one_hot: One-Hot Encoding for Categorical Data

In kbal: Kernel Balancing

One-Hot Encoding for Categorical Data

Description

Converts raw categorical string/factor data matrix into numeric one-hot encoded data matrix. Intended to help prepare data to be passed to kbal argument allx when categorical data is used.

Usage

one_hot(data)

Arguments

data	a dataframe or matrix where columns are string or factor type covariates

Value

onehot_data

a matrix of combined sample and population data with rows corresponding to units and columns one-hot encoded categorical covariates

Examples

#Ex 1. Make up some categorical demographic data
dat = data.frame(pid = c(rep("Rep", 20),
                         rep("Dem", 20), 
                         rep("Ind", 20)), 
                 gender = c(rep("female", 35),
                            rep("male", 25)))
#Convert to one-hot encoded data matrix:
onehot_dat = one_hot(dat)

#Ex 2. lalonde data
set.seed(123)
data("lalonde")
# Select a random subset of 500 rows
lalonde_sample <- sample(1:nrow(lalonde), 500, replace = FALSE)
lalonde <- lalonde[lalonde_sample, ]

cat_vars=c("black","hisp","married","nodegr","u74","u75")
onehot_lalonde = one_hot(lalonde[, cat_vars])

kbal documentation built on April 3, 2025, 6:04 p.m.