dot-bf_case_glm_no_corr: GLM Without GEE: Damped Newton-Raphson + Backfitting (Case d)

.bf_case_glm_no_corrR Documentation

GLM Without GEE: Damped Newton-Raphson + Backfitting (Case d)

Description

Damped Newton-Raphson outer loop wrapping the weighted backfitting inner loop. Each damped Newton-Raphson step computes weights from the current linear predictor, then calls .bf_inner_weighted.

Usage

.bf_case_glm_no_corr(
  X,
  y,
  K,
  p_expansions,
  split,
  family,
  order_list,
  glm_weight_function,
  need_dispersion_for_estimation,
  dispersion_function,
  observation_weights,
  iterate,
  tol,
  max_backfit_iter,
  parallel_eigen,
  parallel_qr = FALSE,
  cl,
  chunk_size,
  num_chunks,
  rem_chunks,
  schur_zero,
  unique_penalty_per_partition,
  VhalfInv,
  verbose,
  constraint_values = list(),
  ...
)

Arguments

X

List of K+1 design matrices (N_k \times p_expansions). Unwhitened even when GEE.

y

List of K+1 response vectors. Unwhitened even when GEE.

K

Integer; number of interior knots.

p_expansions

Integer; number of coefficients per partition.

split

Output of .bf_split_components.

family

GLM family object.

order_list

List of observation index vectors by partition.

glm_weight_function

GLM weight function.

need_dispersion_for_estimation

Logical.

dispersion_function

Dispersion estimation function.

observation_weights

List of observation weights.

iterate

Logical; if FALSE, single pass (no iteration).

tol

Convergence tolerance.

max_backfit_iter

Integer.

parallel_eigen, cl, chunk_size, num_chunks, rem_chunks

Parallel arguments.

schur_zero

List of zeros.

unique_penalty_per_partition

Logical.

VhalfInv

Inverse square root of the working correlation matrix in the original observation ordering. When both are non-NULL, the GEE paths are used.

verbose

Logical.

constraint_values

List of constraint right-hand sides.

...

Additional arguments passed to weight and dispersion functions.

Value

A named list with beta_spline and beta_flat.


lgspline documentation built on Aug. 5, 2026, 1:10 a.m.