crossvalidation {kyotil}

Cross Validation Functions

Description

Cross validation utility functions.

Usage

sample.for.cv (dat, v, seed)
get.kfold.splits (dat, k, seed)
kfold.split (k, n1, n0)
ran.kfold.split(k, n1, n0, replicates)
lpo.split(n1, n0)
get.splits (dat, cv.scheme=c("LPO","5fold","50xrandom4:1"), seed) 

Arguments

dat

a data frame. One of the columns must be named y and y should be 0/1 with 1 for case and 0 for control

v

v-fold cross validation

seed

seed for random number generators

k

var.equal

n1

var.equal

n0

var.equal

replicates

var.equal

cv.scheme

var.equal

Details

sample.for.cv: case and controls are sampled separately.

Value

sample.for.cv returns a list of two vector of integers: train and test, which refer to the rows of dat