get_2ndDerivPenalty_wrapper: Wrapper for Smoothing Spline Penalty Computation

Description

Computes smoothing spline penalty matrix with optional parallel processing. Calls get_2ndDerivPenalty after processing spline vs. nonspline terms and preparing for parallel if desired.

Usage

get_2ndDerivPenalty_wrapper(
  K,
  colnm_expansions,
  C,
  power1_cols,
  power2_cols,
  power3_cols,
  power4_cols,
  interaction_single_cols,
  interaction_quad_cols,
  triplet_cols,
  nonspline_cols,
  nc,
  parallel,
  cl
)

Value

A \(p \times p\) penalty matrix for smoothing spline regularization containing the elementwise sum of the integrated squared second derivative of the fitted function with respect to predictors of interest.

Function is exported for reference purposes - use at your own risk!

Arguments

K: Number of partitions (\(K+1\))
colnm_expansions: Column names of basis expansions
C: Basis expansion matrix of two rows, first of all maximums, second of all minimums, for all variables of interest = rbind(apply(C, 2, max)), rbind(apply(C, 2, min))) for cubic expansions "C"
power1_cols: Linear term columns
power2_cols: Quadratic term columns
power3_cols: Cubic term columns
power4_cols: Quartic term columns
interaction_single_cols: Single interaction columns
interaction_quad_cols: Quadratic interaction columns
triplet_cols: Triplet interaction columns
nonspline_cols: Predictors not treated as spline effects
nc: Number of cubic expansions
parallel: Logical to enable parallel processing
cl: Cluster object for parallel computation