cCV: Generate phenotype vectors for cross validation

In cpgen: Parallelized Genomic Prediction and GWAS

Description

This function takes a phenotype vector and generates folds * reps masked vectors for cross validation. Every vector has as many additional missing values as length(y) / folds.

Usage

cCV(y,folds=5,reps=1,matrix=FALSE,seed=NULL)

Arguments

y	vector of phenotypes - may already contain missing values
folds	integer, number of folds
reps	integer, number of replications
matrix	boolean, if TRUE function returns a matrix rather than a list
seed	numeric scalar, seed for sample

Value

List (matrix) with as many items (columns) as folds * reps

See Also

clmm, get_pred, get_cor

Examples

## Not run: 
# generate random data
rand_data(500,5000)

y_CV <- cCV(y,folds=5,reps=20)

## End(Not run)

Example output

Loading required package: Matrix
Loading required package: pedigreemm
Loading required package: lme4

cpgen documentation built on May 2, 2019, 8:15 a.m.