add_missingness: Add missing values to a predictor matrix

In missPLS: Methods and Reproducible Workflows for Partial Least Squares with Missing Data

Add missing values to a predictor matrix

Description

Create MCAR or MAR missingness on the predictor matrix x. Missingness is generated column-wise so that each predictor receives approximately the same missing-data proportion, matching the simulation strategy used in the original work.

Usage

add_missingness(
  x,
  y,
  mechanism = c("MCAR", "MAR"),
  missing_prop,
  seed = NULL,
  mar_y_bias = 0.8
)

Arguments

x	Predictor matrix or data frame.
y	Numeric response vector.
mechanism	Missingness mechanism: "MCAR" or "MAR".
missing_prop	Missing-data proportion as a fraction (0.05) or a percentage (5).
seed	Optional random seed. If supplied, it is used only for this call.
mar_y_bias	Proportion of missing values assigned to the upper half of the observed y values under the MAR mechanism.

Value

A list with components x_incomplete, missing_mask, missing_prop, mechanism, and seed.

Examples

sim <- simulate_pls_data(n = 20, p = 10, true_ncomp = 2, seed = 1)
miss <- add_missingness(sim$x, sim$y, mechanism = "MCAR", missing_prop = 10, seed = 2)
mean(is.na(miss$x_incomplete))

missPLS documentation built on April 30, 2026, 9:09 a.m.