labelled (version 2.11.0)

to_factor: Convert input to a factor.

Description

The base function base::as.factor() is not a generic, but this variant is. By default, to_factor() is a wrapper for base::as.factor(). Please note that to_factor() differs slightly from haven::as_factor() method provided by haven package.

unlabelled(x) is a shortcut for to_factor(x, strict = TRUE, unclass = TRUE, labelled_only = TRUE).

Usage

to_factor(x, ...)
# S3 method for haven_labelled
to_factor(
  x,
  levels = c("labels", "values", "prefixed"),
  ordered = FALSE,
  nolabel_to_na = FALSE,
  sort_levels = c("auto", "none", "labels", "values"),
  decreasing = FALSE,
  drop_unused_labels = FALSE,
  user_na_to_na = FALSE,
  strict = FALSE,
  unclass = FALSE,
  explicit_tagged_na = FALSE,
  ...
)
# S3 method for data.frame
to_factor(
  x,
  levels = c("labels", "values", "prefixed"),
  ordered = FALSE,
  nolabel_to_na = FALSE,
  sort_levels = c("auto", "none", "labels", "values"),
  decreasing = FALSE,
  labelled_only = TRUE,
  drop_unused_labels = FALSE,
  strict = FALSE,
  unclass = FALSE,
  explicit_tagged_na = FALSE,
  ...
)
unlabelled(x, ...)

Arguments

x

Object to coerce to a factor.

...

Other arguments passed down to method.

levels

What should be used for the factor levels: the labels, the values or labels prefixed with values?

ordered

TRUE for ordinal factors, FALSE (default) for nominal factors.

nolabel_to_na

Should values with no label be converted to NA?

sort_levels

How the factor levels should be sorted? (see Details)

decreasing

Should levels be sorted in decreasing order?

drop_unused_labels

Should unused value labels be dropped? (applied only if strict = FALSE)

user_na_to_na

Convert user defined missing values into NA?

strict

Convert to factor only if all values have a defined label?

unclass

If not converted to a factor (when strict = TRUE), convert to a character or a numeric factor by applying base::unclass()?

explicit_tagged_na

Should tagged NA (cf. haven::tagged_na()) be kept as explicit factor levels?

labelled_only

for a data.frame, convert only labelled variables to factors?

Details

If some values doesn't have a label, automatic labels will be created, except if nolabel_to_na is TRUE.

If sort_levels == 'values', the levels will be sorted according to the values of x. If sort_levels == 'labels', the levels will be sorted according to labels' names. If sort_levels == 'none', the levels will be in the order the value labels are defined in x. If some labels are automatically created, they will be added at the end. If sort_levels == 'auto', sort_levels == 'none' will be used, except if some values doesn't have a defined label. In such case, sort_levels == 'values' will be applied.

When applied to a data.frame, only labelled vectors are converted by default to a factor. Use labelled_only = FALSE to convert all variables to factors.

unlabelled() is a shortcut for quickly removing value labels of a vector or of a data.frame. If all observed values have a value label, then the vector will be converted into a factor. Otherwise, the vector will be unclassed. If you want to remove value labels in all cases, use remove_val_labels().

Examples

v <- labelled(c(1,2,2,2,3,9,1,3,2,NA), c(yes = 1, no = 3, "don't know" = 9))
to_factor(v)
to_factor(v, nolabel_to_na = TRUE)
to_factor(v, 'p')
to_factor(v, sort_levels = 'v')
to_factor(v, sort_levels = 'n')
to_factor(v, sort_levels = 'l')

x <- labelled(c('H', 'M', 'H', 'L'), c(low = 'L', medium = 'M', high = 'H'))
to_factor(x, ordered = TRUE)

# Strict conversion
v <- labelled(c(1, 1, 2, 3), labels = c(No = 1, Yes = 2))
to_factor(v)
to_factor(v, strict = TRUE) # Not converted because 3 does not have a label
to_factor(v, strict = TRUE, unclass = TRUE)

df <- data.frame(
  a = labelled(c(1, 1, 2, 3), labels = c(No = 1, Yes = 2)),
  b = labelled(c(1, 1, 2, 3), labels = c(No = 1, Yes = 2, DK = 3)),
  c = labelled(
    c("a", "a", "b", "c"),
    labels = c(No = "a", Maybe = "b", Yes = "c")
  ),
  d = 1:4,
  e = factor(c("item1", "item2", "item1", "item2")),
  f = c("itemA", "itemA", "itemB", "itemB"),
  stringsAsFactors = FALSE
)
if (require(dplyr)) {
  glimpse(df)
  glimpse(unlabelled(df))
}