scorecard (version 0.4.4)

split_df: Split a Data Frame

Description

Split a data frame into multiple datasets according to the specified ratios.

Usage

split_df(dt, y = NULL, ratios = c(0.7, 0.3), name_dfs = c("train",
  "test"), oot = list(time_col = NULL, time_start = NULL, ratio = NULL),
  seed = 618, ...)

Value

A list of data frames

Arguments

dt

A data frame.

y

Name of y variable, Defaults to NULL. The input data will split based on the predictor y, if it is provide.

ratios

A numeric vector indicating the ratio of total rows contained in each split, defaults to c(0.7, 0.3).

name_dfs

Name of returned data frames. Its length should equals to the ratios'. Defaults to train and test.

oot

The out-of-time validation dataset parameters. The parameters of time_cols and either time_start or ratio need to be supplied.

seed

A random seed, Defaults to 618.

...

Additional parameters.

Examples

# load German credit data
data(germancredit)

# Example I
dt_list = split_df(germancredit, y="creditability")

# dimensions of each split data sets
lapply(dt_list, dim)


# Example II
dt_list2 = split_df(germancredit, y="creditability",
  ratios = c(0.5, 0.3, 0.2),
  name_dfs = c('train', 'test', 'valid'))
lapply(dt_list2, dim)