ezr.split_data: Split Data
In jmp1989/easyr: Helpful wrappers for common EDA, Data Manipulation, & Modeling

Seperate data into train and test datasets. Data can be returned in a single dataset or in multiple

1
2
3

ezr.split_data(dataset, perc = c(0.75), strata = NULL,
  return_as_single_df = FALSE, seed = 2019,
  datasplit_identifiers = c("train", "test", "valid"))

`dataset`	dataset. H2o or regular.
`strata`	Default is NULL. This is for startified sampling. Not valid for h2o dataframes.
`return_as_single_df`	Return as a single dataframe.
`prop`	A vector of values. Only a single valid is needed for standard train/test split. If a 2nd value is entered then a valid dataset will be created.
`datasplit_identifiers.`	Assumed to be in this order: train/test/valid. You may call things otherwise, but the returned dataset may be named differently.