mlr (version 2.17.0)

reimpute: Re-impute a data set

Description

This function accepts a data frame or a task and an imputation description as returned by impute to perform the following actions:

  1. Restore dropped columns, setting them to NA

  2. Add dummy variables for columns as specified in impute

  3. Optionally check factors for new levels to treat them as NAs

  4. Reorder factor levels to ensure identical integer representation as before

  5. Impute missing values using previously collected data

Usage

reimpute(obj, desc)

Arguments

obj

(data.frame | Task) Input data.

desc

(ImputationDesc) Imputation description as returned by impute.

Value

Imputated data.frame or task with imputed data.

See Also

Other impute: imputations, impute(), makeImputeMethod(), makeImputeWrapper()