validate_outcomes_are_binary {hardhat}R Documentation

Ensure that the outcome has binary factors

Description

validate - asserts the following:

check - returns the following:

Usage

validate_outcomes_are_binary(outcomes)

check_outcomes_are_binary(outcomes, ..., call = caller_env())

Arguments

outcomes

An object to check.

...

These dots are for future extensions and must be empty.

call

The call used for errors and warnings.

Details

The expected way to use this validation function is to supply it the ⁠$outcomes⁠ element of the result of a call to mold().

Value

validate_outcomes_are_binary() returns outcomes invisibly.

check_outcomes_are_binary() returns a named list of three components, ok, bad_cols, and num_levels.

Validation

hardhat provides validation functions at two levels.

See Also

Other validation functions: validate_column_names(), validate_no_formula_duplication(), validate_outcomes_are_factors(), validate_outcomes_are_numeric(), validate_outcomes_are_univariate(), validate_prediction_size(), validate_predictors_are_numeric()

Examples

# Not a binary factor. 0 levels
check_outcomes_are_binary(data.frame(x = 1))

# Not a binary factor. 1 level
check_outcomes_are_binary(data.frame(x = factor("A")))

# All good
check_outcomes_are_binary(data.frame(x = factor(c("A", "B"))))

[Package hardhat version 1.4.1 Index]