rNCV.perm: rNCV.perm is identical to rNCV except do instead of dopar

In kforthman/caretStack: Stacks models from the caret package to create a more robust prediction

Description

2018-08-07 version: use ensemble predictions from rNCV for permutation ###

Usage

rNCV.perm(
  data,
  resp.var,
  ref.lv = NULL,
  nRep,
  nFolds.outer,
  ML.methods,
  control,
  tuneL,
  preProc.opt,
  metric,
  dir.path,
  file.root,
  stack.method = "wt.avg",
  weighted.by = NULL,
  stack.wt = NULL,
  control.stack = NULL
)

Details

Parallel at k-fold CV (previous version in Replicates) ##

rNCV.perm is identical to rNCV except (1) do instead of dopar (parallel for permutation replicates) (2) res contains only predictive values/probabilities (i.e. y.pred.comb) (3) use ensemble predictions for model performance in rNCV() and rNCV.perm()

kforthman/caretStack documentation built on June 21, 2021, 8:38 a.m.