imputation.train: Impute training data

In MarcoNiemann/kaggle_house: Analysing House Prices in Ames, Iowa and Build a Sales Price Prediction Model

Description

Imputation function for the training dataset using package mice and manual substitution to remove missing values.

Usage

imputation.train(data = data_train_numeric_clean)

Arguments

data	Input data which is set by default to data_train_numeric_clean

Details

Since some features are not missing at random imputation is not an appropriate approach. Therefore the features MasVnrArea, LotFrontage, Electrical and GarafeYrBlt are manually imputed. The remaining features that still contain at least one missing value are solely imputed using mice with 50 maximum iterations with seed 500 using predictive-mean matching. After the execution one can access the variable data_train_numeric_clean_imputed.

MarcoNiemann/kaggle_house documentation built on May 7, 2019, 2:50 p.m.