pipe_check: Create a pipeline step that learns what the data looks like
In jeroenvdhoven/datapiper: datapiper

Create a pipeline step that learns what the data looks like

1
2
3

pipe_check(train, response, on_missing_column = c("error", "add")[1],
  on_extra_column = c("remove", "error")[1],
  on_type_error = c("ignore", "error")[1])

`train`	Data frame containing the train data.
`response`	The response variable. Will be used as an optional column name. Does not have to exist in the train dataset (useful when it is added in later in the pipeline)
`on_missing_column`	What to do when a new dataset misses columns. Either "error", which causes an error, or "add", which adds the missing columns with only NA's filled in.
`on_extra_column`	What to do when a new dataset has extra column.s Either "error", which causes an error, or "remove", which removes the extra columns.
`on_type_error`	What to do when a new dataset causes warnings or errors on casting columns to new types. Either "error", which causes an error, or "ignore", which ignores warnings but will still allow errors to propagate.