Skip to contents

The universal output of MvaMethod$fit(). Every analysis method must produce this shape, mapping its native output to these fields.

Usage

new_mva_decomposition(
  feature_weights,
  design_weights = NULL,
  importance,
  scores_feature = NULL,
  scores_design = NULL,
  extra = list(),
  method = NULL
)

Arguments

feature_weights

P x K matrix of feature loadings (brain weights)

design_weights

D x K matrix of design/behavior loadings, or NULL

importance

Length-K vector (singular values, eigenvalues, etc.)

scores_feature

N x K matrix of feature-space scores (brain scores)

scores_design

N x K matrix of design-space scores, or NULL

extra

Named list of method-specific extras

method

Character string identifying the method

Value

An mva_decomposition object