Extract a prototype — extract

extract_ptype() extracts a tibble with 0 rows from data. This contains all of the required information about column names, classes, and factor levels that are required to check the structure of new data at prediction time.

Usage

extract_ptype(data, ..., call = current_env())

Arguments

data: A data frame or matrix.
...: These dots are for future extensions and must be empty.
call: The call used for errors and warnings.

Value

A 0 row slice of data after converting it to a tibble.

Details

extract_ptype() is useful when creating a new preprocessing blueprint. It extracts the required information that will be used by the validation checks at prediction time.

Examples


hardhat:::extract_ptype(iris)
#> # A tibble: 0 × 5
#> # ℹ 5 variables: Sepal.Length <dbl>, Sepal.Width <dbl>,
#> #   Petal.Length <dbl>, Petal.Width <dbl>, Species <fct>