Package 'makedummies'

Create Dummy Variables from Categorical Data

Description

Create dummy variables from categorical data. This package can convert categorical data (factor and ordered) into dummy variables and handle multiple columns simultaneously. This package enables to select whether a dummy variable for base group is included (for principal component analysis/factor analysis) or excluded (for regression analysis) by an option. makedummies function accepts data.frame, matrix, and tbl (tibble) class (by tibble package). matrix class data is automatically converted to data.frame class.

Usage

makedummies(dat, ...)

## Default S3 method:
makedummies(dat, basal_level = FALSE, col = NULL,
  numerical = NULL, as.is = NULL, ...)

## S3 method for class 'matrix'
makedummies(dat, ...)

## S3 method for class 'tbl'
makedummies(dat, basal_level = FALSE, col = NULL,
  numerical = NULL, as.is = NULL, ...)

Arguments

dat	data of data.frame, matrix, or tbl class
...	arguments to makedummies.data.frame (tbl class)
basal_level	logical TRUE : include a dummy variable for base group FALSE (default) : exclude a dummy variable for base group
col	Columns vector (all columns are used if NULL is given)
numerical	Columns vector converting from factor/ordered to numeric (ignore if column is numeric)
as.is	Columns vector not converting

Value

return as data.frame or tbl class

Note

Pull Request #1 (add column name when when columns has binary value) (https://github.com/toshi-ara/makedummies/pull/1). Thanks to Kohki YAMAGIWA for the contribution.

Examples

#### 'data.frame' class
## factor
dat <- data.frame(x = factor(rep(c("a", "b", "c"), each = 3)))
dat$x
makedummies(dat)

## ordered
dat <- data.frame(x = factor(rep(c("a", "b", "c"), each = 3)))
dat$x <- ordered(dat$x, levels = c("a" ,"c" ,"b"))
dat$x
makedummies(dat)

## numeric
dat <- data.frame(x = rep(1:3, each = 3))
makedummies(dat)

## factor and numeric
dat <- data.frame(
    x = factor(rep(c("a", "b", "c"), each = 3)),
    y = rep(1:3, each = 3)
)
makedummies(dat)

## factors
dat <- data.frame(
    x = factor(rep(c("a", "b", "c"), each = 3)),
    y = factor(rep(1:3, each = 3))
)
makedummies(dat)

## data including NA

dat <- data.frame(
    x = factor(rep(c("a", "b", "c"), each = 3)),
    y = rep(1:3, each = 3)
)
dat$x[4] <- NA; dat$y[6] <- NA
makedummies(dat)

## "col" option
dat <- data.frame(
    x = factor(rep(c("a", "b", "c"), each = 3)),
    y = factor(rep(1:3, each = 3))
)
makedummies(dat, col = "x")

## "numerical" option
dat <- data.frame(
    x = factor(rep(c("a", "b", "c"), each = 3)),
    y = factor(rep(1:3, each = 3))
)
makedummies(dat, numeric = "x")

dat <- data.frame(
    x = factor(rep(c("a", "b", "c"), each = 3)),
    y = rep(4:6, each = 3)
)
dat$x <- ordered(dat$x, levels = c("a" ,"c" ,"b"))
dat
dat$x
makedummies(dat, numeric = c("x", "y"))

## "as.is" option
dat <- data.frame(
    x = factor(rep(c("a", "b", "c"), each = 3)),
    y = factor(rep(1:3, each = 3))
)
dat
makedummies(dat, as.is = "x")
makedummies(dat, as.is = c("x", "y"))

#### 'tibble' class
if (require(tibble)) {
  dat <- as_tibble(iris)
  makedummies(dat[46:55,], col = "Species", basal_level = TRUE)

  # non-standard variable name
  dat2 <- tibble(
      `1` = factor(rep(c("c", "a", "b"), each = 3)),
      `@` = factor(rep(1:3, each = 3)),
      `&` = rep(4:6, each = 3)
  )
  dat2

  makedummies(dat2, basal_level = TRUE)
  makedummies(dat2, as.is = "@", basal_level = TRUE)
  makedummies(dat2, numerical = "1", basal_level = TRUE)
}

---