Imputation in R

Question

Is there any package in R which helps to replace NA values in a column with different criteria such as 'mean' of column, 'median' of column, 'maximum' value in column and so on.

Answer 1

You can work with the "Hmisc" package, which helps you to impute columns.

Let's take this data-frame:

DF <- data.frame(cost = c(10, 20, NA, 40), fruit = c('apple','guava','banana','orange'))

Imputing with mean value:

with(DF, impute(cost, mean))

Imputing with median value:

with(DF, impute(cost, median))

Imputing with random value:

with(DF, impute(cost, 'random'))

Imputing with minimum value:

with(DF, impute(cost, min))

Imputing with maximum value:

with(DF, impute(cost, max))