Compute one reduced-model information criterion
Usage
compute_ic(
y,
Z,
criterion = c("cicomp", "icomp", "aic", "caic", "bic"),
response_type = c("continuous", "categorical", "survival"),
delta = NULL,
n_active = NULL,
support_penalty = FALSE,
tol = 1e-10
)
Arguments
- y
Response vector.
- Z
Reduced predictors.
- criterion
Criterion name.
- response_type
Response type.
- delta
Optional event indicator.
- n_active
Optional active-support size.
- support_penalty
Include active-support size in effective df.
- tol
Numerical tolerance.
Value
Scalar criterion value.