sterm_control: Controls for smooth terms

In neural-structured-additive-learning/sparsedistreg: Fitting Deep Conditional Transformation Models

Controls for smooth terms

Description

Controls for smooth terms

Usage

sterm_control(
  sterm_default = NULL,
  teterm_default = function(x, y, la) paste0("sparse(ti(", x, ", ", y, ", la = ", la,
    "))"),
  penalize_byterm = NULL,
  spline_type = c("tp", "ps")
)

Arguments

sterm_default	default function for creating univariate effects
teterm_default	default function for creating bivariate effects
penalize_byterm	if not NULL but positive value, the linear effect in by-terms of smooth effects will be penalized

Value

list of options

neural-structured-additive-learning/sparsedistreg documentation built on May 13, 2022, 3:56 a.m.