Struct FofReResult 

#[non_exhaustive]
pub struct FofReResult {

    pub intercept: Vec<f64>,
    pub beta_surface: FdMatrix,
    pub fitted: FdMatrix,
    pub residuals: FdMatrix,
    pub r_squared_t: Vec<f64>,
    pub r_squared: f64,
    pub ncomp_x: usize,
    pub ncomp_y: usize,
    pub fpca_x: FpcaResult,
    pub fpca_y: FpcaResult,
    pub coef_matrix: FdMatrix,
    pub random_effects: FdMatrix,
    pub sigma2_u: Vec<f64>,
    pub sigma2_eps: f64,
    pub n_subjects: usize,
}

Result of a random-effects function-on-function regression.

Extends FofResult with subject-level random intercepts on Y-score components, enabling prediction for grouped/longitudinal functional data.

Divergence from refund::pffr

R’s pffr(y ~ pcre(x)) uses a penalized-spline GAMM (mgcv) backend with functional random effects represented in a spline basis. fof_re_regression instead uses the FPC-score parametrization: both X and Y are decomposed into FPC scores (fdata_to_pc_1d), and for each Y-score component a scalar linear mixed model (REML EM, reusing famm::fit_scalar_mixed_model) is fitted with the X-scores as fixed-effect covariates and subject IDs as the grouping factor. No basis penalties are applied — this matches the locked design decision for Phase 32.

Fields (Non-exhaustive)

Non-exhaustive structs could have additional fields added in future. Therefore, non-exhaustive structs cannot be constructed in external crates using the traditional Struct { .. } syntax; cannot be matched against without a wildcard ..; and struct update syntax will not work.

intercept: Vec<f64>

Intercept function α(s) = mean_y(s) (length m_y)

beta_surface: FdMatrix

Coefficient surface β(s,t) stored as (m_y × m_x) matrix

fitted: FdMatrix

Fitted response curves (n × m_y), including fixed + random effects

residuals: FdMatrix

Residual curves (n × m_y)

r_squared_t: Vec<f64>

R-squared per response grid point (length m_y)

r_squared: f64

Overall R-squared (mean of pointwise R-squared)

ncomp_x: usize

Number of predictor FPC components used

ncomp_y: usize

Number of response FPC components used

fpca_x: FpcaResult

FPCA of predictor (carried for prediction)

fpca_y: FpcaResult

FPCA of response (carried for reconstruction)

coef_matrix: FdMatrix

Fixed-effect coefficient matrix B (ncomp_x × ncomp_y): Y-scores ≈ X-scores * B

random_effects: FdMatrix

Subject-level random intercept functions (n_subjects × m_y)

sigma2_u: Vec<f64>

Per-Y-component random-intercept variance (length ncomp_y)

sigma2_eps: f64

Mean residual variance across Y-score components

n_subjects: usize

Number of unique subjects

Trait Implementations

impl Clone for FofReResult

fn clone(&self) -> FofReResult

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for FofReResult

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl PartialEq for FofReResult

fn eq(&self, other: &FofReResult) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl StructuralPartialEq for FofReResult