Balance samples across groups — balance_samples • blueterra

Down-samples or samples with replacement so each group has the same number of rows.

Usage

balance_samples(data, group, n = NULL, replace = FALSE, seed = NULL)

Arguments

data: A data frame.
group: Character name of the grouping column.
n: Optional number of rows per group. Defaults to the smallest group size when replace = FALSE.
replace: Logical. Sample with replacement.
seed: Optional random seed.

Value

A tibble.

See also

prepare_model_matrix()

Examples

df <- data.frame(group = rep(c("a", "b"), c(2, 5)), value = seq_len(7))
balance_samples(df, group = "group")
#> # A tibble: 4 × 2
#>   group value
#>   <chr> <int>
#> 1 a         1
#> 2 a         2
#> 3 b         6
#> 4 b         5