pre_impute: Basic imputation

In emil: Evaluation of Modeling without Information Leakage

Description

This solution is optimized for the scenario that the dataset is very large but only contains missing values in a small number of columns.

Usage

pre_impute(data, fun, ...)

pre_impute_median(data)

pre_impute_mean(data)

Arguments

data	Fitting and test datasets, as returned by pre_split or any other standard pre-processing function.
fun	Function for calculating imputation values. Should take a vector and return a scalar.
...	Sent to fun.

Value

A pair of fitting and testing datasets.

Author(s)

Christofer Bäcklin

emil documentation built on Aug. 1, 2018, 1:03 a.m.