Skip to content

Instantly share code, notes, and snippets.

@thedatajango

thedatajango/end_to_end_DS_project.ipynb

Last active May 7, 2022 05:07
Show Gist options
  • Star 1 You must be signed in to star a gist
  • Fork 3 You must be signed in to fork a gist
  • Save thedatajango/d5bf238df20dc0cf76bbb9e25ed1ec86 to your computer and use it in GitHub Desktop.
Save thedatajango/d5bf238df20dc0cf76bbb9e25ed1ec86 to your computer and use it in GitHub Desktop.
Download ZIP
Step-by-Step Data Science project execution
Display the source blob
Display the rendered blob
Raw
end_to_end_DS_project.ipynb
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Problem Statement : Predict Melbourne house price using historic data."
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"import pandas as pd\n",
"import numpy as np\n",
"import matplotlib.pyplot as plt\n",
"import seaborn as sns\n",
"from sklearn.linear_model import LinearRegression\n",
"from sklearn.linear_model import SGDRegressor"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Load Housing Price Prediction data from csv file"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"data = pd.read_csv('./data/Melbourne_housing_FULL.csv')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Suburb: Suburb\n",
"\n",
"Address: Address\n",
"\n",
"Rooms: Number of rooms\n",
"\n",
"Price: Price in Australian dollars\n",
"\n",
"Method: S - property sold; SP - property sold prior; PI - property passed in; PN - sold prior not disclosed; SN - sold not disclosed; NB - no bid; VB - vendor bid; W - withdrawn prior to auction; SA - sold after auction; SS - sold after auction price not disclosed. N/A - price or highest bid not available.\n",
"\n",
"Type: br - bedroom(s); h - house,cottage,villa, semi,terrace; u - unit, duplex; t - townhouse; dev site - development site; o res - other residential.\n",
"\n",
"SellerG: Real Estate Agent\n",
"\n",
"Date: Date sold\n",
"\n",
"Distance: Distance from CBD in Kilometres\n",
"\n",
"Regionname: General Region (West, North West, North, North east ...etc)\n",
"\n",
"Propertycount: Number of properties that exist in the suburb.\n",
"\n",
"Bedroom2 : Scraped # of Bedrooms (from different source)\n",
"\n",
"Bathroom: Number of Bathrooms\n",
"\n",
"Car: Number of carspots\n",
"\n",
"Landsize: Land Size in Metres\n",
"\n",
"BuildingArea: Building Size in Metres\n",
"\n",
"YearBuilt: Year the house was built\n",
"\n",
"CouncilArea: Governing council for the area\n",
"\n",
"Lattitude: Self explanitory\n",
"\n",
"Longtitude: Self explanitory\n",
"\n",
"Postcode: Postal Code"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Let's have a look at shape of the dataframe"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(34857, 21)"
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"data.shape"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Let's look at column names in dataset"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Index(['Suburb', 'Address', 'Rooms', 'Type', 'Price', 'Method', 'SellerG',\n",
" 'Date', 'Distance', 'Postcode', 'Bedroom2', 'Bathroom', 'Car',\n",
" 'Landsize', 'BuildingArea', 'YearBuilt', 'CouncilArea', 'Lattitude',\n",
" 'Longtitude', 'Regionname', 'Propertycount'],\n",
" dtype='object')"
]
},
"execution_count": 4,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"data.columns"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Check if all records have \"Price\". Filter records with \"Price\" and take that as training set."
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"27247"
]
},
"execution_count": 5,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"data.Price.count() # Some historic records have missing values in target variable."
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [],
"source": [
"data_no_nulls_in_target = data[~ data.Price.isnull() ] ## Eliminated recores with \"Price\" null"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(27247, 21)"
]
},
"execution_count": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"data_no_nulls_in_target.shape"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [],
"source": [
"data_no_nulls_in_target.reset_index(drop=True, inplace=True)"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [],
"source": [
"data_no_nulls_in_target.to_csv(index=False, path_or_buf='./data/Melbourne_housing.csv')"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.model_selection import ShuffleSplit\n",
"shuffleSplit = ShuffleSplit(n_splits=1,test_size = 0.2 , random_state=42)\n",
"\n",
"for train_index, test_index in shuffleSplit.split(data_no_nulls_in_target):\n",
" training_set = data_no_nulls_in_target.loc[train_index]\n",
" test_set = data_no_nulls_in_target.loc[test_index]"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(21797, 21)"
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set.shape"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Just to get an idea, let's have a look at first 5 records"
]
},
{
"cell_type": "code",
"execution_count": 12,
"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>Suburb</th>\n",
" <th>Address</th>\n",
" <th>Rooms</th>\n",
" <th>Type</th>\n",
" <th>Price</th>\n",
" <th>Method</th>\n",
" <th>SellerG</th>\n",
" <th>Date</th>\n",
" <th>Distance</th>\n",
" <th>Postcode</th>\n",
" <th>...</th>\n",
" <th>Bathroom</th>\n",
" <th>Car</th>\n",
" <th>Landsize</th>\n",
" <th>BuildingArea</th>\n",
" <th>YearBuilt</th>\n",
" <th>CouncilArea</th>\n",
" <th>Lattitude</th>\n",
" <th>Longtitude</th>\n",
" <th>Regionname</th>\n",
" <th>Propertycount</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>17957</th>\n",
" <td>Brunswick East</td>\n",
" <td>2/20 Lyndhurst Cr</td>\n",
" <td>2</td>\n",
" <td>u</td>\n",
" <td>580000.0</td>\n",
" <td>VB</td>\n",
" <td>Nelson</td>\n",
" <td>26/08/2017</td>\n",
" <td>4.0</td>\n",
" <td>3057.0</td>\n",
" <td>...</td>\n",
" <td>1.0</td>\n",
" <td>1.0</td>\n",
" <td>NaN</td>\n",
" <td>61.0</td>\n",
" <td>1970.0</td>\n",
" <td>Moreland City Council</td>\n",
" <td>-37.76134</td>\n",
" <td>144.97796</td>\n",
" <td>Northern Metropolitan</td>\n",
" <td>5533.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>6753</th>\n",
" <td>Richmond</td>\n",
" <td>21 Lambert St</td>\n",
" <td>3</td>\n",
" <td>t</td>\n",
" <td>1025000.0</td>\n",
" <td>SP</td>\n",
" <td>Jellis</td>\n",
" <td>3/12/2016</td>\n",
" <td>2.6</td>\n",
" <td>3121.0</td>\n",
" <td>...</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>Yarra City Council</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>Northern Metropolitan</td>\n",
" <td>14949.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1833</th>\n",
" <td>Brunswick West</td>\n",
" <td>11/7 Egginton St</td>\n",
" <td>2</td>\n",
" <td>t</td>\n",
" <td>380000.0</td>\n",
" <td>VB</td>\n",
" <td>Nelson</td>\n",
" <td>7/11/2016</td>\n",
" <td>5.9</td>\n",
" <td>3055.0</td>\n",
" <td>...</td>\n",
" <td>1.0</td>\n",
" <td>1.0</td>\n",
" <td>0.0</td>\n",
" <td>85.0</td>\n",
" <td>1970.0</td>\n",
" <td>Moreland City Council</td>\n",
" <td>-37.76070</td>\n",
" <td>144.93930</td>\n",
" <td>Northern Metropolitan</td>\n",
" <td>7082.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>19795</th>\n",
" <td>Pascoe Vale</td>\n",
" <td>3/246 Cumberland Rd</td>\n",
" <td>2</td>\n",
" <td>u</td>\n",
" <td>456000.0</td>\n",
" <td>S</td>\n",
" <td>Brad</td>\n",
" <td>21/10/2017</td>\n",
" <td>8.5</td>\n",
" <td>3044.0</td>\n",
" <td>...</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>Moreland City Council</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>Northern Metropolitan</td>\n",
" <td>7485.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>17054</th>\n",
" <td>Glen Waverley</td>\n",
" <td>31 Paxton Dr</td>\n",
" <td>3</td>\n",
" <td>h</td>\n",
" <td>1300000.0</td>\n",
" <td>S</td>\n",
" <td>hockingstuart</td>\n",
" <td>19/08/2017</td>\n",
" <td>16.7</td>\n",
" <td>3150.0</td>\n",
" <td>...</td>\n",
" <td>2.0</td>\n",
" <td>3.0</td>\n",
" <td>733.0</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>Monash City Council</td>\n",
" <td>-37.89014</td>\n",
" <td>145.18269</td>\n",
" <td>Eastern Metropolitan</td>\n",
" <td>15321.0</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"<p>5 rows × 21 columns</p>\n",
"</div>"
],
"text/plain": [
" Suburb Address Rooms Type Price Method \\\n",
"17957 Brunswick East 2/20 Lyndhurst Cr 2 u 580000.0 VB \n",
"6753 Richmond 21 Lambert St 3 t 1025000.0 SP \n",
"1833 Brunswick West 11/7 Egginton St 2 t 380000.0 VB \n",
"19795 Pascoe Vale 3/246 Cumberland Rd 2 u 456000.0 S \n",
"17054 Glen Waverley 31 Paxton Dr 3 h 1300000.0 S \n",
"\n",
" SellerG Date Distance Postcode ... Bathroom Car \\\n",
"17957 Nelson 26/08/2017 4.0 3057.0 ... 1.0 1.0 \n",
"6753 Jellis 3/12/2016 2.6 3121.0 ... NaN NaN \n",
"1833 Nelson 7/11/2016 5.9 3055.0 ... 1.0 1.0 \n",
"19795 Brad 21/10/2017 8.5 3044.0 ... NaN NaN \n",
"17054 hockingstuart 19/08/2017 16.7 3150.0 ... 2.0 3.0 \n",
"\n",
" Landsize BuildingArea YearBuilt CouncilArea Lattitude \\\n",
"17957 NaN 61.0 1970.0 Moreland City Council -37.76134 \n",
"6753 NaN NaN NaN Yarra City Council NaN \n",
"1833 0.0 85.0 1970.0 Moreland City Council -37.76070 \n",
"19795 NaN NaN NaN Moreland City Council NaN \n",
"17054 733.0 NaN NaN Monash City Council -37.89014 \n",
"\n",
" Longtitude Regionname Propertycount \n",
"17957 144.97796 Northern Metropolitan 5533.0 \n",
"6753 NaN Northern Metropolitan 14949.0 \n",
"1833 144.93930 Northern Metropolitan 7082.0 \n",
"19795 NaN Northern Metropolitan 7485.0 \n",
"17054 145.18269 Eastern Metropolitan 15321.0 \n",
"\n",
"[5 rows x 21 columns]"
]
},
"execution_count": 12,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set.head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Exploratory Data Analysis\n",
"### As part of data analysis we refine the data - below are some common activities we do.\n",
"\n",
"- **Missing : ** Check for missing or incomplete data, impute/fillna with appropriate data\n",
"- **Quality : ** Check for duplicates, accuracy, unusual data.\n",
"- **Parse : ** Prase existing data and create new fearures. e.g. Extract year and month from date\n",
"- **Convert : ** Free text to coded value (LabelEncoder, One-Hot-Encoding or LabelBinarizer)\n",
"- **Derive** Derive new feature out of existing featre/fearues e.g. gender from title Mr. Mrs.\n",
"- **Calculate** percentages, proportion\n",
"- **Remove** Remove redundant or not so useful data\n",
"- **Merge** Merge multiple columns e.g. first and surname for full name\n",
"- **Aggregate** e.g. rollup by year, cluster by area\n",
"- **Filter** e.g. exclude based on location\n",
"- **Sample** e.g. extract a representative data\n",
"- **Summary** Pandas describe function or stats like mean"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Check if any columns have got nulls"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Suburb False\n",
"Address False\n",
"Rooms False\n",
"Type False\n",
"Price False\n",
"Method False\n",
"SellerG False\n",
"Date False\n",
"Distance True\n",
"Postcode True\n",
"Bedroom2 True\n",
"Bathroom True\n",
"Car True\n",
"Landsize True\n",
"BuildingArea True\n",
"YearBuilt True\n",
"CouncilArea True\n",
"Lattitude True\n",
"Longtitude True\n",
"Regionname True\n",
"Propertycount True\n",
"dtype: bool"
]
},
"execution_count": 13,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set.isnull().any()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Look at distinct data types in our data"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"float64 12\n",
"object 8\n",
"int64 1\n",
"dtype: int64"
]
},
"execution_count": 14,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set.dtypes.value_counts()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Check if numerical columns have got nulls. \n",
" * If we see below, most of the numerical columns have got nulls. \n",
" * Some features like \"Postcode\", \"Lattitude\", and \"Longtitude\" are related features. These three reatures are related to \"Address\". I am thinking if we can use \"Postcode\" feature to represent all there."
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Rooms False\n",
"Price False\n",
"Distance True\n",
"Postcode True\n",
"Bedroom2 True\n",
"Bathroom True\n",
"Car True\n",
"Landsize True\n",
"BuildingArea True\n",
"YearBuilt True\n",
"Lattitude True\n",
"Longtitude True\n",
"Propertycount True\n",
"dtype: bool"
]
},
"execution_count": 15,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set.select_dtypes(['float64','int64']).isnull().any()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Check if categorical/text columns have got nulls. \n",
" * \"CouncilArea\", \"Regionname\" have got nulls. These two reatures are related to \"Address\". I am thinking if we can use \"Postcode\" feature to represent all there."
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Suburb False\n",
"Address False\n",
"Type False\n",
"Method False\n",
"SellerG False\n",
"Date False\n",
"CouncilArea True\n",
"Regionname True\n",
"dtype: bool"
]
},
"execution_count": 16,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set.select_dtypes(['object']).isnull().any()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Big Decision (importance of Statistical Data Analysis - Bivariate data analysis):\n",
" #### Should we remove \"CouncilArea\", \"Regionname\", \"Lattitude\", and \"Longtitude\" features and have \"Postcode\" or \"Suburb\" as it may represent the address - in other workds, we need to check if \"Postcode\" or \"Suburb\" an important feature in predicting \"Price\" ?\n",
" * Before we make a decision we have to fix some problems with \"Postcode\" column by do some statistical data analysis\n",
"\n",
"#### Step 1: Check \"Postcode\" equal for all the addresses in same \"Suburb\" (a column with no-nulls). If so, we can write some logic to populate \"Postcode\" based on \"Suburb\".\n",
"#### Step 2: Is \"Postcode\", \"Suburb\" an important feature to predict \"Price\" ?."
]
},
{
"cell_type": "code",
"execution_count": 27,
"metadata": {},
"outputs": [],
"source": [
"df_null_postcodes = training_set[training_set['Postcode'].isnull()]"
]
},
{
"cell_type": "code",
"execution_count": 28,
"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>Suburb</th>\n",
" <th>Address</th>\n",
" <th>Rooms</th>\n",
" <th>Type</th>\n",
" <th>Price</th>\n",
" <th>Method</th>\n",
" <th>SellerG</th>\n",
" <th>Date</th>\n",
" <th>Distance</th>\n",
" <th>Postcode</th>\n",
" <th>...</th>\n",
" <th>Bathroom</th>\n",
" <th>Car</th>\n",
" <th>Landsize</th>\n",
" <th>BuildingArea</th>\n",
" <th>YearBuilt</th>\n",
" <th>CouncilArea</th>\n",
" <th>Lattitude</th>\n",
" <th>Longtitude</th>\n",
" <th>Regionname</th>\n",
" <th>Propertycount</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>23051</th>\n",
" <td>Fawkner Lot</td>\n",
" <td>1/3 Brian St</td>\n",
" <td>3</td>\n",
" <td>h</td>\n",
" <td>616000.0</td>\n",
" <td>SP</td>\n",
" <td>Brad</td>\n",
" <td>6/01/2018</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>...</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"<p>1 rows × 21 columns</p>\n",
"</div>"
],
"text/plain": [
" Suburb Address Rooms Type Price Method SellerG \\\n",
"23051 Fawkner Lot 1/3 Brian St 3 h 616000.0 SP Brad \n",
"\n",
" Date Distance Postcode ... Bathroom Car Landsize \\\n",
"23051 6/01/2018 NaN NaN ... NaN NaN NaN \n",
"\n",
" BuildingArea YearBuilt CouncilArea Lattitude Longtitude Regionname \\\n",
"23051 NaN NaN NaN NaN NaN NaN \n",
"\n",
" Propertycount \n",
"23051 NaN \n",
"\n",
"[1 rows x 21 columns]"
]
},
"execution_count": 28,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df_null_postcodes"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Check if any row with non-null \"Suburb\" is present for row in \"Postcode\""
]
},
{
"cell_type": "code",
"execution_count": 29,
"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>Suburb</th>\n",
" <th>Address</th>\n",
" <th>Rooms</th>\n",
" <th>Type</th>\n",
" <th>Price</th>\n",
" <th>Method</th>\n",
" <th>SellerG</th>\n",
" <th>Date</th>\n",
" <th>Distance</th>\n",
" <th>Postcode</th>\n",
" <th>...</th>\n",
" <th>Bathroom</th>\n",
" <th>Car</th>\n",
" <th>Landsize</th>\n",
" <th>BuildingArea</th>\n",
" <th>YearBuilt</th>\n",
" <th>CouncilArea</th>\n",
" <th>Lattitude</th>\n",
" <th>Longtitude</th>\n",
" <th>Regionname</th>\n",
" <th>Propertycount</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>23051</th>\n",
" <td>Fawkner Lot</td>\n",
" <td>1/3 Brian St</td>\n",
" <td>3</td>\n",
" <td>h</td>\n",
" <td>616000.0</td>\n",
" <td>SP</td>\n",
" <td>Brad</td>\n",
" <td>6/01/2018</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>...</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"<p>1 rows × 21 columns</p>\n",
"</div>"
],
"text/plain": [
" Suburb Address Rooms Type Price Method SellerG \\\n",
"23051 Fawkner Lot 1/3 Brian St 3 h 616000.0 SP Brad \n",
"\n",
" Date Distance Postcode ... Bathroom Car Landsize \\\n",
"23051 6/01/2018 NaN NaN ... NaN NaN NaN \n",
"\n",
" BuildingArea YearBuilt CouncilArea Lattitude Longtitude Regionname \\\n",
"23051 NaN NaN NaN NaN NaN NaN \n",
"\n",
" Propertycount \n",
"23051 NaN \n",
"\n",
"[1 rows x 21 columns]"
]
},
"execution_count": 29,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df_null_postcodes[~ df_null_postcodes['Suburb'].isnull()]"
]
},
{
"cell_type": "code",
"execution_count": 30,
"metadata": {},
"outputs": [],
"source": [
"training_set = training_set[~ training_set['Postcode'].isnull()]"
]
},
{
"cell_type": "code",
"execution_count": 31,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(21796, 21)"
]
},
"execution_count": 31,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set.shape"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Check if distict \"Suburb\" have distict \"Postcode\" "
]
},
{
"cell_type": "code",
"execution_count": 35,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(337,)"
]
},
"execution_count": 35,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set['Suburb'].value_counts().shape"
]
},
{
"cell_type": "code",
"execution_count": 36,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(205,)"
]
},
"execution_count": 36,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set['Postcode'].value_counts().shape"
]
},
{
"cell_type": "code",
"execution_count": 38,
"metadata": {},
"outputs": [],
"source": [
"ser_postcode_suburb = training_set.groupby(['Postcode', 'Suburb']).size()"
]
},
{
"cell_type": "code",
"execution_count": 40,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"pandas.core.series.Series"
]
},
"execution_count": 40,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"type(ser_postcode_suburb)"
]
},
{
"cell_type": "code",
"execution_count": 26,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Postcode Suburb \n",
"3000.0 Melbourne 117\n",
"3002.0 East Melbourne 30\n",
"3003.0 West Melbourne 37\n",
"3006.0 Southbank 46\n",
"3008.0 Docklands 6\n",
"3011.0 Footscray 168\n",
" Seddon 76\n",
"3012.0 Brooklyn 31\n",
" Kingsville 50\n",
" Maidstone 118\n",
" West Footscray 128\n",
"3013.0 Yarraville 194\n",
"3015.0 Newport 168\n",
" South Kingsville 24\n",
" Spotswood 38\n",
"3016.0 Williamstown 140\n",
" Williamstown North 19\n",
"3018.0 Altona 87\n",
" Seaholme 17\n",
"3019.0 Braybrook 64\n",
"dtype: int64"
]
},
"execution_count": 26,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"ser_postcode_suburb.head(20)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### reset_index on pandas Series object will create a DataFrame when drop flag is set to \"False\"\n",
"https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.Series.reset_index.html"
]
},
{
"cell_type": "code",
"execution_count": 27,
"metadata": {},
"outputs": [],
"source": [
"df_postcode_suburb = training_set.groupby(['Postcode', 'Suburb']).size().reset_index().rename(columns={0:'count'})"
]
},
{
"cell_type": "code",
"execution_count": 28,
"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>Postcode</th>\n",
" <th>Suburb</th>\n",
" <th>count</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>3000.0</td>\n",
" <td>Melbourne</td>\n",
" <td>117</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>3002.0</td>\n",
" <td>East Melbourne</td>\n",
" <td>30</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>3003.0</td>\n",
" <td>West Melbourne</td>\n",
" <td>37</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>3006.0</td>\n",
" <td>Southbank</td>\n",
" <td>46</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>3008.0</td>\n",
" <td>Docklands</td>\n",
" <td>6</td>\n",
" </tr>\n",
" <tr>\n",
" <th>5</th>\n",
" <td>3011.0</td>\n",
" <td>Footscray</td>\n",
" <td>168</td>\n",
" </tr>\n",
" <tr>\n",
" <th>6</th>\n",
" <td>3011.0</td>\n",
" <td>Seddon</td>\n",
" <td>76</td>\n",
" </tr>\n",
" <tr>\n",
" <th>7</th>\n",
" <td>3012.0</td>\n",
" <td>Brooklyn</td>\n",
" <td>31</td>\n",
" </tr>\n",
" <tr>\n",
" <th>8</th>\n",
" <td>3012.0</td>\n",
" <td>Kingsville</td>\n",
" <td>50</td>\n",
" </tr>\n",
" <tr>\n",
" <th>9</th>\n",
" <td>3012.0</td>\n",
" <td>Maidstone</td>\n",
" <td>118</td>\n",
" </tr>\n",
" <tr>\n",
" <th>10</th>\n",
" <td>3012.0</td>\n",
" <td>West Footscray</td>\n",
" <td>128</td>\n",
" </tr>\n",
" <tr>\n",
" <th>11</th>\n",
" <td>3013.0</td>\n",
" <td>Yarraville</td>\n",
" <td>194</td>\n",
" </tr>\n",
" <tr>\n",
" <th>12</th>\n",
" <td>3015.0</td>\n",
" <td>Newport</td>\n",
" <td>168</td>\n",
" </tr>\n",
" <tr>\n",
" <th>13</th>\n",
" <td>3015.0</td>\n",
" <td>South Kingsville</td>\n",
" <td>24</td>\n",
" </tr>\n",
" <tr>\n",
" <th>14</th>\n",
" <td>3015.0</td>\n",
" <td>Spotswood</td>\n",
" <td>38</td>\n",
" </tr>\n",
" <tr>\n",
" <th>15</th>\n",
" <td>3016.0</td>\n",
" <td>Williamstown</td>\n",
" <td>140</td>\n",
" </tr>\n",
" <tr>\n",
" <th>16</th>\n",
" <td>3016.0</td>\n",
" <td>Williamstown North</td>\n",
" <td>19</td>\n",
" </tr>\n",
" <tr>\n",
" <th>17</th>\n",
" <td>3018.0</td>\n",
" <td>Altona</td>\n",
" <td>87</td>\n",
" </tr>\n",
" <tr>\n",
" <th>18</th>\n",
" <td>3018.0</td>\n",
" <td>Seaholme</td>\n",
" <td>17</td>\n",
" </tr>\n",
" <tr>\n",
" <th>19</th>\n",
" <td>3019.0</td>\n",
" <td>Braybrook</td>\n",
" <td>64</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" Postcode Suburb count\n",
"0 3000.0 Melbourne 117\n",
"1 3002.0 East Melbourne 30\n",
"2 3003.0 West Melbourne 37\n",
"3 3006.0 Southbank 46\n",
"4 3008.0 Docklands 6\n",
"5 3011.0 Footscray 168\n",
"6 3011.0 Seddon 76\n",
"7 3012.0 Brooklyn 31\n",
"8 3012.0 Kingsville 50\n",
"9 3012.0 Maidstone 118\n",
"10 3012.0 West Footscray 128\n",
"11 3013.0 Yarraville 194\n",
"12 3015.0 Newport 168\n",
"13 3015.0 South Kingsville 24\n",
"14 3015.0 Spotswood 38\n",
"15 3016.0 Williamstown 140\n",
"16 3016.0 Williamstown North 19\n",
"17 3018.0 Altona 87\n",
"18 3018.0 Seaholme 17\n",
"19 3019.0 Braybrook 64"
]
},
"execution_count": 28,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df_postcode_suburb.head(20)"
]
},
{
"cell_type": "code",
"execution_count": 29,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Postcode = 3000.0\n",
"*************************\n",
" Suburb count\n",
"0 Melbourne 117\n",
"--------------------------------------------------\n",
"Postcode = 3002.0\n",
"*************************\n",
" Suburb count\n",
"1 East Melbourne 30\n",
"--------------------------------------------------\n",
"Postcode = 3003.0\n",
"*************************\n",
" Suburb count\n",
"2 West Melbourne 37\n",
"--------------------------------------------------\n",
"Postcode = 3006.0\n",
"*************************\n",
" Suburb count\n",
"3 Southbank 46\n",
"--------------------------------------------------\n",
"Postcode = 3008.0\n",
"*************************\n",
" Suburb count\n",
"4 Docklands 6\n",
"--------------------------------------------------\n",
"Postcode = 3011.0\n",
"*************************\n",
" Suburb count\n",
"5 Footscray 168\n",
"6 Seddon 76\n",
"--------------------------------------------------\n",
"Postcode = 3012.0\n",
"*************************\n",
" Suburb count\n",
"7 Brooklyn 31\n",
"8 Kingsville 50\n",
"9 Maidstone 118\n",
"10 West Footscray 128\n",
"--------------------------------------------------\n",
"Postcode = 3013.0\n",
"*************************\n",
" Suburb count\n",
"11 Yarraville 194\n",
"--------------------------------------------------\n",
"Postcode = 3015.0\n",
"*************************\n",
" Suburb count\n",
"12 Newport 168\n",
"13 South Kingsville 24\n",
"14 Spotswood 38\n",
"--------------------------------------------------\n",
"Postcode = 3016.0\n",
"*************************\n",
" Suburb count\n",
"15 Williamstown 140\n",
"16 Williamstown North 19\n",
"--------------------------------------------------\n",
"Postcode = 3018.0\n",
"*************************\n",
" Suburb count\n",
"17 Altona 87\n",
"18 Seaholme 17\n",
"--------------------------------------------------\n",
"Postcode = 3019.0\n",
"*************************\n",
" Suburb count\n",
"19 Braybrook 64\n",
"--------------------------------------------------\n",
"Postcode = 3020.0\n",
"*************************\n",
" Suburb count\n",
"20 Albion 53\n",
"21 Sunshine 137\n",
"22 Sunshine North 97\n",
"23 Sunshine West 131\n",
"--------------------------------------------------\n",
"Postcode = 3021.0\n",
"*************************\n",
" Suburb count\n",
"24 Albanvale 7\n",
"25 Kealba 32\n",
"26 Kings Park 10\n",
"27 St Albans 80\n",
"--------------------------------------------------\n",
"Postcode = 3022.0\n",
"*************************\n",
" Suburb count\n",
"28 Ardeer 17\n",
"--------------------------------------------------\n",
"Postcode = 3023.0\n",
"*************************\n",
" Suburb count\n",
"29 Burnside 4\n",
"30 Burnside Heights 9\n",
"31 Cairnlea 15\n",
"32 Caroline Springs 20\n",
"33 Deer Park 31\n",
"--------------------------------------------------\n",
"Postcode = 3024.0\n",
"*************************\n",
" Suburb count\n",
"34 Wyndham Vale 42\n",
"--------------------------------------------------\n",
"Postcode = 3025.0\n",
"*************************\n",
" Suburb count\n",
"35 Altona North 118\n",
"--------------------------------------------------\n",
"Postcode = 3027.0\n",
"*************************\n",
" Suburb count\n",
"36 Williams Landing 5\n",
"--------------------------------------------------\n",
"Postcode = 3028.0\n",
"*************************\n",
" Suburb count\n",
"37 Altona Meadows 21\n",
"38 Laverton 6\n",
"39 Seabrook 3\n",
"--------------------------------------------------\n",
"Postcode = 3029.0\n",
"*************************\n",
" Suburb count\n",
"40 Hoppers Crossing 107\n",
"41 Tarneit 29\n",
"42 Truganina 18\n",
"--------------------------------------------------\n",
"Postcode = 3030.0\n",
"*************************\n",
" Suburb count\n",
"43 Derrimut 14\n",
"44 Point Cook 68\n",
"45 Werribee 120\n",
"46 Werribee South 2\n",
"--------------------------------------------------\n",
"Postcode = 3031.0\n",
"*************************\n",
" Suburb count\n",
"47 Flemington 80\n",
"48 Kensington 150\n",
"--------------------------------------------------\n",
"Postcode = 3032.0\n",
"*************************\n",
" Suburb count\n",
"49 Ascot Vale 167\n",
"50 Maribyrnong 146\n",
"51 Travancore 10\n",
"--------------------------------------------------\n",
"Postcode = 3033.0\n",
"*************************\n",
" Suburb count\n",
"52 Keilor East 172\n",
"--------------------------------------------------\n",
"Postcode = 3034.0\n",
"*************************\n",
" Suburb count\n",
"53 Avondale Heights 117\n",
"--------------------------------------------------\n",
"Postcode = 3036.0\n",
"*************************\n",
" Suburb count\n",
"54 Keilor 13\n",
"--------------------------------------------------\n",
"Postcode = 3037.0\n",
"*************************\n",
" Suburb count\n",
"55 Delahey 17\n",
"56 Hillside 54\n",
"57 Sydenham 21\n",
"58 Taylors Hill 28\n",
"--------------------------------------------------\n",
"Postcode = 3038.0\n",
"*************************\n",
" Suburb count\n",
"59 Keilor Downs 35\n",
"60 Keilor Lodge 5\n",
"61 Taylors Lakes 51\n",
"--------------------------------------------------\n",
"Postcode = 3039.0\n",
"*************************\n",
" Suburb count\n",
"62 Moonee Ponds 201\n",
"--------------------------------------------------\n",
"Postcode = 3040.0\n",
"*************************\n",
" Suburb count\n",
"63 Aberfeldie 58\n",
"64 Essendon 291\n",
"65 Essendon West 31\n",
"--------------------------------------------------\n",
"Postcode = 3041.0\n",
"*************************\n",
" Suburb count\n",
"66 Essendon North 33\n",
"67 Strathmore 115\n",
"68 Strathmore Heights 11\n",
"--------------------------------------------------\n",
"Postcode = 3042.0\n",
"*************************\n",
" Suburb count\n",
"69 Airport West 125\n",
"70 Keilor Park 29\n",
"71 Niddrie 100\n",
"--------------------------------------------------\n",
"Postcode = 3043.0\n",
"*************************\n",
" Suburb count\n",
"72 Gladstone Park 44\n",
"73 Gowanbrae 33\n",
"74 Tullamarine 41\n",
"--------------------------------------------------\n",
"Postcode = 3044.0\n",
"*************************\n",
" Suburb count\n",
"75 Pascoe Vale 253\n",
"--------------------------------------------------\n",
"Postcode = 3046.0\n",
"*************************\n",
" Suburb count\n",
"76 Glenroy 275\n",
"77 Hadfield 66\n",
"78 Oak Park 87\n",
"--------------------------------------------------\n",
"Postcode = 3047.0\n",
"*************************\n",
" Suburb count\n",
"79 Broadmeadows 62\n",
"80 Dallas 27\n",
"81 Jacana 44\n",
"--------------------------------------------------\n",
"Postcode = 3048.0\n",
"*************************\n",
" Suburb count\n",
"82 Coolaroo 7\n",
"83 Meadow Heights 38\n",
"--------------------------------------------------\n",
"Postcode = 3049.0\n",
"*************************\n",
" Suburb count\n",
"84 Attwood 12\n",
"85 Westmeadows 33\n",
"--------------------------------------------------\n",
"Postcode = 3051.0\n",
"*************************\n",
" Suburb count\n",
"86 North Melbourne 95\n",
"--------------------------------------------------\n",
"Postcode = 3052.0\n",
"*************************\n",
" Suburb count\n",
"87 Parkville 42\n",
"--------------------------------------------------\n",
"Postcode = 3053.0\n",
"*************************\n",
" Suburb count\n",
"88 Carlton 57\n",
"--------------------------------------------------\n",
"Postcode = 3054.0\n",
"*************************\n",
" Suburb count\n",
"89 Carlton North 57\n",
"90 Princes Hill 10\n",
"--------------------------------------------------\n",
"Postcode = 3055.0\n",
"*************************\n",
" Suburb count\n",
"91 Brunswick West 169\n",
"--------------------------------------------------\n",
"Postcode = 3056.0\n",
"*************************\n",
" Suburb count\n",
"92 Brunswick 323\n",
"--------------------------------------------------\n",
"Postcode = 3057.0\n",
"*************************\n",
" Suburb count\n",
"93 Brunswick East 127\n",
"--------------------------------------------------\n",
"Postcode = 3058.0\n",
"*************************\n",
" Suburb count\n",
"94 Coburg 243\n",
"95 Coburg North 99\n",
"--------------------------------------------------\n",
"Postcode = 3059.0\n",
"*************************\n",
" Suburb count\n",
"96 Greenvale 63\n",
"--------------------------------------------------\n",
"Postcode = 3060.0\n",
"*************************\n",
" Suburb count\n",
"97 Fawkner 125\n",
"--------------------------------------------------\n",
"Postcode = 3061.0\n",
"*************************\n",
" Suburb count\n",
"98 Campbellfield 15\n",
"--------------------------------------------------\n",
"Postcode = 3064.0\n",
"*************************\n",
" Suburb count\n",
"99 Craigieburn 168\n",
"100 Kalkallo 1\n",
"101 Mickleham 7\n",
"102 Roxburgh Park 66\n",
"--------------------------------------------------\n",
"Postcode = 3065.0\n",
"*************************\n",
" Suburb count\n",
"103 Fitzroy 87\n",
"--------------------------------------------------\n",
"Postcode = 3066.0\n",
"*************************\n",
" Suburb count\n",
"104 Collingwood 69\n",
"--------------------------------------------------\n",
"Postcode = 3067.0\n",
"*************************\n",
" Suburb count\n",
"105 Abbotsford 84\n",
"--------------------------------------------------\n",
"Postcode = 3068.0\n",
"*************************\n",
" Suburb count\n",
"106 Clifton Hill 91\n",
"107 Fitzroy North 119\n",
"--------------------------------------------------\n",
"Postcode = 3070.0\n",
"*************************\n",
" Suburb count\n",
"108 Northcote 270\n",
"--------------------------------------------------\n",
"Postcode = 3071.0\n",
"*************************\n",
" Suburb count\n",
"109 Thornbury 215\n",
"--------------------------------------------------\n",
"Postcode = 3072.0\n",
"*************************\n",
" Suburb count\n",
"110 Preston 332\n",
"--------------------------------------------------\n",
"Postcode = 3073.0\n",
"*************************\n",
" Suburb count\n",
"111 Reservoir 592\n",
"--------------------------------------------------\n",
"Postcode = 3074.0\n",
"*************************\n",
" Suburb count\n",
"112 Thomastown 70\n",
"--------------------------------------------------\n",
"Postcode = 3075.0\n",
"*************************\n",
" Suburb count\n",
"113 Lalor 67\n",
"--------------------------------------------------\n",
"Postcode = 3076.0\n",
"*************************\n",
" Suburb count\n",
"114 Epping 122\n",
"--------------------------------------------------\n",
"Postcode = 3078.0\n",
"*************************\n",
" Suburb count\n",
"115 Alphington 52\n",
"116 Fairfield 57\n",
"--------------------------------------------------\n",
"Postcode = 3079.0\n",
"*************************\n",
" Suburb count\n",
"117 Ivanhoe 136\n",
"118 Ivanhoe East 38\n",
"--------------------------------------------------\n",
"Postcode = 3081.0\n",
"*************************\n",
" Suburb count\n",
"119 Bellfield 21\n",
"120 Heidelberg Heights 97\n",
"121 Heidelberg West 87\n",
"--------------------------------------------------\n",
"Postcode = 3082.0\n",
"*************************\n",
" Suburb count\n",
"122 Mill Park 116\n",
"--------------------------------------------------\n",
"Postcode = 3083.0\n",
"*************************\n",
" Suburb count\n",
"123 Bundoora 111\n",
"124 Kingsbury 34\n",
"--------------------------------------------------\n",
"Postcode = 3084.0\n",
"*************************\n",
" Suburb count\n",
"125 Eaglemont 28\n",
"126 Heidelberg 63\n",
"127 Rosanna 105\n",
"128 Viewbank 56\n",
"129 viewbank 1\n",
"--------------------------------------------------\n",
"Postcode = 3085.0\n",
"*************************\n",
" Suburb count\n",
"130 MacLeod 53\n",
"131 Yallambie 36\n",
"--------------------------------------------------\n",
"Postcode = 3087.0\n",
"*************************\n",
" Suburb count\n",
"132 Watsonia 67\n",
"133 Watsonia North 21\n",
"--------------------------------------------------\n",
"Postcode = 3088.0\n",
"*************************\n",
" Suburb count\n",
"134 Briar Hill 16\n",
"135 Greensborough 90\n",
"136 St Helena 9\n",
"--------------------------------------------------\n",
"Postcode = 3089.0\n",
"*************************\n",
" Suburb count\n",
"137 Diamond Creek 19\n",
"--------------------------------------------------\n",
"Postcode = 3093.0\n",
"*************************\n",
" Suburb count\n",
"138 Lower Plenty 17\n",
"--------------------------------------------------\n",
"Postcode = 3094.0\n",
"*************************\n",
" Suburb count\n",
"139 Montmorency 48\n",
"--------------------------------------------------\n",
"Postcode = 3095.0\n",
"*************************\n",
" Suburb count\n",
"140 Eltham 61\n",
"141 Eltham North 16\n",
"142 Research 4\n",
"--------------------------------------------------\n",
"Postcode = 3096.0\n",
"*************************\n",
" Suburb count\n",
"143 Wattle Glen 1\n",
"--------------------------------------------------\n",
"Postcode = 3099.0\n",
"*************************\n",
" Suburb count\n",
"144 Hurstbridge 2\n",
"--------------------------------------------------\n",
"Postcode = 3101.0\n",
"*************************\n",
" Suburb count\n",
"145 Kew 264\n",
"--------------------------------------------------\n",
"Postcode = 3102.0\n",
"*************************\n",
" Suburb count\n",
"146 Kew East 77\n",
"--------------------------------------------------\n",
"Postcode = 3103.0\n",
"*************************\n",
" Suburb count\n",
"147 Balwyn 187\n",
"148 Deepdene 6\n",
"--------------------------------------------------\n",
"Postcode = 3104.0\n",
"*************************\n",
" Suburb count\n",
"149 Balwyn North 245\n",
"--------------------------------------------------\n",
"Postcode = 3105.0\n",
"*************************\n",
" Suburb count\n",
"150 Bulleen 106\n",
"--------------------------------------------------\n",
"Postcode = 3106.0\n",
"*************************\n",
" Suburb count\n",
"151 Templestowe 54\n",
"--------------------------------------------------\n",
"Postcode = 3107.0\n",
"*************************\n",
" Suburb count\n",
"152 Templestowe Lower 141\n",
"--------------------------------------------------\n",
"Postcode = 3108.0\n",
"*************************\n",
" Suburb count\n",
"153 Doncaster 164\n",
"--------------------------------------------------\n",
"Postcode = 3109.0\n",
"*************************\n",
" Suburb count\n",
"154 Doncaster East 95\n",
"--------------------------------------------------\n",
"Postcode = 3111.0\n",
"*************************\n",
" Suburb count\n",
"155 Donvale 34\n",
"--------------------------------------------------\n",
"Postcode = 3113.0\n",
"*************************\n",
" Suburb count\n",
"156 North Warrandyte 5\n",
"157 Warrandyte 8\n",
"--------------------------------------------------\n",
"Postcode = 3115.0\n",
"*************************\n",
" Suburb count\n",
"158 Wonga Park 2\n",
"--------------------------------------------------\n",
"Postcode = 3116.0\n",
"*************************\n",
" Suburb count\n",
"159 Chirnside Park 8\n",
"--------------------------------------------------\n",
"Postcode = 3121.0\n",
"*************************\n",
" Suburb count\n",
"160 Burnley 10\n",
"161 Cremorne 31\n",
"162 Richmond 349\n",
"--------------------------------------------------\n",
"Postcode = 3122.0\n",
"*************************\n",
" Suburb count\n",
"163 Hawthorn 253\n",
"--------------------------------------------------\n",
"Postcode = 3123.0\n",
"*************************\n",
" Suburb count\n",
"164 Hawthorn East 159\n",
"--------------------------------------------------\n",
"Postcode = 3124.0\n",
"*************************\n",
" Suburb count\n",
"165 Camberwell 222\n",
"--------------------------------------------------\n",
"Postcode = 3125.0\n",
"*************************\n",
" Suburb count\n",
"166 Burwood 131\n",
"--------------------------------------------------\n",
"Postcode = 3126.0\n",
"*************************\n",
" Suburb count\n",
"167 Canterbury 67\n",
"--------------------------------------------------\n",
"Postcode = 3127.0\n",
"*************************\n",
" Suburb count\n",
"168 Mont Albert 63\n",
"169 Surrey Hills 162\n",
"--------------------------------------------------\n",
"Postcode = 3128.0\n",
"*************************\n",
" Suburb count\n",
"170 Box Hill 67\n",
"--------------------------------------------------\n",
"Postcode = 3130.0\n",
"*************************\n",
" Suburb count\n",
"171 Blackburn 49\n",
"172 Blackburn North 30\n",
"173 Blackburn South 32\n",
"--------------------------------------------------\n",
"Postcode = 3131.0\n",
"*************************\n",
" Suburb count\n",
"174 Forest Hill 44\n",
"175 Nunawading 41\n",
"--------------------------------------------------\n",
"Postcode = 3132.0\n",
"*************************\n",
" Suburb count\n",
"176 Mitcham 75\n",
"--------------------------------------------------\n",
"Postcode = 3133.0\n",
"*************************\n",
" Suburb count\n",
"177 Vermont 30\n",
"178 Vermont South 27\n",
"--------------------------------------------------\n",
"Postcode = 3134.0\n",
"*************************\n",
" Suburb count\n",
"179 Ringwood 44\n",
"180 Ringwood North 23\n",
"181 Warranwood 3\n",
"--------------------------------------------------\n",
"Postcode = 3135.0\n",
"*************************\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
" Suburb count\n",
"182 Heathmont 32\n",
"183 Ringwood East 27\n",
"--------------------------------------------------\n",
"Postcode = 3136.0\n",
"*************************\n",
" Suburb count\n",
"184 Croydon 83\n",
"185 Croydon Hills 11\n",
"186 Croydon North 16\n",
"187 Croydon South 10\n",
"188 croydon 1\n",
"--------------------------------------------------\n",
"Postcode = 3137.0\n",
"*************************\n",
" Suburb count\n",
"189 Kilsyth 17\n",
"--------------------------------------------------\n",
"Postcode = 3138.0\n",
"*************************\n",
" Suburb count\n",
"190 Mooroolbark 29\n",
"--------------------------------------------------\n",
"Postcode = 3140.0\n",
"*************************\n",
" Suburb count\n",
"191 Lilydale 4\n",
"--------------------------------------------------\n",
"Postcode = 3141.0\n",
"*************************\n",
" Suburb count\n",
"192 South Yarra 276\n",
"--------------------------------------------------\n",
"Postcode = 3142.0\n",
"*************************\n",
" Suburb count\n",
"193 Toorak 135\n",
"--------------------------------------------------\n",
"Postcode = 3143.0\n",
"*************************\n",
" Suburb count\n",
"194 Armadale 118\n",
"--------------------------------------------------\n",
"Postcode = 3144.0\n",
"*************************\n",
" Suburb count\n",
"195 Kooyong 5\n",
"196 Malvern 100\n",
"--------------------------------------------------\n",
"Postcode = 3145.0\n",
"*************************\n",
" Suburb count\n",
"197 Caulfield East 15\n",
"198 Malvern East 183\n",
"--------------------------------------------------\n",
"Postcode = 3146.0\n",
"*************************\n",
" Suburb count\n",
"199 Glen Iris 255\n",
"--------------------------------------------------\n",
"Postcode = 3147.0\n",
"*************************\n",
" Suburb count\n",
"200 Ashburton 90\n",
"201 Ashwood 83\n",
"--------------------------------------------------\n",
"Postcode = 3148.0\n",
"*************************\n",
" Suburb count\n",
"202 Chadstone 61\n",
"--------------------------------------------------\n",
"Postcode = 3149.0\n",
"*************************\n",
" Suburb count\n",
"203 Mount Waverley 135\n",
"--------------------------------------------------\n",
"Postcode = 3150.0\n",
"*************************\n",
" Suburb count\n",
"204 Glen Waverley 116\n",
"205 Wheelers Hill 52\n",
"--------------------------------------------------\n",
"Postcode = 3151.0\n",
"*************************\n",
" Suburb count\n",
"206 Burwood East 33\n",
"--------------------------------------------------\n",
"Postcode = 3152.0\n",
"*************************\n",
" Suburb count\n",
"207 Wantirna 35\n",
"208 Wantirna South 35\n",
"--------------------------------------------------\n",
"Postcode = 3153.0\n",
"*************************\n",
" Suburb count\n",
"209 Bayswater 27\n",
"210 Bayswater North 17\n",
"--------------------------------------------------\n",
"Postcode = 3154.0\n",
"*************************\n",
" Suburb count\n",
"211 The Basin 3\n",
"--------------------------------------------------\n",
"Postcode = 3155.0\n",
"*************************\n",
" Suburb count\n",
"212 Boronia 33\n",
"--------------------------------------------------\n",
"Postcode = 3156.0\n",
"*************************\n",
" Suburb count\n",
"213 Ferntree Gully 46\n",
"214 Lysterfield 1\n",
"--------------------------------------------------\n",
"Postcode = 3158.0\n",
"*************************\n",
" Suburb count\n",
"215 Upwey 2\n",
"--------------------------------------------------\n",
"Postcode = 3160.0\n",
"*************************\n",
" Suburb count\n",
"216 Tecoma 1\n",
"--------------------------------------------------\n",
"Postcode = 3161.0\n",
"*************************\n",
" Suburb count\n",
"217 Caulfield North 48\n",
"--------------------------------------------------\n",
"Postcode = 3162.0\n",
"*************************\n",
" Suburb count\n",
"218 Caulfield 14\n",
"219 Caulfield South 63\n",
"--------------------------------------------------\n",
"Postcode = 3163.0\n",
"*************************\n",
" Suburb count\n",
"220 Carnegie 183\n",
"221 Glen Huntly 40\n",
"222 Murrumbeena 110\n",
"--------------------------------------------------\n",
"Postcode = 3165.0\n",
"*************************\n",
" Suburb count\n",
"223 Bentleigh East 405\n",
"--------------------------------------------------\n",
"Postcode = 3166.0\n",
"*************************\n",
" Suburb count\n",
"224 Hughesdale 42\n",
"225 Huntingdale 7\n",
"226 Oakleigh 64\n",
"227 Oakleigh East 23\n",
"--------------------------------------------------\n",
"Postcode = 3167.0\n",
"*************************\n",
" Suburb count\n",
"228 Oakleigh South 87\n",
"--------------------------------------------------\n",
"Postcode = 3168.0\n",
"*************************\n",
" Suburb count\n",
"229 Clayton 31\n",
"230 Notting Hill 7\n",
"--------------------------------------------------\n",
"Postcode = 3169.0\n",
"*************************\n",
" Suburb count\n",
"231 Clarinda 12\n",
"232 Clayton South 29\n",
"--------------------------------------------------\n",
"Postcode = 3170.0\n",
"*************************\n",
" Suburb count\n",
"233 Mulgrave 64\n",
"--------------------------------------------------\n",
"Postcode = 3171.0\n",
"*************************\n",
" Suburb count\n",
"234 Springvale 26\n",
"--------------------------------------------------\n",
"Postcode = 3172.0\n",
"*************************\n",
" Suburb count\n",
"235 Dingley Village 36\n",
"236 Springvale South 8\n",
"--------------------------------------------------\n",
"Postcode = 3173.0\n",
"*************************\n",
" Suburb count\n",
"237 Keysborough 23\n",
"--------------------------------------------------\n",
"Postcode = 3174.0\n",
"*************************\n",
" Suburb count\n",
"238 Noble Park 41\n",
"--------------------------------------------------\n",
"Postcode = 3175.0\n",
"*************************\n",
" Suburb count\n",
"239 Dandenong 33\n",
"240 Dandenong North 33\n",
"--------------------------------------------------\n",
"Postcode = 3177.0\n",
"*************************\n",
" Suburb count\n",
"241 Doveton 15\n",
"242 Eumemmerring 1\n",
"--------------------------------------------------\n",
"Postcode = 3178.0\n",
"*************************\n",
" Suburb count\n",
"243 Rowville 25\n",
"--------------------------------------------------\n",
"Postcode = 3179.0\n",
"*************************\n",
" Suburb count\n",
"244 Scoresby 9\n",
"--------------------------------------------------\n",
"Postcode = 3180.0\n",
"*************************\n",
" Suburb count\n",
"245 Knoxfield 14\n",
"--------------------------------------------------\n",
"Postcode = 3181.0\n",
"*************************\n",
" Suburb count\n",
"246 Prahran 156\n",
"247 Windsor 63\n",
"--------------------------------------------------\n",
"Postcode = 3182.0\n",
"*************************\n",
" Suburb count\n",
"248 St Kilda 234\n",
"--------------------------------------------------\n",
"Postcode = 3183.0\n",
"*************************\n",
" Suburb count\n",
"249 Balaclava 47\n",
"--------------------------------------------------\n",
"Postcode = 3184.0\n",
"*************************\n",
" Suburb count\n",
"250 Elwood 162\n",
"--------------------------------------------------\n",
"Postcode = 3185.0\n",
"*************************\n",
" Suburb count\n",
"251 Elsternwick 77\n",
"252 Gardenvale 10\n",
"253 Ripponlea 12\n",
"--------------------------------------------------\n",
"Postcode = 3186.0\n",
"*************************\n",
" Suburb count\n",
"254 Brighton 254\n",
"--------------------------------------------------\n",
"Postcode = 3187.0\n",
"*************************\n",
" Suburb count\n",
"255 Brighton East 213\n",
"--------------------------------------------------\n",
"Postcode = 3188.0\n",
"*************************\n",
" Suburb count\n",
"256 Hampton 179\n",
"257 Hampton East 67\n",
"--------------------------------------------------\n",
"Postcode = 3189.0\n",
"*************************\n",
" Suburb count\n",
"258 Moorabbin 88\n",
"--------------------------------------------------\n",
"Postcode = 3190.0\n",
"*************************\n",
" Suburb count\n",
"259 Highett 69\n",
"--------------------------------------------------\n",
"Postcode = 3191.0\n",
"*************************\n",
" Suburb count\n",
"260 Sandringham 59\n",
"--------------------------------------------------\n",
"Postcode = 3192.0\n",
"*************************\n",
" Suburb count\n",
"261 Cheltenham 120\n",
"--------------------------------------------------\n",
"Postcode = 3193.0\n",
"*************************\n",
" Suburb count\n",
"262 Beaumaris 69\n",
"263 Black Rock 19\n",
"--------------------------------------------------\n",
"Postcode = 3194.0\n",
"*************************\n",
" Suburb count\n",
"264 Mentone 74\n",
"--------------------------------------------------\n",
"Postcode = 3195.0\n",
"*************************\n",
" Suburb count\n",
"265 Aspendale 25\n",
"266 Aspendale Gardens 10\n",
"267 Mordialloc 46\n",
"268 Parkdale 62\n",
"269 Waterways 1\n",
"--------------------------------------------------\n",
"Postcode = 3196.0\n",
"*************************\n",
" Suburb count\n",
"270 Bonbeach 10\n",
"271 Chelsea 26\n",
"272 Chelsea Heights 14\n",
"273 Edithvale 25\n",
"--------------------------------------------------\n",
"Postcode = 3197.0\n",
"*************************\n",
" Suburb count\n",
"274 Carrum 21\n",
"275 Patterson Lakes 5\n",
"--------------------------------------------------\n",
"Postcode = 3198.0\n",
"*************************\n",
" Suburb count\n",
"276 Seaford 34\n",
"--------------------------------------------------\n",
"Postcode = 3199.0\n",
"*************************\n",
" Suburb count\n",
"277 Frankston 71\n",
"278 Frankston South 34\n",
"--------------------------------------------------\n",
"Postcode = 3200.0\n",
"*************************\n",
" Suburb count\n",
"279 Frankston North 13\n",
"--------------------------------------------------\n",
"Postcode = 3201.0\n",
"*************************\n",
" Suburb count\n",
"280 Carrum Downs 20\n",
"--------------------------------------------------\n",
"Postcode = 3202.0\n",
"*************************\n",
" Suburb count\n",
"281 Heatherton 2\n",
"--------------------------------------------------\n",
"Postcode = 3204.0\n",
"*************************\n",
" Suburb count\n",
"282 Bentleigh 223\n",
"283 McKinnon 22\n",
"284 Ormond 102\n",
"--------------------------------------------------\n",
"Postcode = 3205.0\n",
"*************************\n",
" Suburb count\n",
"285 South Melbourne 118\n",
"--------------------------------------------------\n",
"Postcode = 3206.0\n",
"*************************\n",
" Suburb count\n",
"286 Albert Park 79\n",
"287 Middle Park 45\n",
"--------------------------------------------------\n",
"Postcode = 3207.0\n",
"*************************\n",
" Suburb count\n",
"288 Port Melbourne 214\n",
"--------------------------------------------------\n",
"Postcode = 3335.0\n",
"*************************\n",
" Suburb count\n",
"289 Plumpton 6\n",
"290 Rockbank 3\n",
"--------------------------------------------------\n",
"Postcode = 3337.0\n",
"*************************\n",
" Suburb count\n",
"291 Kurunjang 13\n",
"292 Melton 25\n",
"293 Melton West 26\n",
"--------------------------------------------------\n",
"Postcode = 3338.0\n",
"*************************\n",
" Suburb count\n",
"294 Brookfield 3\n",
"295 Melton South 42\n",
"--------------------------------------------------\n",
"Postcode = 3340.0\n",
"*************************\n",
" Suburb count\n",
"296 Bacchus Marsh 3\n",
"297 Darley 1\n",
"--------------------------------------------------\n",
"Postcode = 3427.0\n",
"*************************\n",
" Suburb count\n",
"298 Diggers Rest 6\n",
"--------------------------------------------------\n",
"Postcode = 3428.0\n",
"*************************\n",
" Suburb count\n",
"299 Bulla 1\n",
"--------------------------------------------------\n",
"Postcode = 3429.0\n",
"*************************\n",
" Suburb count\n",
"300 Sunbury 112\n",
"--------------------------------------------------\n",
"Postcode = 3431.0\n",
"*************************\n",
" Suburb count\n",
"301 Riddells Creek 1\n",
"--------------------------------------------------\n",
"Postcode = 3437.0\n",
"*************************\n",
" Suburb count\n",
"302 Bullengarook 4\n",
"303 Gisborne 19\n",
"304 Gisborne South 2\n",
"--------------------------------------------------\n",
"Postcode = 3438.0\n",
"*************************\n",
" Suburb count\n",
"305 New Gisborne 3\n",
"--------------------------------------------------\n",
"Postcode = 3750.0\n",
"*************************\n",
" Suburb count\n",
"306 Wollert 39\n",
"--------------------------------------------------\n",
"Postcode = 3752.0\n",
"*************************\n",
" Suburb count\n",
"307 South Morang 71\n",
"--------------------------------------------------\n",
"Postcode = 3754.0\n",
"*************************\n",
" Suburb count\n",
"308 Doreen 28\n",
"309 Mernda 66\n",
"--------------------------------------------------\n",
"Postcode = 3756.0\n",
"*************************\n",
" Suburb count\n",
"310 Wallan 8\n",
"--------------------------------------------------\n",
"Postcode = 3757.0\n",
"*************************\n",
" Suburb count\n",
"311 Whittlesea 4\n",
"--------------------------------------------------\n",
"Postcode = 3775.0\n",
"*************************\n",
" Suburb count\n",
"312 Yarra Glen 1\n",
"--------------------------------------------------\n",
"Postcode = 3777.0\n",
"*************************\n",
" Suburb count\n",
"313 Healesville 1\n",
"--------------------------------------------------\n",
"Postcode = 3782.0\n",
"*************************\n",
" Suburb count\n",
"314 Emerald 3\n",
"--------------------------------------------------\n",
"Postcode = 3786.0\n",
"*************************\n",
" Suburb count\n",
"315 Ferny Creek 1\n",
"--------------------------------------------------\n",
"Postcode = 3793.0\n",
"*************************\n",
" Suburb count\n",
"316 Monbulk 1\n",
"--------------------------------------------------\n",
"Postcode = 3795.0\n",
"*************************\n",
" Suburb count\n",
"317 Silvan 2\n",
"--------------------------------------------------\n",
"Postcode = 3796.0\n",
"*************************\n",
" Suburb count\n",
"318 Mount Evelyn 3\n",
"--------------------------------------------------\n",
"Postcode = 3802.0\n",
"*************************\n",
" Suburb count\n",
"319 Endeavour Hills 16\n",
"--------------------------------------------------\n",
"Postcode = 3803.0\n",
"*************************\n",
" Suburb count\n",
"320 Hallam 7\n",
"--------------------------------------------------\n",
"Postcode = 3805.0\n",
"*************************\n",
" Suburb count\n",
"321 Narre Warren 14\n",
"--------------------------------------------------\n",
"Postcode = 3806.0\n",
"*************************\n",
" Suburb count\n",
"322 Berwick 34\n",
"--------------------------------------------------\n",
"Postcode = 3807.0\n",
"*************************\n",
" Suburb count\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"323 Beaconsfield 2\n",
"--------------------------------------------------\n",
"Postcode = 3808.0\n",
"*************************\n",
" Suburb count\n",
"324 Beaconsfield Upper 2\n",
"--------------------------------------------------\n",
"Postcode = 3809.0\n",
"*************************\n",
" Suburb count\n",
"325 Officer 3\n",
"--------------------------------------------------\n",
"Postcode = 3810.0\n",
"*************************\n",
" Suburb count\n",
"326 Pakenham 10\n",
"--------------------------------------------------\n",
"Postcode = 3910.0\n",
"*************************\n",
" Suburb count\n",
"327 Langwarrin 14\n",
"--------------------------------------------------\n",
"Postcode = 3975.0\n",
"*************************\n",
" Suburb count\n",
"328 Lynbrook 2\n",
"--------------------------------------------------\n",
"Postcode = 3976.0\n",
"*************************\n",
" Suburb count\n",
"329 Hampton Park 5\n",
"--------------------------------------------------\n",
"Postcode = 3977.0\n",
"*************************\n",
" Suburb count\n",
"330 Cranbourne 9\n",
"331 Cranbourne East 1\n",
"332 Cranbourne North 3\n",
"333 Cranbourne West 4\n",
"334 Sandhurst 1\n",
"335 Skye 3\n",
"--------------------------------------------------\n",
"Postcode = 3978.0\n",
"*************************\n",
" Suburb count\n",
"336 Clyde North 3\n",
"--------------------------------------------------\n"
]
}
],
"source": [
"for postcode , group in df_postcode_suburb.groupby(['Postcode']):\n",
" print(\"Postcode = %s\" %(postcode))\n",
" print(25*'*')\n",
" print(group[['Suburb', 'count']])\n",
" print(50*'-')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Is Postalcode an important feature in predicting Price ?"
]
},
{
"cell_type": "code",
"execution_count": 30,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 3600x720 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"gr = sns.catplot(data=training_set, x=\"Postcode\", y=\"Price\", kind=\"bar\", ci=None,height=10, aspect=5)\n",
"plt.xlabel(\"Postcode\")\n",
"plt.ylabel(\"Average Price\")\n",
"l1 = gr.set_yticklabels(rotation = 45)\n",
"l2 = gr.set_xticklabels(rotation = 45)"
]
},
{
"cell_type": "code",
"execution_count": 31,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 3600x720 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"gr = sns.catplot(data=training_set, x=\"Suburb\", y=\"Price\", kind=\"bar\", ci=None,height=10, aspect=5)\n",
"plt.xlabel(\"Suburb\")\n",
"plt.ylabel(\"Average Price\")\n",
"l1 = gr.set_yticklabels(rotation = 45)\n",
"l2 = gr.set_xticklabels(rotation = 45)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Drop \"CouncilArea\", \"Regionname\", \"Lattitude\", \"Suburb\", \"Longtitude\""
]
},
{
"cell_type": "code",
"execution_count": 32,
"metadata": {},
"outputs": [],
"source": [
"training_set.drop([\"Address\", \"CouncilArea\", \"Regionname\", \"Lattitude\", \"Suburb\", \"Longtitude\"], axis=1, inplace=True)"
]
},
{
"cell_type": "code",
"execution_count": 33,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Index(['Rooms', 'Type', 'Price', 'Method', 'SellerG', 'Date', 'Distance',\n",
" 'Postcode', 'Bedroom2', 'Bathroom', 'Car', 'Landsize', 'BuildingArea',\n",
" 'YearBuilt', 'Propertycount'],\n",
" dtype='object')"
]
},
"execution_count": 33,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set.columns"
]
},
{
"cell_type": "code",
"execution_count": 34,
"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>Rooms</th>\n",
" <th>Type</th>\n",
" <th>Price</th>\n",
" <th>Method</th>\n",
" <th>SellerG</th>\n",
" <th>Date</th>\n",
" <th>Distance</th>\n",
" <th>Postcode</th>\n",
" <th>Bedroom2</th>\n",
" <th>Bathroom</th>\n",
" <th>Car</th>\n",
" <th>Landsize</th>\n",
" <th>BuildingArea</th>\n",
" <th>YearBuilt</th>\n",
" <th>Propertycount</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>17957</th>\n",
" <td>2</td>\n",
" <td>u</td>\n",
" <td>580000.0</td>\n",
" <td>VB</td>\n",
" <td>Nelson</td>\n",
" <td>26/08/2017</td>\n",
" <td>4.0</td>\n",
" <td>3057.0</td>\n",
" <td>2.0</td>\n",
" <td>1.0</td>\n",
" <td>1.0</td>\n",
" <td>NaN</td>\n",
" <td>61.0</td>\n",
" <td>1970.0</td>\n",
" <td>5533.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>6753</th>\n",
" <td>3</td>\n",
" <td>t</td>\n",
" <td>1025000.0</td>\n",
" <td>SP</td>\n",
" <td>Jellis</td>\n",
" <td>3/12/2016</td>\n",
" <td>2.6</td>\n",
" <td>3121.0</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>14949.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1833</th>\n",
" <td>2</td>\n",
" <td>t</td>\n",
" <td>380000.0</td>\n",
" <td>VB</td>\n",
" <td>Nelson</td>\n",
" <td>7/11/2016</td>\n",
" <td>5.9</td>\n",
" <td>3055.0</td>\n",
" <td>2.0</td>\n",
" <td>1.0</td>\n",
" <td>1.0</td>\n",
" <td>0.0</td>\n",
" <td>85.0</td>\n",
" <td>1970.0</td>\n",
" <td>7082.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>19795</th>\n",
" <td>2</td>\n",
" <td>u</td>\n",
" <td>456000.0</td>\n",
" <td>S</td>\n",
" <td>Brad</td>\n",
" <td>21/10/2017</td>\n",
" <td>8.5</td>\n",
" <td>3044.0</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>7485.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>17054</th>\n",
" <td>3</td>\n",
" <td>h</td>\n",
" <td>1300000.0</td>\n",
" <td>S</td>\n",
" <td>hockingstuart</td>\n",
" <td>19/08/2017</td>\n",
" <td>16.7</td>\n",
" <td>3150.0</td>\n",
" <td>3.0</td>\n",
" <td>2.0</td>\n",
" <td>3.0</td>\n",
" <td>733.0</td>\n",
" <td>NaN</td>\n",
" <td>NaN</td>\n",
" <td>15321.0</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" Rooms Type Price Method SellerG Date Distance \\\n",
"17957 2 u 580000.0 VB Nelson 26/08/2017 4.0 \n",
"6753 3 t 1025000.0 SP Jellis 3/12/2016 2.6 \n",
"1833 2 t 380000.0 VB Nelson 7/11/2016 5.9 \n",
"19795 2 u 456000.0 S Brad 21/10/2017 8.5 \n",
"17054 3 h 1300000.0 S hockingstuart 19/08/2017 16.7 \n",
"\n",
" Postcode Bedroom2 Bathroom Car Landsize BuildingArea YearBuilt \\\n",
"17957 3057.0 2.0 1.0 1.0 NaN 61.0 1970.0 \n",
"6753 3121.0 NaN NaN NaN NaN NaN NaN \n",
"1833 3055.0 2.0 1.0 1.0 0.0 85.0 1970.0 \n",
"19795 3044.0 NaN NaN NaN NaN NaN NaN \n",
"17054 3150.0 3.0 2.0 3.0 733.0 NaN NaN \n",
"\n",
" Propertycount \n",
"17957 5533.0 \n",
"6753 14949.0 \n",
"1833 7082.0 \n",
"19795 7485.0 \n",
"17054 15321.0 "
]
},
"execution_count": 34,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set.head()"
]
},
{
"cell_type": "code",
"execution_count": 35,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Type False\n",
"Method False\n",
"SellerG False\n",
"Date False\n",
"dtype: bool"
]
},
"execution_count": 35,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set.select_dtypes(['object']).isnull().any()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Date field has got date in string format, I think it is good to extract sold year as we have YearBuilt in the dataset"
]
},
{
"cell_type": "code",
"execution_count": 36,
"metadata": {},
"outputs": [],
"source": [
"training_set.Date = pd.to_datetime(training_set.Date)"
]
},
{
"cell_type": "code",
"execution_count": 37,
"metadata": {},
"outputs": [],
"source": [
"training_set[\"sold_year\"] = training_set.Date.apply(lambda x: x.year)"
]
},
{
"cell_type": "code",
"execution_count": 38,
"metadata": {},
"outputs": [],
"source": [
"training_set.drop(\"Date\", axis=1, inplace=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### EDA (Statistical data analysis) of numerical columns and fixing problems."
]
},
{
"cell_type": "code",
"execution_count": 39,
"metadata": {
"scrolled": true
},
"outputs": [
{
"data": {
"text/plain": [
"Rooms False\n",
"Price False\n",
"Distance False\n",
"Postcode False\n",
"Bedroom2 True\n",
"Bathroom True\n",
"Car True\n",
"Landsize True\n",
"BuildingArea True\n",
"YearBuilt True\n",
"Propertycount True\n",
"sold_year False\n",
"dtype: bool"
]
},
"execution_count": 39,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set.select_dtypes(['float64','int64']).isnull().any()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# --------------------------------------------------------------------"
]
},
{
"cell_type": "code",
"execution_count": 40,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(5163, 15)"
]
},
"execution_count": 40,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set[training_set.Bedroom2.isnull()].shape"
]
},
{
"cell_type": "code",
"execution_count": 41,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"<matplotlib.axes._subplots.AxesSubplot at 0x19bc14d7d68>"
]
},
"execution_count": 41,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"training_set.Bedroom2.hist(bins=20)"
]
},
{
"cell_type": "code",
"execution_count": 42,
"metadata": {},
"outputs": [],
"source": [
"training_set.Bedroom2.fillna(value=training_set.Bedroom2.mean(), inplace=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# --------------------------------------------------------------------"
]
},
{
"cell_type": "code",
"execution_count": 43,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(5169, 15)"
]
},
"execution_count": 43,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set[training_set.Bathroom.isnull()].shape"
]
},
{
"cell_type": "code",
"execution_count": 44,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"<matplotlib.axes._subplots.AxesSubplot at 0x19bc16ab2e8>"
]
},
"execution_count": 44,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAD4CAYAAAAAczaOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjAsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+17YcXAAAWeUlEQVR4nO3df4xdZZ3H8fdnqUjB1RaRG7ZtdmqcqGgDshOokpi71C0FjOUPSWq6MpAms390Fc0kbvGfZkESTESURMlObN3iumK3SmiEiE3hZrN/UPm5VKikI9R2bKXqlOrI+mP0u3/cZ+y9ZWbunbk/zvQ+n1cyued8z3POec7T28+9c+6ZexQRmJlZHv6q6A6YmVn3OPTNzDLi0Dczy4hD38wsIw59M7OMLCq6A7O54IILoq+vb97r//a3v+W8885rX4fOYB6Leh6PUzwW9XphPJ566qlfRsTbplu2oEO/r6+PJ598ct7rVyoVyuVy+zp0BvNY1PN4nOKxqNcL4yHppzMt8+kdM7OMOPTNzDLi0Dczy4hD38wsIw59M7OMOPTNzDLi0Dczy4hD38wsIw59M7OMLOi/yD1T9W15qJD9HrrzukL2a2ZnDr/TNzPLiEPfzCwjDn0zs4w49M3MMuLQNzPLiEPfzCwjTYW+pE9Lel7SjyR9S9I5klZK2ifpoKRvSzo7tX1jmh9Ny/tqtnNrqr8o6erOHJKZmc2kYehLWgZ8EhiIiPcCZwEbgM8Dd0dEP3AC2JRW2QSciIh3AHendki6OK33HmAd8FVJZ7X3cMzMbDbNnt5ZBCyWtAg4FzgGXAXsSst3ANen6fVpnrR8jSSl+v0R8fuIeBkYBS5v/RDMzKxZDf8iNyJ+JukLwGHg/4AfAE8Br0bEZGo2BixL08uAI2ndSUkngbem+uM1m65d5y8kDQFDAKVSiUqlMvejSiYmJlpaf76GV002btQBsx1rUWOxUHk8TvFY1Ov18WgY+pKWUn2XvhJ4Ffgv4JppmsbUKjMsm6leX4gYAUYABgYGopUbFBd1g+Obivoaho3lGZf1ws2e28njcYrHol6vj0czp3c+BLwcEb+IiD8C3wU+ACxJp3sAlgNH0/QYsAIgLX8LMF5bn2YdMzPrgmZC/zCwWtK56dz8GuAF4DHgo6nNIPBgmt6d5knLH42ISPUN6eqelUA/8MP2HIaZmTWjmXP6+yTtAp4GJoFnqJ5+eQi4X9LnUm1bWmUb8A1Jo1Tf4W9I23le0k6qLxiTwOaI+FObj8fMzGbR1FcrR8RWYOtp5ZeY5uqbiPgdcMMM27kDuGOOfTQzszbxX+SamWXEoW9mlhGHvplZRhz6ZmYZceibmWXEoW9mlhGHvplZRhz6ZmYZceibmWXEoW9mlhGHvplZRhz6ZmYZceibmWXEoW9mlhGHvplZRhz6ZmYZaRj6kt4p6dman19L+pSk8yXtkXQwPS5N7SXpHkmjkp6TdFnNtgZT+4OSBmfeq5mZdULD0I+IFyPi0oi4FPg74DXgAWALsDci+oG9aR7gGqr3v+0HhoB7ASSdT/XuW1dQvePW1qkXCjMz6465nt5ZA/wkIn4KrAd2pPoO4Po0vR64L6oeB5ZIugi4GtgTEeMRcQLYA6xr+QjMzKxpTd0jt8YG4FtpuhQRxwAi4pikC1N9GXCkZp2xVJupXkfSENXfECiVSlQqlTl28ZSJiYmW1p+v4VWTXd8nMOuxFjUWC5XH4xSPRb1eH4+mQ1/S2cBHgFsbNZ2mFrPU6wsRI8AIwMDAQJTL5Wa7+DqVSoVW1p+vm7Y81PV9AhzaWJ5xWVFjsVB5PE7xWNTr9fGYy+mda4CnI+KVNP9KOm1Dejye6mPAipr1lgNHZ6mbmVmXzCX0P8apUzsAu4GpK3AGgQdr6jemq3hWAyfTaaBHgLWSlqYPcNemmpmZdUlTp3cknQv8A/BPNeU7gZ2SNgGHgRtS/WHgWmCU6pU+NwNExLik24EnUrvbImK85SMwM7OmNRX6EfEa8NbTar+iejXP6W0D2DzDdrYD2+feTTMza4e5Xr1jC1jfLB8gD6+a7OgHzIfuvK5j2zaz9vHXMJiZZcShb2aWEYe+mVlGHPpmZhlx6JuZZcShb2aWEYe+mVlGHPpmZhlx6JuZZcShb2aWEYe+mVlGHPpmZhlx6JuZZcShb2aWkaZCX9ISSbsk/VjSAUnvl3S+pD2SDqbHpamtJN0jaVTSc5Iuq9nOYGp/UNLgzHs0M7NOaPad/peB70fEu4BLgAPAFmBvRPQDe9M8VO+l259+hoB7ASSdD2wFrgAuB7ZOvVCYmVl3NAx9SW8GPghsA4iIP0TEq8B6YEdqtgO4Pk2vB+6LqseBJenG6VcDeyJiPCJOAHuAdW09GjMzm1Uzd856O/AL4OuSLgGeAm4BSumG50TEMUkXpvbLgCM164+l2kz1OpKGqP6GQKlUolKpzOV46kxMTLS0/nwNr5rs+j4bKS3ubL+KGOdWFPXcWIg8FvV6fTyaCf1FwGXAJyJin6Qvc+pUznQ0TS1mqdcXIkaAEYCBgYEol8tNdHF6lUqFVtafr07elnC+hldNctf+zt0d89DGcse23QlFPTcWIo9FvV4fj2bO6Y8BYxGxL83vovoi8Eo6bUN6PF7TfkXN+suBo7PUzcysSxqGfkT8HDgi6Z2ptAZ4AdgNTF2BMwg8mKZ3Azemq3hWAyfTaaBHgLWSlqYPcNemmpmZdUmzv+9/AvimpLOBl4Cbqb5g7JS0CTgM3JDaPgxcC4wCr6W2RMS4pNuBJ1K72yJivC1HYWZmTWkq9CPiWWBgmkVrpmkbwOYZtrMd2D6XDpqZWfv4L3LNzDLi0Dczy4hD38wsIw59M7OMOPTNzDLi0Dczy4hD38wsIw59M7OMOPTNzDLi0Dczy4hD38wsIw59M7OMOPTNzDLi0Dczy4hD38wsIw59M7OMNBX6kg5J2i/pWUlPptr5kvZIOpgel6a6JN0jaVTSc5Iuq9nOYGp/UNLgTPszM7POmMs7/b+PiEsjYuoOWluAvRHRD+xN8wDXAP3pZwi4F6ovEsBW4ArgcmDr1AuFmZl1Ryund9YDO9L0DuD6mvp9UfU4sETSRcDVwJ6IGI+IE8AeYF0L+zczszlq9sboAfxAUgD/FhEjQCkijgFExDFJF6a2y4AjNeuOpdpM9TqShqj+hkCpVKJSqTR/NKeZmJhoaf35Gl412fV9NlJa3Nl+FTHOrSjqubEQeSzq9fp4NBv6V0bE0RTseyT9eJa2mqYWs9TrC9UXlBGAgYGBKJfLTXbx9SqVCq2sP183bXmo6/tsZHjVJHftb/afe+4ObSx3bNudUNRzYyHyWNTr9fFo6vRORBxNj8eBB6iek38lnbYhPR5PzceAFTWrLweOzlI3M7MuaRj6ks6T9NdT08Ba4EfAbmDqCpxB4ME0vRu4MV3Fsxo4mU4DPQKslbQ0fYC7NtXMzKxLmvl9vwQ8IGmq/X9GxPclPQHslLQJOAzckNo/DFwLjAKvATcDRMS4pNuBJ1K72yJivG1HYmZmDTUM/Yh4CbhkmvqvgDXT1APYPMO2tgPb595NMzNrB/9FrplZRhz6ZmYZceibmWXEoW9mlhGHvplZRhz6ZmYZceibmWXEoW9mlhGHvplZRhz6ZmYZceibmWXEoW9mlhGHvplZRhz6ZmYZceibmWWk6dCXdJakZyR9L82vlLRP0kFJ35Z0dqq/Mc2PpuV9Ndu4NdVflHR1uw/GzMxmN5d3+rcAB2rmPw/cHRH9wAlgU6pvAk5ExDuAu1M7JF0MbADeA6wDvirprNa6b2Zmc9FU6EtaDlwHfC3NC7gK2JWa7ACuT9Pr0zxp+ZrUfj1wf0T8PiJepno7xcvbcRBmZtacZt/pfwn4DPDnNP9W4NWImEzzY8CyNL0MOAKQlp9M7f9Sn2YdMzPrgob3yJX0YeB4RDwlqTxVnqZpNFg22zq1+xsChgBKpRKVSqVRF2c0MTHR0vrzNbxqsnGjList7my/ihjnVhT13FiIPBb1en08GoY+cCXwEUnXAucAb6b6zn+JpEXp3fxy4GhqPwasAMYkLQLeAozX1KfUrvMXETECjAAMDAxEuVyex2FVVSoVWll/vm7a8lDX99nI8KpJ7trfzD/3/BzaWO7YtjuhqOfGQuSxqNfr49Hw9E5E3BoRyyOij+oHsY9GxEbgMeCjqdkg8GCa3p3mScsfjYhI9Q3p6p6VQD/ww7YdiZmZNdTKW79/Ae6X9DngGWBbqm8DviFplOo7/A0AEfG8pJ3AC8AksDki/tTC/s3MbI7mFPoRUQEqafolprn6JiJ+B9www/p3AHfMtZNmZtYe/otcM7OMOPTNzDLi0Dczy4hD38wsIw59M7OMOPTNzDLi0Dczy4hD38wsIw59M7OMOPTNzDLi0Dczy4hD38wsIw59M7OMOPTNzDLi0Dczy4hD38wsIw1DX9I5kn4o6X8lPS/pX1N9paR9kg5K+raks1P9jWl+NC3vq9nWran+oqSrO3VQZmY2vWbe6f8euCoiLgEuBdZJWg18Hrg7IvqBE8Cm1H4TcCIi3gHcndoh6WKqt058D7AO+Kqks9p5MGZmNrtmboweETGRZt+QfgK4CtiV6juA69P0+jRPWr5GklL9/oj4fUS8DIwyze0Wzcysc5q6R256R/4U8A7gK8BPgFcjYjI1GQOWpellwBGAiJiUdBJ4a6o/XrPZ2nVq9zUEDAGUSiUqlcrcjqjGxMRES+vP1/CqycaNuqy0uLP9KmKcW1HUc2Mh8ljU6/XxaCr0I+JPwKWSlgAPAO+erll61AzLZqqfvq8RYARgYGAgyuVyM12cVqVSoZX15+umLQ91fZ+NDK+a5K79Tf1zz8uhjeWObbsTinpuLEQei3q9Ph5zunonIl4FKsBqYImkqRRZDhxN02PACoC0/C3AeG19mnXMzKwLmrl6523pHT6SFgMfAg4AjwEfTc0GgQfT9O40T1r+aEREqm9IV/esBPqBH7brQMzMrLFmft+/CNiRzuv/FbAzIr4n6QXgfkmfA54BtqX224BvSBql+g5/A0BEPC9pJ/ACMAlsTqeNzMysSxqGfkQ8B7xvmvpLTHP1TUT8Drhhhm3dAdwx926amVk7+C9yzcwy4tA3M8uIQ9/MLCMOfTOzjDj0zcwy4tA3M8uIQ9/MLCMOfTOzjDj0zcwy4tA3M8uIQ9/MLCMOfTOzjDj0zcwy4tA3M8uIQ9/MLCPN3DlrhaTHJB2Q9LykW1L9fEl7JB1Mj0tTXZLukTQq6TlJl9VsazC1PyhpcKZ9mplZZzTzTn8SGI6Id1O9N+5mSRcDW4C9EdEP7E3zANdQvRViPzAE3AvVFwlgK3AF1ZuvbJ16oTAzs+5oGPoRcSwink7Tv6F6f9xlwHpgR2q2A7g+Ta8H7ouqx6neQP0i4GpgT0SMR8QJYA+wrq1HY2Zms5rTOX1JfVRvnbgPKEXEMai+MAAXpmbLgCM1q42l2kx1MzPrkmZujA6ApDcB3wE+FRG/ljRj02lqMUv99P0MUT0tRKlUolKpNNvF15mYmGhp/fkaXjXZ9X02Ulrc2X4VMc6tKOq5sRB5LOr1+ng0FfqS3kA18L8ZEd9N5VckXRQRx9Lpm+OpPgasqFl9OXA01cun1Sun7ysiRoARgIGBgSiXy6c3aVqlUqGV9efrpi0PdX2fjQyvmuSu/U2/xs/ZoY3ljm27E4p6bixEHot6vT4ezVy9I2AbcCAivlizaDcwdQXOIPBgTf3GdBXPauBkOv3zCLBW0tL0Ae7aVDMzsy5p5q3flcDHgf2Snk21zwJ3AjslbQIOAzekZQ8D1wKjwGvAzQARMS7pduCJ1O62iBhvy1GYmVlTGoZ+RPwP05+PB1gzTfsANs+wre3A9rl00MzM2sd/kWtmlhGHvplZRhz6ZmYZceibmWXEoW9mlhGHvplZRhz6ZmYZceibmWXEoW9mlhGHvplZRhz6ZmYZceibmWXEoW9mlhGHvplZRjp3KyXLSl9Bdws7dOd1hezX7Ezld/pmZhlp5naJ2yUdl/Sjmtr5kvZIOpgel6a6JN0jaVTSc5Iuq1lnMLU/KGlwun2ZmVlnNfNO/9+BdafVtgB7I6If2JvmAa4B+tPPEHAvVF8kgK3AFcDlwNapFwozM+uehqEfEf8NnH4v2/XAjjS9A7i+pn5fVD0OLJF0EXA1sCcixiPiBLCH17+QmJlZh833g9xSRBwDiIhjki5M9WXAkZp2Y6k2U/11JA1R/S2BUqlEpVKZZxdhYmKipfXna3jVZNf32Uhp8cLsV6vm++9b1HNjIfJY1Ov18Wj31TvT3UA9Zqm/vhgxAowADAwMRLlcnndnKpUKraw/XzcVdCXLbIZXTXLX/t67WOvQxvK81ivqubEQeSzq9fp4zPfqnVfSaRvS4/FUHwNW1LRbDhydpW5mZl0039DfDUxdgTMIPFhTvzFdxbMaOJlOAz0CrJW0NH2AuzbVzMysixr+vi/pW0AZuEDSGNWrcO4EdkraBBwGbkjNHwauBUaB14CbASJiXNLtwBOp3W0RcfqHw2Zm1mENQz8iPjbDojXTtA1g8wzb2Q5sn1PvzMysrfwXuWZmGXHom5llxKFvZpYRh76ZWUYc+mZmGXHom5llxKFvZpYRh76ZWUYc+mZmGXHom5llxKFvZpYRh76ZWUYc+mZmGXHom5llxKFvZpaR3rtpqmWlb573Ix5eNdnyvYwP3XldS+ubFaHr7/QlrZP0oqRRSVu6vX8zs5x1NfQlnQV8BbgGuBj4mKSLu9kHM7Ocdfv0zuXAaES8BCDpfmA98EIndrb/Zydb/hXebCbzPbXUKp9WslZ0O/SXAUdq5seAK2obSBoChtLshKQXW9jfBcAvW1i/Z3zSY1HnTB4Pfb7tmzxjx6JDemE8/namBd0OfU1Ti7qZiBFgpC07k56MiIF2bOtM57Go5/E4xWNRr9fHo9sf5I4BK2rmlwNHu9wHM7NsdTv0nwD6Ja2UdDawAdjd5T6YmWWrq6d3ImJS0j8DjwBnAdsj4vkO7rItp4l6hMeinsfjFI9FvZ4eD0VE41ZmZtYT/DUMZmYZceibmWWkJ0PfX/VwiqQVkh6TdEDS85JuKbpPRZN0lqRnJH2v6L4UTdISSbsk/Tg9R95fdJ+KJOnT6f/JjyR9S9I5Rfep3Xou9P1VD68zCQxHxLuB1cDmzMcD4BbgQNGdWCC+DHw/It4FXELG4yJpGfBJYCAi3kv1YpMNxfaq/Xou9Kn5qoeI+AMw9VUPWYqIYxHxdJr+DdX/1MuK7VVxJC0HrgO+VnRfiibpzcAHgW0AEfGHiHi12F4VbhGwWNIi4Fx68O+IejH0p/uqh2xDrpakPuB9wL5ie1KoLwGfAf5cdEcWgLcDvwC+nk53fU3SeUV3qigR8TPgC8Bh4BhwMiJ+UGyv2q8XQ7/hVz3kSNKbgO8An4qIXxfdnyJI+jBwPCKeKrovC8Qi4DLg3oh4H/BbINvPwCQtpXpWYCXwN8B5kv6x2F61Xy+Gvr/q4TSS3kA18L8ZEd8tuj8FuhL4iKRDVE/7XSXpP4rtUqHGgLGImPrNbxfVF4FcfQh4OSJ+ERF/BL4LfKDgPrVdL4a+v+qhhiRRPWd7ICK+WHR/ihQRt0bE8ojoo/q8eDQieu6dXLMi4ufAEUnvTKU1dOhrzs8Qh4HVks5N/2/W0IMfbPfc7RIL+KqHhe5K4OPAfknPptpnI+LhAvtkC8cngG+mN0gvATcX3J/CRMQ+SbuAp6le9fYMPfiVDP4aBjOzjPTi6R0zM5uBQ9/MLCMOfTOzjDj0zcwy4tA3M8uIQ9/MLCMOfTOzjPw/o2MWUIhcE2QAAAAASUVORK5CYII=\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"training_set.Bathroom.hist(bins=10)"
]
},
{
"cell_type": "code",
"execution_count": 45,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"1.0"
]
},
"execution_count": 45,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set.Bathroom.mode()[0]"
]
},
{
"cell_type": "code",
"execution_count": 46,
"metadata": {},
"outputs": [],
"source": [
"training_set.Bathroom.fillna(value=training_set.Bathroom.mode()[0], inplace=True)"
]
},
{
"cell_type": "code",
"execution_count": 47,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(0, 15)"
]
},
"execution_count": 47,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set[training_set.Bathroom.isnull()].shape"
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": true
},
"source": [
"# --------------------------------------------------------------------"
]
},
{
"cell_type": "code",
"execution_count": 48,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(5472, 15)"
]
},
"execution_count": 48,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set[training_set.Car.isnull()].shape"
]
},
{
"cell_type": "code",
"execution_count": 49,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"<matplotlib.axes._subplots.AxesSubplot at 0x19bc051ad30>"
]
},
"execution_count": 49,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"training_set.Car.hist(bins=10)"
]
},
{
"cell_type": "code",
"execution_count": 50,
"metadata": {},
"outputs": [],
"source": [
"training_set.Car.fillna(value=training_set.Car.median(), inplace=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# --------------------------------------------------------------------"
]
},
{
"cell_type": "code",
"execution_count": 51,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(7443, 15)"
]
},
"execution_count": 51,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set[training_set.Landsize.isnull()].shape"
]
},
{
"cell_type": "code",
"execution_count": 52,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"<matplotlib.axes._subplots.AxesSubplot at 0x19bc1268208>"
]
},
"execution_count": 52,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"training_set.Landsize.hist(bins=20)"
]
},
{
"cell_type": "code",
"execution_count": 53,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"count 14353.000000\n",
"mean 570.962447\n",
"std 1956.911573\n",
"min 0.000000\n",
"25% 220.000000\n",
"50% 513.000000\n",
"75% 664.000000\n",
"max 146699.000000\n",
"Name: Landsize, dtype: float64"
]
},
"execution_count": 53,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set.Landsize.describe()"
]
},
{
"cell_type": "code",
"execution_count": 54,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(1540, 15)"
]
},
"execution_count": 54,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set[training_set.Landsize == 0].shape"
]
},
{
"cell_type": "code",
"execution_count": 55,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(3778, 15)"
]
},
"execution_count": 55,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set[training_set.Landsize > 656].shape"
]
},
{
"cell_type": "code",
"execution_count": 56,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(3571, 15)"
]
},
"execution_count": 56,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set[training_set.Landsize < 219].shape"
]
},
{
"cell_type": "code",
"execution_count": 57,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(0, 15)"
]
},
"execution_count": 57,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set[training_set.Landsize < 0].shape"
]
},
{
"cell_type": "code",
"execution_count": 58,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(12813, 15)"
]
},
"execution_count": 58,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set[~training_set.Landsize.isnull() & training_set.Landsize > 0].shape"
]
},
{
"cell_type": "code",
"execution_count": 59,
"metadata": {},
"outputs": [],
"source": [
"training_set[\"Landsize_log\"] = np.log(training_set[~training_set.Landsize.isnull() & training_set.Landsize > 0]['Landsize'])"
]
},
{
"cell_type": "code",
"execution_count": 60,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"count 12813.000000\n",
"mean 6.127162\n",
"std 0.705189\n",
"min 0.000000\n",
"25% 5.717028\n",
"50% 6.315358\n",
"75% 6.530878\n",
"max 11.896138\n",
"Name: Landsize_log, dtype: float64"
]
},
"execution_count": 60,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set[\"Landsize_log\"].describe()"
]
},
{
"cell_type": "code",
"execution_count": 61,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"<matplotlib.axes._subplots.AxesSubplot at 0x19bc1661ba8>"
]
},
"execution_count": 61,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"training_set[\"Landsize_log\"].hist(bins=15)"
]
},
{
"cell_type": "code",
"execution_count": 62,
"metadata": {},
"outputs": [],
"source": [
"Landsize_log_mean = training_set[\"Landsize_log\"].mean()\n",
"training_set[\"Landsize_log\"].fillna(value=Landsize_log_mean, inplace=True)\n",
"training_set[\"Landsize_log\"] = training_set[\"Landsize_log\"].apply(lambda x: Landsize_log_mean if x == 0 else x) "
]
},
{
"cell_type": "code",
"execution_count": 63,
"metadata": {},
"outputs": [],
"source": [
"training_set.drop('Landsize', axis=1, inplace=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# --------------------------------------------------------------------"
]
},
{
"cell_type": "code",
"execution_count": 64,
"metadata": {},
"outputs": [],
"source": [
"training_set.fillna(value= training_set.mean()[[\"BuildingArea\", \"YearBuilt\", \"Propertycount\"]], inplace=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# --------------------------------------------------------------------"
]
},
{
"cell_type": "code",
"execution_count": 65,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Rooms False\n",
"Type False\n",
"Price False\n",
"Method False\n",
"SellerG False\n",
"Distance False\n",
"Postcode False\n",
"Bedroom2 False\n",
"Bathroom False\n",
"Car False\n",
"BuildingArea False\n",
"YearBuilt False\n",
"Propertycount False\n",
"sold_year False\n",
"Landsize_log False\n",
"dtype: bool"
]
},
"execution_count": 65,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set.isnull().any()"
]
},
{
"cell_type": "code",
"execution_count": 66,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Rooms int64\n",
"Type object\n",
"Price float64\n",
"Method object\n",
"SellerG object\n",
"Distance float64\n",
"Postcode float64\n",
"Bedroom2 float64\n",
"Bathroom float64\n",
"Car float64\n",
"BuildingArea float64\n",
"YearBuilt float64\n",
"Propertycount float64\n",
"sold_year int64\n",
"Landsize_log float64\n",
"dtype: object"
]
},
"execution_count": 66,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set.dtypes"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Interested in knowing more about catagorical/text columns. Once after analysis we have to convert the data to numerical.\n",
" * Before we convert catagorical data to numbers, we must see if there is any effect of the \"individual catagories\" on the mean of the target variable.\n",
" * Example if there are 10 houses beloning to Method **\"S\"**, 100 houses belonging to Method **\"SP\"**.\n",
" Total Price of \"S\" houses in 100000, Total Price of \"SP\" houses is 1000000. **The avarage (mean) Price of \"S\" and \"SP\" houses is same - 10000**. That means \"statistically\" **there is no direct effect of Method \"S\", \"SP\" on Price variable** (<font color='red'>note: in business point of view, there might be an effect. Need to check statistical understanding with business analyst</font>). In this scenario we should go with **One-Hot-Encoding or Dummy Variable** creation. Otherwise we can go with **Label Encoding**."
]
},
{
"cell_type": "code",
"execution_count": 67,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Index(['Type', 'Method', 'SellerG'], dtype='object')"
]
},
"execution_count": 67,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set.select_dtypes(['object']).columns"
]
},
{
"cell_type": "code",
"execution_count": 68,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"S 13955\n",
"SP 2882\n",
"PI 2626\n",
"VB 2178\n",
"SA 155\n",
"Name: Method, dtype: int64"
]
},
"execution_count": 68,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set['Method'].value_counts()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Is there a significant effect of different categories in Method on Price"
]
},
{
"cell_type": "code",
"execution_count": 69,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Method\n",
"PI 1.122780e+06\n",
"S 1.049267e+06\n",
"SA 9.966541e+05\n",
"SP 8.802131e+05\n",
"VB 1.193416e+06\n",
"Name: Price, dtype: float64"
]
},
"execution_count": 69,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set.groupby(['Method'])['Price'].mean()"
]
},
{
"cell_type": "code",
"execution_count": 70,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Text(-21.325000000000003, 0.5, 'Mean Price')"
]
},
"execution_count": 70,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 360x360 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"sns.catplot(data=training_set, x=\"Method\", y=\"Price\", kind=\"bar\", ci=None)\n",
"plt.ylabel(\"Mean Price\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Above visualization depects that there is a little difference in mean price for different catagories of Method. Need to do hypothesis testing to make sure that this holds true in population"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### ANOVA: Tukey's Honestly Singnificant Difference test"
]
},
{
"cell_type": "code",
"execution_count": 71,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" Multiple Comparison of Means - Tukey HSD, FWER=0.05 \n",
"==================================================================\n",
"group1 group2 meandiff p-adj lower upper reject\n",
"------------------------------------------------------------------\n",
" PI S -73512.4251 0.001 -110261.7338 -36763.1163 True\n",
" PI SA -126125.7769 0.1129 -268930.766 16679.2122 False\n",
" PI SP -242566.7525 0.001 -289174.5642 -195958.9408 True\n",
" PI VB 70636.5441 0.0011 20566.1169 120706.9713 True\n",
" S SA -52613.3518 0.8216 -192150.1757 86923.4721 False\n",
" S SP -169054.3274 0.001 -204403.2795 -133705.3754 True\n",
" S VB 144148.9692 0.001 104345.6027 183952.3356 True\n",
" SA SP -116440.9756 0.1688 -258892.0255 26010.0743 False\n",
" SA VB 196762.321 0.0017 53141.0788 340383.5631 True\n",
" SP VB 313203.2966 0.001 264151.4424 362255.1507 True\n",
"------------------------------------------------------------------\n"
]
}
],
"source": [
"import statsmodels.stats.multicomp as multi\n",
"\n",
"mc1 = multi.MultiComparison(training_set[\"Price\"],training_set[\"Method\"])\n",
"res1 = mc1.tukeyhsd()\n",
"print(res1.summary())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Most of the pair wise ANOVA tests show that there is a significance \"Price\" difference for different catagories of \"Mehod\". Hence we can do LavelEncoding"
]
},
{
"cell_type": "code",
"execution_count": 72,
"metadata": {},
"outputs": [],
"source": [
"lst_all_method_cats = ['S', 'SP','PI','PN','SN','NB','VB','W','SA','SS','N/A']"
]
},
{
"cell_type": "code",
"execution_count": 73,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"S 13955\n",
"SP 2882\n",
"PI 2626\n",
"VB 2178\n",
"SA 155\n",
"N/A 0\n",
"SS 0\n",
"W 0\n",
"NB 0\n",
"SN 0\n",
"PN 0\n",
"Name: Method, dtype: int64"
]
},
"execution_count": 73,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set.Method = pd.Categorical(training_set.Method, categories=lst_all_method_cats)\n",
"training_set.Method.value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 74,
"metadata": {},
"outputs": [],
"source": [
"training_set = pd.get_dummies(training_set, columns=[\"Method\"], prefix=[\"Method\"], drop_first=True)"
]
},
{
"cell_type": "code",
"execution_count": 75,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"['Method_SP',\n",
" 'Method_PI',\n",
" 'Method_PN',\n",
" 'Method_SN',\n",
" 'Method_NB',\n",
" 'Method_VB',\n",
" 'Method_W',\n",
" 'Method_SA',\n",
" 'Method_SS',\n",
" 'Method_N/A']"
]
},
"execution_count": 75,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"[col for col in training_set.columns if 'Method' in col ]"
]
},
{
"cell_type": "code",
"execution_count": 76,
"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>Method_SP</th>\n",
" <th>Method_PI</th>\n",
" <th>Method_PN</th>\n",
" <th>Method_SN</th>\n",
" <th>Method_NB</th>\n",
" <th>Method_VB</th>\n",
" <th>Method_W</th>\n",
" <th>Method_SA</th>\n",
" <th>Method_SS</th>\n",
" <th>Method_N/A</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>17957</th>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>6753</th>\n",
" <td>1</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1833</th>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>1</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>19795</th>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>17054</th>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" <td>0</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" Method_SP Method_PI Method_PN Method_SN Method_NB Method_VB \\\n",
"17957 0 0 0 0 0 1 \n",
"6753 1 0 0 0 0 0 \n",
"1833 0 0 0 0 0 1 \n",
"19795 0 0 0 0 0 0 \n",
"17054 0 0 0 0 0 0 \n",
"\n",
" Method_W Method_SA Method_SS Method_N/A \n",
"17957 0 0 0 0 \n",
"6753 0 0 0 0 \n",
"1833 0 0 0 0 \n",
"19795 0 0 0 0 \n",
"17054 0 0 0 0 "
]
},
"execution_count": 76,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set[[col for col in training_set.columns if 'Method' in col]].head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# --------------------------------------------------------------------"
]
},
{
"cell_type": "code",
"execution_count": 77,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"h 14781\n",
"u 4733\n",
"t 2282\n",
"Name: Type, dtype: int64"
]
},
"execution_count": 77,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set['Type'].value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 78,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Type\n",
"h 1.203264e+06\n",
"t 9.375948e+05\n",
"u 6.246408e+05\n",
"Name: Price, dtype: float64"
]
},
"execution_count": 78,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set.groupby(['Type'])['Price'].mean()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### We see significant difference, hence we can go with Label Encoding"
]
},
{
"cell_type": "code",
"execution_count": 79,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"0 14781\n",
"2 4733\n",
"1 2282\n",
"Name: Type, dtype: int64"
]
},
"execution_count": 79,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set['Type'] = pd.Categorical(training_set['Type'])\n",
"training_set['Type'] = training_set['Type'].cat.codes\n",
"training_set['Type'].value_counts()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# --------------------------------------------------------------------"
]
},
{
"cell_type": "code",
"execution_count": 80,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(330,)"
]
},
"execution_count": 80,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set['SellerG'].value_counts().shape"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Check the behaviour of the top sellers."
]
},
{
"cell_type": "code",
"execution_count": 81,
"metadata": {},
"outputs": [],
"source": [
"top_sellers = list(training_set['SellerG'].value_counts()[training_set['SellerG'].value_counts().values > 100].index)"
]
},
{
"cell_type": "code",
"execution_count": 82,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"['Nelson', 'Jellis', 'Barry', 'hockingstuart', 'Ray', 'Buxton', 'Marshall', 'Biggin', 'Fletchers', 'Brad', 'Woodards', 'McGrath', 'Greg', 'YPA', 'Noel', 'Jas', 'Stockdale', 'Miles', 'Sweeney', 'RT', 'Harcourts', 'Gary', 'Hodges', 'Raine', 'HAR', 'Love', 'RW', 'Kay', \"O'Brien\", 'Williams', 'Village', 'Douglas', 'C21', 'Chisholm', 'Purplebricks']\n"
]
}
],
"source": [
"print(top_sellers)"
]
},
{
"cell_type": "code",
"execution_count": 83,
"metadata": {},
"outputs": [],
"source": [
"top_sellers_train_set = training_set[training_set['SellerG'].isin(top_sellers)]"
]
},
{
"cell_type": "code",
"execution_count": 84,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 720x720 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"top_sel = sns.catplot(data=top_sellers_train_set, x=\"SellerG\", y=\"Price\", kind=\"bar\", height=10, ci=None)\n",
"plt.ylabel(\"Mean Price\")\n",
"l1 = top_sel.set_yticklabels(rotation = 45)\n",
"l2 = top_sel.set_xticklabels(rotation = 45)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### We see significant difference, hence we can go with Label Encoding"
]
},
{
"cell_type": "code",
"execution_count": 85,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"199 2174\n",
"138 2047\n",
"21 1919\n",
"319 1694\n",
"245 1263\n",
"Name: SellerG, dtype: int64"
]
},
"execution_count": 85,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set['SellerG'] = pd.Categorical(training_set['SellerG'])\n",
"training_set['SellerG'] = training_set['SellerG'].cat.codes\n",
"training_set['SellerG'].value_counts().head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Now data is ready, lets build a Linear Regression model with SGDRegressor"
]
},
{
"cell_type": "code",
"execution_count": 86,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Rooms int64\n",
"Type int8\n",
"Price float64\n",
"SellerG int16\n",
"Distance float64\n",
"Postcode float64\n",
"Bedroom2 float64\n",
"Bathroom float64\n",
"Car float64\n",
"BuildingArea float64\n",
"YearBuilt float64\n",
"Propertycount float64\n",
"sold_year int64\n",
"Landsize_log float64\n",
"Method_SP uint8\n",
"Method_PI uint8\n",
"Method_PN uint8\n",
"Method_SN uint8\n",
"Method_NB uint8\n",
"Method_VB uint8\n",
"Method_W uint8\n",
"Method_SA uint8\n",
"Method_SS uint8\n",
"Method_N/A uint8\n",
"dtype: object"
]
},
"execution_count": 86,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"training_set.dtypes"
]
},
{
"cell_type": "code",
"execution_count": 87,
"metadata": {},
"outputs": [],
"source": [
"input_features = [x for x in training_set.columns if x not in ['Price']]"
]
},
{
"cell_type": "code",
"execution_count": 88,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"['Rooms',\n",
" 'Type',\n",
" 'SellerG',\n",
" 'Distance',\n",
" 'Postcode',\n",
" 'Bedroom2',\n",
" 'Bathroom',\n",
" 'Car',\n",
" 'BuildingArea',\n",
" 'YearBuilt',\n",
" 'Propertycount',\n",
" 'sold_year',\n",
" 'Landsize_log',\n",
" 'Method_SP',\n",
" 'Method_PI',\n",
" 'Method_PN',\n",
" 'Method_SN',\n",
" 'Method_NB',\n",
" 'Method_VB',\n",
" 'Method_W',\n",
" 'Method_SA',\n",
" 'Method_SS',\n",
" 'Method_N/A']"
]
},
"execution_count": 88,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"input_features"
]
},
{
"cell_type": "code",
"execution_count": 89,
"metadata": {},
"outputs": [],
"source": [
"X_train = training_set[input_features].values\n",
"y_train = training_set['Price'].values"
]
},
{
"cell_type": "code",
"execution_count": 90,
"metadata": {},
"outputs": [],
"source": [
"lr = LinearRegression()\n",
"lr_model = lr.fit(X_train, y_train)"
]
},
{
"cell_type": "code",
"execution_count": 91,
"metadata": {},
"outputs": [],
"source": [
"y_train_pred = lr_model.predict(X_train)"
]
},
{
"cell_type": "code",
"execution_count": 92,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"0.48467536024042435\n"
]
}
],
"source": [
"from sklearn.metrics import r2_score\n",
"r2 = r2_score(y_train, y_train_pred)\n",
"print(r2)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Error analysis and finetuning the model"
]
},
{
"cell_type": "code",
"execution_count": 93,
"metadata": {},
"outputs": [],
"source": [
"residuals = y_train - y_train_pred"
]
},
{
"cell_type": "code",
"execution_count": 94,
"metadata": {},
"outputs": [],
"source": [
"def standard_scale(val):\n",
" mean_vals = np.mean(val)\n",
" std_vals = np.std(val)\n",
" z_vals = [(x - mean_vals)/std_vals for x in val]\n",
" return z_vals"
]
},
{
"cell_type": "code",
"execution_count": 95,
"metadata": {},
"outputs": [],
"source": [
"def plotScatterPlot(residuals, y):\n",
" std_residuals = standard_scale(residuals)\n",
" std_predicted = standard_scale(y)\n",
" plt.scatter(std_predicted, std_residuals, alpha = 0.6, color='y')\n",
" l = plt.axhline(y=0, color='r')\n",
" plt.xlabel('STD Predicted values')\n",
" plt.ylabel('STD Residual values')\n",
" plt.axis([-6, 6, -10, 10])\n",
" plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### plotScatterPlot(residuals, y_train_pred)"
]
},
{
"cell_type": "code",
"execution_count": 96,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"plotScatterPlot(residuals, y_train_pred)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### We found heteroscedasticity in the dataset, one of the solution is the apply log transormation on y variable, build model. Let's try that."
]
},
{
"cell_type": "code",
"execution_count": 97,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"numpy.ndarray"
]
},
"execution_count": 97,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"type(y_train)"
]
},
{
"cell_type": "code",
"execution_count": 98,
"metadata": {},
"outputs": [],
"source": [
"y_train_log = np.log(y_train)"
]
},
{
"cell_type": "code",
"execution_count": 99,
"metadata": {},
"outputs": [],
"source": [
"lr_model = lr_model.fit(X_train, y_train_log)"
]
},
{
"cell_type": "code",
"execution_count": 100,
"metadata": {},
"outputs": [],
"source": [
"y_train_log_pred = lr_model.predict(X_train)"
]
},
{
"cell_type": "code",
"execution_count": 101,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"0.5775719437885163\n"
]
}
],
"source": [
"r2 = r2_score(y_train_log, y_train_log_pred)\n",
"print(r2)"
]
},
{
"cell_type": "code",
"execution_count": 102,
"metadata": {},
"outputs": [],
"source": [
"res = y_train_log - y_train_log_pred"
]
},
{
"cell_type": "code",
"execution_count": 103,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"plotScatterPlot(res, y_train_log_pred)"
]
},
{
"cell_type": "code",
"execution_count": 104,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.ensemble import GradientBoostingRegressor\n",
"gbrt = GradientBoostingRegressor(max_depth=4, n_estimators=300, learning_rate=0.1, random_state=42)\n",
"gbrt.fit(X_train, y_train)\n",
"y_pred_gbrt = gbrt.predict(X_train)"
]
},
{
"cell_type": "code",
"execution_count": 105,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"0.8431136902758644\n"
]
}
],
"source": [
"r2 = r2_score(y_train, y_pred_gbrt)\n",
"print(r2)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Apply model on TEST set and finetune model"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Apply EDA outcome on TEST set"
]
},
{
"cell_type": "code",
"execution_count": 106,
"metadata": {},
"outputs": [],
"source": [
"test_set = test_set.drop(test_set[test_set['Postcode'].isnull()].index)\n",
"\n",
"test_set.drop([\"Address\", \"CouncilArea\", \"Regionname\", \"Lattitude\", \"Suburb\", \"Longtitude\"], axis=1, inplace=True)\n",
"\n",
"test_set.Date = pd.to_datetime(test_set.Date)\n",
"test_set[\"sold_year\"] = test_set.Date.apply(lambda x: x.year)\n",
"test_set.drop(\"Date\", axis=1, inplace=True)\n",
"\n",
"test_set.Bedroom2.fillna(value=test_set.Bedroom2.mean(), inplace=True)\n",
"test_set.Bathroom.fillna(value=test_set.Bathroom.mean(), inplace=True)\n",
"test_set.Car.fillna(value=test_set.Car.median(), inplace=True)\n",
"\n",
"test_set[\"Landsize_log\"] = np.log(test_set[~test_set.Landsize.isnull() & test_set.Landsize > 0]['Landsize'])\n",
"Landsize_log_mean = test_set[\"Landsize_log\"].mean()\n",
"test_set[\"Landsize_log\"].fillna(value=Landsize_log_mean, inplace=True)\n",
"test_set[\"Landsize_log\"] = test_set[\"Landsize_log\"].apply(lambda x: Landsize_log_mean if x == 0 else x) \n",
"test_set.drop('Landsize', axis=1, inplace=True)\n",
"\n",
"test_set.fillna(value= test_set.mean()[[\"BuildingArea\", \"YearBuilt\", \"Propertycount\"]], inplace=True)\n",
"\n",
"lst_all_method_cats = ['S', 'SP','PI','PN','SN','NB','VB','W','SA','SS','N/A']\n",
"test_set.Method = pd.Categorical(test_set.Method, categories=lst_all_method_cats)\n",
"test_set = pd.get_dummies(test_set, columns=[\"Method\"], prefix=[\"Method\"], drop_first=True)\n",
"\n",
"test_set['Type'] = pd.Categorical(test_set['Type'])\n",
"test_set['Type'] = test_set['Type'].cat.codes\n",
"\n",
"test_set['SellerG'] = pd.Categorical(test_set['SellerG'])\n",
"test_set['SellerG'] = test_set['SellerG'].cat.codes"
]
},
{
"cell_type": "code",
"execution_count": 107,
"metadata": {},
"outputs": [],
"source": [
"input_features = [x for x in training_set.columns if x not in ['Price']]\n",
"\n",
"X_test = test_set[input_features].values\n",
"y_test = test_set['Price'].values"
]
},
{
"cell_type": "code",
"execution_count": 108,
"metadata": {},
"outputs": [],
"source": [
"y_test_pred_gbrt = gbrt.predict(X_test)"
]
},
{
"cell_type": "code",
"execution_count": 109,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"0.7504911229602481\n"
]
}
],
"source": [
"r2 = r2_score(y_test, y_test_pred_gbrt)\n",
"print(r2)"
]
},
{
"cell_type": "code",
"execution_count": 127,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"GridSearchCV(cv=3, error_score='raise-deprecating',\n",
" estimator=GradientBoostingRegressor(alpha=0.9,\n",
" criterion='friedman_mse',\n",
" init=None, learning_rate=0.1,\n",
" loss='ls', max_depth=3,\n",
" max_features=None,\n",
" max_leaf_nodes=None,\n",
" min_impurity_decrease=0.0,\n",
" min_impurity_split=None,\n",
" min_samples_leaf=1,\n",
" min_samples_split=2,\n",
" min_weight_fraction_leaf=0.0,\n",
" n_estimators=100,\n",
" n_iter_no_change=None,\n",
" presort='auto',\n",
" random_state=15, subsample=1.0,\n",
" tol=0.0001,\n",
" validation_fraction=0.1,\n",
" verbose=0, warm_start=False),\n",
" iid='warn', n_jobs=None,\n",
" param_grid=[{'learning_rate': [0.009, 0.1, 0.11, 0.12],\n",
" 'max_depth': [6, 7, 8], 'n_estimators': [300, 350]}],\n",
" pre_dispatch='2*n_jobs', refit=True, return_train_score=False,\n",
" scoring='neg_mean_squared_error', verbose=0)"
]
},
"execution_count": 127,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"from sklearn.model_selection import GridSearchCV\n",
"param_grid = [\n",
" {'max_depth':[6,7,8], \n",
" 'n_estimators':[300, 350], \n",
" 'learning_rate':[0.09, 0.1, 0.11, 0.12]} ]\n",
"grd_gbr_model = GradientBoostingRegressor(random_state=15)\n",
"grid_search = GridSearchCV(grd_gbr_model, param_grid, cv=3,\n",
" scoring='neg_mean_squared_error')\n",
"grid_search.fit(X_train, y_train)"
]
},
{
"cell_type": "code",
"execution_count": 128,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"{'learning_rate': 0.1, 'max_depth': 7, 'n_estimators': 300}"
]
},
"execution_count": 128,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"grid_search.best_params_"
]
},
{
"cell_type": "code",
"execution_count": 129,
"metadata": {},
"outputs": [],
"source": [
"best_y_pred_gbrt = grid_search.best_estimator_.predict(X_train)"
]
},
{
"cell_type": "code",
"execution_count": 130,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"0.9320388408375453\n"
]
}
],
"source": [
"r2 = r2_score(y_train, best_y_pred_gbrt)\n",
"print(r2)"
]
},
{
"cell_type": "code",
"execution_count": 131,
"metadata": {},
"outputs": [],
"source": [
"best_y_test_pred_gbrt = grid_search.best_estimator_.predict(X_test)"
]
},
{
"cell_type": "code",
"execution_count": 132,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"0.7575570409693281\n"
]
}
],
"source": [
"r2 = r2_score(y_test, best_y_test_pred_gbrt)\n",
"print(r2)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Use Log transformed Y - Variable to train"
]
},
{
"cell_type": "code",
"execution_count": 133,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"GridSearchCV(cv=3, error_score='raise-deprecating',\n",
" estimator=GradientBoostingRegressor(alpha=0.9,\n",
" criterion='friedman_mse',\n",
" init=None, learning_rate=0.1,\n",
" loss='ls', max_depth=3,\n",
" max_features=None,\n",
" max_leaf_nodes=None,\n",
" min_impurity_decrease=0.0,\n",
" min_impurity_split=None,\n",
" min_samples_leaf=1,\n",
" min_samples_split=2,\n",
" min_weight_fraction_leaf=0.0,\n",
" n_estimators=100,\n",
" n_iter_no_change=None,\n",
" presort='auto',\n",
" random_state=15, subsample=1.0,\n",
" tol=0.0001,\n",
" validation_fraction=0.1,\n",
" verbose=0, warm_start=False),\n",
" iid='warn', n_jobs=None,\n",
" param_grid=[{'learning_rate': [0.009, 0.1, 0.11, 0.12],\n",
" 'max_depth': [6, 7, 8], 'n_estimators': [300, 350]}],\n",
" pre_dispatch='2*n_jobs', refit=True, return_train_score=False,\n",
" scoring='neg_mean_squared_error', verbose=0)"
]
},
"execution_count": 133,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"grd_gbr_model = GradientBoostingRegressor(random_state=15)\n",
"grid_search = GridSearchCV(grd_gbr_model, param_grid, cv=3,\n",
" scoring='neg_mean_squared_error')\n",
"grid_search.fit(X_train, y_train_log)"
]
},
{
"cell_type": "code",
"execution_count": 134,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"{'learning_rate': 0.1, 'max_depth': 6, 'n_estimators': 350}"
]
},
"execution_count": 134,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"grid_search.best_params_"
]
},
{
"cell_type": "code",
"execution_count": 135,
"metadata": {},
"outputs": [],
"source": [
"best_y_pred_gbrt_log = grid_search.best_estimator_.predict(X_train)"
]
},
{
"cell_type": "code",
"execution_count": 136,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"0.9206247970542722\n"
]
}
],
"source": [
"r2 = r2_score(y_train_log, best_y_pred_gbrt_log)\n",
"print(r2)"
]
},
{
"cell_type": "code",
"execution_count": 137,
"metadata": {},
"outputs": [],
"source": [
"y_test_log = np.log(y_test)"
]
},
{
"cell_type": "code",
"execution_count": 138,
"metadata": {},
"outputs": [],
"source": [
"best_y_test_pred_gbrt_log = grid_search.best_estimator_.predict(X_test)"
]
},
{
"cell_type": "code",
"execution_count": 139,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"0.8348759436378475\n"
]
}
],
"source": [
"r2 = r2_score(y_test_log, best_y_test_pred_gbrt_log)\n",
"print(r2)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Play with model Hyper-parameters to fix overfitting"
]
},
{
"cell_type": "code",
"execution_count": 180,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.ensemble import GradientBoostingRegressor\n",
"gbrt1 = GradientBoostingRegressor(max_depth=2, n_estimators=550, learning_rate=0.1, random_state=42)\n",
"gbrt1.fit(X_train, y_train_log)\n",
"y_pred_gbrt_log1 = gbrt1.predict(X_train)"
]
},
{
"cell_type": "code",
"execution_count": 181,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"0.8448632507182897\n"
]
}
],
"source": [
"r2 = r2_score(y_train_log, y_pred_gbrt_log1)\n",
"print(r2)"
]
},
{
"cell_type": "code",
"execution_count": 182,
"metadata": {},
"outputs": [],
"source": [
"y_test_pred_gbrt_log1 = gbrt1.predict(X_test)"
]
},
{
"cell_type": "code",
"execution_count": 183,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"0.8202256090463653\n"
]
}
],
"source": [
"r2 = r2_score(y_test_log, y_test_pred_gbrt_log1)\n",
"print(r2)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.7.3"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
@thedatajango
Copy link
Author

first cut

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment