crossv_kfold: Generate cross-validated K-fold test-training pairs
In jrnold/resamplr: Resampling methods

Description Usage Arguments Value Methods (by class) References See Also Examples

Generate cross-validated K-fold test-training pairs. In addition to ordinary K-fold cross-validation, this supports stratified K-fold cross validation if data is a grouped data frame and stratify = TRUE, and Group K-fold if data is a grouped data frame and stratify = FALSE.

crossv_kfold(data, K, ...)

## S3 method for class 'data.frame'
crossv_kfold(data, K = 5L, shuffle = TRUE, ...)

## S3 method for class 'grouped_df'
crossv_kfold(data, K = 5L, shuffle = TRUE,
  stratify = FALSE, ...)

`data`	A data frame
`K`	The number of folds
`...`	Arguments passed to methods
`shuffle`	If `TRUE`, randomly assign observations to folds. Otherwise, observations are sequentially assigned to folds.
`stratify`	If `TRUE`, within each group observations are split into folds, and those folds combined. If `FALSE`, groups are assigned into folds.

A data frame with K rows and the following columns:

sample: A list of resample objects. Training sets.
.id: An integer vector of identifiers.

data.frame: Splits rows in a data frame into folds.
grouped_df: Partitions rows within each group of a grouped data frame into folds if stratify = FALSE. This ensures that the test and training sets will have approximately equal proportions of each group. If stratify = TRUE, then the groups are partitioned into folds.

Breiman, L., Friedman, J.H., Olshen, R.A. and Stone, C.J. (1984) Classification and Regression Trees. Wadsworth.
Burman, P. (1989) A comparative study of ordinary cross-validation, v-fold cross-validation and repeated learning-testing methods. Biometrika, 76, 503–514
Davison, A.C. and Hinkley, D.V. (1997) Bootstrap Methods and Their Application. Cambridge University Press.
Efron, B. (1986) How biased is the apparent error rate of a prediction rule? Journal of the American Statistical Association, 81, 461–470.
Stone, M. (1974) Cross-validation choice and assessment of statistical predictions (with Discussion). Journal of the Royal Statistical Society, B, 36, 111–147.

This function has more features than the modelr function crossv_kfold.

# Example originally from modelr::crossv_mc
library("purrr")
library("dplyr")

# 5-fold cross-validation
cv1 <- crossv_kfold(mtcars, K = 5)
models <- map(cv1$train, ~ lm(mpg ~ wt, data = .))
summary(map2_dbl(models, cv1$test, modelr::rmse))

# k-fold by group
cv2 <- crossv_kfold(group_by(mtcars, cyl), K = 2)
models <- map(cv2$train, ~ lm(mpg ~ wt, data = .))
summary(map2_dbl(models, cv2$test, modelr::rmse))

# stratified k-fold
cv3 <- crossv_kfold(group_by(mtcars, am), K = 3, stratify = TRUE)
models <- map(cv3$train, ~ lm(mpg ~ wt, data = .))
summary(map2_dbl(models, cv3$test, modelr::rmse))