sold < 2005:\n",
    "        return labels[0]\n",
    "    elif 2005 <= year_sold <= 2015:\n",
    "        return labels[1]\n",
    "    else:\n",
    "        return labels[2]\n",
    "\n",
    "# Apply the custom function to create the new categorical column\n",
    "df['YrSold'] = df['YrSold'].apply(categorize_year_sold)\n",
    "df['YrSold']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 524,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define a custom function to assign seasons based on the 'MoSold' values\n",
    "def categorize_season(month_sold):\n",
    "    if month_sold in [12, 1, 2]:\n",
    "        return 'Winter'\n",
    "    elif month_sold in [3, 4, 5]:\n",
    "        return 'Spring'\n",
    "    elif month_sold in [6, 7, 8]:\n",
    "        return 'Summer'\n",
    "    else:\n",
    "        return 'Fall'\n",
    "\n",
    "# Apply the custom function to create the new categorical column\n",
    "df['MoSold'] = df['MoSold'].apply(categorize_season)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 525,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "43 Index(['LotArea', 'LotFrontage', 'TotalBsmtSF', 'BedroomAbvGr', 'Fireplaces',\n",
      "       'GarageCars', 'WoodDeckSF', 'MSSubClass', 'OverallQual', 'YearBuilt',\n",
      "       'YearRemodAdd', 'MasVnrArea', 'BsmtFinSF1', 'BsmtUnfSF', '1stFlrSF',\n",
      "       '2ndFlrSF', 'GrLivArea', 'BsmtFullBath', 'FullBath', 'HalfBath',\n",
      "       'TotRmsAbvGrd', 'GarageYrBlt', 'GarageArea', 'OpenPorchSF',\n",
      "       'Foundation_PConc', 'ExterQual_Gd', 'ExterQual_TA', 'BsmtQual_Gd',\n",
      "       'BsmtExposure_Gd', 'BsmtExposure_No', 'BsmtFinType1_GLQ',\n",
      "       'KitchenQual_Gd', 'KitchenQual_TA', 'FireplaceQu_Gd', 'FireplaceQu_NaN',\n",
      "       'Neighborhood_NoRidge', 'Neighborhood_NridgHt', 'RoofStyle_Gable',\n",
      "       'RoofStyle_Hip', 'Exterior1st_CemntBd', 'Exterior2nd_CmentBd',\n",
      "       'GarageType_BuiltIn', 'SaleType_New'],\n",
      "      dtype='object')\n"
     ]
    }
   ],
   "source": [
    "from sklearn.feature_selection import SelectFromModel\n",
    "\n",
    "df_cat = df.select_dtypes(exclude=\"number\")\n",
    "df_cat = df_cat.fillna(value=\"NaN\")\n",
    "df_cat = pd.get_dummies(df_cat, drop_first=True)\n",
    "\n",
    "df_num = df.select_dtypes(include=\"number\")\n",
    "df_num = df_num.fillna(df_num.mean())\n",
    "\n",
    "data = pd.concat([df_num, df_cat], axis=1)\n",
    "X = data.drop(\"Expensive\", axis=1)\n",
    "y = data[\"Expensive\"]\n",
    "X_train,X_test,y_train,y_test = train_test_split(X,y, test_size=0.3)\n",
    "\n",
    "sel = SelectFromModel(RandomForestClassifier())\n",
    "sel.fit(X_train, y_train)\n",
    "selected_feat= X_train.columns[(sel.get_support())]\n",
    "\n",
    "print(len(selected_feat), selected_feat)\n",
    "selected_features = ['LotArea', 'LotFrontage', 'TotalBsmtSF', 'BedroomAbvGr', 'Fireplaces',\n",
    "                     'GarageCars', 'WoodDeckSF', 'MSSubClass', 'OverallQual', 'YearBuilt',\n",
    "                     'YearRemodAdd', 'MasVnrArea', 'BsmtFinSF1', 'BsmtUnfSF', '1stFlrSF',\n",
    "                     '2ndFlrSF', 'GrLivArea', 'BsmtFullBath', 'TotRmsAbvGrd', 'GarageYrBlt',\n",
    "                     'GarageArea', 'OpenPorchSF', 'Foundation', 'ExterQual', 'BsmtQual',\n",
    "                     'KitchenQual', 'FireplaceQu', 'Neighborhood', 'RoofStyle', 'MasVnrType', 'SaleType']\n",
    "\n",
    "df = df[selected_features + [\"Expensive\"]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 526,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['LotArea', 'LotFrontage', 'TotalBsmtSF', 'BedroomAbvGr', 'Fireplaces',\n",
       "       'GarageCars', 'WoodDeckSF', 'MSSubClass', 'OverallQual', 'YearBuilt',\n",
       "       'YearRemodAdd', 'MasVnrArea', 'BsmtFinSF1', 'BsmtUnfSF', '1stFlrSF',\n",
       "       '2ndFlrSF', 'GrLivArea', 'BsmtFullBath', 'FullBath', 'HalfBath',\n",
       "       'TotRmsAbvGrd', 'GarageYrBlt', 'GarageArea', 'OpenPorchSF',\n",
       "       'Foundation_PConc', 'ExterQual_Gd', 'ExterQual_TA', 'BsmtQual_Gd',\n",
       "       'BsmtExposure_Gd', 'BsmtExposure_No', 'BsmtFinType1_GLQ',\n",
       "       'KitchenQual_Gd', 'KitchenQual_TA', 'FireplaceQu_Gd', 'FireplaceQu_NaN',\n",
       "       'Neighborhood_NoRidge', 'Neighborhood_NridgHt', 'RoofStyle_Gable',\n",
       "       'RoofStyle_Hip', 'Exterior1st_CemntBd', 'Exterior2nd_CmentBd',\n",
       "       'GarageType_BuiltIn', 'SaleType_New'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 526,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "selected_feat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 527,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create train/test split\n",
    "X = df.drop([\"Expensive\"], axis=1)\n",
    "y = df[\"Expensive\"]\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.1, random_state=101)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 528,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1314, 31) (146, 31) (1314,) (146,)\n"
     ]
    }
   ],
   "source": [
    "print(X_train.shape, X_test.shape, y_train.shape, y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 529,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Divide columns by data_types\n",
    "X_train_cat = X_train.select_dtypes(exclude=\"number\")\n",
    "X_train_num = X_train.select_dtypes(include=\"number\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 530,
   "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>Foundation</th>\n",
       "      <th>ExterQual</th>\n",
       "      <th>BsmtQual</th>\n",
       "      <th>KitchenQual</th>\n",
       "      <th>FireplaceQu</th>\n",
       "      <th>Neighborhood</th>\n",
       "      <th>RoofStyle</th>\n",
       "      <th>MasVnrType</th>\n",
       "      <th>SaleType</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1314</td>\n",
       "      <td>1314</td>\n",
       "      <td>1283</td>\n",
       "      <td>1314</td>\n",
       "      <td>706</td>\n",
       "      <td>1314</td>\n",
       "      <td>1314</td>\n",
       "      <td>1306</td>\n",
       "      <td>1314</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>6</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>25</td>\n",
       "      <td>6</td>\n",
       "      <td>4</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>PConc</td>\n",
       "      <td>TA</td>\n",
       "      <td>TA</td>\n",
       "      <td>TA</td>\n",
       "      <td>Gd</td>\n",
       "      <td>NAmes</td>\n",
       "      <td>Gable</td>\n",
       "      <td>None</td>\n",
       "      <td>WD</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>581</td>\n",
       "      <td>810</td>\n",
       "      <td>587</td>\n",
       "      <td>656</td>\n",
       "      <td>350</td>\n",
       "      <td>198</td>\n",
       "      <td>1028</td>\n",
       "      <td>775</td>\n",
       "      <td>1139</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Foundation ExterQual BsmtQual KitchenQual FireplaceQu Neighborhood  \\\n",
       "count        1314      1314     1283        1314         706         1314   \n",
       "unique          6         4        4           4           5           25   \n",
       "top         PConc        TA       TA          TA          Gd        NAmes   \n",
       "freq          581       810      587         656         350          198   \n",
       "\n",
       "       RoofStyle MasVnrType SaleType  \n",
       "count       1314       1306     1314  \n",
       "unique         6          4        9  \n",
       "top        Gable       None       WD  \n",
       "freq        1028        775     1139  "
      ]
     },
     "execution_count": 530,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train_cat.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 531,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-16 {color: black;}#sk-container-id-16 pre{padding: 0;}#sk-container-id-16 div.sk-toggleable {background-color: white;}#sk-container-id-16 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-16 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-16 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-16 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-16 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-16 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-16 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-16 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-16 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-16 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-16 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-16 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-16 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-16 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-16 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-16 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-16 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-16 div.sk-item {position: relative;z-index: 1;}#sk-container-id-16 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-16 div.sk-item::before, #sk-container-id-16 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-16 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-16 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-16 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-16 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-16 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-16 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-16 div.sk-label-container {text-align: center;}#sk-container-id-16 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-16 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-16\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>ColumnTransformer(transformers=[(&#x27;num_pipe&#x27;,\n",
       "                                 Pipeline(steps=[(&#x27;simpleimputer&#x27;,\n",
       "                                                  SimpleImputer())]),\n",
       "                                 Index([&#x27;LotArea&#x27;, &#x27;LotFrontage&#x27;, &#x27;TotalBsmtSF&#x27;, &#x27;BedroomAbvGr&#x27;, &#x27;Fireplaces&#x27;,\n",
       "       &#x27;GarageCars&#x27;, &#x27;WoodDeckSF&#x27;, &#x27;MSSubClass&#x27;, &#x27;OverallQual&#x27;, &#x27;YearBuilt&#x27;,\n",
       "       &#x27;YearRemodAdd&#x27;, &#x27;MasVnrArea&#x27;, &#x27;BsmtFinSF1&#x27;, &#x27;BsmtUnfSF&#x27;, &#x27;1stFlrSF&#x27;,\n",
       "       &#x27;2ndFlrSF&#x27;, &#x27;GrLivArea&#x27;, &#x27;BsmtFullBath&#x27;, &#x27;TotRmsAbvGrd&#x27;...\n",
       "                                (&#x27;cat_pipe&#x27;,\n",
       "                                 Pipeline(steps=[(&#x27;simpleimputer&#x27;,\n",
       "                                                  SimpleImputer(fill_value=&#x27;NaN&#x27;,\n",
       "                                                                strategy=&#x27;constant&#x27;)),\n",
       "                                                 (&#x27;onehotencoder&#x27;,\n",
       "                                                  OneHotEncoder(handle_unknown=&#x27;infrequent_if_exist&#x27;,\n",
       "                                                                min_frequency=0.03,\n",
       "                                                                sparse_output=False))]),\n",
       "                                 Index([&#x27;Foundation&#x27;, &#x27;ExterQual&#x27;, &#x27;BsmtQual&#x27;, &#x27;KitchenQual&#x27;, &#x27;FireplaceQu&#x27;,\n",
       "       &#x27;Neighborhood&#x27;, &#x27;RoofStyle&#x27;, &#x27;MasVnrType&#x27;, &#x27;SaleType&#x27;],\n",
       "      dtype=&#x27;object&#x27;))])</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-91\" type=\"checkbox\" ><label for=\"sk-estimator-id-91\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">ColumnTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnTransformer(transformers=[(&#x27;num_pipe&#x27;,\n",
       "                                 Pipeline(steps=[(&#x27;simpleimputer&#x27;,\n",
       "                                                  SimpleImputer())]),\n",
       "                                 Index([&#x27;LotArea&#x27;, &#x27;LotFrontage&#x27;, &#x27;TotalBsmtSF&#x27;, &#x27;BedroomAbvGr&#x27;, &#x27;Fireplaces&#x27;,\n",
       "       &#x27;GarageCars&#x27;, &#x27;WoodDeckSF&#x27;, &#x27;MSSubClass&#x27;, &#x27;OverallQual&#x27;, &#x27;YearBuilt&#x27;,\n",
       "       &#x27;YearRemodAdd&#x27;, &#x27;MasVnrArea&#x27;, &#x27;BsmtFinSF1&#x27;, &#x27;BsmtUnfSF&#x27;, &#x27;1stFlrSF&#x27;,\n",
       "       &#x27;2ndFlrSF&#x27;, &#x27;GrLivArea&#x27;, &#x27;BsmtFullBath&#x27;, &#x27;TotRmsAbvGrd&#x27;...\n",
       "                                (&#x27;cat_pipe&#x27;,\n",
       "                                 Pipeline(steps=[(&#x27;simpleimputer&#x27;,\n",
       "                                                  SimpleImputer(fill_value=&#x27;NaN&#x27;,\n",
       "                                                                strategy=&#x27;constant&#x27;)),\n",
       "                                                 (&#x27;onehotencoder&#x27;,\n",
       "                                                  OneHotEncoder(handle_unknown=&#x27;infrequent_if_exist&#x27;,\n",
       "                                                                min_frequency=0.03,\n",
       "                                                                sparse_output=False))]),\n",
       "                                 Index([&#x27;Foundation&#x27;, &#x27;ExterQual&#x27;, &#x27;BsmtQual&#x27;, &#x27;KitchenQual&#x27;, &#x27;FireplaceQu&#x27;,\n",
       "       &#x27;Neighborhood&#x27;, &#x27;RoofStyle&#x27;, &#x27;MasVnrType&#x27;, &#x27;SaleType&#x27;],\n",
       "      dtype=&#x27;object&#x27;))])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-92\" type=\"checkbox\" ><label for=\"sk-estimator-id-92\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">num_pipe</label><div class=\"sk-toggleable__content\"><pre>Index([&#x27;LotArea&#x27;, &#x27;LotFrontage&#x27;, &#x27;TotalBsmtSF&#x27;, &#x27;BedroomAbvGr&#x27;, &#x27;Fireplaces&#x27;,\n",
       "       &#x27;GarageCars&#x27;, &#x27;WoodDeckSF&#x27;, &#x27;MSSubClass&#x27;, &#x27;OverallQual&#x27;, &#x27;YearBuilt&#x27;,\n",
       "       &#x27;YearRemodAdd&#x27;, &#x27;MasVnrArea&#x27;, &#x27;BsmtFinSF1&#x27;, &#x27;BsmtUnfSF&#x27;, &#x27;1stFlrSF&#x27;,\n",
       "       &#x27;2ndFlrSF&#x27;, &#x27;GrLivArea&#x27;, &#x27;BsmtFullBath&#x27;, &#x27;TotRmsAbvGrd&#x27;, &#x27;GarageYrBlt&#x27;,\n",
       "       &#x27;GarageArea&#x27;, &#x27;OpenPorchSF&#x27;],\n",
       "      dtype=&#x27;object&#x27;)</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-93\" type=\"checkbox\" ><label for=\"sk-estimator-id-93\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer()</pre></div></div></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-94\" type=\"checkbox\" ><label for=\"sk-estimator-id-94\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">cat_pipe</label><div class=\"sk-toggleable__content\"><pre>Index([&#x27;Foundation&#x27;, &#x27;ExterQual&#x27;, &#x27;BsmtQual&#x27;, &#x27;KitchenQual&#x27;, &#x27;FireplaceQu&#x27;,\n",
       "       &#x27;Neighborhood&#x27;, &#x27;RoofStyle&#x27;, &#x27;MasVnrType&#x27;, &#x27;SaleType&#x27;],\n",
       "      dtype=&#x27;object&#x27;)</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-95\" type=\"checkbox\" ><label for=\"sk-estimator-id-95\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer(fill_value=&#x27;NaN&#x27;, strategy=&#x27;constant&#x27;)</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-96\" type=\"checkbox\" ><label for=\"sk-estimator-id-96\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">OneHotEncoder</label><div class=\"sk-toggleable__content\"><pre>OneHotEncoder(handle_unknown=&#x27;infrequent_if_exist&#x27;, min_frequency=0.03,\n",
       "              sparse_output=False)</pre></div></div></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "ColumnTransformer(transformers=[('num_pipe',\n",
       "                                 Pipeline(steps=[('simpleimputer',\n",
       "                                                  SimpleImputer())]),\n",
       "                                 Index(['LotArea', 'LotFrontage', 'TotalBsmtSF', 'BedroomAbvGr', 'Fireplaces',\n",
       "       'GarageCars', 'WoodDeckSF', 'MSSubClass', 'OverallQual', 'YearBuilt',\n",
       "       'YearRemodAdd', 'MasVnrArea', 'BsmtFinSF1', 'BsmtUnfSF', '1stFlrSF',\n",
       "       '2ndFlrSF', 'GrLivArea', 'BsmtFullBath', 'TotRmsAbvGrd'...\n",
       "                                ('cat_pipe',\n",
       "                                 Pipeline(steps=[('simpleimputer',\n",
       "                                                  SimpleImputer(fill_value='NaN',\n",
       "                                                                strategy='constant')),\n",
       "                                                 ('onehotencoder',\n",
       "                                                  OneHotEncoder(handle_unknown='infrequent_if_exist',\n",
       "                                                                min_frequency=0.03,\n",
       "                                                                sparse_output=False))]),\n",
       "                                 Index(['Foundation', 'ExterQual', 'BsmtQual', 'KitchenQual', 'FireplaceQu',\n",
       "       'Neighborhood', 'RoofStyle', 'MasVnrType', 'SaleType'],\n",
       "      dtype='object'))])"
      ]
     },
     "execution_count": 531,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "num_pipe = make_pipeline(SimpleImputer(strategy=\"mean\"))\n",
    "\n",
    "# # 3.3. categorical pipeline = \"N_A\" imputer + categorical encoder\n",
    "categorical_pipe = make_pipeline(SimpleImputer(strategy=\"constant\", fill_value=\"NaN\"),\n",
    "                                 OneHotEncoder(sparse_output=False,handle_unknown='infrequent_if_exist',min_frequency=0.03))\n",
    "\n",
    "full_preprocessing = ColumnTransformer(\n",
    "    transformers=[\n",
    "        (\"num_pipe\", num_pipe, X_train_num.columns),\n",
    "        (\"cat_pipe\", categorical_pipe, X_train_cat.columns),\n",
    "    ]\n",
    ")\n",
    "\n",
    "full_preprocessing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Remove outliers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 532,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define IQRTransformer\n",
    "class IQRTransformer(BaseEstimator, TransformerMixin):\n",
    "    def __init__(self, multiplier=1.5):\n",
    "        self.multiplier = multiplier\n",
    "        self.num_outliers_removed = 0\n",
    "\n",
    "    def fit(self, X, y=None):\n",
    "        # Calculate the IQR for each feature\n",
    "        self.lower_bound = np.percentile(X, 25, axis=0) - self.multiplier * np.subtract(*np.percentile(X, [75, 25], axis=0))\n",
    "        self.upper_bound = np.percentile(X, 75, axis=0) + self.multiplier * np.subtract(*np.percentile(X, [75, 25], axis=0))\n",
    "        return self\n",
    "\n",
    "    def transform(self, X):\n",
    "        # Count the number of outliers removed\n",
    "        self.num_outliers_removed += np.sum((X < self.lower_bound) | (X > self.upper_bound))\n",
    "\n",
    "        # Apply the IQR bounds to each feature\n",
    "        X_transformed = np.clip(X, self.lower_bound, self.upper_bound)\n",
    "        return X_transformed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 533,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ColumnTransformer(transformers=[('num_pipe',\n",
      "                                 Pipeline(steps=[('simpleimputer',\n",
      "                                                  SimpleImputer())]),\n",
      "                                 Index(['LotArea', 'LotFrontage', 'TotalBsmtSF', 'BedroomAbvGr', 'Fireplaces',\n",
      "       'GarageCars', 'WoodDeckSF', 'MSSubClass', 'OverallQual', 'YearBuilt',\n",
      "       'YearRemodAdd', 'MasVnrArea', 'BsmtFinSF1', 'BsmtUnfSF', '1stFlrSF',\n",
      "       '2ndFlrSF', 'GrLivArea', 'BsmtFullBath', 'TotRmsAbvGrd'...\n",
      "                                ('cat_pipe',\n",
      "                                 Pipeline(steps=[('simpleimputer',\n",
      "                                                  SimpleImputer(fill_value='NaN',\n",
      "                                                                strategy='constant')),\n",
      "                                                 ('onehotencoder',\n",
      "                                                  OneHotEncoder(handle_unknown='infrequent_if_exist',\n",
      "                                                                min_frequency=0.03,\n",
      "                                                                sparse_output=False))]),\n",
      "                                 Index(['Foundation', 'ExterQual', 'BsmtQual', 'KitchenQual', 'FireplaceQu',\n",
      "       'Neighborhood', 'RoofStyle', 'MasVnrType', 'SaleType'],\n",
      "      dtype='object'))])\n",
      "['LotArea' 'LotFrontage' 'TotalBsmtSF' 'BedroomAbvGr' 'Fireplaces'\n",
      " 'GarageCars' 'WoodDeckSF' 'MSSubClass' 'OverallQual' 'YearBuilt'\n",
      " 'YearRemodAdd' 'MasVnrArea' 'BsmtFinSF1' 'BsmtUnfSF' '1stFlrSF'\n",
      " '2ndFlrSF' 'GrLivArea' 'BsmtFullBath' 'TotRmsAbvGrd' 'GarageYrBlt'\n",
      " 'GarageArea' 'OpenPorchSF']\n"
     ]
    }
   ],
   "source": [
    "full_pipeline_Dtree = make_pipeline(full_preprocessing,\n",
    "                                         IQRTransformer(),\n",
    "                                         StandardScaler(),\n",
    "                                         DecisionTreeClassifier())\n",
    "full_pipeline_Dtree.fit(X_train,y_train)\n",
    "print(full_pipeline_Dtree.named_steps.columntransformer)\n",
    "print(full_pipeline_Dtree.named_steps.columntransformer.named_transformers_[\"num_pipe\"].named_steps.simpleimputer.get_feature_names_out())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### DecisionTreeClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 534,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 1080 candidates, totalling 5400 fits\n"
     ]
    }
   ],
   "source": [
    "# Full pipeline: preprocessor + model\n",
    "full_pipeline_Dtree_iqr = make_pipeline(full_preprocessing,\n",
    "                                         IQRTransformer(),\n",
    "                                         StandardScaler(),\n",
    "                                         DecisionTreeClassifier())\n",
    "\n",
    "# Define parameter grid\n",
    "param_grid_iqr = {\n",
    "    \"columntransformer__num_pipe__simpleimputer__strategy\":[\"mean\", \"median\", \"most_frequent\"],\n",
    "    \"standardscaler__with_mean\":[True, False],\n",
    "    \"standardscaler__with_std\":[True, False],\n",
    "    \"decisiontreeclassifier__max_depth\": range(2, 14, 2),\n",
    "    \"decisiontreeclassifier__min_samples_leaf\": range(3, 12, 2),\n",
    "    \"decisiontreeclassifier__criterion\": [\"gini\", \"entropy\", \"log_loss\"],\n",
    "}\n",
    "\n",
    "# Define GridSearchCV\n",
    "search_dtree_iqr = GridSearchCV(full_pipeline_Dtree_iqr,\n",
    "                                param_grid_iqr,\n",
    "                                cv=5,\n",
    "                                verbose=1)\n",
    "\n",
    "search_dtree_iqr.fit(X_train, y_train)\n",
    "\n",
    "\n",
    "# Create a dictionary to store the best scores from the different models\n",
    "scores = {\"dtree_iqr\": search_dtree_iqr.best_score_}\n",
    "\n",
    "best_param = search_dtree_iqr.best_params_\n",
    "\n",
    "print(best_param, scores)\n",
    "\n",
    "# Test accuracy\n",
    "print(accuracy_score(search_dtree_iqr.predict(X_test), y_test))\n",
    "\n",
    "# Plot confusion matrix\n",
    "print(ConfusionMatrixDisplay.from_estimator(search_dtree_iqr,\n",
    "                                             X_test,\n",
    "                                             y_test,\n",
    "                                             display_labels=['Not Survived', 'Survived']))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### KNNClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 576 candidates, totalling 2880 fits\n",
      "{'columntransformer__num_pipe__simpleimputer__strategy': 'mean', 'kneighborsclassifier__n_neighbors': 4, 'kneighborsclassifier__weights': 'distance', 'standardscaler__with_mean': True, 'standardscaler__with_std': True} {'dtree_iqr': 0.9398717092851129, 'knn_iqr': 0.9398775142948365}\n",
      "0.958904109589041\n",
      "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay object at 0x000002EF133970A0>\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Full pipeline: preprocessor + model\n",
    "full_pipeline_knn_iqr = make_pipeline(full_preprocessing,\n",
    "                                         IQRTransformer(),\n",
    "                                         StandardScaler(),\n",
    "                                         KNeighborsClassifier())\n",
    "\n",
    "# Define parameter grid\n",
    "param_grid_knn = {\n",
    "    \"columntransformer__num_pipe__simpleimputer__strategy\":[\"mean\", \"median\", \"most_frequent\"],\n",
    "    \"standardscaler__with_mean\":[True, False],\n",
    "    \"standardscaler__with_std\":[True, False],\n",
    "    \"kneighborsclassifier__n_neighbors\": range(2, 50, 2),\n",
    "    \"kneighborsclassifier__weights\": [\"uniform\", \"distance\"],\n",
    "}\n",
    "\n",
    "# Define GridSearchCV\n",
    "search_knn_iqr = GridSearchCV(full_pipeline_knn_iqr,\n",
    "                                param_grid_knn,\n",
    "                                cv=5,\n",
    "                                verbose=1)\n",
    "\n",
    "search_knn_iqr.fit(X_train, y_train)\n",
    "\n",
    "\n",
    "# Create a dictionary to store the best scores from the different models\n",
    "scores[\"knn_iqr\"]= search_knn_iqr.best_score_\n",
    "\n",
    "best_param = search_knn_iqr.best_params_\n",
    "\n",
    "print(best_param, scores)\n",
    "\n",
    "# Test accuracy\n",
    "print(accuracy_score(search_knn_iqr.predict(X_test), y_test))\n",
    "\n",
    "# Plot confusion matrix\n",
    "print(ConfusionMatrixDisplay.from_estimator(search_knn_iqr,\n",
    "                                             X_test,\n",
    "                                             y_test,\n",
    "                                             display_labels=['Not Survived', 'Survived']))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Logistic regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 96 candidates, totalling 480 fits\n",
      "{'columntransformer__num_pipe__simpleimputer__strategy': 'mean', 'logisticregression__C': 10, 'logisticregression__penalty': 'l1', 'standardscaler__with_mean': True, 'standardscaler__with_std': True} {'dtree_iqr': 0.9398717092851129, 'knn_iqr': 0.9398775142948365, 'lr_iqr': 0.9421443705918208}\n",
      "0.952054794520548\n",
      "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay object at 0x000002EF18E09DE0>\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "# Full pipeline: preprocessor + model\n",
    "full_pipeline_lr_iqr = make_pipeline(full_preprocessing,\n",
    "                                         IQRTransformer(),\n",
    "                                         StandardScaler(),\n",
    "                                         LogisticRegression(max_iter=1000, solver='saga', tol=0.1))\n",
    "\n",
    "# Define parameter grid\n",
    "param_grid_lr = {\n",
    "    \"columntransformer__num_pipe__simpleimputer__strategy\":[\"mean\", \"median\", \"most_frequent\"],\n",
    "    \"standardscaler__with_mean\":[True, False],\n",
    "    \"standardscaler__with_std\":[True, False],\n",
    "    'logisticregression__C': [0.1, 1, 10, 100],\n",
    "    'logisticregression__penalty': ['l1', 'l2']\n",
    "}\n",
    "\n",
    "# Define GridSearchCV\n",
    "search_lr = GridSearchCV(full_pipeline_lr_iqr,\n",
    "                                param_grid_lr,\n",
    "                                cv=5,\n",
    "                                verbose=1)\n",
    "\n",
    "search_lr.fit(X_train, y_train)\n",
    "\n",
    "\n",
    "# Create a dictionary to store the best scores from the different models\n",
    "scores[\"lr_iqr\"]= search_lr.best_score_\n",
    "\n",
    "best_param = search_lr.best_params_\n",
    "\n",
    "print(best_param, scores)\n",
    "\n",
    "# Test accuracy\n",
    "print(accuracy_score(search_lr.predict(X_test), y_test))\n",
    "\n",
    "# Plot confusion matrix\n",
    "print(ConfusionMatrixDisplay.from_estimator(search_lr,\n",
    "                                             X_test,\n",
    "                                             y_test,\n",
    "                                             display_labels=['Not Survived', 'Survived']))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### RandomForestClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 336 candidates, totalling 1680 fits\n",
      "{'columntransformer__num_pipe__simpleimputer__strategy': 'most_frequent', 'randomforestclassifier__max_depth': 30, 'randomforestclassifier__max_features': 9, 'standardscaler__with_mean': True, 'standardscaler__with_std': False} {'dtree_iqr': 0.9398717092851129, 'knn_iqr': 0.9398775142948365, 'lr_iqr': 0.9421443705918208, 'rf_iqr': 0.9543348910109424, 'svc_iqr': 0.95660755231765}\n",
      "0.958904109589041\n",
      "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay object at 0x000002EF197B2170>\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Full pipeline: preprocessor + model\n",
    "full_pipeline_rf_iqr = make_pipeline(full_preprocessing,\n",
    "                                         IQRTransformer(),\n",
    "                                         StandardScaler(),\n",
    "                                         RandomForestClassifier())\n",
    "\n",
    "# Define parameter grid\n",
    "param_grid_rf = {\"columntransformer__num_pipe__simpleimputer__strategy\":[\"mean\", \"median\", \"most_frequent\"],\n",
    "                 \"standardscaler__with_mean\":[True, False],\n",
    "                 \"standardscaler__with_std\":[True, False],\n",
    "                 'randomforestclassifier__max_depth': list(range(15, 35, 5)), \n",
    "                 'randomforestclassifier__max_features': list(range(3,10))}\n",
    "\n",
    "\n",
    "# Define GridSearchCV\n",
    "search_rf = GridSearchCV(full_pipeline_rf_iqr,\n",
    "                                param_grid_rf,\n",
    "                                cv=5,\n",
    "                                verbose=1)\n",
    "\n",
    "search_rf.fit(X_train, y_train)\n",
    "\n",
    "\n",
    "# Create a dictionary to store the best scores from the different models\n",
    "scores[\"rf_iqr\"]= search_rf.best_score_\n",
    "\n",
    "best_param = search_rf.best_params_\n",
    "\n",
    "print(best_param, scores)\n",
    "\n",
    "# Test accuracy\n",
    "print(accuracy_score(search_rf.predict(X_test), y_test))\n",
    "\n",
    "# Plot confusion matrix\n",
    "print(ConfusionMatrixDisplay.from_estimator(search_rf,\n",
    "                                             X_test,\n",
    "                                             y_test,\n",
    "                                             display_labels=['Not Survived', 'Survived']))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### SVC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 300 candidates, totalling 1500 fits\n",
      "{'columntransformer__num_pipe__simpleimputer__strategy': 'median', 'standardscaler__with_mean': True, 'standardscaler__with_std': True, 'svc__C': 10, 'svc__gamma': 0.01, 'svc__kernel': 'rbf'}\n",
      "0.9452054794520548\n",
      "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay object at 0x000002EF22622800>\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.svm import SVC\n",
    "\n",
    "# Full pipeline: preprocessor + model\n",
    "full_pipeline_svc_iqr = make_pipeline(full_preprocessing,\n",
    "                                         IQRTransformer(),\n",
    "                                         StandardScaler(),\n",
    "                                         SVC())\n",
    "\n",
    "# Define parameter grid\n",
    "param_grid_svc = {\"columntransformer__num_pipe__simpleimputer__strategy\":[\"mean\", \"median\", \"most_frequent\"],\n",
    "                 \"standardscaler__with_mean\":[True, False],\n",
    "                 \"standardscaler__with_std\":[True, False],\n",
    "                  'svc__C': [0.1, 1, 10, 100, 1000],  \n",
    "                  'svc__gamma': [1, 0.1, 0.01, 0.001, 0.0001], \n",
    "                  'svc__kernel': ['rbf']}  \n",
    "\n",
    "# Define GridSearchCV\n",
    "search_svc = GridSearchCV(full_pipeline_svc_iqr,\n",
    "                                param_grid_svc,\n",
    "                                cv=5,\n",
    "                                verbose=1)\n",
    "\n",
    "search_svc.fit(X_train, y_train)\n",
    "\n",
    "\n",
    "# Create a dictionary to store the best scores from the different models\n",
    "scores[\"svc_iqr\"]= search_svc.best_score_\n",
    "\n",
    "best_param = search_svc.best_params_\n",
    "\n",
    "print(best_param)\n",
    "\n",
    "# Test accuracy\n",
    "print(accuracy_score(search_svc.predict(X_test), y_test))\n",
    "\n",
    "# Plot confusion matrix\n",
    "print(ConfusionMatrixDisplay.from_estimator(search_svc,\n",
    "                                             X_test,\n",
    "                                             y_test,\n",
    "                                             display_labels=['Not Survived', 'Survived']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'dtree_iqr': 0.9398717092851129,\n",
       " 'knn_iqr': 0.9398775142948365,\n",
       " 'lr_iqr': 0.9421443705918208,\n",
       " 'rf_iqr': 0.956616259832235,\n",
       " 'svc_iqr': 0.95660755231765}"
      ]
     },
     "execution_count": 493,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scores"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "{'dtree_iqr': 0.9426323319027181,\n",
    " 'knn_iqr': 0.9426323319027181,\n",
    " 'lr_iqr': 0.9563222185539783,\n",
    " 'rf_iqr': 0.9546128168445766,\n",
    " 'svc_iqr': 0.9537691207219104}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Submit the best model to the competition"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#loading the new test data for the competition\n",
    "url = \"https://drive.google.com/file/d/1HMsEK5Kq7CcGFz7_GyQyI4ziEr3iAu8x/view?usp=drive_link\"\n",
    "path = 'https://drive.google.com/uc?export=download&id='+url.split('/')[-2]\n",
    "competition_test = pd.read_csv(path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Id = competition_test.pop(\"Id\")\n",
    "\n",
    "competition_test = competition_test.drop([\"PoolQC\", \"MiscFeature\", \"Alley\", \"Fence\"], axis=1)\n",
    "\n",
    "# Assuming df is your DataFrame\n",
    "labels = ['SoldBefore05', 'SoldBetween05-15', 'SoldAfter15']\n",
    "\n",
    "# Define a custom function to assign categories based on the 'YrSold' values\n",
    "def categorize_year_sold(year_sold):\n",
    "    if year_sold < 2005:\n",
    "        return labels[0]\n",
    "    elif 2005 <= year_sold <= 2015:\n",
    "        return labels[1]\n",
    "    else:\n",
    "        return labels[2]\n",
    "\n",
    "# Apply the custom function to create the new categorical column\n",
    "competition_test['YrSold'] = competition_test['YrSold'].apply(categorize_year_sold)\n",
    "\n",
    "\n",
    "# Define a custom function to assign seasons based on the 'MoSold' values\n",
    "def categorize_season(month_sold):\n",
    "    if month_sold in [12, 1, 2]:\n",
    "        return 'Winter'\n",
    "    elif month_sold in [3, 4, 5]:\n",
    "        return 'Spring'\n",
    "    elif month_sold in [6, 7, 8]:\n",
    "        return 'Summer'\n",
    "    else:\n",
    "        return 'Fall'\n",
    "\n",
    "# Apply the custom function to create the new categorical column\n",
    "competition_test['MoSold'] = competition_test['MoSold'].apply(categorize_season)\n",
    "\n",
    "# Restrict the columns\n",
    "competition_test = competition_test[selected_features]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# make predictions with the brand new (unlabelled) dataset for the competition!\n",
    "submission_prediction = search_rf.predict(competition_test)\n",
    "submission_prediction = list(submission_prediction)\n",
    "\n",
    "# Create the Dataframe for the submission. It must contain two columns: \"Id\" and \"Expensive\"\n",
    "submission_df = pd.DataFrame()\n",
    "submission_df['Id'] = Id\n",
    "submission_df['Expensive'] = submission_prediction\n",
    "\n",
    "# create the csv for the competition and download it\n",
    "submission_df.to_csv(\"classification_1.csv\", index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# make predictions with the brand new (unlabelled) dataset for the competition!\n",
    "submission_prediction = search_lr.predict(competition_test)\n",
    "submission_prediction = list(submission_prediction)\n",
    "\n",
    "# Create the Dataframe for the submission. It must contain two columns: \"Id\" and \"Expensive\"\n",
    "submission_df = pd.DataFrame()\n",
    "submission_df['Id'] = Id\n",
    "submission_df['Expensive'] = submission_prediction\n",
    "\n",
    "# create the csv for the competition and download it\n",
    "submission_df.to_csv(\"classification_2.csv\", index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# make predictions with the brand new (unlabelled) dataset for the competition!\n",
    "submission_prediction = search_svc.predict(competition_test)\n",
    "submission_prediction = list(submission_prediction)\n",
    "\n",
    "# Create the Dataframe for the submission. It must contain two columns: \"Id\" and \"Expensive\"\n",
    "submission_df = pd.DataFrame()\n",
    "submission_df['Id'] = Id\n",
    "submission_df['Expensive'] = submission_prediction\n",
    "\n",
    "# create the csv for the competition and download it\n",
    "submission_df.to_csv(\"classification_3.csv\", index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# make predictions with the brand new (unlabelled) dataset for the competition!\n",
    "submission_prediction = search_svc.predict(competition_test)\n",
    "submission_prediction = list(submission_prediction)\n",
    "\n",
    "# Create the Dataframe for the submission. It must contain two columns: \"Id\" and \"Expensive\"\n",
    "submission_df = pd.DataFrame()\n",
    "submission_df['Id'] = Id\n",
    "submission_df['Expensive'] = submission_prediction\n",
    "\n",
    "# create the csv for the competition and download it\n",
    "submission_df.to_csv(\"classification_3.csv\", index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "wbs_env",
   "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": "3.10.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}