Skip to main content

model_selection_ncomp

fdars_core::scalar_on_function

Function model_selection_ncomp 

Source 
pub fn model_selection_ncomp(
    data: &FdMatrix,
    y: &[f64],
    scalar_covariates: Option<&FdMatrix>,
    max_ncomp: usize,
    criterion: SelectionCriterion,
) -> Result<ModelSelectionResult, FdarError>
Expand description

Select optimal ncomp for fregre_lm using AIC, BIC, or GCV.

Fits models for ncomp = 1..=max_ncomp, collects information criteria, and returns the best ncomp under the chosen criterion.

§Arguments

  • data — Functional predictor matrix (n × m)
  • y — Scalar responses (length n)
  • scalar_covariates — Optional scalar covariates (n × p)
  • max_ncomp — Maximum number of FPC components to try
  • criterion — Which criterion to minimise

§Errors

Returns FdarError::InvalidParameter if max_ncomp is zero. Returns FdarError::ComputationFailed if no candidate ncomp produces a valid fitted model (all calls to fregre_lm fail).