mse: Mean squared error of factor model

In zdebruine/RcppML: Rcpp Machine Learning Library

Mean squared error of factor model

Description

Same as the evaluate S4 method for the nmf class, but allows one to input the 'w', 'd', 'h', and 'data' independently.

Usage

mse(w, d = NULL, h, data, mask = NULL, missing_only = FALSE, ...)

Arguments

w	feature factor matrix (features as rows)
d	scaling diagonal vector (if applicable)
h	sample factor matrix (samples as columns)
data	dense or sparse matrix of features in rows and samples in columns. Prefer matrix or Matrix::dgCMatrix, respectively
mask	dense or sparse matrix of values in data to handle as missing. Prefer Matrix::dgCMatrix. Alternatively, specify "zeros" or "NA" to mask either all zeros or NA values.
missing_only	only calculate mean squared error at masked values
...	additional arguments

zdebruine/RcppML documentation built on Sept. 13, 2023, 11:44 p.m.