Practical Predictive Analytics

Imputing missing values using the 'mice' package

In this example we will use the mice package to impute some missing values for the age variable in the all.df dataframe. The value of the age variable will be imputed by two other existing variables: gender and education.

To begin, install and load the mice package:

install.packages("mice") library(mice)  

We will now run the md.pattern() function, which will show you the distribution of the missing values over the other columns in the dataframe. The md.pattern() function output is useful for suggesting which variables might be good candidates to use for imputing the missing values:

md.pattern(all.df) 

The output from md.pattern() function is shown later. Each row shows a count of observation ...