Skip to contents

Project a Covariance Matrix (generic)

Source: R/all_generic.R, R/covstatis.R
project_cov.Rd

Generic for projecting a new covariance matrix onto a previously fitted `covstatis` model.

Usage

project_cov(x, new_data, ...)

# S3 method for class 'covstatis'
project_cov(x, new_data, ...)

Arguments

x

A model object for which a `project_cov` method exists (e.g. an object of class `covstatis`).

new_data

A symmetric numeric matrix to be projected.

...

Additional arguments passed to methods.

Value

A numeric matrix of projected scores.

A numeric matrix of projected scores with dimensions nrow(new_data) × ncomp.

Details

For covstatis objects, a new covariance/correlation matrix is transformed using the same preprocessing steps (centering and normalization) as were applied during model fitting, then projected onto the model space. This returns the "partial factor scores" or ROI-level coordinates for the new subject.