fct_encode_one_hot {hardhat}

Encode a factor as a one-hot indicator matrix

Description

fct_encode_one_hot() encodes a factor as a one-hot indicator matrix.

This matrix consists of length(x) rows and length(levels(x)) columns. Every value in row i of the matrix is filled with 0L except for the column that has the same name as x[[i]], which is instead filled with 1L.

Usage

fct_encode_one_hot(x)

Arguments

x

A factor.

x can't contain missing values.

x is allowed to be an ordered factor.

Details

The columns are returned in the same order as levels(x).

If x has names, the names are propagated onto the result as the row names.

Value

An integer matrix with length(x) rows and length(levels(x)) columns.

Examples

fct_encode_one_hot(factor(letters))

fct_encode_one_hot(factor(letters[1:2], levels = letters))

set.seed(1234)
fct_encode_one_hot(factor(sample(letters[1:4], 10, TRUE)))