predict.nmfre: Predict method for nmfre objects

In nmfkc: Non-Negative Matrix Factorization with Kernel Covariates

Predict method for nmfre objects

Description

Predicts the response from a fitted NMF-RE model. With newA supplied, returns the fixed-effect prediction X\,\Theta\,A_{new} (new units carry no estimated random effect U, so only the population mean is predicted). Without newA, returns the in-sample BLUP fit X(\Theta A + U) (object$XB.blup), matching fitted().

Usage

## S3 method for class 'nmfre'
predict(object, newA = NULL, ...)

Arguments

object	An object of class "nmfre" from nmfre.
newA	Optional covariate matrix (K x M) for new units. If NULL (default), the in-sample BLUP fit is returned.
...	Not used.

Value

A matrix of predicted (P x M) values.

See Also

nmfre, fitted.nmf

nmfkc documentation built on July 14, 2026, 1:07 a.m.