frequency_weights: Frequency weights

In hardhat: Construct Modeling Packages

Frequency weights

Description

frequency_weights() creates a vector of frequency weights which allow you to compactly repeat an observation a set number of times. Frequency weights are supplied as a non-negative integer vector, where only whole numbers are allowed.

Usage

frequency_weights(x)

Arguments

x	An integer vector.

Details

Frequency weights are integers that denote how many times a particular row of the data has been observed. They help compress redundant rows into a single entry.

In tidymodels, frequency weights are used for all parts of the preprocessing, model fitting, and performance estimation operations.

Value

A new frequency weights vector.

See Also

importance_weights()

Examples

# Record that the first observation has 10 replicates, the second has 12
# replicates, and so on
frequency_weights(c(10, 12, 2, 1))

# Fractional values are not allowed
try(frequency_weights(c(1.5, 2.3, 10)))

hardhat documentation built on March 31, 2023, 10:21 p.m.