group_vfold_cv: Group V-Fold Cross-Validation

In rsample: General Resampling Infrastructure

Group V-Fold Cross-Validation

Description

Group V-fold cross-validation creates splits of the data based on some grouping variable (which may have more than a single row associated with it). The function can create as many splits as there are unique values of the grouping variable or it can create a smaller set of splits where more than one group is left out at a time. A common use of this kind of resampling is when you have repeated measures of the same subject.

Usage

group_vfold_cv(
  data,
  group = NULL,
  v = NULL,
  repeats = 1,
  balance = c("groups", "observations"),
  ...,
  strata = NULL,
  pool = 0.1
)

Arguments

data	A data frame.
group	A variable in data (single character or name) used for grouping observations with the same value to either the analysis or assessment set within a fold.
v	The number of partitions of the data set. If left as NULL (the default), v will be set to the number of unique values in the grouping variable, creating "leave-one-group-out" splits.
repeats	The number of times to repeat the V-fold partitioning.
balance	If v is less than the number of unique groups, how should groups be combined into folds? Should be one of "groups", which will assign roughly the same number of groups to each fold, or "observations", which will assign roughly the same number of observations to each fold.
...	These dots are for future extensions and must be empty.
strata	A variable in data (single character or name) used to conduct stratified sampling. When not NULL, each resample is created within the stratification variable. Numeric strata are binned into quartiles.
pool	A proportion of data used to determine if a particular group is too small and should be pooled into another group. We do not recommend decreasing this argument below its default of 0.1 because of the dangers of stratifying groups that are too small.

Value

A tibble with classes group_vfold_cv, rset, tbl_df, tbl, and data.frame. The results include a column for the data split objects and an identification variable.

Examples

data(ames, package = "modeldata")

set.seed(123)
group_vfold_cv(ames, group = Neighborhood, v = 5)
group_vfold_cv(
  ames,
  group = Neighborhood,
  v = 5,
  balance = "observations"
)
group_vfold_cv(ames, group = Neighborhood, v = 5, repeats = 2)

# Leave-one-group-out CV
group_vfold_cv(ames, group = Neighborhood)

library(dplyr)
data(Sacramento, package = "modeldata")

city_strata <- Sacramento %>%
  group_by(city) %>%
  summarize(strata = mean(price)) %>%
  summarize(city = city,
            strata = cut(strata, quantile(strata), include.lowest = TRUE))

sacramento_data <- Sacramento %>%
  full_join(city_strata, by = "city")

group_vfold_cv(sacramento_data, city, strata = strata)

rsample documentation built on May 29, 2024, 11:03 a.m.