predict.fairadapt: Prediction function for new data from a saved 'fairadapt'...

In fairadapt: Fair Data Adaptation with Quantile Preservation

Prediction function for new data from a saved fairadapt object

Description

Prediction function for new data from a saved fairadapt object

Usage

## S3 method for class 'fairadapt'
predict(object, newdata, ...)

Arguments

object	Object of class fairadapt, a result of an adaptation procedure.
newdata	A data.frame containing the new data.
...	Additional arguments forwarded to computeQuants().

Details

The newdata argument should be compatible with adapt.test argument that was used when constructing the fairadapt object. In particular, newdata should contain column names that appear in the formula argument that was used when calling fairadapt() (apart from the outcome variable on the LHS of the formula).

Value

A data.frame containing the adapted version of the new data.

Examples

n_samp <- 200
uni_dim <- c(       "gender", "edu", "test", "score")
uni_adj <- matrix(c(       0,     1,      1,       0,
                           0,     0,      1,       1,
                           0,     0,      0,       1,
                           0,     0,      0,       0),
                  ncol = length(uni_dim),
                  dimnames = rep(list(uni_dim), 2),
                  byrow = TRUE)

uni_ada <- fairadapt(score ~ .,
  train.data = head(uni_admission, n = n_samp),
  adj.mat = uni_adj,
  prot.attr = "gender"
)

predict(object = uni_ada, newdata = tail(uni_admission, n = n_samp))

fairadapt documentation built on Sept. 11, 2024, 5:51 p.m.