AIC: AIC calculation for the output of 'SIMP()'

View source: R/model_selection.R

AICR Documentation

AIC calculation for the output of SIMP()

Description

AIC calculation for the output of SIMP()

Usage

AIC(SIMP.fit, r, pc, pd, p2, dx, dy, combined = T, Chain.no = 1)

Arguments

SIMP.fit

The output of SIMP().

r

Dimension of response Y.

pc

Dimension of X1C, the continuous part of the predictors of interest.

pd

Dimension of X1D, the discrete part of the predictors of interest.

p2

Dimension of X2, predictors of not main interest.

dx

Partial predictor envelope dimension of SIMP.

dy

Partial response envelope dimension of SIMP.

combined

Logical. Indicate whether posterior samples (after burn-in) from all chains should be combined to calculate AIC. If there is only one chain, then there is no difference by indicating this parameter to be TRUE or FALSE.

Chain.no

If combined = FALSE, we need to indicate which chain should be used for the calculation of AIC. Chain.no must be an integer between 1 and the number of chains.


yanbowisc/SIMP documentation built on Oct. 30, 2022, 1:33 a.m.