From d25ed9661e285ca0a3def56c89909ff26f1542e1 Mon Sep 17 00:00:00 2001 From: Alexander Hess Date: Tue, 25 May 2021 08:08:54 +0200 Subject: [PATCH] Rename notebook 1 --- 00_data_cleaning.ipynb | 4353 ++++++++++++++++++++++++++++++++++++++++ 1_data_cleaning.ipynb | 4353 ---------------------------------------- README.md | 2 +- 3 files changed, 4354 insertions(+), 4354 deletions(-) create mode 100644 00_data_cleaning.ipynb delete mode 100644 1_data_cleaning.ipynb diff --git a/00_data_cleaning.ipynb b/00_data_cleaning.ipynb new file mode 100644 index 0000000..547dc52 --- /dev/null +++ b/00_data_cleaning.ipynb @@ -0,0 +1,4353 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# Data Cleaning" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## \"Housekeeping\"" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Import all the third-party (scientific) libraries needed." + ] + }, + { + "cell_type": "code", + "execution_count": 1, + "metadata": {}, + "outputs": [], + "source": [ + "import missingno as msno\n", + "import numpy as np\n", + "import pandas as pd" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "The *utils.py* module defines helper dictionaries, lists, and functions that help with parsing the data types correctly, look up column descriptions, and refer to groups of data columns.\n", + "\n", + "**Note:** the suffix \\_*COLUMNS* indicates a dictionary with all meta information on the provided data file and \\_*VARIABLES* a list with only the column names (i.e., the keys of the respective \\_*COLUMNS* dictionary)." + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "metadata": {}, + "outputs": [], + "source": [ + "from utils import (\n", + " ALL_COLUMNS,\n", + " ALL_VARIABLES,\n", + " CONTINUOUS_COLUMNS,\n", + " CONTINUOUS_VARIABLES,\n", + " DISCRETE_COLUMNS,\n", + " DISCRETE_VARIABLES,\n", + " INDEX_COLUMNS,\n", + " LABEL_COLUMNS, # groups nominal and ordinal\n", + " LABEL_TYPES,\n", + " NOMINAL_COLUMNS,\n", + " NOMINAL_VARIABLES,\n", + " NUMERIC_VARIABLES, # groups continuous and discrete\n", + " ORDINAL_COLUMNS,\n", + " ORDINAL_VARIABLES,\n", + " TARGET_VARIABLES, # = Sale Price\n", + " correct_column_names,\n", + " print_column_list,\n", + " update_column_descriptions,\n", + ")" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "metadata": {}, + "outputs": [], + "source": [ + "# Show all data columns.\n", + "pd.set_option(\"display.max_columns\", 100)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Load the Data" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "The original data are available for [download](https://www.amstat.org/publications/jse/v19n3/decock/AmesHousing.xls) and a detailed description of the data types for each column can be found [here](https://www.amstat.org/publications/jse/v19n3/decock/DataDocumentation.txt). These meta data go into the `dtype` argument of the `read_excel` function below to parse the data correctly. There are four different generic data types defined that are casted as follows:\n", + "\n", + "- continuous -> np.float64\n", + "- discrete -> actually np.int64 but np.float64 because of missing values\n", + "- nominal -> object (str)\n", + "- ordinal -> object (str), the order can be looked up in the above mentioned *ALL_COLUMNS* dictionary\n", + "\n", + "**Note 1:** the data come with a lot of \"NA\" text strings that do **not** indicate missing data but, for example, the absence of a basement or a parking lot (see the linked data description).\n", + "\n", + "**Note 2:** the mappings from column names to data types are encoded in the \"utils.py\" module that defines the aforementioned helper dictionaries / lists.\n", + "\n", + "**Note 3:** the Excel file with all the data is either loaded from the local dictionary (= \"cache\") or obtained fresh from the source." + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "metadata": {}, + "outputs": [], + "source": [ + "# To avoid redundancy.\n", + "kwargs = {\n", + " \"dtype\": { # Ensure each column is parsed as the correct data type.\n", + " column: ( # This creates a mapping from column name to data type.\n", + " object if mapping_info[\"type\"] in LABEL_TYPES else np.float64\n", + " )\n", + " for (column, mapping_info) in ALL_COLUMNS.items()\n", + " },\n", + " \"na_values\": \"\", # By default, pandas treats NA strings as missing,\n", + " \"keep_default_na\": False, # which is not the correct meaning here.\n", + "}\n", + "\n", + "try:\n", + " df = pd.read_excel(\"data/data_raw.xls\", **kwargs)\n", + "except FileNotFoundError:\n", + " df = pd.read_excel(\n", + " \"https://www.amstat.org/publications/jse/v19n3/decock/AmesHousing.xls\", **kwargs\n", + " )\n", + " # Cache the obtained file.\n", + " df.to_excel(\"data/data_raw.xls\")" + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "metadata": {}, + "outputs": [], + "source": [ + "# Some columns names differ between the Excel file and\n", + "# the data description file. Correct that with the values\n", + "# in the Excel file.\n", + "correct_column_names(df.columns)\n", + "# Use a compound index and keep both\n", + "# identifying columns in the DataFrame.\n", + "df = df.set_index(INDEX_COLUMNS)\n", + "# Put the provided columns into the same\n", + "# order as in the encoded description file.\n", + "# Note that the target variable \"SalePrice\"\n", + "# is not in the description file.\n", + "df = df[ALL_VARIABLES + TARGET_VARIABLES]" + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
1st Flr SF2nd Flr SF3Ssn PorchAlleyBedroom AbvGrBldg TypeBsmt CondBsmt ExposureBsmt Full BathBsmt Half BathBsmt QualBsmt Unf SFBsmtFin SF 1BsmtFin SF 2BsmtFin Type 1BsmtFin Type 2Central AirCondition 1Condition 2ElectricalEnclosed PorchExter CondExter QualExterior 1stExterior 2ndFenceFireplace QuFireplacesFoundationFull BathFunctionalGarage AreaGarage CarsGarage CondGarage FinishGarage QualGarage TypeGarage Yr BltGr Liv AreaHalf BathHeatingHeating QCHouse StyleKitchen AbvGrKitchen QualLand ContourLand SlopeLot AreaLot ConfigLot FrontageLot ShapeLow Qual Fin SFMS SubClassMS ZoningMas Vnr AreaMas Vnr TypeMisc FeatureMisc ValMo SoldNeighborhoodOpen Porch SFOverall CondOverall QualPaved DrivePool AreaPool QCRoof MatlRoof StyleSale ConditionSale TypeScreen PorchStreetTotRms AbvGrdTotal Bsmt SFUtilitiesWood Deck SFYear BuiltYear Remod/AddYr SoldSalePrice
OrderPID
15263011001656.00.00NA31FamGdGd1.00.0TA441.0639.00.0BLQUnfYNormNormSBrkr0.0TATABrkFacePlywoodNAGd2.0CBlock1.0Typ528.02.0TAFinTAAttchd1960.01656.00.0GasAFa1Story1TALvlGtl31770.0Corner141.0IR10.0020RL112.0StoneNA0.05.0NAmes62.056P0.0NACompShgHipNormalWD0.0Pave71080.0AllPub210.01960.01960.02010.0215000
2526350040896.00.00NA21FamTANo0.00.0TA270.0468.0144.0RecLwQYFeedrNormSBrkr0.0TATAVinylSdVinylSdMnPrvNA0.0CBlock1.0Typ730.01.0TAUnfTAAttchd1961.0896.00.0GasATA1Story1TALvlGtl11622.0Inside80.0Reg0.0020RH0.0NoneNA0.06.0NAmes0.065Y0.0NACompShgGableNormalWD120.0Pave5882.0AllPub140.01961.01961.02010.0105000
35263510101329.00.00NA31FamTANo0.00.0TA406.0923.00.0ALQUnfYNormNormSBrkr0.0TATAWd SdngWd SdngNANA0.0CBlock1.0Typ312.01.0TAUnfTAAttchd1958.01329.01.0GasATA1Story1GdLvlGtl14267.0Corner81.0IR10.0020RL108.0BrkFaceGar212500.06.0NAmes36.066Y0.0NACompShgHipNormalWD0.0Pave61329.0AllPub393.01958.01958.02010.0172000
45263530302110.00.00NA31FamTANo1.00.0TA1045.01065.00.0ALQUnfYNormNormSBrkr0.0TAGdBrkFaceBrkFaceNATA2.0CBlock2.0Typ522.02.0TAFinTAAttchd1968.02110.01.0GasAEx1Story1ExLvlGtl11160.0Corner93.0Reg0.0020RL0.0NoneNA0.04.0NAmes0.057Y0.0NACompShgHipNormalWD0.0Pave82110.0AllPub0.01968.01968.02010.0244000
5527105010928.0701.00NA31FamTANo0.00.0Gd137.0791.00.0GLQUnfYNormNormSBrkr0.0TATAVinylSdVinylSdMnPrvTA1.0PConc2.0Typ482.02.0TAFinTAAttchd1997.01629.01.0GasAGd2Story1TALvlGtl13830.0Inside74.0IR10.0060RL0.0NoneNA0.03.0Gilbert34.055Y0.0NACompShgGableNormalWD0.0Pave6928.0AllPub212.01997.01998.02010.0189900
6527105030926.0678.00NA31FamTANo0.00.0TA324.0602.00.0GLQUnfYNormNormSBrkr0.0TATAVinylSdVinylSdNAGd1.0PConc2.0Typ470.02.0TAFinTAAttchd1998.01604.01.0GasAEx2Story1GdLvlGtl9978.0Inside78.0IR10.0060RL20.0BrkFaceNA0.06.0Gilbert36.066Y0.0NACompShgGableNormalWD0.0Pave7926.0AllPub360.01998.01998.02010.0195500
75271271501338.00.00NA2TwnhsETAMn1.00.0Gd722.0616.00.0GLQUnfYNormNormSBrkr170.0TAGdCemntBdCmentBdNANA0.0PConc2.0Typ582.02.0TAFinTAAttchd2001.01338.00.0GasAEx1Story1GdLvlGtl4920.0Inside41.0Reg0.0120RL0.0NoneNA0.04.0StoneBr0.058Y0.0NACompShgGableNormalWD0.0Pave61338.0AllPub0.02001.02001.02010.0213500
85271450801280.00.00NA2TwnhsETANo0.00.0Gd1017.0263.00.0ALQUnfYNormNormSBrkr0.0TAGdHdBoardHdBoardNANA0.0PConc2.0Typ506.02.0TARFnTAAttchd1992.01280.00.0GasAEx1Story1GdHLSGtl5005.0Inside43.0IR10.0120RL0.0NoneNA0.01.0StoneBr82.058Y0.0NACompShgGableNormalWD144.0Pave51280.0AllPub0.01992.01992.02010.0191500
95271460301616.00.00NA2TwnhsETANo1.00.0Gd415.01180.00.0GLQUnfYNormNormSBrkr0.0TAGdCemntBdCmentBdNATA1.0PConc2.0Typ608.02.0TARFnTAAttchd1995.01616.00.0GasAEx1Story1GdLvlGtl5389.0Inside39.0IR10.0120RL0.0NoneNA0.03.0StoneBr152.058Y0.0NACompShgGableNormalWD0.0Pave51595.0AllPub237.01995.01996.02010.0236500
105271621301028.0776.00NA31FamTANo0.00.0TA994.00.00.0UnfUnfYNormNormSBrkr0.0TATAVinylSdVinylSdNATA1.0PConc2.0Typ442.02.0TAFinTAAttchd1999.01804.01.0GasAGd2Story1GdLvlGtl7500.0Inside60.0Reg0.0060RL0.0NoneNA0.06.0Gilbert60.057Y0.0NACompShgGableNormalWD0.0Pave7994.0AllPub140.01999.01999.02010.0189000
\n", + "
" + ], + "text/plain": [ + " 1st Flr SF 2nd Flr SF 3Ssn Porch Alley Bedroom AbvGr \\\n", + "Order PID \n", + "1 526301100 1656.0 0.0 0 NA 3 \n", + "2 526350040 896.0 0.0 0 NA 2 \n", + "3 526351010 1329.0 0.0 0 NA 3 \n", + "4 526353030 2110.0 0.0 0 NA 3 \n", + "5 527105010 928.0 701.0 0 NA 3 \n", + "6 527105030 926.0 678.0 0 NA 3 \n", + "7 527127150 1338.0 0.0 0 NA 2 \n", + "8 527145080 1280.0 0.0 0 NA 2 \n", + "9 527146030 1616.0 0.0 0 NA 2 \n", + "10 527162130 1028.0 776.0 0 NA 3 \n", + "\n", + " Bldg Type Bsmt Cond Bsmt Exposure Bsmt Full Bath \\\n", + "Order PID \n", + "1 526301100 1Fam Gd Gd 1.0 \n", + "2 526350040 1Fam TA No 0.0 \n", + "3 526351010 1Fam TA No 0.0 \n", + "4 526353030 1Fam TA No 1.0 \n", + "5 527105010 1Fam TA No 0.0 \n", + "6 527105030 1Fam TA No 0.0 \n", + "7 527127150 TwnhsE TA Mn 1.0 \n", + "8 527145080 TwnhsE TA No 0.0 \n", + "9 527146030 TwnhsE TA No 1.0 \n", + "10 527162130 1Fam TA No 0.0 \n", + "\n", + " Bsmt Half Bath Bsmt Qual Bsmt Unf SF BsmtFin SF 1 \\\n", + "Order PID \n", + "1 526301100 0.0 TA 441.0 639.0 \n", + "2 526350040 0.0 TA 270.0 468.0 \n", + "3 526351010 0.0 TA 406.0 923.0 \n", + "4 526353030 0.0 TA 1045.0 1065.0 \n", + "5 527105010 0.0 Gd 137.0 791.0 \n", + "6 527105030 0.0 TA 324.0 602.0 \n", + "7 527127150 0.0 Gd 722.0 616.0 \n", + "8 527145080 0.0 Gd 1017.0 263.0 \n", + "9 527146030 0.0 Gd 415.0 1180.0 \n", + "10 527162130 0.0 TA 994.0 0.0 \n", + "\n", + " BsmtFin SF 2 BsmtFin Type 1 BsmtFin Type 2 Central Air \\\n", + "Order PID \n", + "1 526301100 0.0 BLQ Unf Y \n", + "2 526350040 144.0 Rec LwQ Y \n", + "3 526351010 0.0 ALQ Unf Y \n", + "4 526353030 0.0 ALQ Unf Y \n", + "5 527105010 0.0 GLQ Unf Y \n", + "6 527105030 0.0 GLQ Unf Y \n", + "7 527127150 0.0 GLQ Unf Y \n", + "8 527145080 0.0 ALQ Unf Y \n", + "9 527146030 0.0 GLQ Unf Y \n", + "10 527162130 0.0 Unf Unf Y \n", + "\n", + " Condition 1 Condition 2 Electrical Enclosed Porch Exter Cond \\\n", + "Order PID \n", + "1 526301100 Norm Norm SBrkr 0.0 TA \n", + "2 526350040 Feedr Norm SBrkr 0.0 TA \n", + "3 526351010 Norm Norm SBrkr 0.0 TA \n", + "4 526353030 Norm Norm SBrkr 0.0 TA \n", + "5 527105010 Norm Norm SBrkr 0.0 TA \n", + "6 527105030 Norm Norm SBrkr 0.0 TA \n", + "7 527127150 Norm Norm SBrkr 170.0 TA \n", + "8 527145080 Norm Norm SBrkr 0.0 TA \n", + "9 527146030 Norm Norm SBrkr 0.0 TA \n", + "10 527162130 Norm Norm SBrkr 0.0 TA \n", + "\n", + " Exter Qual Exterior 1st Exterior 2nd Fence Fireplace Qu \\\n", + "Order PID \n", + "1 526301100 TA BrkFace Plywood NA Gd \n", + "2 526350040 TA VinylSd VinylSd MnPrv NA \n", + "3 526351010 TA Wd Sdng Wd Sdng NA NA \n", + "4 526353030 Gd BrkFace BrkFace NA TA \n", + "5 527105010 TA VinylSd VinylSd MnPrv TA \n", + "6 527105030 TA VinylSd VinylSd NA Gd \n", + "7 527127150 Gd CemntBd CmentBd NA NA \n", + "8 527145080 Gd HdBoard HdBoard NA NA \n", + "9 527146030 Gd CemntBd CmentBd NA TA \n", + "10 527162130 TA VinylSd VinylSd NA TA \n", + "\n", + " Fireplaces Foundation Full Bath Functional Garage Area \\\n", + "Order PID \n", + "1 526301100 2.0 CBlock 1.0 Typ 528.0 \n", + "2 526350040 0.0 CBlock 1.0 Typ 730.0 \n", + "3 526351010 0.0 CBlock 1.0 Typ 312.0 \n", + "4 526353030 2.0 CBlock 2.0 Typ 522.0 \n", + "5 527105010 1.0 PConc 2.0 Typ 482.0 \n", + "6 527105030 1.0 PConc 2.0 Typ 470.0 \n", + "7 527127150 0.0 PConc 2.0 Typ 582.0 \n", + "8 527145080 0.0 PConc 2.0 Typ 506.0 \n", + "9 527146030 1.0 PConc 2.0 Typ 608.0 \n", + "10 527162130 1.0 PConc 2.0 Typ 442.0 \n", + "\n", + " Garage Cars Garage Cond Garage Finish Garage Qual \\\n", + "Order PID \n", + "1 526301100 2.0 TA Fin TA \n", + "2 526350040 1.0 TA Unf TA \n", + "3 526351010 1.0 TA Unf TA \n", + "4 526353030 2.0 TA Fin TA \n", + "5 527105010 2.0 TA Fin TA \n", + "6 527105030 2.0 TA Fin TA \n", + "7 527127150 2.0 TA Fin TA \n", + "8 527145080 2.0 TA RFn TA \n", + "9 527146030 2.0 TA RFn TA \n", + "10 527162130 2.0 TA Fin TA \n", + "\n", + " Garage Type Garage Yr Blt Gr Liv Area Half Bath Heating \\\n", + "Order PID \n", + "1 526301100 Attchd 1960.0 1656.0 0.0 GasA \n", + "2 526350040 Attchd 1961.0 896.0 0.0 GasA \n", + "3 526351010 Attchd 1958.0 1329.0 1.0 GasA \n", + "4 526353030 Attchd 1968.0 2110.0 1.0 GasA \n", + "5 527105010 Attchd 1997.0 1629.0 1.0 GasA \n", + "6 527105030 Attchd 1998.0 1604.0 1.0 GasA \n", + "7 527127150 Attchd 2001.0 1338.0 0.0 GasA \n", + "8 527145080 Attchd 1992.0 1280.0 0.0 GasA \n", + "9 527146030 Attchd 1995.0 1616.0 0.0 GasA \n", + "10 527162130 Attchd 1999.0 1804.0 1.0 GasA \n", + "\n", + " Heating QC House Style Kitchen AbvGr Kitchen Qual \\\n", + "Order PID \n", + "1 526301100 Fa 1Story 1 TA \n", + "2 526350040 TA 1Story 1 TA \n", + "3 526351010 TA 1Story 1 Gd \n", + "4 526353030 Ex 1Story 1 Ex \n", + "5 527105010 Gd 2Story 1 TA \n", + "6 527105030 Ex 2Story 1 Gd \n", + "7 527127150 Ex 1Story 1 Gd \n", + "8 527145080 Ex 1Story 1 Gd \n", + "9 527146030 Ex 1Story 1 Gd \n", + "10 527162130 Gd 2Story 1 Gd \n", + "\n", + " Land Contour Land Slope Lot Area Lot Config Lot Frontage \\\n", + "Order PID \n", + "1 526301100 Lvl Gtl 31770.0 Corner 141.0 \n", + "2 526350040 Lvl Gtl 11622.0 Inside 80.0 \n", + "3 526351010 Lvl Gtl 14267.0 Corner 81.0 \n", + "4 526353030 Lvl Gtl 11160.0 Corner 93.0 \n", + "5 527105010 Lvl Gtl 13830.0 Inside 74.0 \n", + "6 527105030 Lvl Gtl 9978.0 Inside 78.0 \n", + "7 527127150 Lvl Gtl 4920.0 Inside 41.0 \n", + "8 527145080 HLS Gtl 5005.0 Inside 43.0 \n", + "9 527146030 Lvl Gtl 5389.0 Inside 39.0 \n", + "10 527162130 Lvl Gtl 7500.0 Inside 60.0 \n", + "\n", + " Lot Shape Low Qual Fin SF MS SubClass MS Zoning \\\n", + "Order PID \n", + "1 526301100 IR1 0.0 020 RL \n", + "2 526350040 Reg 0.0 020 RH \n", + "3 526351010 IR1 0.0 020 RL \n", + "4 526353030 Reg 0.0 020 RL \n", + "5 527105010 IR1 0.0 060 RL \n", + "6 527105030 IR1 0.0 060 RL \n", + "7 527127150 Reg 0.0 120 RL \n", + "8 527145080 IR1 0.0 120 RL \n", + "9 527146030 IR1 0.0 120 RL \n", + "10 527162130 Reg 0.0 060 RL \n", + "\n", + " Mas Vnr Area Mas Vnr Type Misc Feature Misc Val Mo Sold \\\n", + "Order PID \n", + "1 526301100 112.0 Stone NA 0.0 5.0 \n", + "2 526350040 0.0 None NA 0.0 6.0 \n", + "3 526351010 108.0 BrkFace Gar2 12500.0 6.0 \n", + "4 526353030 0.0 None NA 0.0 4.0 \n", + "5 527105010 0.0 None NA 0.0 3.0 \n", + "6 527105030 20.0 BrkFace NA 0.0 6.0 \n", + "7 527127150 0.0 None NA 0.0 4.0 \n", + "8 527145080 0.0 None NA 0.0 1.0 \n", + "9 527146030 0.0 None NA 0.0 3.0 \n", + "10 527162130 0.0 None NA 0.0 6.0 \n", + "\n", + " Neighborhood Open Porch SF Overall Cond Overall Qual \\\n", + "Order PID \n", + "1 526301100 NAmes 62.0 5 6 \n", + "2 526350040 NAmes 0.0 6 5 \n", + "3 526351010 NAmes 36.0 6 6 \n", + "4 526353030 NAmes 0.0 5 7 \n", + "5 527105010 Gilbert 34.0 5 5 \n", + "6 527105030 Gilbert 36.0 6 6 \n", + "7 527127150 StoneBr 0.0 5 8 \n", + "8 527145080 StoneBr 82.0 5 8 \n", + "9 527146030 StoneBr 152.0 5 8 \n", + "10 527162130 Gilbert 60.0 5 7 \n", + "\n", + " Paved Drive Pool Area Pool QC Roof Matl Roof Style \\\n", + "Order PID \n", + "1 526301100 P 0.0 NA CompShg Hip \n", + "2 526350040 Y 0.0 NA CompShg Gable \n", + "3 526351010 Y 0.0 NA CompShg Hip \n", + "4 526353030 Y 0.0 NA CompShg Hip \n", + "5 527105010 Y 0.0 NA CompShg Gable \n", + "6 527105030 Y 0.0 NA CompShg Gable \n", + "7 527127150 Y 0.0 NA CompShg Gable \n", + "8 527145080 Y 0.0 NA CompShg Gable \n", + "9 527146030 Y 0.0 NA CompShg Gable \n", + "10 527162130 Y 0.0 NA CompShg Gable \n", + "\n", + " Sale Condition Sale Type Screen Porch Street TotRms AbvGrd \\\n", + "Order PID \n", + "1 526301100 Normal WD 0.0 Pave 7 \n", + "2 526350040 Normal WD 120.0 Pave 5 \n", + "3 526351010 Normal WD 0.0 Pave 6 \n", + "4 526353030 Normal WD 0.0 Pave 8 \n", + "5 527105010 Normal WD 0.0 Pave 6 \n", + "6 527105030 Normal WD 0.0 Pave 7 \n", + "7 527127150 Normal WD 0.0 Pave 6 \n", + "8 527145080 Normal WD 144.0 Pave 5 \n", + "9 527146030 Normal WD 0.0 Pave 5 \n", + "10 527162130 Normal WD 0.0 Pave 7 \n", + "\n", + " Total Bsmt SF Utilities Wood Deck SF Year Built \\\n", + "Order PID \n", + "1 526301100 1080.0 AllPub 210.0 1960.0 \n", + "2 526350040 882.0 AllPub 140.0 1961.0 \n", + "3 526351010 1329.0 AllPub 393.0 1958.0 \n", + "4 526353030 2110.0 AllPub 0.0 1968.0 \n", + "5 527105010 928.0 AllPub 212.0 1997.0 \n", + "6 527105030 926.0 AllPub 360.0 1998.0 \n", + "7 527127150 1338.0 AllPub 0.0 2001.0 \n", + "8 527145080 1280.0 AllPub 0.0 1992.0 \n", + "9 527146030 1595.0 AllPub 237.0 1995.0 \n", + "10 527162130 994.0 AllPub 140.0 1999.0 \n", + "\n", + " Year Remod/Add Yr Sold SalePrice \n", + "Order PID \n", + "1 526301100 1960.0 2010.0 215000 \n", + "2 526350040 1961.0 2010.0 105000 \n", + "3 526351010 1958.0 2010.0 172000 \n", + "4 526353030 1968.0 2010.0 244000 \n", + "5 527105010 1998.0 2010.0 189900 \n", + "6 527105030 1998.0 2010.0 195500 \n", + "7 527127150 2001.0 2010.0 213500 \n", + "8 527145080 1992.0 2010.0 191500 \n", + "9 527146030 1996.0 2010.0 236500 \n", + "10 527162130 1999.0 2010.0 189000 " + ] + }, + "execution_count": 6, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "df.head(10)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Spelling Mistakes & Data Types" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Some textual values appear differently in the provided data file as compared to the specification. These inconsistencies are manually repaired." + ] + }, + { + "cell_type": "code", + "execution_count": 7, + "metadata": {}, + "outputs": [], + "source": [ + "# Repair spelling and whitespace mistakes.\n", + "df[\"Bldg Type\"] = df[\"Bldg Type\"].replace(to_replace=\"2fmCon\", value=\"2FmCon\")\n", + "df[\"Bldg Type\"] = df[\"Bldg Type\"].replace(to_replace=\"Duplex\", value=\"Duplx\")\n", + "df[\"Bldg Type\"] = df[\"Bldg Type\"].replace(to_replace=\"Twnhs\", value=\"TwnhsI\")\n", + "df[\"Exterior 2nd\"] = df[\"Exterior 2nd\"].replace(to_replace=\"Brk Cmn\", value=\"BrkComm\")\n", + "df[\"Exterior 2nd\"] = df[\"Exterior 2nd\"].replace(to_replace=\"CmentBd\", value=\"CemntBd\")\n", + "df[\"Exterior 2nd\"] = df[\"Exterior 2nd\"].replace(to_replace=\"Wd Shng\", value=\"WdShing\")\n", + "df[\"MS Zoning\"] = df[\"MS Zoning\"].replace(to_replace=\"A (agr)\", value=\"A\")\n", + "df[\"MS Zoning\"] = df[\"MS Zoning\"].replace(to_replace=\"C (all)\", value=\"C\")\n", + "df[\"MS Zoning\"] = df[\"MS Zoning\"].replace(to_replace=\"I (all)\", value=\"I\")\n", + "df[\"Neighborhood\"] = df[\"Neighborhood\"].replace(to_replace=\"NAmes\", value=\"Names\")\n", + "df[\"Sale Type\"] = df[\"Sale Type\"].replace(to_replace=\"WD \", value=\"WD\")" + ] + }, + { + "cell_type": "code", + "execution_count": 8, + "metadata": {}, + "outputs": [], + "source": [ + "# Ensure that the remaining textual values in the data file are a subset\n", + "# of the values allowed in the specification.\n", + "for column, mapping_info in LABEL_COLUMNS.items():\n", + " # Note that .unique() returns a numpy array with integer dtype in cases\n", + " # where the provided data can be casted as such (e.g., \"Overall Qual\" column).\n", + " values_in_data = set(str(x) for x in df[column].unique() if x is not np.NaN)\n", + " values_in_description = set(mapping_info[\"lookups\"].keys())\n", + " assert values_in_data <= values_in_description" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Interestingly, all numeric columns (i.e. also \"continuous\" variables) come with only integer values." + ] + }, + { + "cell_type": "code", + "execution_count": 9, + "metadata": {}, + "outputs": [], + "source": [ + "# Show that all \"continuous\" variables come as integers.\n", + "for column in NUMERIC_VARIABLES + TARGET_VARIABLES:\n", + " not_null = df[column].notnull()\n", + " mask = (\n", + " df.loc[not_null, column].astype(np.int64)\n", + " != df.loc[not_null, column].astype(np.float64)\n", + " )\n", + " assert not mask.any()\n", + "# Cast discrete fields as integers where possible,\n", + "# i.e., all columns without missing values.\n", + "for column in DISCRETE_VARIABLES:\n", + " try:\n", + " df[column] = df[column].astype(np.int64)\n", + " except ValueError:\n", + " mask = df[column].notnull()\n", + " df.loc[mask, column].astype(np.int64)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Raw Data Overview" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "The overall shape of the data is a 2930 rows x 80 columns matrix." + ] + }, + { + "cell_type": "code", + "execution_count": 10, + "metadata": {}, + "outputs": [ + { + "data": { + "text/plain": [ + "(2930, 80)" + ] + }, + "execution_count": 10, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "df.shape" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Continuous Variables" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "The continuous columns are truly continuous in the sense that each column has at least 14 unique value realizations." + ] + }, + { + "cell_type": "code", + "execution_count": 11, + "metadata": {}, + "outputs": [], + "source": [ + "for column in CONTINUOUS_VARIABLES:\n", + " mask = df[column].notnull()\n", + " num_realizations = len(list(x for x in df.loc[mask, column].unique()))\n", + " assert num_realizations > 13" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "A brief description of the variables:" + ] + }, + { + "cell_type": "code", + "execution_count": 12, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "1st Flr SF First Floor square feet\n", + "2nd Flr SF Second floor square feet\n", + "3Ssn Porch Three season porch area in square feet\n", + "Bsmt Unf SF Unfinished square feet of basement area\n", + "BsmtFin SF 1 Type 1 finished square feet\n", + "BsmtFin SF 2 Type 2 finished square feet\n", + "Enclosed Porch Enclosed porch area in square feet\n", + "Garage Area Size of garage in square feet\n", + "Gr Liv Area Above grade (ground) living area square feet\n", + "Lot Area Lot size in square feet\n", + "Lot Frontage Linear feet of street connected to property\n", + "Low Qual Fin SF Low quality finished square feet (all floors)\n", + "Mas Vnr Area Masonry veneer area in square feet\n", + "Misc Val $Value of miscellaneous feature\n", + "Open Porch SF Open porch area in square feet\n", + "Pool Area Pool area in square feet\n", + "Screen Porch Screen porch area in square feet\n", + "Total Bsmt SF Total square feet of basement area\n", + "Wood Deck SF Wood deck area in square feet\n" + ] + } + ], + "source": [ + "print_column_list(CONTINUOUS_COLUMNS)" + ] + }, + { + "cell_type": "code", + "execution_count": 13, + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
1st Flr SF2nd Flr SF3Ssn PorchBsmt Unf SFBsmtFin SF 1BsmtFin SF 2Enclosed PorchGarage AreaGr Liv AreaLot AreaLot FrontageLow Qual Fin SFMas Vnr AreaMisc ValOpen Porch SFPool AreaScreen PorchTotal Bsmt SFWood Deck SF
OrderPID
15263011001656.00.00441.0639.00.00.0528.01656.031770.0141.00.0112.00.062.00.00.01080.0210.0
2526350040896.00.00270.0468.0144.00.0730.0896.011622.080.00.00.00.00.00.0120.0882.0140.0
35263510101329.00.00406.0923.00.00.0312.01329.014267.081.00.0108.012500.036.00.00.01329.0393.0
45263530302110.00.001045.01065.00.00.0522.02110.011160.093.00.00.00.00.00.00.02110.00.0
5527105010928.0701.00137.0791.00.00.0482.01629.013830.074.00.00.00.034.00.00.0928.0212.0
\n", + "
" + ], + "text/plain": [ + " 1st Flr SF 2nd Flr SF 3Ssn Porch Bsmt Unf SF \\\n", + "Order PID \n", + "1 526301100 1656.0 0.0 0 441.0 \n", + "2 526350040 896.0 0.0 0 270.0 \n", + "3 526351010 1329.0 0.0 0 406.0 \n", + "4 526353030 2110.0 0.0 0 1045.0 \n", + "5 527105010 928.0 701.0 0 137.0 \n", + "\n", + " BsmtFin SF 1 BsmtFin SF 2 Enclosed Porch Garage Area \\\n", + "Order PID \n", + "1 526301100 639.0 0.0 0.0 528.0 \n", + "2 526350040 468.0 144.0 0.0 730.0 \n", + "3 526351010 923.0 0.0 0.0 312.0 \n", + "4 526353030 1065.0 0.0 0.0 522.0 \n", + "5 527105010 791.0 0.0 0.0 482.0 \n", + "\n", + " Gr Liv Area Lot Area Lot Frontage Low Qual Fin SF \\\n", + "Order PID \n", + "1 526301100 1656.0 31770.0 141.0 0.0 \n", + "2 526350040 896.0 11622.0 80.0 0.0 \n", + "3 526351010 1329.0 14267.0 81.0 0.0 \n", + "4 526353030 2110.0 11160.0 93.0 0.0 \n", + "5 527105010 1629.0 13830.0 74.0 0.0 \n", + "\n", + " Mas Vnr Area Misc Val Open Porch SF Pool Area \\\n", + "Order PID \n", + "1 526301100 112.0 0.0 62.0 0.0 \n", + "2 526350040 0.0 0.0 0.0 0.0 \n", + "3 526351010 108.0 12500.0 36.0 0.0 \n", + "4 526353030 0.0 0.0 0.0 0.0 \n", + "5 527105010 0.0 0.0 34.0 0.0 \n", + "\n", + " Screen Porch Total Bsmt SF Wood Deck SF \n", + "Order PID \n", + "1 526301100 0.0 1080.0 210.0 \n", + "2 526350040 120.0 882.0 140.0 \n", + "3 526351010 0.0 1329.0 393.0 \n", + "4 526353030 0.0 2110.0 0.0 \n", + "5 527105010 0.0 928.0 212.0 " + ] + }, + "execution_count": 13, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "df[CONTINUOUS_VARIABLES].head()" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Except for the column *Lot Frontage* the columns with missing data only have a couple of missing values (i.e., < 1% of all the rows)." + ] + }, + { + "cell_type": "code", + "execution_count": 14, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "MultiIndex: 2930 entries, (1, 526301100) to (2930, 924151050)\n", + "Data columns (total 19 columns):\n", + " # Column Non-Null Count Dtype \n", + "--- ------ -------------- ----- \n", + " 0 1st Flr SF 2930 non-null float64\n", + " 1 2nd Flr SF 2930 non-null float64\n", + " 2 3Ssn Porch 2930 non-null int64 \n", + " 3 Bsmt Unf SF 2929 non-null float64\n", + " 4 BsmtFin SF 1 2929 non-null float64\n", + " 5 BsmtFin SF 2 2929 non-null float64\n", + " 6 Enclosed Porch 2930 non-null float64\n", + " 7 Garage Area 2929 non-null float64\n", + " 8 Gr Liv Area 2930 non-null float64\n", + " 9 Lot Area 2930 non-null float64\n", + " 10 Lot Frontage 2440 non-null float64\n", + " 11 Low Qual Fin SF 2930 non-null float64\n", + " 12 Mas Vnr Area 2907 non-null float64\n", + " 13 Misc Val 2930 non-null float64\n", + " 14 Open Porch SF 2930 non-null float64\n", + " 15 Pool Area 2930 non-null float64\n", + " 16 Screen Porch 2930 non-null float64\n", + " 17 Total Bsmt SF 2929 non-null float64\n", + " 18 Wood Deck SF 2930 non-null float64\n", + "dtypes: float64(18), int64(1)\n", + "memory usage: 621.3 KB\n" + ] + } + ], + "source": [ + "df[CONTINUOUS_VARIABLES].info()" + ] + }, + { + "cell_type": "code", + "execution_count": 15, + "metadata": {}, + "outputs": [], + "source": [ + "# The columns with a lot of missing\n", + "# values will be treated seperately below.\n", + "missing_a_lot = [\"Lot Frontage\"]" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Discrete Variables" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "The discrete columns have between 2 and 15 unique realizations each if year numbers are excluded from the analysis." + ] + }, + { + "cell_type": "code", + "execution_count": 16, + "metadata": {}, + "outputs": [], + "source": [ + "for column in DISCRETE_VARIABLES:\n", + " mask = df[column].notnull()\n", + " num_realizations = len(list(x for x in df.loc[mask, column].unique()))\n", + " if column not in (\"Year Built\", \"Year Remod/Add\", \"Garage Yr Blt\"):\n", + " assert num_realizations < 15\n", + " assert num_realizations > 2" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "A brief description of the variables:" + ] + }, + { + "cell_type": "code", + "execution_count": 17, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Bedroom AbvGr Bedrooms above grade (does NOT include basement bedrooms)\n", + "Bsmt Full Bath Basement full bathrooms\n", + "Bsmt Half Bath Basement half bathrooms\n", + "Fireplaces Number of fireplaces\n", + "Full Bath Full bathrooms above grade\n", + "Garage Cars Size of garage in car capacity\n", + "Garage Yr Blt Year garage was built\n", + "Half Bath Half baths above grade\n", + "Kitchen AbvGr Kitchens above grade\n", + "Mo Sold Month Sold (MM)\n", + "TotRms AbvGrd Total rooms above grade (does not include bathrooms)\n", + "Year Built Original construction date\n", + "Year Remod/Add Remodel date (same as construction date if no remodeling or additions)\n", + "Yr Sold Year Sold (YYYY)\n" + ] + } + ], + "source": [ + "print_column_list(DISCRETE_COLUMNS)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "**Note:** columns with missing values are implicitly casted to a *float64* type an the *int64* type has no concept of a NaN (=\"Not a number\") value." + ] + }, + { + "cell_type": "code", + "execution_count": 18, + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
Bedroom AbvGrBsmt Full BathBsmt Half BathFireplacesFull BathGarage CarsGarage Yr BltHalf BathKitchen AbvGrMo SoldTotRms AbvGrdYear BuiltYear Remod/AddYr Sold
OrderPID
152630110031.00.0212.01960.00157196019602010
252635004020.00.0011.01961.00165196119612010
352635101030.00.0011.01958.01166195819582010
452635303031.00.0222.01968.01148196819682010
552710501030.00.0122.01997.01136199719982010
\n", + "
" + ], + "text/plain": [ + " Bedroom AbvGr Bsmt Full Bath Bsmt Half Bath Fireplaces \\\n", + "Order PID \n", + "1 526301100 3 1.0 0.0 2 \n", + "2 526350040 2 0.0 0.0 0 \n", + "3 526351010 3 0.0 0.0 0 \n", + "4 526353030 3 1.0 0.0 2 \n", + "5 527105010 3 0.0 0.0 1 \n", + "\n", + " Full Bath Garage Cars Garage Yr Blt Half Bath \\\n", + "Order PID \n", + "1 526301100 1 2.0 1960.0 0 \n", + "2 526350040 1 1.0 1961.0 0 \n", + "3 526351010 1 1.0 1958.0 1 \n", + "4 526353030 2 2.0 1968.0 1 \n", + "5 527105010 2 2.0 1997.0 1 \n", + "\n", + " Kitchen AbvGr Mo Sold TotRms AbvGrd Year Built \\\n", + "Order PID \n", + "1 526301100 1 5 7 1960 \n", + "2 526350040 1 6 5 1961 \n", + "3 526351010 1 6 6 1958 \n", + "4 526353030 1 4 8 1968 \n", + "5 527105010 1 3 6 1997 \n", + "\n", + " Year Remod/Add Yr Sold \n", + "Order PID \n", + "1 526301100 1960 2010 \n", + "2 526350040 1961 2010 \n", + "3 526351010 1958 2010 \n", + "4 526353030 1968 2010 \n", + "5 527105010 1998 2010 " + ] + }, + "execution_count": 18, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "df[DISCRETE_VARIABLES].head()" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Except for the *Garage Yr Blt* column no variable has a significant number of missing values (i.e., > 1% of all rows)." + ] + }, + { + "cell_type": "code", + "execution_count": 19, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "MultiIndex: 2930 entries, (1, 526301100) to (2930, 924151050)\n", + "Data columns (total 14 columns):\n", + " # Column Non-Null Count Dtype \n", + "--- ------ -------------- ----- \n", + " 0 Bedroom AbvGr 2930 non-null int64 \n", + " 1 Bsmt Full Bath 2928 non-null float64\n", + " 2 Bsmt Half Bath 2928 non-null float64\n", + " 3 Fireplaces 2930 non-null int64 \n", + " 4 Full Bath 2930 non-null int64 \n", + " 5 Garage Cars 2929 non-null float64\n", + " 6 Garage Yr Blt 2771 non-null float64\n", + " 7 Half Bath 2930 non-null int64 \n", + " 8 Kitchen AbvGr 2930 non-null int64 \n", + " 9 Mo Sold 2930 non-null int64 \n", + " 10 TotRms AbvGrd 2930 non-null int64 \n", + " 11 Year Built 2930 non-null int64 \n", + " 12 Year Remod/Add 2930 non-null int64 \n", + " 13 Yr Sold 2930 non-null int64 \n", + "dtypes: float64(4), int64(10)\n", + "memory usage: 506.9 KB\n" + ] + } + ], + "source": [ + "df[DISCRETE_VARIABLES].info()" + ] + }, + { + "cell_type": "code", + "execution_count": 20, + "metadata": {}, + "outputs": [], + "source": [ + "missing_a_lot.append(\"Garage Yr Blt\")" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Nominal Variables" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Except for the total of 28 neighborhoods, the nominal columns come with anywhere between 1 and 18 different labels each." + ] + }, + { + "cell_type": "code", + "execution_count": 21, + "metadata": {}, + "outputs": [], + "source": [ + "for column in NOMINAL_VARIABLES:\n", + " mask = df[column].notnull()\n", + " num_realizations = len(list(x for x in df.loc[mask, column].unique()))\n", + " if column not in (\"Neighborhood\"):\n", + " assert num_realizations < 18\n", + " assert num_realizations > 1" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "A brief description of the variables:" + ] + }, + { + "cell_type": "code", + "execution_count": 22, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Alley Type of alley access to property\n", + "Bldg Type Type of dwelling\n", + "Central Air Central air conditioning\n", + "Condition 1 Proximity to various conditions\n", + "Condition 2 Proximity to various conditions (if more than one is present)\n", + "Exterior 1st Exterior covering on house\n", + "Exterior 2nd Exterior covering on house (if more than one material)\n", + "Foundation Type of foundation\n", + "Garage Type Garage location\n", + "Heating Type of heating\n", + "House Style Style of dwelling\n", + "Land Contour Flatness of the property\n", + "Lot Config Lot configuration\n", + "MS SubClass Identifies the type of dwelling involved in the sale.\n", + "MS Zoning Identifies the general zoning classification of the sale.\n", + "Mas Vnr Type Masonry veneer type\n", + "Misc Feature Miscellaneous feature not covered in other categories\n", + "Neighborhood Physical locations within Ames city limits (map available)\n", + "Roof Matl Roof material\n", + "Roof Style Type of roof\n", + "Sale Condition Condition of sale\n", + "Sale Type Type of sale\n", + "Street Type of road access to property\n" + ] + } + ], + "source": [ + "print_column_list(NOMINAL_COLUMNS)" + ] + }, + { + "cell_type": "code", + "execution_count": 23, + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
AlleyBldg TypeCentral AirCondition 1Condition 2Exterior 1stExterior 2ndFoundationGarage TypeHeatingHouse StyleLand ContourLot ConfigMS SubClassMS ZoningMas Vnr TypeMisc FeatureNeighborhoodRoof MatlRoof StyleSale ConditionSale TypeStreet
OrderPID
1526301100NA1FamYNormNormBrkFacePlywoodCBlockAttchdGasA1StoryLvlCorner020RLStoneNANamesCompShgHipNormalWDPave
2526350040NA1FamYFeedrNormVinylSdVinylSdCBlockAttchdGasA1StoryLvlInside020RHNoneNANamesCompShgGableNormalWDPave
3526351010NA1FamYNormNormWd SdngWd SdngCBlockAttchdGasA1StoryLvlCorner020RLBrkFaceGar2NamesCompShgHipNormalWDPave
4526353030NA1FamYNormNormBrkFaceBrkFaceCBlockAttchdGasA1StoryLvlCorner020RLNoneNANamesCompShgHipNormalWDPave
5527105010NA1FamYNormNormVinylSdVinylSdPConcAttchdGasA2StoryLvlInside060RLNoneNAGilbertCompShgGableNormalWDPave
6527105030NA1FamYNormNormVinylSdVinylSdPConcAttchdGasA2StoryLvlInside060RLBrkFaceNAGilbertCompShgGableNormalWDPave
7527127150NATwnhsEYNormNormCemntBdCemntBdPConcAttchdGasA1StoryLvlInside120RLNoneNAStoneBrCompShgGableNormalWDPave
8527145080NATwnhsEYNormNormHdBoardHdBoardPConcAttchdGasA1StoryHLSInside120RLNoneNAStoneBrCompShgGableNormalWDPave
9527146030NATwnhsEYNormNormCemntBdCemntBdPConcAttchdGasA1StoryLvlInside120RLNoneNAStoneBrCompShgGableNormalWDPave
10527162130NA1FamYNormNormVinylSdVinylSdPConcAttchdGasA2StoryLvlInside060RLNoneNAGilbertCompShgGableNormalWDPave
\n", + "
" + ], + "text/plain": [ + " Alley Bldg Type Central Air Condition 1 Condition 2 \\\n", + "Order PID \n", + "1 526301100 NA 1Fam Y Norm Norm \n", + "2 526350040 NA 1Fam Y Feedr Norm \n", + "3 526351010 NA 1Fam Y Norm Norm \n", + "4 526353030 NA 1Fam Y Norm Norm \n", + "5 527105010 NA 1Fam Y Norm Norm \n", + "6 527105030 NA 1Fam Y Norm Norm \n", + "7 527127150 NA TwnhsE Y Norm Norm \n", + "8 527145080 NA TwnhsE Y Norm Norm \n", + "9 527146030 NA TwnhsE Y Norm Norm \n", + "10 527162130 NA 1Fam Y Norm Norm \n", + "\n", + " Exterior 1st Exterior 2nd Foundation Garage Type Heating \\\n", + "Order PID \n", + "1 526301100 BrkFace Plywood CBlock Attchd GasA \n", + "2 526350040 VinylSd VinylSd CBlock Attchd GasA \n", + "3 526351010 Wd Sdng Wd Sdng CBlock Attchd GasA \n", + "4 526353030 BrkFace BrkFace CBlock Attchd GasA \n", + "5 527105010 VinylSd VinylSd PConc Attchd GasA \n", + "6 527105030 VinylSd VinylSd PConc Attchd GasA \n", + "7 527127150 CemntBd CemntBd PConc Attchd GasA \n", + "8 527145080 HdBoard HdBoard PConc Attchd GasA \n", + "9 527146030 CemntBd CemntBd PConc Attchd GasA \n", + "10 527162130 VinylSd VinylSd PConc Attchd GasA \n", + "\n", + " House Style Land Contour Lot Config MS SubClass MS Zoning \\\n", + "Order PID \n", + "1 526301100 1Story Lvl Corner 020 RL \n", + "2 526350040 1Story Lvl Inside 020 RH \n", + "3 526351010 1Story Lvl Corner 020 RL \n", + "4 526353030 1Story Lvl Corner 020 RL \n", + "5 527105010 2Story Lvl Inside 060 RL \n", + "6 527105030 2Story Lvl Inside 060 RL \n", + "7 527127150 1Story Lvl Inside 120 RL \n", + "8 527145080 1Story HLS Inside 120 RL \n", + "9 527146030 1Story Lvl Inside 120 RL \n", + "10 527162130 2Story Lvl Inside 060 RL \n", + "\n", + " Mas Vnr Type Misc Feature Neighborhood Roof Matl Roof Style \\\n", + "Order PID \n", + "1 526301100 Stone NA Names CompShg Hip \n", + "2 526350040 None NA Names CompShg Gable \n", + "3 526351010 BrkFace Gar2 Names CompShg Hip \n", + "4 526353030 None NA Names CompShg Hip \n", + "5 527105010 None NA Gilbert CompShg Gable \n", + "6 527105030 BrkFace NA Gilbert CompShg Gable \n", + "7 527127150 None NA StoneBr CompShg Gable \n", + "8 527145080 None NA StoneBr CompShg Gable \n", + "9 527146030 None NA StoneBr CompShg Gable \n", + "10 527162130 None NA Gilbert CompShg Gable \n", + "\n", + " Sale Condition Sale Type Street \n", + "Order PID \n", + "1 526301100 Normal WD Pave \n", + "2 526350040 Normal WD Pave \n", + "3 526351010 Normal WD Pave \n", + "4 526353030 Normal WD Pave \n", + "5 527105010 Normal WD Pave \n", + "6 527105030 Normal WD Pave \n", + "7 527127150 Normal WD Pave \n", + "8 527145080 Normal WD Pave \n", + "9 527146030 Normal WD Pave \n", + "10 527162130 Normal WD Pave " + ] + }, + "execution_count": 23, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "df[NOMINAL_VARIABLES].head(10)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "For the nominal variables there is only a neglectable number of missing values." + ] + }, + { + "cell_type": "code", + "execution_count": 24, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "MultiIndex: 2930 entries, (1, 526301100) to (2930, 924151050)\n", + "Data columns (total 23 columns):\n", + " # Column Non-Null Count Dtype \n", + "--- ------ -------------- ----- \n", + " 0 Alley 2930 non-null object\n", + " 1 Bldg Type 2930 non-null object\n", + " 2 Central Air 2930 non-null object\n", + " 3 Condition 1 2930 non-null object\n", + " 4 Condition 2 2930 non-null object\n", + " 5 Exterior 1st 2930 non-null object\n", + " 6 Exterior 2nd 2930 non-null object\n", + " 7 Foundation 2930 non-null object\n", + " 8 Garage Type 2930 non-null object\n", + " 9 Heating 2930 non-null object\n", + " 10 House Style 2930 non-null object\n", + " 11 Land Contour 2930 non-null object\n", + " 12 Lot Config 2930 non-null object\n", + " 13 MS SubClass 2930 non-null object\n", + " 14 MS Zoning 2930 non-null object\n", + " 15 Mas Vnr Type 2907 non-null object\n", + " 16 Misc Feature 2930 non-null object\n", + " 17 Neighborhood 2930 non-null object\n", + " 18 Roof Matl 2930 non-null object\n", + " 19 Roof Style 2930 non-null object\n", + " 20 Sale Condition 2930 non-null object\n", + " 21 Sale Type 2930 non-null object\n", + " 22 Street 2930 non-null object\n", + "dtypes: object(23)\n", + "memory usage: 712.9+ KB\n" + ] + } + ], + "source": [ + "df[NOMINAL_VARIABLES].info()" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Ordinal Variables" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "The ordinal columns come with anywhere between 2 and 11 distinct labels each." + ] + }, + { + "cell_type": "code", + "execution_count": 25, + "metadata": {}, + "outputs": [], + "source": [ + "for column in ORDINAL_VARIABLES:\n", + " mask = df[column].notnull()\n", + " num_realizations = len(list(x for x in df.loc[mask, column].unique()))\n", + " assert 2 < num_realizations < 11" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "A brief description of the variables:" + ] + }, + { + "cell_type": "code", + "execution_count": 26, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Bsmt Cond Evaluates the general condition of the basement\n", + "Bsmt Exposure Refers to walkout or garden level walls\n", + "Bsmt Qual Evaluates the height of the basement\n", + "BsmtFin Type 1 Rating of basement finished area\n", + "BsmtFin Type 2 Rating of basement finished area (if multiple types)\n", + "Electrical Electrical system\n", + "Exter Cond Evaluates the present condition of the material on the exterior\n", + "Exter Qual Evaluates the quality of the material on the exterior\n", + "Fence Fence quality\n", + "Fireplace Qu Fireplace quality\n", + "Functional Home functionality (Assume typical unless deductions are warranted)\n", + "Garage Cond Garage condition\n", + "Garage Finish Interior finish of the garage\n", + "Garage Qual Garage quality\n", + "Heating QC Heating quality and condition\n", + "Kitchen Qual Kitchen quality\n", + "Land Slope Slope of property\n", + "Lot Shape General shape of property\n", + "Overall Cond Rates the overall condition of the house\n", + "Overall Qual Rates the overall material and finish of the house\n", + "Paved Drive Paved driveway\n", + "Pool QC Pool quality\n", + "Utilities Type of utilities available\n" + ] + } + ], + "source": [ + "print_column_list(ORDINAL_COLUMNS)" + ] + }, + { + "cell_type": "code", + "execution_count": 27, + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
Bsmt CondBsmt ExposureBsmt QualBsmtFin Type 1BsmtFin Type 2ElectricalExter CondExter QualFenceFireplace QuFunctionalGarage CondGarage FinishGarage QualHeating QCKitchen QualLand SlopeLot ShapeOverall CondOverall QualPaved DrivePool QCUtilities
OrderPID
1526301100GdGdTABLQUnfSBrkrTATANAGdTypTAFinTAFaTAGtlIR156PNAAllPub
2526350040TANoTARecLwQSBrkrTATAMnPrvNATypTAUnfTATATAGtlReg65YNAAllPub
3526351010TANoTAALQUnfSBrkrTATANANATypTAUnfTATAGdGtlIR166YNAAllPub
4526353030TANoTAALQUnfSBrkrTAGdNATATypTAFinTAExExGtlReg57YNAAllPub
5527105010TANoGdGLQUnfSBrkrTATAMnPrvTATypTAFinTAGdTAGtlIR155YNAAllPub
6527105030TANoTAGLQUnfSBrkrTATANAGdTypTAFinTAExGdGtlIR166YNAAllPub
7527127150TAMnGdGLQUnfSBrkrTAGdNANATypTAFinTAExGdGtlReg58YNAAllPub
8527145080TANoGdALQUnfSBrkrTAGdNANATypTARFnTAExGdGtlIR158YNAAllPub
9527146030TANoGdGLQUnfSBrkrTAGdNATATypTARFnTAExGdGtlIR158YNAAllPub
10527162130TANoTAUnfUnfSBrkrTATANATATypTAFinTAGdGdGtlReg57YNAAllPub
\n", + "
" + ], + "text/plain": [ + " Bsmt Cond Bsmt Exposure Bsmt Qual BsmtFin Type 1 \\\n", + "Order PID \n", + "1 526301100 Gd Gd TA BLQ \n", + "2 526350040 TA No TA Rec \n", + "3 526351010 TA No TA ALQ \n", + "4 526353030 TA No TA ALQ \n", + "5 527105010 TA No Gd GLQ \n", + "6 527105030 TA No TA GLQ \n", + "7 527127150 TA Mn Gd GLQ \n", + "8 527145080 TA No Gd ALQ \n", + "9 527146030 TA No Gd GLQ \n", + "10 527162130 TA No TA Unf \n", + "\n", + " BsmtFin Type 2 Electrical Exter Cond Exter Qual Fence \\\n", + "Order PID \n", + "1 526301100 Unf SBrkr TA TA NA \n", + "2 526350040 LwQ SBrkr TA TA MnPrv \n", + "3 526351010 Unf SBrkr TA TA NA \n", + "4 526353030 Unf SBrkr TA Gd NA \n", + "5 527105010 Unf SBrkr TA TA MnPrv \n", + "6 527105030 Unf SBrkr TA TA NA \n", + "7 527127150 Unf SBrkr TA Gd NA \n", + "8 527145080 Unf SBrkr TA Gd NA \n", + "9 527146030 Unf SBrkr TA Gd NA \n", + "10 527162130 Unf SBrkr TA TA NA \n", + "\n", + " Fireplace Qu Functional Garage Cond Garage Finish Garage Qual \\\n", + "Order PID \n", + "1 526301100 Gd Typ TA Fin TA \n", + "2 526350040 NA Typ TA Unf TA \n", + "3 526351010 NA Typ TA Unf TA \n", + "4 526353030 TA Typ TA Fin TA \n", + "5 527105010 TA Typ TA Fin TA \n", + "6 527105030 Gd Typ TA Fin TA \n", + "7 527127150 NA Typ TA Fin TA \n", + "8 527145080 NA Typ TA RFn TA \n", + "9 527146030 TA Typ TA RFn TA \n", + "10 527162130 TA Typ TA Fin TA \n", + "\n", + " Heating QC Kitchen Qual Land Slope Lot Shape Overall Cond \\\n", + "Order PID \n", + "1 526301100 Fa TA Gtl IR1 5 \n", + "2 526350040 TA TA Gtl Reg 6 \n", + "3 526351010 TA Gd Gtl IR1 6 \n", + "4 526353030 Ex Ex Gtl Reg 5 \n", + "5 527105010 Gd TA Gtl IR1 5 \n", + "6 527105030 Ex Gd Gtl IR1 6 \n", + "7 527127150 Ex Gd Gtl Reg 5 \n", + "8 527145080 Ex Gd Gtl IR1 5 \n", + "9 527146030 Ex Gd Gtl IR1 5 \n", + "10 527162130 Gd Gd Gtl Reg 5 \n", + "\n", + " Overall Qual Paved Drive Pool QC Utilities \n", + "Order PID \n", + "1 526301100 6 P NA AllPub \n", + "2 526350040 5 Y NA AllPub \n", + "3 526351010 6 Y NA AllPub \n", + "4 526353030 7 Y NA AllPub \n", + "5 527105010 5 Y NA AllPub \n", + "6 527105030 6 Y NA AllPub \n", + "7 527127150 8 Y NA AllPub \n", + "8 527145080 8 Y NA AllPub \n", + "9 527146030 8 Y NA AllPub \n", + "10 527162130 7 Y NA AllPub " + ] + }, + "execution_count": 27, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "df[ORDINAL_VARIABLES].head(10)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "For the ordinal variables there is only a neglectable number of missing values." + ] + }, + { + "cell_type": "code", + "execution_count": 28, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "MultiIndex: 2930 entries, (1, 526301100) to (2930, 924151050)\n", + "Data columns (total 23 columns):\n", + " # Column Non-Null Count Dtype \n", + "--- ------ -------------- ----- \n", + " 0 Bsmt Cond 2929 non-null object\n", + " 1 Bsmt Exposure 2926 non-null object\n", + " 2 Bsmt Qual 2929 non-null object\n", + " 3 BsmtFin Type 1 2929 non-null object\n", + " 4 BsmtFin Type 2 2928 non-null object\n", + " 5 Electrical 2929 non-null object\n", + " 6 Exter Cond 2930 non-null object\n", + " 7 Exter Qual 2930 non-null object\n", + " 8 Fence 2930 non-null object\n", + " 9 Fireplace Qu 2930 non-null object\n", + " 10 Functional 2930 non-null object\n", + " 11 Garage Cond 2929 non-null object\n", + " 12 Garage Finish 2928 non-null object\n", + " 13 Garage Qual 2929 non-null object\n", + " 14 Heating QC 2930 non-null object\n", + " 15 Kitchen Qual 2930 non-null object\n", + " 16 Land Slope 2930 non-null object\n", + " 17 Lot Shape 2930 non-null object\n", + " 18 Overall Cond 2930 non-null object\n", + " 19 Overall Qual 2930 non-null object\n", + " 20 Paved Drive 2930 non-null object\n", + " 21 Pool QC 2930 non-null object\n", + " 22 Utilities 2930 non-null object\n", + "dtypes: object(23)\n", + "memory usage: 712.9+ KB\n" + ] + } + ], + "source": [ + "df[ORDINAL_VARIABLES].info()" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Missing Data" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Visualizations" + ] + }, + { + "cell_type": "code", + "execution_count": 29, + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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\n", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], + "source": [ + "msno.matrix(df[CONTINUOUS_VARIABLES]);" + ] + }, + { + "cell_type": "code", + "execution_count": 30, + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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\n", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], + "source": [ + "msno.matrix(df[DISCRETE_VARIABLES]);" + ] + }, + { + "cell_type": "code", + "execution_count": 31, + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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\n", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], + "source": [ + "msno.matrix(df[NOMINAL_VARIABLES]);" + ] + }, + { + "cell_type": "code", + "execution_count": 32, + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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\n", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], + "source": [ + "msno.matrix(df[ORDINAL_VARIABLES]);" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### Cleansing" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "Since only about 1% of the overall number of observations exhibit variables with missing data (disregarding the columns *Lot Frontage* and *Garage Yr Blt*), the decision is made to discard these rows entirely to not have to deal with interpolating meaningful replacements for the missing values." + ] + }, + { + "cell_type": "code", + "execution_count": 33, + "metadata": {}, + "outputs": [], + "source": [ + "remaining_columns = sorted(set(ALL_VARIABLES) - set(missing_a_lot)) + TARGET_VARIABLES\n", + "mask = df[remaining_columns].isnull().any(axis=1)\n", + "assert (100 * mask.sum() / df.shape[0]) < 1.1 # percent\n", + "df = df[~mask]" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "The two columns with a lot of missing values regard the age of a house's optional garage and the length of the intersection with the street where the house is located. The first is assumed as not important for the house appraisal and the second is assumed to be captured in other variables (e.g. overall size of the house). Therefore, for sake of simplicity both columns are dropped from the DataFrame." + ] + }, + { + "cell_type": "code", + "execution_count": 34, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Garage Yr Blt Year garage was built\n", + "Lot Frontage Linear feet of street connected to property\n" + ] + } + ], + "source": [ + "print_column_list(missing_a_lot)" + ] + }, + { + "cell_type": "code", + "execution_count": 35, + "metadata": {}, + "outputs": [], + "source": [ + "df = df[remaining_columns]" + ] + }, + { + "cell_type": "code", + "execution_count": 36, + "metadata": {}, + "outputs": [], + "source": [ + "# Remove the discarded columns from the helper dictionaries / lists.\n", + "update_column_descriptions(df.columns)\n", + "# Without any more missing data, cast all numeric\n", + "# columns as floats or integers respectively.\n", + "for column in CONTINUOUS_VARIABLES + TARGET_VARIABLES:\n", + " df[column] = df[column].astype(np.float64)\n", + "for column in DISCRETE_VARIABLES:\n", + " df[column] = df[column].astype(np.int64)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Clean Data" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "The cleaned data comes as a 2898 rows x 78 columns matrix." + ] + }, + { + "cell_type": "code", + "execution_count": 37, + "metadata": {}, + "outputs": [ + { + "data": { + "text/plain": [ + "(2898, 78)" + ] + }, + "execution_count": 37, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "df.shape" + ] + }, + { + "cell_type": "code", + "execution_count": 38, + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
1st Flr SF2nd Flr SF3Ssn PorchAlleyBedroom AbvGrBldg TypeBsmt CondBsmt ExposureBsmt Full BathBsmt Half BathBsmt QualBsmt Unf SFBsmtFin SF 1BsmtFin SF 2BsmtFin Type 1BsmtFin Type 2Central AirCondition 1Condition 2ElectricalEnclosed PorchExter CondExter QualExterior 1stExterior 2ndFenceFireplace QuFireplacesFoundationFull BathFunctionalGarage AreaGarage CarsGarage CondGarage FinishGarage QualGarage TypeGr Liv AreaHalf BathHeatingHeating QCHouse StyleKitchen AbvGrKitchen QualLand ContourLand SlopeLot AreaLot ConfigLot ShapeLow Qual Fin SFMS SubClassMS ZoningMas Vnr AreaMas Vnr TypeMisc FeatureMisc ValMo SoldNeighborhoodOpen Porch SFOverall CondOverall QualPaved DrivePool AreaPool QCRoof MatlRoof StyleSale ConditionSale TypeScreen PorchStreetTotRms AbvGrdTotal Bsmt SFUtilitiesWood Deck SFYear BuiltYear Remod/AddYr SoldSalePrice
OrderPID
15263011001656.00.00.0NA31FamGdGd10TA441.0639.00.0BLQUnfYNormNormSBrkr0.0TATABrkFacePlywoodNAGd2CBlock1Typ528.02TAFinTAAttchd1656.00GasAFa1Story1TALvlGtl31770.0CornerIR10.0020RL112.0StoneNA0.05Names62.056P0.0NACompShgHipNormalWD0.0Pave71080.0AllPub210.0196019602010215000.0
2526350040896.00.00.0NA21FamTANo00TA270.0468.0144.0RecLwQYFeedrNormSBrkr0.0TATAVinylSdVinylSdMnPrvNA0CBlock1Typ730.01TAUnfTAAttchd896.00GasATA1Story1TALvlGtl11622.0InsideReg0.0020RH0.0NoneNA0.06Names0.065Y0.0NACompShgGableNormalWD120.0Pave5882.0AllPub140.0196119612010105000.0
35263510101329.00.00.0NA31FamTANo00TA406.0923.00.0ALQUnfYNormNormSBrkr0.0TATAWd SdngWd SdngNANA0CBlock1Typ312.01TAUnfTAAttchd1329.01GasATA1Story1GdLvlGtl14267.0CornerIR10.0020RL108.0BrkFaceGar212500.06Names36.066Y0.0NACompShgHipNormalWD0.0Pave61329.0AllPub393.0195819582010172000.0
45263530302110.00.00.0NA31FamTANo10TA1045.01065.00.0ALQUnfYNormNormSBrkr0.0TAGdBrkFaceBrkFaceNATA2CBlock2Typ522.02TAFinTAAttchd2110.01GasAEx1Story1ExLvlGtl11160.0CornerReg0.0020RL0.0NoneNA0.04Names0.057Y0.0NACompShgHipNormalWD0.0Pave82110.0AllPub0.0196819682010244000.0
5527105010928.0701.00.0NA31FamTANo00Gd137.0791.00.0GLQUnfYNormNormSBrkr0.0TATAVinylSdVinylSdMnPrvTA1PConc2Typ482.02TAFinTAAttchd1629.01GasAGd2Story1TALvlGtl13830.0InsideIR10.0060RL0.0NoneNA0.03Gilbert34.055Y0.0NACompShgGableNormalWD0.0Pave6928.0AllPub212.0199719982010189900.0
\n", + "
" + ], + "text/plain": [ + " 1st Flr SF 2nd Flr SF 3Ssn Porch Alley Bedroom AbvGr \\\n", + "Order PID \n", + "1 526301100 1656.0 0.0 0.0 NA 3 \n", + "2 526350040 896.0 0.0 0.0 NA 2 \n", + "3 526351010 1329.0 0.0 0.0 NA 3 \n", + "4 526353030 2110.0 0.0 0.0 NA 3 \n", + "5 527105010 928.0 701.0 0.0 NA 3 \n", + "\n", + " Bldg Type Bsmt Cond Bsmt Exposure Bsmt Full Bath \\\n", + "Order PID \n", + "1 526301100 1Fam Gd Gd 1 \n", + "2 526350040 1Fam TA No 0 \n", + "3 526351010 1Fam TA No 0 \n", + "4 526353030 1Fam TA No 1 \n", + "5 527105010 1Fam TA No 0 \n", + "\n", + " Bsmt Half Bath Bsmt Qual Bsmt Unf SF BsmtFin SF 1 \\\n", + "Order PID \n", + "1 526301100 0 TA 441.0 639.0 \n", + "2 526350040 0 TA 270.0 468.0 \n", + "3 526351010 0 TA 406.0 923.0 \n", + "4 526353030 0 TA 1045.0 1065.0 \n", + "5 527105010 0 Gd 137.0 791.0 \n", + "\n", + " BsmtFin SF 2 BsmtFin Type 1 BsmtFin Type 2 Central Air \\\n", + "Order PID \n", + "1 526301100 0.0 BLQ Unf Y \n", + "2 526350040 144.0 Rec LwQ Y \n", + "3 526351010 0.0 ALQ Unf Y \n", + "4 526353030 0.0 ALQ Unf Y \n", + "5 527105010 0.0 GLQ Unf Y \n", + "\n", + " Condition 1 Condition 2 Electrical Enclosed Porch Exter Cond \\\n", + "Order PID \n", + "1 526301100 Norm Norm SBrkr 0.0 TA \n", + "2 526350040 Feedr Norm SBrkr 0.0 TA \n", + "3 526351010 Norm Norm SBrkr 0.0 TA \n", + "4 526353030 Norm Norm SBrkr 0.0 TA \n", + "5 527105010 Norm Norm SBrkr 0.0 TA \n", + "\n", + " Exter Qual Exterior 1st Exterior 2nd Fence Fireplace Qu \\\n", + "Order PID \n", + "1 526301100 TA BrkFace Plywood NA Gd \n", + "2 526350040 TA VinylSd VinylSd MnPrv NA \n", + "3 526351010 TA Wd Sdng Wd Sdng NA NA \n", + "4 526353030 Gd BrkFace BrkFace NA TA \n", + "5 527105010 TA VinylSd VinylSd MnPrv TA \n", + "\n", + " Fireplaces Foundation Full Bath Functional Garage Area \\\n", + "Order PID \n", + "1 526301100 2 CBlock 1 Typ 528.0 \n", + "2 526350040 0 CBlock 1 Typ 730.0 \n", + "3 526351010 0 CBlock 1 Typ 312.0 \n", + "4 526353030 2 CBlock 2 Typ 522.0 \n", + "5 527105010 1 PConc 2 Typ 482.0 \n", + "\n", + " Garage Cars Garage Cond Garage Finish Garage Qual \\\n", + "Order PID \n", + "1 526301100 2 TA Fin TA \n", + "2 526350040 1 TA Unf TA \n", + "3 526351010 1 TA Unf TA \n", + "4 526353030 2 TA Fin TA \n", + "5 527105010 2 TA Fin TA \n", + "\n", + " Garage Type Gr Liv Area Half Bath Heating Heating QC \\\n", + "Order PID \n", + "1 526301100 Attchd 1656.0 0 GasA Fa \n", + "2 526350040 Attchd 896.0 0 GasA TA \n", + "3 526351010 Attchd 1329.0 1 GasA TA \n", + "4 526353030 Attchd 2110.0 1 GasA Ex \n", + "5 527105010 Attchd 1629.0 1 GasA Gd \n", + "\n", + " House Style Kitchen AbvGr Kitchen Qual Land Contour \\\n", + "Order PID \n", + "1 526301100 1Story 1 TA Lvl \n", + "2 526350040 1Story 1 TA Lvl \n", + "3 526351010 1Story 1 Gd Lvl \n", + "4 526353030 1Story 1 Ex Lvl \n", + "5 527105010 2Story 1 TA Lvl \n", + "\n", + " Land Slope Lot Area Lot Config Lot Shape Low Qual Fin SF \\\n", + "Order PID \n", + "1 526301100 Gtl 31770.0 Corner IR1 0.0 \n", + "2 526350040 Gtl 11622.0 Inside Reg 0.0 \n", + "3 526351010 Gtl 14267.0 Corner IR1 0.0 \n", + "4 526353030 Gtl 11160.0 Corner Reg 0.0 \n", + "5 527105010 Gtl 13830.0 Inside IR1 0.0 \n", + "\n", + " MS SubClass MS Zoning Mas Vnr Area Mas Vnr Type Misc Feature \\\n", + "Order PID \n", + "1 526301100 020 RL 112.0 Stone NA \n", + "2 526350040 020 RH 0.0 None NA \n", + "3 526351010 020 RL 108.0 BrkFace Gar2 \n", + "4 526353030 020 RL 0.0 None NA \n", + "5 527105010 060 RL 0.0 None NA \n", + "\n", + " Misc Val Mo Sold Neighborhood Open Porch SF Overall Cond \\\n", + "Order PID \n", + "1 526301100 0.0 5 Names 62.0 5 \n", + "2 526350040 0.0 6 Names 0.0 6 \n", + "3 526351010 12500.0 6 Names 36.0 6 \n", + "4 526353030 0.0 4 Names 0.0 5 \n", + "5 527105010 0.0 3 Gilbert 34.0 5 \n", + "\n", + " Overall Qual Paved Drive Pool Area Pool QC Roof Matl \\\n", + "Order PID \n", + "1 526301100 6 P 0.0 NA CompShg \n", + "2 526350040 5 Y 0.0 NA CompShg \n", + "3 526351010 6 Y 0.0 NA CompShg \n", + "4 526353030 7 Y 0.0 NA CompShg \n", + "5 527105010 5 Y 0.0 NA CompShg \n", + "\n", + " Roof Style Sale Condition Sale Type Screen Porch Street \\\n", + "Order PID \n", + "1 526301100 Hip Normal WD 0.0 Pave \n", + "2 526350040 Gable Normal WD 120.0 Pave \n", + "3 526351010 Hip Normal WD 0.0 Pave \n", + "4 526353030 Hip Normal WD 0.0 Pave \n", + "5 527105010 Gable Normal WD 0.0 Pave \n", + "\n", + " TotRms AbvGrd Total Bsmt SF Utilities Wood Deck SF \\\n", + "Order PID \n", + "1 526301100 7 1080.0 AllPub 210.0 \n", + "2 526350040 5 882.0 AllPub 140.0 \n", + "3 526351010 6 1329.0 AllPub 393.0 \n", + "4 526353030 8 2110.0 AllPub 0.0 \n", + "5 527105010 6 928.0 AllPub 212.0 \n", + "\n", + " Year Built Year Remod/Add Yr Sold SalePrice \n", + "Order PID \n", + "1 526301100 1960 1960 2010 215000.0 \n", + "2 526350040 1961 1961 2010 105000.0 \n", + "3 526351010 1958 1958 2010 172000.0 \n", + "4 526353030 1968 1968 2010 244000.0 \n", + "5 527105010 1997 1998 2010 189900.0 " + ] + }, + "execution_count": 38, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "df.head()" + ] + }, + { + "cell_type": "code", + "execution_count": 39, + "metadata": {}, + "outputs": [], + "source": [ + "df.to_csv(\"data/data_clean.csv\")" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.8.5" + } + }, + "nbformat": 4, + "nbformat_minor": 4 +} diff --git a/1_data_cleaning.ipynb b/1_data_cleaning.ipynb deleted file mode 100644 index 3ea742d..0000000 --- a/1_data_cleaning.ipynb +++ /dev/null @@ -1,4353 +0,0 @@ -{ - "cells": [ - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# Data Cleaning" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## \"Housekeeping\"" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Import all the third-party (scientific) libraries needed." - ] - }, - { - "cell_type": "code", - "execution_count": 1, - "metadata": {}, - "outputs": [], - "source": [ - "import missingno as msno\n", - "import numpy as np\n", - "import pandas as pd" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "The *utils.py* module defines helper dictionaries, lists, and functions that help with parsing the data types correctly, look up column descriptions, and refer to groups of data columns.\n", - "\n", - "**Note:** the suffix \\_*COLUMNS* indicates a dictionary with all meta information on the provided data file and \\_*VARIABLES* a list with only the column names (i.e., the keys of the respective \\_*COLUMNS* dictionary)." - ] - }, - { - "cell_type": "code", - "execution_count": 2, - "metadata": {}, - "outputs": [], - "source": [ - "from utils import (\n", - " ALL_COLUMNS,\n", - " ALL_VARIABLES,\n", - " CONTINUOUS_COLUMNS,\n", - " CONTINUOUS_VARIABLES,\n", - " DISCRETE_COLUMNS,\n", - " DISCRETE_VARIABLES,\n", - " INDEX_COLUMNS,\n", - " LABEL_COLUMNS, # groups nominal and ordinal\n", - " LABEL_TYPES,\n", - " NOMINAL_COLUMNS,\n", - " NOMINAL_VARIABLES,\n", - " NUMERIC_VARIABLES, # groups continuous and discrete\n", - " ORDINAL_COLUMNS,\n", - " ORDINAL_VARIABLES,\n", - " TARGET_VARIABLES, # = Sale Price\n", - " correct_column_names,\n", - " print_column_list,\n", - " update_column_descriptions,\n", - ")" - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "metadata": {}, - "outputs": [], - "source": [ - "# Show all data columns.\n", - "pd.set_option(\"display.max_columns\", 100)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Load the Data" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "The original data are available for [download](https://www.amstat.org/publications/jse/v19n3/decock/AmesHousing.xls) and a detailed description of the data types for each column can be found [here](https://www.amstat.org/publications/jse/v19n3/decock/DataDocumentation.txt). These meta data go into the `dtype` argument of the `read_excel` function below to parse the data correctly. There are four different generic data types defined that are casted as follows:\n", - "\n", - "- continuous -> np.float64\n", - "- discrete -> actually np.int64 but np.float64 because of missing values\n", - "- nominal -> object (str)\n", - "- ordinal -> object (str), the order can be looked up in the above mentioned *ALL_COLUMNS* dictionary\n", - "\n", - "**Note 1:** the data come with a lot of \"NA\" text strings that do **not** indicate missing data but, for example, the absence of a basement or a parking lot (see the linked data description).\n", - "\n", - "**Note 2:** the mappings from column names to data types are encoded in the \"utils.py\" module that defines the aforementioned helper dictionaries / lists.\n", - "\n", - "**Note 3:** the Excel file with all the data is either loaded from the local dictionary (= \"cache\") or obtained fresh from the source." - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "metadata": {}, - "outputs": [], - "source": [ - "# To avoid redundancy.\n", - "kwargs = {\n", - " \"dtype\": { # Ensure each column is parsed as the correct data type.\n", - " column: ( # This creates a mapping from column name to data type.\n", - " object if mapping_info[\"type\"] in LABEL_TYPES else np.float64\n", - " )\n", - " for (column, mapping_info) in ALL_COLUMNS.items()\n", - " },\n", - " \"na_values\": \"\", # By default, pandas treats NA strings as missing,\n", - " \"keep_default_na\": False, # which is not the correct meaning here.\n", - "}\n", - "\n", - "try:\n", - " df = pd.read_excel(\"data/data_raw.xls\", **kwargs)\n", - "except FileNotFoundError:\n", - " df = pd.read_excel(\n", - " \"https://www.amstat.org/publications/jse/v19n3/decock/AmesHousing.xls\", **kwargs\n", - " )\n", - " # Cache the obtained file.\n", - " df.to_excel(\"data/data_raw.xls\")" - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "metadata": {}, - "outputs": [], - "source": [ - "# Some columns names differ between the Excel file and\n", - "# the data description file. Correct that with the values\n", - "# in the Excel file.\n", - "correct_column_names(df.columns)\n", - "# Use a compound index and keep both\n", - "# identifying columns in the DataFrame.\n", - "df = df.set_index(INDEX_COLUMNS)\n", - "# Put the provided columns into the same\n", - "# order as in the encoded description file.\n", - "# Note that the target variable \"SalePrice\"\n", - "# is not in the description file.\n", - "df = df[ALL_VARIABLES + TARGET_VARIABLES]" - ] - }, - { - "cell_type": "code", - "execution_count": 6, - "metadata": {}, - "outputs": [ - { - "data": { - "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
1st Flr SF2nd Flr SF3Ssn PorchAlleyBedroom AbvGrBldg TypeBsmt CondBsmt ExposureBsmt Full BathBsmt Half BathBsmt QualBsmt Unf SFBsmtFin SF 1BsmtFin SF 2BsmtFin Type 1BsmtFin Type 2Central AirCondition 1Condition 2ElectricalEnclosed PorchExter CondExter QualExterior 1stExterior 2ndFenceFireplace QuFireplacesFoundationFull BathFunctionalGarage AreaGarage CarsGarage CondGarage FinishGarage QualGarage TypeGarage Yr BltGr Liv AreaHalf BathHeatingHeating QCHouse StyleKitchen AbvGrKitchen QualLand ContourLand SlopeLot AreaLot ConfigLot FrontageLot ShapeLow Qual Fin SFMS SubClassMS ZoningMas Vnr AreaMas Vnr TypeMisc FeatureMisc ValMo SoldNeighborhoodOpen Porch SFOverall CondOverall QualPaved DrivePool AreaPool QCRoof MatlRoof StyleSale ConditionSale TypeScreen PorchStreetTotRms AbvGrdTotal Bsmt SFUtilitiesWood Deck SFYear BuiltYear Remod/AddYr SoldSalePrice
OrderPID
15263011001656.00.00NA31FamGdGd1.00.0TA441.0639.00.0BLQUnfYNormNormSBrkr0.0TATABrkFacePlywoodNAGd2.0CBlock1.0Typ528.02.0TAFinTAAttchd1960.01656.00.0GasAFa1Story1TALvlGtl31770.0Corner141.0IR10.0020RL112.0StoneNA0.05.0NAmes62.056P0.0NACompShgHipNormalWD0.0Pave71080.0AllPub210.01960.01960.02010.0215000
2526350040896.00.00NA21FamTANo0.00.0TA270.0468.0144.0RecLwQYFeedrNormSBrkr0.0TATAVinylSdVinylSdMnPrvNA0.0CBlock1.0Typ730.01.0TAUnfTAAttchd1961.0896.00.0GasATA1Story1TALvlGtl11622.0Inside80.0Reg0.0020RH0.0NoneNA0.06.0NAmes0.065Y0.0NACompShgGableNormalWD120.0Pave5882.0AllPub140.01961.01961.02010.0105000
35263510101329.00.00NA31FamTANo0.00.0TA406.0923.00.0ALQUnfYNormNormSBrkr0.0TATAWd SdngWd SdngNANA0.0CBlock1.0Typ312.01.0TAUnfTAAttchd1958.01329.01.0GasATA1Story1GdLvlGtl14267.0Corner81.0IR10.0020RL108.0BrkFaceGar212500.06.0NAmes36.066Y0.0NACompShgHipNormalWD0.0Pave61329.0AllPub393.01958.01958.02010.0172000
45263530302110.00.00NA31FamTANo1.00.0TA1045.01065.00.0ALQUnfYNormNormSBrkr0.0TAGdBrkFaceBrkFaceNATA2.0CBlock2.0Typ522.02.0TAFinTAAttchd1968.02110.01.0GasAEx1Story1ExLvlGtl11160.0Corner93.0Reg0.0020RL0.0NoneNA0.04.0NAmes0.057Y0.0NACompShgHipNormalWD0.0Pave82110.0AllPub0.01968.01968.02010.0244000
5527105010928.0701.00NA31FamTANo0.00.0Gd137.0791.00.0GLQUnfYNormNormSBrkr0.0TATAVinylSdVinylSdMnPrvTA1.0PConc2.0Typ482.02.0TAFinTAAttchd1997.01629.01.0GasAGd2Story1TALvlGtl13830.0Inside74.0IR10.0060RL0.0NoneNA0.03.0Gilbert34.055Y0.0NACompShgGableNormalWD0.0Pave6928.0AllPub212.01997.01998.02010.0189900
6527105030926.0678.00NA31FamTANo0.00.0TA324.0602.00.0GLQUnfYNormNormSBrkr0.0TATAVinylSdVinylSdNAGd1.0PConc2.0Typ470.02.0TAFinTAAttchd1998.01604.01.0GasAEx2Story1GdLvlGtl9978.0Inside78.0IR10.0060RL20.0BrkFaceNA0.06.0Gilbert36.066Y0.0NACompShgGableNormalWD0.0Pave7926.0AllPub360.01998.01998.02010.0195500
75271271501338.00.00NA2TwnhsETAMn1.00.0Gd722.0616.00.0GLQUnfYNormNormSBrkr170.0TAGdCemntBdCmentBdNANA0.0PConc2.0Typ582.02.0TAFinTAAttchd2001.01338.00.0GasAEx1Story1GdLvlGtl4920.0Inside41.0Reg0.0120RL0.0NoneNA0.04.0StoneBr0.058Y0.0NACompShgGableNormalWD0.0Pave61338.0AllPub0.02001.02001.02010.0213500
85271450801280.00.00NA2TwnhsETANo0.00.0Gd1017.0263.00.0ALQUnfYNormNormSBrkr0.0TAGdHdBoardHdBoardNANA0.0PConc2.0Typ506.02.0TARFnTAAttchd1992.01280.00.0GasAEx1Story1GdHLSGtl5005.0Inside43.0IR10.0120RL0.0NoneNA0.01.0StoneBr82.058Y0.0NACompShgGableNormalWD144.0Pave51280.0AllPub0.01992.01992.02010.0191500
95271460301616.00.00NA2TwnhsETANo1.00.0Gd415.01180.00.0GLQUnfYNormNormSBrkr0.0TAGdCemntBdCmentBdNATA1.0PConc2.0Typ608.02.0TARFnTAAttchd1995.01616.00.0GasAEx1Story1GdLvlGtl5389.0Inside39.0IR10.0120RL0.0NoneNA0.03.0StoneBr152.058Y0.0NACompShgGableNormalWD0.0Pave51595.0AllPub237.01995.01996.02010.0236500
105271621301028.0776.00NA31FamTANo0.00.0TA994.00.00.0UnfUnfYNormNormSBrkr0.0TATAVinylSdVinylSdNATA1.0PConc2.0Typ442.02.0TAFinTAAttchd1999.01804.01.0GasAGd2Story1GdLvlGtl7500.0Inside60.0Reg0.0060RL0.0NoneNA0.06.0Gilbert60.057Y0.0NACompShgGableNormalWD0.0Pave7994.0AllPub140.01999.01999.02010.0189000
\n", - "
" - ], - "text/plain": [ - " 1st Flr SF 2nd Flr SF 3Ssn Porch Alley Bedroom AbvGr \\\n", - "Order PID \n", - "1 526301100 1656.0 0.0 0 NA 3 \n", - "2 526350040 896.0 0.0 0 NA 2 \n", - "3 526351010 1329.0 0.0 0 NA 3 \n", - "4 526353030 2110.0 0.0 0 NA 3 \n", - "5 527105010 928.0 701.0 0 NA 3 \n", - "6 527105030 926.0 678.0 0 NA 3 \n", - "7 527127150 1338.0 0.0 0 NA 2 \n", - "8 527145080 1280.0 0.0 0 NA 2 \n", - "9 527146030 1616.0 0.0 0 NA 2 \n", - "10 527162130 1028.0 776.0 0 NA 3 \n", - "\n", - " Bldg Type Bsmt Cond Bsmt Exposure Bsmt Full Bath \\\n", - "Order PID \n", - "1 526301100 1Fam Gd Gd 1.0 \n", - "2 526350040 1Fam TA No 0.0 \n", - "3 526351010 1Fam TA No 0.0 \n", - "4 526353030 1Fam TA No 1.0 \n", - "5 527105010 1Fam TA No 0.0 \n", - "6 527105030 1Fam TA No 0.0 \n", - "7 527127150 TwnhsE TA Mn 1.0 \n", - "8 527145080 TwnhsE TA No 0.0 \n", - "9 527146030 TwnhsE TA No 1.0 \n", - "10 527162130 1Fam TA No 0.0 \n", - "\n", - " Bsmt Half Bath Bsmt Qual Bsmt Unf SF BsmtFin SF 1 \\\n", - "Order PID \n", - "1 526301100 0.0 TA 441.0 639.0 \n", - "2 526350040 0.0 TA 270.0 468.0 \n", - "3 526351010 0.0 TA 406.0 923.0 \n", - "4 526353030 0.0 TA 1045.0 1065.0 \n", - "5 527105010 0.0 Gd 137.0 791.0 \n", - "6 527105030 0.0 TA 324.0 602.0 \n", - "7 527127150 0.0 Gd 722.0 616.0 \n", - "8 527145080 0.0 Gd 1017.0 263.0 \n", - "9 527146030 0.0 Gd 415.0 1180.0 \n", - "10 527162130 0.0 TA 994.0 0.0 \n", - "\n", - " BsmtFin SF 2 BsmtFin Type 1 BsmtFin Type 2 Central Air \\\n", - "Order PID \n", - "1 526301100 0.0 BLQ Unf Y \n", - "2 526350040 144.0 Rec LwQ Y \n", - "3 526351010 0.0 ALQ Unf Y \n", - "4 526353030 0.0 ALQ Unf Y \n", - "5 527105010 0.0 GLQ Unf Y \n", - "6 527105030 0.0 GLQ Unf Y \n", - "7 527127150 0.0 GLQ Unf Y \n", - "8 527145080 0.0 ALQ Unf Y \n", - "9 527146030 0.0 GLQ Unf Y \n", - "10 527162130 0.0 Unf Unf Y \n", - "\n", - " Condition 1 Condition 2 Electrical Enclosed Porch Exter Cond \\\n", - "Order PID \n", - "1 526301100 Norm Norm SBrkr 0.0 TA \n", - "2 526350040 Feedr Norm SBrkr 0.0 TA \n", - "3 526351010 Norm Norm SBrkr 0.0 TA \n", - "4 526353030 Norm Norm SBrkr 0.0 TA \n", - "5 527105010 Norm Norm SBrkr 0.0 TA \n", - "6 527105030 Norm Norm SBrkr 0.0 TA \n", - "7 527127150 Norm Norm SBrkr 170.0 TA \n", - "8 527145080 Norm Norm SBrkr 0.0 TA \n", - "9 527146030 Norm Norm SBrkr 0.0 TA \n", - "10 527162130 Norm Norm SBrkr 0.0 TA \n", - "\n", - " Exter Qual Exterior 1st Exterior 2nd Fence Fireplace Qu \\\n", - "Order PID \n", - "1 526301100 TA BrkFace Plywood NA Gd \n", - "2 526350040 TA VinylSd VinylSd MnPrv NA \n", - "3 526351010 TA Wd Sdng Wd Sdng NA NA \n", - "4 526353030 Gd BrkFace BrkFace NA TA \n", - "5 527105010 TA VinylSd VinylSd MnPrv TA \n", - "6 527105030 TA VinylSd VinylSd NA Gd \n", - "7 527127150 Gd CemntBd CmentBd NA NA \n", - "8 527145080 Gd HdBoard HdBoard NA NA \n", - "9 527146030 Gd CemntBd CmentBd NA TA \n", - "10 527162130 TA VinylSd VinylSd NA TA \n", - "\n", - " Fireplaces Foundation Full Bath Functional Garage Area \\\n", - "Order PID \n", - "1 526301100 2.0 CBlock 1.0 Typ 528.0 \n", - "2 526350040 0.0 CBlock 1.0 Typ 730.0 \n", - "3 526351010 0.0 CBlock 1.0 Typ 312.0 \n", - "4 526353030 2.0 CBlock 2.0 Typ 522.0 \n", - "5 527105010 1.0 PConc 2.0 Typ 482.0 \n", - "6 527105030 1.0 PConc 2.0 Typ 470.0 \n", - "7 527127150 0.0 PConc 2.0 Typ 582.0 \n", - "8 527145080 0.0 PConc 2.0 Typ 506.0 \n", - "9 527146030 1.0 PConc 2.0 Typ 608.0 \n", - "10 527162130 1.0 PConc 2.0 Typ 442.0 \n", - "\n", - " Garage Cars Garage Cond Garage Finish Garage Qual \\\n", - "Order PID \n", - "1 526301100 2.0 TA Fin TA \n", - "2 526350040 1.0 TA Unf TA \n", - "3 526351010 1.0 TA Unf TA \n", - "4 526353030 2.0 TA Fin TA \n", - "5 527105010 2.0 TA Fin TA \n", - "6 527105030 2.0 TA Fin TA \n", - "7 527127150 2.0 TA Fin TA \n", - "8 527145080 2.0 TA RFn TA \n", - "9 527146030 2.0 TA RFn TA \n", - "10 527162130 2.0 TA Fin TA \n", - "\n", - " Garage Type Garage Yr Blt Gr Liv Area Half Bath Heating \\\n", - "Order PID \n", - "1 526301100 Attchd 1960.0 1656.0 0.0 GasA \n", - "2 526350040 Attchd 1961.0 896.0 0.0 GasA \n", - "3 526351010 Attchd 1958.0 1329.0 1.0 GasA \n", - "4 526353030 Attchd 1968.0 2110.0 1.0 GasA \n", - "5 527105010 Attchd 1997.0 1629.0 1.0 GasA \n", - "6 527105030 Attchd 1998.0 1604.0 1.0 GasA \n", - "7 527127150 Attchd 2001.0 1338.0 0.0 GasA \n", - "8 527145080 Attchd 1992.0 1280.0 0.0 GasA \n", - "9 527146030 Attchd 1995.0 1616.0 0.0 GasA \n", - "10 527162130 Attchd 1999.0 1804.0 1.0 GasA \n", - "\n", - " Heating QC House Style Kitchen AbvGr Kitchen Qual \\\n", - "Order PID \n", - "1 526301100 Fa 1Story 1 TA \n", - "2 526350040 TA 1Story 1 TA \n", - "3 526351010 TA 1Story 1 Gd \n", - "4 526353030 Ex 1Story 1 Ex \n", - "5 527105010 Gd 2Story 1 TA \n", - "6 527105030 Ex 2Story 1 Gd \n", - "7 527127150 Ex 1Story 1 Gd \n", - "8 527145080 Ex 1Story 1 Gd \n", - "9 527146030 Ex 1Story 1 Gd \n", - "10 527162130 Gd 2Story 1 Gd \n", - "\n", - " Land Contour Land Slope Lot Area Lot Config Lot Frontage \\\n", - "Order PID \n", - "1 526301100 Lvl Gtl 31770.0 Corner 141.0 \n", - "2 526350040 Lvl Gtl 11622.0 Inside 80.0 \n", - "3 526351010 Lvl Gtl 14267.0 Corner 81.0 \n", - "4 526353030 Lvl Gtl 11160.0 Corner 93.0 \n", - "5 527105010 Lvl Gtl 13830.0 Inside 74.0 \n", - "6 527105030 Lvl Gtl 9978.0 Inside 78.0 \n", - "7 527127150 Lvl Gtl 4920.0 Inside 41.0 \n", - "8 527145080 HLS Gtl 5005.0 Inside 43.0 \n", - "9 527146030 Lvl Gtl 5389.0 Inside 39.0 \n", - "10 527162130 Lvl Gtl 7500.0 Inside 60.0 \n", - "\n", - " Lot Shape Low Qual Fin SF MS SubClass MS Zoning \\\n", - "Order PID \n", - "1 526301100 IR1 0.0 020 RL \n", - "2 526350040 Reg 0.0 020 RH \n", - "3 526351010 IR1 0.0 020 RL \n", - "4 526353030 Reg 0.0 020 RL \n", - "5 527105010 IR1 0.0 060 RL \n", - "6 527105030 IR1 0.0 060 RL \n", - "7 527127150 Reg 0.0 120 RL \n", - "8 527145080 IR1 0.0 120 RL \n", - "9 527146030 IR1 0.0 120 RL \n", - "10 527162130 Reg 0.0 060 RL \n", - "\n", - " Mas Vnr Area Mas Vnr Type Misc Feature Misc Val Mo Sold \\\n", - "Order PID \n", - "1 526301100 112.0 Stone NA 0.0 5.0 \n", - "2 526350040 0.0 None NA 0.0 6.0 \n", - "3 526351010 108.0 BrkFace Gar2 12500.0 6.0 \n", - "4 526353030 0.0 None NA 0.0 4.0 \n", - "5 527105010 0.0 None NA 0.0 3.0 \n", - "6 527105030 20.0 BrkFace NA 0.0 6.0 \n", - "7 527127150 0.0 None NA 0.0 4.0 \n", - "8 527145080 0.0 None NA 0.0 1.0 \n", - "9 527146030 0.0 None NA 0.0 3.0 \n", - "10 527162130 0.0 None NA 0.0 6.0 \n", - "\n", - " Neighborhood Open Porch SF Overall Cond Overall Qual \\\n", - "Order PID \n", - "1 526301100 NAmes 62.0 5 6 \n", - "2 526350040 NAmes 0.0 6 5 \n", - "3 526351010 NAmes 36.0 6 6 \n", - "4 526353030 NAmes 0.0 5 7 \n", - "5 527105010 Gilbert 34.0 5 5 \n", - "6 527105030 Gilbert 36.0 6 6 \n", - "7 527127150 StoneBr 0.0 5 8 \n", - "8 527145080 StoneBr 82.0 5 8 \n", - "9 527146030 StoneBr 152.0 5 8 \n", - "10 527162130 Gilbert 60.0 5 7 \n", - "\n", - " Paved Drive Pool Area Pool QC Roof Matl Roof Style \\\n", - "Order PID \n", - "1 526301100 P 0.0 NA CompShg Hip \n", - "2 526350040 Y 0.0 NA CompShg Gable \n", - "3 526351010 Y 0.0 NA CompShg Hip \n", - "4 526353030 Y 0.0 NA CompShg Hip \n", - "5 527105010 Y 0.0 NA CompShg Gable \n", - "6 527105030 Y 0.0 NA CompShg Gable \n", - "7 527127150 Y 0.0 NA CompShg Gable \n", - "8 527145080 Y 0.0 NA CompShg Gable \n", - "9 527146030 Y 0.0 NA CompShg Gable \n", - "10 527162130 Y 0.0 NA CompShg Gable \n", - "\n", - " Sale Condition Sale Type Screen Porch Street TotRms AbvGrd \\\n", - "Order PID \n", - "1 526301100 Normal WD 0.0 Pave 7 \n", - "2 526350040 Normal WD 120.0 Pave 5 \n", - "3 526351010 Normal WD 0.0 Pave 6 \n", - "4 526353030 Normal WD 0.0 Pave 8 \n", - "5 527105010 Normal WD 0.0 Pave 6 \n", - "6 527105030 Normal WD 0.0 Pave 7 \n", - "7 527127150 Normal WD 0.0 Pave 6 \n", - "8 527145080 Normal WD 144.0 Pave 5 \n", - "9 527146030 Normal WD 0.0 Pave 5 \n", - "10 527162130 Normal WD 0.0 Pave 7 \n", - "\n", - " Total Bsmt SF Utilities Wood Deck SF Year Built \\\n", - "Order PID \n", - "1 526301100 1080.0 AllPub 210.0 1960.0 \n", - "2 526350040 882.0 AllPub 140.0 1961.0 \n", - "3 526351010 1329.0 AllPub 393.0 1958.0 \n", - "4 526353030 2110.0 AllPub 0.0 1968.0 \n", - "5 527105010 928.0 AllPub 212.0 1997.0 \n", - "6 527105030 926.0 AllPub 360.0 1998.0 \n", - "7 527127150 1338.0 AllPub 0.0 2001.0 \n", - "8 527145080 1280.0 AllPub 0.0 1992.0 \n", - "9 527146030 1595.0 AllPub 237.0 1995.0 \n", - "10 527162130 994.0 AllPub 140.0 1999.0 \n", - "\n", - " Year Remod/Add Yr Sold SalePrice \n", - "Order PID \n", - "1 526301100 1960.0 2010.0 215000 \n", - "2 526350040 1961.0 2010.0 105000 \n", - "3 526351010 1958.0 2010.0 172000 \n", - "4 526353030 1968.0 2010.0 244000 \n", - "5 527105010 1998.0 2010.0 189900 \n", - "6 527105030 1998.0 2010.0 195500 \n", - "7 527127150 2001.0 2010.0 213500 \n", - "8 527145080 1992.0 2010.0 191500 \n", - "9 527146030 1996.0 2010.0 236500 \n", - "10 527162130 1999.0 2010.0 189000 " - ] - }, - "execution_count": 6, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "df.head(10)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Spelling Mistakes & Data Types" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Some textual values appear differently in the provided data file as compared to the specification. These inconsistencies are manually repaired." - ] - }, - { - "cell_type": "code", - "execution_count": 7, - "metadata": {}, - "outputs": [], - "source": [ - "# Repair spelling and whitespace mistakes.\n", - "df[\"Bldg Type\"] = df[\"Bldg Type\"].replace(to_replace=\"2fmCon\", value=\"2FmCon\")\n", - "df[\"Bldg Type\"] = df[\"Bldg Type\"].replace(to_replace=\"Duplex\", value=\"Duplx\")\n", - "df[\"Bldg Type\"] = df[\"Bldg Type\"].replace(to_replace=\"Twnhs\", value=\"TwnhsI\")\n", - "df[\"Exterior 2nd\"] = df[\"Exterior 2nd\"].replace(to_replace=\"Brk Cmn\", value=\"BrkComm\")\n", - "df[\"Exterior 2nd\"] = df[\"Exterior 2nd\"].replace(to_replace=\"CmentBd\", value=\"CemntBd\")\n", - "df[\"Exterior 2nd\"] = df[\"Exterior 2nd\"].replace(to_replace=\"Wd Shng\", value=\"WdShing\")\n", - "df[\"MS Zoning\"] = df[\"MS Zoning\"].replace(to_replace=\"A (agr)\", value=\"A\")\n", - "df[\"MS Zoning\"] = df[\"MS Zoning\"].replace(to_replace=\"C (all)\", value=\"C\")\n", - "df[\"MS Zoning\"] = df[\"MS Zoning\"].replace(to_replace=\"I (all)\", value=\"I\")\n", - "df[\"Neighborhood\"] = df[\"Neighborhood\"].replace(to_replace=\"NAmes\", value=\"Names\")\n", - "df[\"Sale Type\"] = df[\"Sale Type\"].replace(to_replace=\"WD \", value=\"WD\")" - ] - }, - { - "cell_type": "code", - "execution_count": 8, - "metadata": {}, - "outputs": [], - "source": [ - "# Ensure that the remaining textual values in the data file are a subset\n", - "# of the values allowed in the specification.\n", - "for column, mapping_info in LABEL_COLUMNS.items():\n", - " # Note that .unique() returns a numpy array with integer dtype in cases\n", - " # where the provided data can be casted as such (e.g., \"Overall Qual\" column).\n", - " values_in_data = set(str(x) for x in df[column].unique() if x is not np.NaN)\n", - " values_in_description = set(mapping_info[\"lookups\"].keys())\n", - " assert values_in_data <= values_in_description" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Interestingly, all numeric columns (i.e. also \"continuous\" variables) come with only integer values." - ] - }, - { - "cell_type": "code", - "execution_count": 9, - "metadata": {}, - "outputs": [], - "source": [ - "# Show that all \"continuous\" variables come as integers.\n", - "for column in NUMERIC_VARIABLES + TARGET_VARIABLES:\n", - " not_null = df[column].notnull()\n", - " mask = (\n", - " df.loc[not_null, column].astype(np.int64)\n", - " != df.loc[not_null, column].astype(np.float64)\n", - " )\n", - " assert not mask.any()\n", - "# Cast discrete fields as integers where possible,\n", - "# i.e., all columns without missing values.\n", - "for column in DISCRETE_VARIABLES:\n", - " try:\n", - " df[column] = df[column].astype(np.int64)\n", - " except ValueError:\n", - " mask = df[column].notnull()\n", - " df.loc[mask, column].astype(np.int64)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Raw Data Overview" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "The overall shape of the data is a 2930 rows x 80 columns matrix." - ] - }, - { - "cell_type": "code", - "execution_count": 10, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "(2930, 80)" - ] - }, - "execution_count": 10, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "df.shape" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Continuous Variables" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "The continuous columns are truly continuous in the sense that each column has at least 14 unique value realizations." - ] - }, - { - "cell_type": "code", - "execution_count": 11, - "metadata": {}, - "outputs": [], - "source": [ - "for column in CONTINUOUS_VARIABLES:\n", - " mask = df[column].notnull()\n", - " num_realizations = len(list(x for x in df.loc[mask, column].unique()))\n", - " assert num_realizations > 13" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "A brief description of the variables:" - ] - }, - { - "cell_type": "code", - "execution_count": 12, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "1st Flr SF First Floor square feet\n", - "2nd Flr SF Second floor square feet\n", - "3Ssn Porch Three season porch area in square feet\n", - "Bsmt Unf SF Unfinished square feet of basement area\n", - "BsmtFin SF 1 Type 1 finished square feet\n", - "BsmtFin SF 2 Type 2 finished square feet\n", - "Enclosed Porch Enclosed porch area in square feet\n", - "Garage Area Size of garage in square feet\n", - "Gr Liv Area Above grade (ground) living area square feet\n", - "Lot Area Lot size in square feet\n", - "Lot Frontage Linear feet of street connected to property\n", - "Low Qual Fin SF Low quality finished square feet (all floors)\n", - "Mas Vnr Area Masonry veneer area in square feet\n", - "Misc Val $Value of miscellaneous feature\n", - "Open Porch SF Open porch area in square feet\n", - "Pool Area Pool area in square feet\n", - "Screen Porch Screen porch area in square feet\n", - "Total Bsmt SF Total square feet of basement area\n", - "Wood Deck SF Wood deck area in square feet\n" - ] - } - ], - "source": [ - "print_column_list(CONTINUOUS_COLUMNS)" - ] - }, - { - "cell_type": "code", - "execution_count": 13, - "metadata": {}, - "outputs": [ - { - "data": { - "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
1st Flr SF2nd Flr SF3Ssn PorchBsmt Unf SFBsmtFin SF 1BsmtFin SF 2Enclosed PorchGarage AreaGr Liv AreaLot AreaLot FrontageLow Qual Fin SFMas Vnr AreaMisc ValOpen Porch SFPool AreaScreen PorchTotal Bsmt SFWood Deck SF
OrderPID
15263011001656.00.00441.0639.00.00.0528.01656.031770.0141.00.0112.00.062.00.00.01080.0210.0
2526350040896.00.00270.0468.0144.00.0730.0896.011622.080.00.00.00.00.00.0120.0882.0140.0
35263510101329.00.00406.0923.00.00.0312.01329.014267.081.00.0108.012500.036.00.00.01329.0393.0
45263530302110.00.001045.01065.00.00.0522.02110.011160.093.00.00.00.00.00.00.02110.00.0
5527105010928.0701.00137.0791.00.00.0482.01629.013830.074.00.00.00.034.00.00.0928.0212.0
\n", - "
" - ], - "text/plain": [ - " 1st Flr SF 2nd Flr SF 3Ssn Porch Bsmt Unf SF \\\n", - "Order PID \n", - "1 526301100 1656.0 0.0 0 441.0 \n", - "2 526350040 896.0 0.0 0 270.0 \n", - "3 526351010 1329.0 0.0 0 406.0 \n", - "4 526353030 2110.0 0.0 0 1045.0 \n", - "5 527105010 928.0 701.0 0 137.0 \n", - "\n", - " BsmtFin SF 1 BsmtFin SF 2 Enclosed Porch Garage Area \\\n", - "Order PID \n", - "1 526301100 639.0 0.0 0.0 528.0 \n", - "2 526350040 468.0 144.0 0.0 730.0 \n", - "3 526351010 923.0 0.0 0.0 312.0 \n", - "4 526353030 1065.0 0.0 0.0 522.0 \n", - "5 527105010 791.0 0.0 0.0 482.0 \n", - "\n", - " Gr Liv Area Lot Area Lot Frontage Low Qual Fin SF \\\n", - "Order PID \n", - "1 526301100 1656.0 31770.0 141.0 0.0 \n", - "2 526350040 896.0 11622.0 80.0 0.0 \n", - "3 526351010 1329.0 14267.0 81.0 0.0 \n", - "4 526353030 2110.0 11160.0 93.0 0.0 \n", - "5 527105010 1629.0 13830.0 74.0 0.0 \n", - "\n", - " Mas Vnr Area Misc Val Open Porch SF Pool Area \\\n", - "Order PID \n", - "1 526301100 112.0 0.0 62.0 0.0 \n", - "2 526350040 0.0 0.0 0.0 0.0 \n", - "3 526351010 108.0 12500.0 36.0 0.0 \n", - "4 526353030 0.0 0.0 0.0 0.0 \n", - "5 527105010 0.0 0.0 34.0 0.0 \n", - "\n", - " Screen Porch Total Bsmt SF Wood Deck SF \n", - "Order PID \n", - "1 526301100 0.0 1080.0 210.0 \n", - "2 526350040 120.0 882.0 140.0 \n", - "3 526351010 0.0 1329.0 393.0 \n", - "4 526353030 0.0 2110.0 0.0 \n", - "5 527105010 0.0 928.0 212.0 " - ] - }, - "execution_count": 13, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "df[CONTINUOUS_VARIABLES].head()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Except for the column *Lot Frontage* the columns with missing data only have a couple of missing values (i.e., < 1% of all the rows)." - ] - }, - { - "cell_type": "code", - "execution_count": 14, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "\n", - "MultiIndex: 2930 entries, (1, 526301100) to (2930, 924151050)\n", - "Data columns (total 19 columns):\n", - " # Column Non-Null Count Dtype \n", - "--- ------ -------------- ----- \n", - " 0 1st Flr SF 2930 non-null float64\n", - " 1 2nd Flr SF 2930 non-null float64\n", - " 2 3Ssn Porch 2930 non-null int64 \n", - " 3 Bsmt Unf SF 2929 non-null float64\n", - " 4 BsmtFin SF 1 2929 non-null float64\n", - " 5 BsmtFin SF 2 2929 non-null float64\n", - " 6 Enclosed Porch 2930 non-null float64\n", - " 7 Garage Area 2929 non-null float64\n", - " 8 Gr Liv Area 2930 non-null float64\n", - " 9 Lot Area 2930 non-null float64\n", - " 10 Lot Frontage 2440 non-null float64\n", - " 11 Low Qual Fin SF 2930 non-null float64\n", - " 12 Mas Vnr Area 2907 non-null float64\n", - " 13 Misc Val 2930 non-null float64\n", - " 14 Open Porch SF 2930 non-null float64\n", - " 15 Pool Area 2930 non-null float64\n", - " 16 Screen Porch 2930 non-null float64\n", - " 17 Total Bsmt SF 2929 non-null float64\n", - " 18 Wood Deck SF 2930 non-null float64\n", - "dtypes: float64(18), int64(1)\n", - "memory usage: 572.3 KB\n" - ] - } - ], - "source": [ - "df[CONTINUOUS_VARIABLES].info()" - ] - }, - { - "cell_type": "code", - "execution_count": 15, - "metadata": {}, - "outputs": [], - "source": [ - "# The columns with a lot of missing\n", - "# values will be treated seperately below.\n", - "missing_a_lot = [\"Lot Frontage\"]" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Discrete Variables" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "The discrete columns have between 2 and 15 unique realizations each if year numbers are excluded from the analysis." - ] - }, - { - "cell_type": "code", - "execution_count": 16, - "metadata": {}, - "outputs": [], - "source": [ - "for column in DISCRETE_VARIABLES:\n", - " mask = df[column].notnull()\n", - " num_realizations = len(list(x for x in df.loc[mask, column].unique()))\n", - " if column not in (\"Year Built\", \"Year Remod/Add\", \"Garage Yr Blt\"):\n", - " assert num_realizations < 15\n", - " assert num_realizations > 2" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "A brief description of the variables:" - ] - }, - { - "cell_type": "code", - "execution_count": 17, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Bedroom AbvGr Bedrooms above grade (does NOT include basement bedrooms)\n", - "Bsmt Full Bath Basement full bathrooms\n", - "Bsmt Half Bath Basement half bathrooms\n", - "Fireplaces Number of fireplaces\n", - "Full Bath Full bathrooms above grade\n", - "Garage Cars Size of garage in car capacity\n", - "Garage Yr Blt Year garage was built\n", - "Half Bath Half baths above grade\n", - "Kitchen AbvGr Kitchens above grade\n", - "Mo Sold Month Sold (MM)\n", - "TotRms AbvGrd Total rooms above grade (does not include bathrooms)\n", - "Year Built Original construction date\n", - "Year Remod/Add Remodel date (same as construction date if no remodeling or additions)\n", - "Yr Sold Year Sold (YYYY)\n" - ] - } - ], - "source": [ - "print_column_list(DISCRETE_COLUMNS)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "**Note:** columns with missing values are implicitly casted to a *float64* type an the *int64* type has no concept of a NaN (=\"Not a number\") value." - ] - }, - { - "cell_type": "code", - "execution_count": 18, - "metadata": {}, - "outputs": [ - { - "data": { - "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
Bedroom AbvGrBsmt Full BathBsmt Half BathFireplacesFull BathGarage CarsGarage Yr BltHalf BathKitchen AbvGrMo SoldTotRms AbvGrdYear BuiltYear Remod/AddYr Sold
OrderPID
152630110031.00.0212.01960.00157196019602010
252635004020.00.0011.01961.00165196119612010
352635101030.00.0011.01958.01166195819582010
452635303031.00.0222.01968.01148196819682010
552710501030.00.0122.01997.01136199719982010
\n", - "
" - ], - "text/plain": [ - " Bedroom AbvGr Bsmt Full Bath Bsmt Half Bath Fireplaces \\\n", - "Order PID \n", - "1 526301100 3 1.0 0.0 2 \n", - "2 526350040 2 0.0 0.0 0 \n", - "3 526351010 3 0.0 0.0 0 \n", - "4 526353030 3 1.0 0.0 2 \n", - "5 527105010 3 0.0 0.0 1 \n", - "\n", - " Full Bath Garage Cars Garage Yr Blt Half Bath \\\n", - "Order PID \n", - "1 526301100 1 2.0 1960.0 0 \n", - "2 526350040 1 1.0 1961.0 0 \n", - "3 526351010 1 1.0 1958.0 1 \n", - "4 526353030 2 2.0 1968.0 1 \n", - "5 527105010 2 2.0 1997.0 1 \n", - "\n", - " Kitchen AbvGr Mo Sold TotRms AbvGrd Year Built \\\n", - "Order PID \n", - "1 526301100 1 5 7 1960 \n", - "2 526350040 1 6 5 1961 \n", - "3 526351010 1 6 6 1958 \n", - "4 526353030 1 4 8 1968 \n", - "5 527105010 1 3 6 1997 \n", - "\n", - " Year Remod/Add Yr Sold \n", - "Order PID \n", - "1 526301100 1960 2010 \n", - "2 526350040 1961 2010 \n", - "3 526351010 1958 2010 \n", - "4 526353030 1968 2010 \n", - "5 527105010 1998 2010 " - ] - }, - "execution_count": 18, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "df[DISCRETE_VARIABLES].head()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Except for the *Garage Yr Blt* column no variable has a significant number of missing values (i.e., > 1% of all rows)." - ] - }, - { - "cell_type": "code", - "execution_count": 19, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "\n", - "MultiIndex: 2930 entries, (1, 526301100) to (2930, 924151050)\n", - "Data columns (total 14 columns):\n", - " # Column Non-Null Count Dtype \n", - "--- ------ -------------- ----- \n", - " 0 Bedroom AbvGr 2930 non-null int64 \n", - " 1 Bsmt Full Bath 2928 non-null float64\n", - " 2 Bsmt Half Bath 2928 non-null float64\n", - " 3 Fireplaces 2930 non-null int64 \n", - " 4 Full Bath 2930 non-null int64 \n", - " 5 Garage Cars 2929 non-null float64\n", - " 6 Garage Yr Blt 2771 non-null float64\n", - " 7 Half Bath 2930 non-null int64 \n", - " 8 Kitchen AbvGr 2930 non-null int64 \n", - " 9 Mo Sold 2930 non-null int64 \n", - " 10 TotRms AbvGrd 2930 non-null int64 \n", - " 11 Year Built 2930 non-null int64 \n", - " 12 Year Remod/Add 2930 non-null int64 \n", - " 13 Yr Sold 2930 non-null int64 \n", - "dtypes: float64(4), int64(10)\n", - "memory usage: 457.8 KB\n" - ] - } - ], - "source": [ - "df[DISCRETE_VARIABLES].info()" - ] - }, - { - "cell_type": "code", - "execution_count": 20, - "metadata": {}, - "outputs": [], - "source": [ - "missing_a_lot.append(\"Garage Yr Blt\")" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Nominal Variables" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Except for the total of 28 neighborhoods, the nominal columns come with anywhere between 1 and 18 different labels each." - ] - }, - { - "cell_type": "code", - "execution_count": 21, - "metadata": {}, - "outputs": [], - "source": [ - "for column in NOMINAL_VARIABLES:\n", - " mask = df[column].notnull()\n", - " num_realizations = len(list(x for x in df.loc[mask, column].unique()))\n", - " if column not in (\"Neighborhood\"):\n", - " assert num_realizations < 18\n", - " assert num_realizations > 1" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "A brief description of the variables:" - ] - }, - { - "cell_type": "code", - "execution_count": 22, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Alley Type of alley access to property\n", - "Bldg Type Type of dwelling\n", - "Central Air Central air conditioning\n", - "Condition 1 Proximity to various conditions\n", - "Condition 2 Proximity to various conditions (if more than one is present)\n", - "Exterior 1st Exterior covering on house\n", - "Exterior 2nd Exterior covering on house (if more than one material)\n", - "Foundation Type of foundation\n", - "Garage Type Garage location\n", - "Heating Type of heating\n", - "House Style Style of dwelling\n", - "Land Contour Flatness of the property\n", - "Lot Config Lot configuration\n", - "MS SubClass Identifies the type of dwelling involved in the sale.\n", - "MS Zoning Identifies the general zoning classification of the sale.\n", - "Mas Vnr Type Masonry veneer type\n", - "Misc Feature Miscellaneous feature not covered in other categories\n", - "Neighborhood Physical locations within Ames city limits (map available)\n", - "Roof Matl Roof material\n", - "Roof Style Type of roof\n", - "Sale Condition Condition of sale\n", - "Sale Type Type of sale\n", - "Street Type of road access to property\n" - ] - } - ], - "source": [ - "print_column_list(NOMINAL_COLUMNS)" - ] - }, - { - "cell_type": "code", - "execution_count": 23, - "metadata": {}, - "outputs": [ - { - "data": { - "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
AlleyBldg TypeCentral AirCondition 1Condition 2Exterior 1stExterior 2ndFoundationGarage TypeHeatingHouse StyleLand ContourLot ConfigMS SubClassMS ZoningMas Vnr TypeMisc FeatureNeighborhoodRoof MatlRoof StyleSale ConditionSale TypeStreet
OrderPID
1526301100NA1FamYNormNormBrkFacePlywoodCBlockAttchdGasA1StoryLvlCorner020RLStoneNANamesCompShgHipNormalWDPave
2526350040NA1FamYFeedrNormVinylSdVinylSdCBlockAttchdGasA1StoryLvlInside020RHNoneNANamesCompShgGableNormalWDPave
3526351010NA1FamYNormNormWd SdngWd SdngCBlockAttchdGasA1StoryLvlCorner020RLBrkFaceGar2NamesCompShgHipNormalWDPave
4526353030NA1FamYNormNormBrkFaceBrkFaceCBlockAttchdGasA1StoryLvlCorner020RLNoneNANamesCompShgHipNormalWDPave
5527105010NA1FamYNormNormVinylSdVinylSdPConcAttchdGasA2StoryLvlInside060RLNoneNAGilbertCompShgGableNormalWDPave
6527105030NA1FamYNormNormVinylSdVinylSdPConcAttchdGasA2StoryLvlInside060RLBrkFaceNAGilbertCompShgGableNormalWDPave
7527127150NATwnhsEYNormNormCemntBdCemntBdPConcAttchdGasA1StoryLvlInside120RLNoneNAStoneBrCompShgGableNormalWDPave
8527145080NATwnhsEYNormNormHdBoardHdBoardPConcAttchdGasA1StoryHLSInside120RLNoneNAStoneBrCompShgGableNormalWDPave
9527146030NATwnhsEYNormNormCemntBdCemntBdPConcAttchdGasA1StoryLvlInside120RLNoneNAStoneBrCompShgGableNormalWDPave
10527162130NA1FamYNormNormVinylSdVinylSdPConcAttchdGasA2StoryLvlInside060RLNoneNAGilbertCompShgGableNormalWDPave
\n", - "
" - ], - "text/plain": [ - " Alley Bldg Type Central Air Condition 1 Condition 2 \\\n", - "Order PID \n", - "1 526301100 NA 1Fam Y Norm Norm \n", - "2 526350040 NA 1Fam Y Feedr Norm \n", - "3 526351010 NA 1Fam Y Norm Norm \n", - "4 526353030 NA 1Fam Y Norm Norm \n", - "5 527105010 NA 1Fam Y Norm Norm \n", - "6 527105030 NA 1Fam Y Norm Norm \n", - "7 527127150 NA TwnhsE Y Norm Norm \n", - "8 527145080 NA TwnhsE Y Norm Norm \n", - "9 527146030 NA TwnhsE Y Norm Norm \n", - "10 527162130 NA 1Fam Y Norm Norm \n", - "\n", - " Exterior 1st Exterior 2nd Foundation Garage Type Heating \\\n", - "Order PID \n", - "1 526301100 BrkFace Plywood CBlock Attchd GasA \n", - "2 526350040 VinylSd VinylSd CBlock Attchd GasA \n", - "3 526351010 Wd Sdng Wd Sdng CBlock Attchd GasA \n", - "4 526353030 BrkFace BrkFace CBlock Attchd GasA \n", - "5 527105010 VinylSd VinylSd PConc Attchd GasA \n", - "6 527105030 VinylSd VinylSd PConc Attchd GasA \n", - "7 527127150 CemntBd CemntBd PConc Attchd GasA \n", - "8 527145080 HdBoard HdBoard PConc Attchd GasA \n", - "9 527146030 CemntBd CemntBd PConc Attchd GasA \n", - "10 527162130 VinylSd VinylSd PConc Attchd GasA \n", - "\n", - " House Style Land Contour Lot Config MS SubClass MS Zoning \\\n", - "Order PID \n", - "1 526301100 1Story Lvl Corner 020 RL \n", - "2 526350040 1Story Lvl Inside 020 RH \n", - "3 526351010 1Story Lvl Corner 020 RL \n", - "4 526353030 1Story Lvl Corner 020 RL \n", - "5 527105010 2Story Lvl Inside 060 RL \n", - "6 527105030 2Story Lvl Inside 060 RL \n", - "7 527127150 1Story Lvl Inside 120 RL \n", - "8 527145080 1Story HLS Inside 120 RL \n", - "9 527146030 1Story Lvl Inside 120 RL \n", - "10 527162130 2Story Lvl Inside 060 RL \n", - "\n", - " Mas Vnr Type Misc Feature Neighborhood Roof Matl Roof Style \\\n", - "Order PID \n", - "1 526301100 Stone NA Names CompShg Hip \n", - "2 526350040 None NA Names CompShg Gable \n", - "3 526351010 BrkFace Gar2 Names CompShg Hip \n", - "4 526353030 None NA Names CompShg Hip \n", - "5 527105010 None NA Gilbert CompShg Gable \n", - "6 527105030 BrkFace NA Gilbert CompShg Gable \n", - "7 527127150 None NA StoneBr CompShg Gable \n", - "8 527145080 None NA StoneBr CompShg Gable \n", - "9 527146030 None NA StoneBr CompShg Gable \n", - "10 527162130 None NA Gilbert CompShg Gable \n", - "\n", - " Sale Condition Sale Type Street \n", - "Order PID \n", - "1 526301100 Normal WD Pave \n", - "2 526350040 Normal WD Pave \n", - "3 526351010 Normal WD Pave \n", - "4 526353030 Normal WD Pave \n", - "5 527105010 Normal WD Pave \n", - "6 527105030 Normal WD Pave \n", - "7 527127150 Normal WD Pave \n", - "8 527145080 Normal WD Pave \n", - "9 527146030 Normal WD Pave \n", - "10 527162130 Normal WD Pave " - ] - }, - "execution_count": 23, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "df[NOMINAL_VARIABLES].head(10)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "For the nominal variables there is only a neglectable number of missing values." - ] - }, - { - "cell_type": "code", - "execution_count": 24, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "\n", - "MultiIndex: 2930 entries, (1, 526301100) to (2930, 924151050)\n", - "Data columns (total 23 columns):\n", - " # Column Non-Null Count Dtype \n", - "--- ------ -------------- ----- \n", - " 0 Alley 2930 non-null object\n", - " 1 Bldg Type 2930 non-null object\n", - " 2 Central Air 2930 non-null object\n", - " 3 Condition 1 2930 non-null object\n", - " 4 Condition 2 2930 non-null object\n", - " 5 Exterior 1st 2930 non-null object\n", - " 6 Exterior 2nd 2930 non-null object\n", - " 7 Foundation 2930 non-null object\n", - " 8 Garage Type 2930 non-null object\n", - " 9 Heating 2930 non-null object\n", - " 10 House Style 2930 non-null object\n", - " 11 Land Contour 2930 non-null object\n", - " 12 Lot Config 2930 non-null object\n", - " 13 MS SubClass 2930 non-null object\n", - " 14 MS Zoning 2930 non-null object\n", - " 15 Mas Vnr Type 2907 non-null object\n", - " 16 Misc Feature 2930 non-null object\n", - " 17 Neighborhood 2930 non-null object\n", - " 18 Roof Matl 2930 non-null object\n", - " 19 Roof Style 2930 non-null object\n", - " 20 Sale Condition 2930 non-null object\n", - " 21 Sale Type 2930 non-null object\n", - " 22 Street 2930 non-null object\n", - "dtypes: object(23)\n", - "memory usage: 663.8+ KB\n" - ] - } - ], - "source": [ - "df[NOMINAL_VARIABLES].info()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Ordinal Variables" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "The ordinal columns come with anywhere between 2 and 11 distinct labels each." - ] - }, - { - "cell_type": "code", - "execution_count": 25, - "metadata": {}, - "outputs": [], - "source": [ - "for column in ORDINAL_VARIABLES:\n", - " mask = df[column].notnull()\n", - " num_realizations = len(list(x for x in df.loc[mask, column].unique()))\n", - " assert 2 < num_realizations < 11" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "A brief description of the variables:" - ] - }, - { - "cell_type": "code", - "execution_count": 26, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Bsmt Cond Evaluates the general condition of the basement\n", - "Bsmt Exposure Refers to walkout or garden level walls\n", - "Bsmt Qual Evaluates the height of the basement\n", - "BsmtFin Type 1 Rating of basement finished area\n", - "BsmtFin Type 2 Rating of basement finished area (if multiple types)\n", - "Electrical Electrical system\n", - "Exter Cond Evaluates the present condition of the material on the exterior\n", - "Exter Qual Evaluates the quality of the material on the exterior\n", - "Fence Fence quality\n", - "Fireplace Qu Fireplace quality\n", - "Functional Home functionality (Assume typical unless deductions are warranted)\n", - "Garage Cond Garage condition\n", - "Garage Finish Interior finish of the garage\n", - "Garage Qual Garage quality\n", - "Heating QC Heating quality and condition\n", - "Kitchen Qual Kitchen quality\n", - "Land Slope Slope of property\n", - "Lot Shape General shape of property\n", - "Overall Cond Rates the overall condition of the house\n", - "Overall Qual Rates the overall material and finish of the house\n", - "Paved Drive Paved driveway\n", - "Pool QC Pool quality\n", - "Utilities Type of utilities available\n" - ] - } - ], - "source": [ - "print_column_list(ORDINAL_COLUMNS)" - ] - }, - { - "cell_type": "code", - "execution_count": 27, - "metadata": {}, - "outputs": [ - { - "data": { - "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
Bsmt CondBsmt ExposureBsmt QualBsmtFin Type 1BsmtFin Type 2ElectricalExter CondExter QualFenceFireplace QuFunctionalGarage CondGarage FinishGarage QualHeating QCKitchen QualLand SlopeLot ShapeOverall CondOverall QualPaved DrivePool QCUtilities
OrderPID
1526301100GdGdTABLQUnfSBrkrTATANAGdTypTAFinTAFaTAGtlIR156PNAAllPub
2526350040TANoTARecLwQSBrkrTATAMnPrvNATypTAUnfTATATAGtlReg65YNAAllPub
3526351010TANoTAALQUnfSBrkrTATANANATypTAUnfTATAGdGtlIR166YNAAllPub
4526353030TANoTAALQUnfSBrkrTAGdNATATypTAFinTAExExGtlReg57YNAAllPub
5527105010TANoGdGLQUnfSBrkrTATAMnPrvTATypTAFinTAGdTAGtlIR155YNAAllPub
6527105030TANoTAGLQUnfSBrkrTATANAGdTypTAFinTAExGdGtlIR166YNAAllPub
7527127150TAMnGdGLQUnfSBrkrTAGdNANATypTAFinTAExGdGtlReg58YNAAllPub
8527145080TANoGdALQUnfSBrkrTAGdNANATypTARFnTAExGdGtlIR158YNAAllPub
9527146030TANoGdGLQUnfSBrkrTAGdNATATypTARFnTAExGdGtlIR158YNAAllPub
10527162130TANoTAUnfUnfSBrkrTATANATATypTAFinTAGdGdGtlReg57YNAAllPub
\n", - "
" - ], - "text/plain": [ - " Bsmt Cond Bsmt Exposure Bsmt Qual BsmtFin Type 1 \\\n", - "Order PID \n", - "1 526301100 Gd Gd TA BLQ \n", - "2 526350040 TA No TA Rec \n", - "3 526351010 TA No TA ALQ \n", - "4 526353030 TA No TA ALQ \n", - "5 527105010 TA No Gd GLQ \n", - "6 527105030 TA No TA GLQ \n", - "7 527127150 TA Mn Gd GLQ \n", - "8 527145080 TA No Gd ALQ \n", - "9 527146030 TA No Gd GLQ \n", - "10 527162130 TA No TA Unf \n", - "\n", - " BsmtFin Type 2 Electrical Exter Cond Exter Qual Fence \\\n", - "Order PID \n", - "1 526301100 Unf SBrkr TA TA NA \n", - "2 526350040 LwQ SBrkr TA TA MnPrv \n", - "3 526351010 Unf SBrkr TA TA NA \n", - "4 526353030 Unf SBrkr TA Gd NA \n", - "5 527105010 Unf SBrkr TA TA MnPrv \n", - "6 527105030 Unf SBrkr TA TA NA \n", - "7 527127150 Unf SBrkr TA Gd NA \n", - "8 527145080 Unf SBrkr TA Gd NA \n", - "9 527146030 Unf SBrkr TA Gd NA \n", - "10 527162130 Unf SBrkr TA TA NA \n", - "\n", - " Fireplace Qu Functional Garage Cond Garage Finish Garage Qual \\\n", - "Order PID \n", - "1 526301100 Gd Typ TA Fin TA \n", - "2 526350040 NA Typ TA Unf TA \n", - "3 526351010 NA Typ TA Unf TA \n", - "4 526353030 TA Typ TA Fin TA \n", - "5 527105010 TA Typ TA Fin TA \n", - "6 527105030 Gd Typ TA Fin TA \n", - "7 527127150 NA Typ TA Fin TA \n", - "8 527145080 NA Typ TA RFn TA \n", - "9 527146030 TA Typ TA RFn TA \n", - "10 527162130 TA Typ TA Fin TA \n", - "\n", - " Heating QC Kitchen Qual Land Slope Lot Shape Overall Cond \\\n", - "Order PID \n", - "1 526301100 Fa TA Gtl IR1 5 \n", - "2 526350040 TA TA Gtl Reg 6 \n", - "3 526351010 TA Gd Gtl IR1 6 \n", - "4 526353030 Ex Ex Gtl Reg 5 \n", - "5 527105010 Gd TA Gtl IR1 5 \n", - "6 527105030 Ex Gd Gtl IR1 6 \n", - "7 527127150 Ex Gd Gtl Reg 5 \n", - "8 527145080 Ex Gd Gtl IR1 5 \n", - "9 527146030 Ex Gd Gtl IR1 5 \n", - "10 527162130 Gd Gd Gtl Reg 5 \n", - "\n", - " Overall Qual Paved Drive Pool QC Utilities \n", - "Order PID \n", - "1 526301100 6 P NA AllPub \n", - "2 526350040 5 Y NA AllPub \n", - "3 526351010 6 Y NA AllPub \n", - "4 526353030 7 Y NA AllPub \n", - "5 527105010 5 Y NA AllPub \n", - "6 527105030 6 Y NA AllPub \n", - "7 527127150 8 Y NA AllPub \n", - "8 527145080 8 Y NA AllPub \n", - "9 527146030 8 Y NA AllPub \n", - "10 527162130 7 Y NA AllPub " - ] - }, - "execution_count": 27, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "df[ORDINAL_VARIABLES].head(10)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "For the ordinal variables there is only a neglectable number of missing values." - ] - }, - { - "cell_type": "code", - "execution_count": 28, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "\n", - "MultiIndex: 2930 entries, (1, 526301100) to (2930, 924151050)\n", - "Data columns (total 23 columns):\n", - " # Column Non-Null Count Dtype \n", - "--- ------ -------------- ----- \n", - " 0 Bsmt Cond 2929 non-null object\n", - " 1 Bsmt Exposure 2926 non-null object\n", - " 2 Bsmt Qual 2929 non-null object\n", - " 3 BsmtFin Type 1 2929 non-null object\n", - " 4 BsmtFin Type 2 2928 non-null object\n", - " 5 Electrical 2929 non-null object\n", - " 6 Exter Cond 2930 non-null object\n", - " 7 Exter Qual 2930 non-null object\n", - " 8 Fence 2930 non-null object\n", - " 9 Fireplace Qu 2930 non-null object\n", - " 10 Functional 2930 non-null object\n", - " 11 Garage Cond 2929 non-null object\n", - " 12 Garage Finish 2928 non-null object\n", - " 13 Garage Qual 2929 non-null object\n", - " 14 Heating QC 2930 non-null object\n", - " 15 Kitchen Qual 2930 non-null object\n", - " 16 Land Slope 2930 non-null object\n", - " 17 Lot Shape 2930 non-null object\n", - " 18 Overall Cond 2930 non-null object\n", - " 19 Overall Qual 2930 non-null object\n", - " 20 Paved Drive 2930 non-null object\n", - " 21 Pool QC 2930 non-null object\n", - " 22 Utilities 2930 non-null object\n", - "dtypes: object(23)\n", - "memory usage: 663.8+ KB\n" - ] - } - ], - "source": [ - "df[ORDINAL_VARIABLES].info()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Missing Data" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Visualizations" - ] - }, - { - "cell_type": "code", - "execution_count": 29, - "metadata": {}, - "outputs": [ - { - "data": { - "image/png": "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\n", - "text/plain": [ - "
" - ] - }, - "metadata": { - "needs_background": "light" - }, - "output_type": "display_data" - } - ], - "source": [ - "msno.matrix(df[CONTINUOUS_VARIABLES]);" - ] - }, - { - "cell_type": "code", - "execution_count": 30, - "metadata": {}, - "outputs": [ - { - "data": { - "image/png": "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\n", - "text/plain": [ - "
" - ] - }, - "metadata": { - "needs_background": "light" - }, - "output_type": "display_data" - } - ], - "source": [ - "msno.matrix(df[DISCRETE_VARIABLES]);" - ] - }, - { - "cell_type": "code", - "execution_count": 31, - "metadata": {}, - "outputs": [ - { - "data": { - "image/png": "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\n", - "text/plain": [ - "
" - ] - }, - "metadata": { - "needs_background": "light" - }, - "output_type": "display_data" - } - ], - "source": [ - "msno.matrix(df[NOMINAL_VARIABLES]);" - ] - }, - { - "cell_type": "code", - "execution_count": 32, - "metadata": {}, - "outputs": [ - { - "data": { - "image/png": "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\n", - "text/plain": [ - "
" - ] - }, - "metadata": { - "needs_background": "light" - }, - "output_type": "display_data" - } - ], - "source": [ - "msno.matrix(df[ORDINAL_VARIABLES]);" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Cleansing" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Since only about 1% of the overall number of observations exhibit variables with missing data (disregarding the columns *Lot Frontage* and *Garage Yr Blt*), the decision is made to discard these rows entirely to not have to deal with interpolating meaningful replacements for the missing values." - ] - }, - { - "cell_type": "code", - "execution_count": 33, - "metadata": {}, - "outputs": [], - "source": [ - "remaining_columns = sorted(set(ALL_VARIABLES) - set(missing_a_lot)) + TARGET_VARIABLES\n", - "mask = df[remaining_columns].isnull().any(axis=1)\n", - "assert (100 * mask.sum() / df.shape[0]) < 1.1 # percent\n", - "df = df[~mask]" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "The two columns with a lot of missing values regard the age of a house's optional garage and the length of the intersection with the street where the house is located. The first is assumed as not important for the house appraisal and the second is assumed to be captured in other variables (e.g. overall size of the house). Therefore, for sake of simplicity both columns are dropped from the DataFrame." - ] - }, - { - "cell_type": "code", - "execution_count": 34, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Garage Yr Blt Year garage was built\n", - "Lot Frontage Linear feet of street connected to property\n" - ] - } - ], - "source": [ - "print_column_list(missing_a_lot)" - ] - }, - { - "cell_type": "code", - "execution_count": 35, - "metadata": {}, - "outputs": [], - "source": [ - "df = df[remaining_columns]" - ] - }, - { - "cell_type": "code", - "execution_count": 36, - "metadata": {}, - "outputs": [], - "source": [ - "# Remove the discarded columns from the helper dictionaries / lists.\n", - "update_column_descriptions(df.columns)\n", - "# Without any more missing data, cast all numeric\n", - "# columns as floats or integers respectively.\n", - "for column in CONTINUOUS_VARIABLES + TARGET_VARIABLES:\n", - " df[column] = df[column].astype(np.float64)\n", - "for column in DISCRETE_VARIABLES:\n", - " df[column] = df[column].astype(np.int64)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Clean Data" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "The cleaned data comes as a 2898 rows x 78 columns matrix." - ] - }, - { - "cell_type": "code", - "execution_count": 37, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "(2898, 78)" - ] - }, - "execution_count": 37, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "df.shape" - ] - }, - { - "cell_type": "code", - "execution_count": 38, - "metadata": {}, - "outputs": [ - { - "data": { - "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
1st Flr SF2nd Flr SF3Ssn PorchAlleyBedroom AbvGrBldg TypeBsmt CondBsmt ExposureBsmt Full BathBsmt Half BathBsmt QualBsmt Unf SFBsmtFin SF 1BsmtFin SF 2BsmtFin Type 1BsmtFin Type 2Central AirCondition 1Condition 2ElectricalEnclosed PorchExter CondExter QualExterior 1stExterior 2ndFenceFireplace QuFireplacesFoundationFull BathFunctionalGarage AreaGarage CarsGarage CondGarage FinishGarage QualGarage TypeGr Liv AreaHalf BathHeatingHeating QCHouse StyleKitchen AbvGrKitchen QualLand ContourLand SlopeLot AreaLot ConfigLot ShapeLow Qual Fin SFMS SubClassMS ZoningMas Vnr AreaMas Vnr TypeMisc FeatureMisc ValMo SoldNeighborhoodOpen Porch SFOverall CondOverall QualPaved DrivePool AreaPool QCRoof MatlRoof StyleSale ConditionSale TypeScreen PorchStreetTotRms AbvGrdTotal Bsmt SFUtilitiesWood Deck SFYear BuiltYear Remod/AddYr SoldSalePrice
OrderPID
15263011001656.00.00.0NA31FamGdGd10TA441.0639.00.0BLQUnfYNormNormSBrkr0.0TATABrkFacePlywoodNAGd2CBlock1Typ528.02TAFinTAAttchd1656.00GasAFa1Story1TALvlGtl31770.0CornerIR10.0020RL112.0StoneNA0.05Names62.056P0.0NACompShgHipNormalWD0.0Pave71080.0AllPub210.0196019602010215000.0
2526350040896.00.00.0NA21FamTANo00TA270.0468.0144.0RecLwQYFeedrNormSBrkr0.0TATAVinylSdVinylSdMnPrvNA0CBlock1Typ730.01TAUnfTAAttchd896.00GasATA1Story1TALvlGtl11622.0InsideReg0.0020RH0.0NoneNA0.06Names0.065Y0.0NACompShgGableNormalWD120.0Pave5882.0AllPub140.0196119612010105000.0
35263510101329.00.00.0NA31FamTANo00TA406.0923.00.0ALQUnfYNormNormSBrkr0.0TATAWd SdngWd SdngNANA0CBlock1Typ312.01TAUnfTAAttchd1329.01GasATA1Story1GdLvlGtl14267.0CornerIR10.0020RL108.0BrkFaceGar212500.06Names36.066Y0.0NACompShgHipNormalWD0.0Pave61329.0AllPub393.0195819582010172000.0
45263530302110.00.00.0NA31FamTANo10TA1045.01065.00.0ALQUnfYNormNormSBrkr0.0TAGdBrkFaceBrkFaceNATA2CBlock2Typ522.02TAFinTAAttchd2110.01GasAEx1Story1ExLvlGtl11160.0CornerReg0.0020RL0.0NoneNA0.04Names0.057Y0.0NACompShgHipNormalWD0.0Pave82110.0AllPub0.0196819682010244000.0
5527105010928.0701.00.0NA31FamTANo00Gd137.0791.00.0GLQUnfYNormNormSBrkr0.0TATAVinylSdVinylSdMnPrvTA1PConc2Typ482.02TAFinTAAttchd1629.01GasAGd2Story1TALvlGtl13830.0InsideIR10.0060RL0.0NoneNA0.03Gilbert34.055Y0.0NACompShgGableNormalWD0.0Pave6928.0AllPub212.0199719982010189900.0
\n", - "
" - ], - "text/plain": [ - " 1st Flr SF 2nd Flr SF 3Ssn Porch Alley Bedroom AbvGr \\\n", - "Order PID \n", - "1 526301100 1656.0 0.0 0.0 NA 3 \n", - "2 526350040 896.0 0.0 0.0 NA 2 \n", - "3 526351010 1329.0 0.0 0.0 NA 3 \n", - "4 526353030 2110.0 0.0 0.0 NA 3 \n", - "5 527105010 928.0 701.0 0.0 NA 3 \n", - "\n", - " Bldg Type Bsmt Cond Bsmt Exposure Bsmt Full Bath \\\n", - "Order PID \n", - "1 526301100 1Fam Gd Gd 1 \n", - "2 526350040 1Fam TA No 0 \n", - "3 526351010 1Fam TA No 0 \n", - "4 526353030 1Fam TA No 1 \n", - "5 527105010 1Fam TA No 0 \n", - "\n", - " Bsmt Half Bath Bsmt Qual Bsmt Unf SF BsmtFin SF 1 \\\n", - "Order PID \n", - "1 526301100 0 TA 441.0 639.0 \n", - "2 526350040 0 TA 270.0 468.0 \n", - "3 526351010 0 TA 406.0 923.0 \n", - "4 526353030 0 TA 1045.0 1065.0 \n", - "5 527105010 0 Gd 137.0 791.0 \n", - "\n", - " BsmtFin SF 2 BsmtFin Type 1 BsmtFin Type 2 Central Air \\\n", - "Order PID \n", - "1 526301100 0.0 BLQ Unf Y \n", - "2 526350040 144.0 Rec LwQ Y \n", - "3 526351010 0.0 ALQ Unf Y \n", - "4 526353030 0.0 ALQ Unf Y \n", - "5 527105010 0.0 GLQ Unf Y \n", - "\n", - " Condition 1 Condition 2 Electrical Enclosed Porch Exter Cond \\\n", - "Order PID \n", - "1 526301100 Norm Norm SBrkr 0.0 TA \n", - "2 526350040 Feedr Norm SBrkr 0.0 TA \n", - "3 526351010 Norm Norm SBrkr 0.0 TA \n", - "4 526353030 Norm Norm SBrkr 0.0 TA \n", - "5 527105010 Norm Norm SBrkr 0.0 TA \n", - "\n", - " Exter Qual Exterior 1st Exterior 2nd Fence Fireplace Qu \\\n", - "Order PID \n", - "1 526301100 TA BrkFace Plywood NA Gd \n", - "2 526350040 TA VinylSd VinylSd MnPrv NA \n", - "3 526351010 TA Wd Sdng Wd Sdng NA NA \n", - "4 526353030 Gd BrkFace BrkFace NA TA \n", - "5 527105010 TA VinylSd VinylSd MnPrv TA \n", - "\n", - " Fireplaces Foundation Full Bath Functional Garage Area \\\n", - "Order PID \n", - "1 526301100 2 CBlock 1 Typ 528.0 \n", - "2 526350040 0 CBlock 1 Typ 730.0 \n", - "3 526351010 0 CBlock 1 Typ 312.0 \n", - "4 526353030 2 CBlock 2 Typ 522.0 \n", - "5 527105010 1 PConc 2 Typ 482.0 \n", - "\n", - " Garage Cars Garage Cond Garage Finish Garage Qual \\\n", - "Order PID \n", - "1 526301100 2 TA Fin TA \n", - "2 526350040 1 TA Unf TA \n", - "3 526351010 1 TA Unf TA \n", - "4 526353030 2 TA Fin TA \n", - "5 527105010 2 TA Fin TA \n", - "\n", - " Garage Type Gr Liv Area Half Bath Heating Heating QC \\\n", - "Order PID \n", - "1 526301100 Attchd 1656.0 0 GasA Fa \n", - "2 526350040 Attchd 896.0 0 GasA TA \n", - "3 526351010 Attchd 1329.0 1 GasA TA \n", - "4 526353030 Attchd 2110.0 1 GasA Ex \n", - "5 527105010 Attchd 1629.0 1 GasA Gd \n", - "\n", - " House Style Kitchen AbvGr Kitchen Qual Land Contour \\\n", - "Order PID \n", - "1 526301100 1Story 1 TA Lvl \n", - "2 526350040 1Story 1 TA Lvl \n", - "3 526351010 1Story 1 Gd Lvl \n", - "4 526353030 1Story 1 Ex Lvl \n", - "5 527105010 2Story 1 TA Lvl \n", - "\n", - " Land Slope Lot Area Lot Config Lot Shape Low Qual Fin SF \\\n", - "Order PID \n", - "1 526301100 Gtl 31770.0 Corner IR1 0.0 \n", - "2 526350040 Gtl 11622.0 Inside Reg 0.0 \n", - "3 526351010 Gtl 14267.0 Corner IR1 0.0 \n", - "4 526353030 Gtl 11160.0 Corner Reg 0.0 \n", - "5 527105010 Gtl 13830.0 Inside IR1 0.0 \n", - "\n", - " MS SubClass MS Zoning Mas Vnr Area Mas Vnr Type Misc Feature \\\n", - "Order PID \n", - "1 526301100 020 RL 112.0 Stone NA \n", - "2 526350040 020 RH 0.0 None NA \n", - "3 526351010 020 RL 108.0 BrkFace Gar2 \n", - "4 526353030 020 RL 0.0 None NA \n", - "5 527105010 060 RL 0.0 None NA \n", - "\n", - " Misc Val Mo Sold Neighborhood Open Porch SF Overall Cond \\\n", - "Order PID \n", - "1 526301100 0.0 5 Names 62.0 5 \n", - "2 526350040 0.0 6 Names 0.0 6 \n", - "3 526351010 12500.0 6 Names 36.0 6 \n", - "4 526353030 0.0 4 Names 0.0 5 \n", - "5 527105010 0.0 3 Gilbert 34.0 5 \n", - "\n", - " Overall Qual Paved Drive Pool Area Pool QC Roof Matl \\\n", - "Order PID \n", - "1 526301100 6 P 0.0 NA CompShg \n", - "2 526350040 5 Y 0.0 NA CompShg \n", - "3 526351010 6 Y 0.0 NA CompShg \n", - "4 526353030 7 Y 0.0 NA CompShg \n", - "5 527105010 5 Y 0.0 NA CompShg \n", - "\n", - " Roof Style Sale Condition Sale Type Screen Porch Street \\\n", - "Order PID \n", - "1 526301100 Hip Normal WD 0.0 Pave \n", - "2 526350040 Gable Normal WD 120.0 Pave \n", - "3 526351010 Hip Normal WD 0.0 Pave \n", - "4 526353030 Hip Normal WD 0.0 Pave \n", - "5 527105010 Gable Normal WD 0.0 Pave \n", - "\n", - " TotRms AbvGrd Total Bsmt SF Utilities Wood Deck SF \\\n", - "Order PID \n", - "1 526301100 7 1080.0 AllPub 210.0 \n", - "2 526350040 5 882.0 AllPub 140.0 \n", - "3 526351010 6 1329.0 AllPub 393.0 \n", - "4 526353030 8 2110.0 AllPub 0.0 \n", - "5 527105010 6 928.0 AllPub 212.0 \n", - "\n", - " Year Built Year Remod/Add Yr Sold SalePrice \n", - "Order PID \n", - "1 526301100 1960 1960 2010 215000.0 \n", - "2 526350040 1961 1961 2010 105000.0 \n", - "3 526351010 1958 1958 2010 172000.0 \n", - "4 526353030 1968 1968 2010 244000.0 \n", - "5 527105010 1997 1998 2010 189900.0 " - ] - }, - "execution_count": 38, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "df.head()" - ] - }, - { - "cell_type": "code", - "execution_count": 39, - "metadata": {}, - "outputs": [], - "source": [ - "df.to_csv(\"data/data_clean.csv\")" - ] - } - ], - "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": "3.7.8" - } - }, - "nbformat": 4, - "nbformat_minor": 4 -} diff --git a/README.md b/README.md index dbf207d..4ff2676 100644 --- a/README.md +++ b/README.md @@ -15,7 +15,7 @@ A video presentation of the case study is available on [YouTube