impute_cells: Impute missing cells of grid

In iml: Interpretable Machine Learning

Impute missing cells of grid

Description

by default assumes first column of cell.dat is x1 and second is x2 leave grid1 NULL if feature x1 is a factor the difference variable has to be named .yhat.diff

Usage

impute_cells(cell.dat, grid1 = NULL, grid2, x1.ind = 1, x2.ind = 2)

Arguments

cell.dat	data.table with at least 4 columns: .yhat.diff and the two interval indices. Make sure that empty cells are also included and cell.dat is not the sparse representation.
grid1	data.frame where each row is the actual value for a given interval index for feature 1. If empty impute_cells assumes that the feature is categorical (factor).
grid2	data.frame where each row is the actual value for a given interval index for feature 2
x1.ind	column number or name of cell.dat for feature 1. If one feature is categorical, has to be x1
x2.ind	column number or name of cell.dat for feature 2

iml documentation built on May 29, 2024, 1:59 a.m.