bootnComp: Bootstrap to find optimal number of components
Description
Combined bootstrap and cross validation procedure to find optimal number of
PLS components
Usage
bootnComp(X, y, R = 99, plotting = FALSE)
Value
Including other information in a list, the optimal number of
components
Arguments
- X
predictors as a matrix
- y
response
- R
number of bootstrap replicates
- plotting
if TRUE, a diagnostic plot is drawn for each bootstrap
replicate
Details
Heavily used internally in function impRZilr.