cv_grouped_dat: Create (train,test) grouped data sets from a grouped data set

In RomeroLab/pudms: Positive-Unlabeled Learning for the Analysis of Deep Mutational Scanning Datasets

Description

Create (train,test) grouped data sets from a grouped data set

Usage

cv_grouped_dat(
  grouped_dat,
  test_idx,
  nfolds = 10,
  seed = 1,
  folds = NULL,
  return.datasets = TRUE
)

Arguments

grouped_dat	a grouped data set
test_idx	an integer from 1 to nfolds
nfolds	number of folds
seed	seed for the reproducibility
folds	NULL or a list of (cv_labeled,cv_unlabeled). If NULL, cv_labeled and cv_unlabeled are generated randomly, then train/test sets are created. If not, train/test data sets are created from previous cv_labeled and cv_unlabeled list
return.datasets	a logical value; if TRUE, test_grouped_dat / train_grouped_dat will be returned

Value

a list containing (train_grouped_dat, test_grouped_dat,folds = (cv_labeled, cv_unlabeled))

RomeroLab/pudms documentation built on Jan. 2, 2021, 5:10 a.m.