Skip to contents

Per-effect coefficient curves on the original X scale

Source: R/mfsusie_methods.R
coef.mfsusie.Rd

Returns the per-effect, per-outcome coefficient curves on the original (unstandardized) X scale, projected back through the inverse wavelet transform to the original position grid.

Usage

# S3 method for class 'mfsusie'
coef(object, smooth_method = NULL, ...)

Arguments

object

an mfsusie fit.

smooth_method

optional character. When supplied, returns the post-smoothed effect curves stored at object$smoothed[[smooth_method]] instead of the raw wavelet-inverse curves. The named smoother must have been applied via mf_post_smooth(object, method = smooth_method) beforehand. When NULL (default), coef.mfsusie returns the raw curves; smoothing is an explicit opt-in.

...

ignored.

Value

list of length M; each element an L x T_m matrix whose row l is effect l's coefficient curve for outcome m. When smooth_method is supplied, the returned curves are on the same shape but smoothed by the named method; the returned object carries an attribute attr(., "smooth_method") recording the choice.