to_labelled: Convert to labelled data

In larmarange/labelled: Manipulating Labelled Data

Convert to labelled data

Description

Convert a factor or data imported with foreign or memisc to labelled data.

Usage

to_labelled(x, ...)

## S3 method for class 'data.frame'
to_labelled(x, ...)

## S3 method for class 'list'
to_labelled(x, ...)

## S3 method for class 'data.set'
to_labelled(x, ...)

## S3 method for class 'importer'
to_labelled(x, ...)

foreign_to_labelled(x)

memisc_to_labelled(x)

## S3 method for class 'factor'
to_labelled(x, labels = NULL, .quiet = FALSE, ...)

Arguments

x	Factor or dataset to convert to labelled data frame
...	Not used
labels	When converting a factor only: an optional named vector indicating how factor levels should be coded. If a factor level is not found in labels, it will be converted to NA.
.quiet	do not display warnings for prefixed factors with duplicated codes

Details

to_labelled() is a general wrapper calling the appropriate sub-functions.

memisc_to_labelled() converts a memisc::data.set()]' object created with memisc package to a labelled data frame.

foreign_to_labelled() converts data imported with foreign::read.spss() or foreign::read.dta() from foreign package to a labelled data frame, i.e. using haven::labelled(). Factors will not be converted. Therefore, you should use use.value.labels = FALSE when importing with foreign::read.spss() or convert.factors = FALSE when importing with foreign::read.dta().

To convert correctly defined missing values imported with foreign::read.spss(), you should have used to.data.frame = FALSE and use.missings = FALSE. If you used the option to.data.frame = TRUE, meta data describing missing values will not be attached to the import. If you used use.missings = TRUE, missing values would have been converted to NA.

So far, missing values defined in Stata are always imported as NA by foreign::read.dta() and could not be retrieved by foreign_to_labelled().

If you convert a labelled vector into a factor with prefix, i.e. by using to_factor(levels = "prefixed"), to_labelled.factor() is able to reconvert it to a labelled vector with same values and labels.

Value

A tbl data frame or a labelled vector.

See Also

haven::labelled(), foreign::read.spss(), foreign::read.dta(), memisc::data.set(), memisc::importer, to_factor().

Examples

## Not run: 
# from foreign
library(foreign)
sav <- system.file("files", "electric.sav", package = "foreign")
df <- to_labelled(read.spss(
  sav,
  to.data.frame = FALSE,
  use.value.labels = FALSE,
  use.missings = FALSE
))

# from memisc
library(memisc)
nes1948.por <- UnZip("anes/NES1948.ZIP", "NES1948.POR", package = "memisc")
nes1948 <- spss.portable.file(nes1948.por)
ds <- as.data.set(nes1948)
df <- to_labelled(ds)

## End(Not run)

# Converting factors to labelled vectors
f <- factor(
  c("yes", "yes", "no", "no", "don't know", "no", "yes", "don't know")
)
to_labelled(f)
to_labelled(f, c("yes" = 1, "no" = 2, "don't know" = 9))
to_labelled(f, c("yes" = 1, "no" = 2))
to_labelled(f, c("yes" = "Y", "no" = "N", "don't know" = "DK"))

s1 <- labelled(c("M", "M", "F"), c(Male = "M", Female = "F"))
labels <- val_labels(s1)
f1 <- to_factor(s1)
f1

to_labelled(f1)
identical(s1, to_labelled(f1))
to_labelled(f1, labels)
identical(s1, to_labelled(f1, labels))

l <- labelled(
  c(1, 1, 2, 2, 9, 2, 1, 9),
  c("yes" = 1, "no" = 2, "don't know" = 9)
)
f <- to_factor(l, levels = "p")
f
to_labelled(f)
identical(to_labelled(f), l)

larmarange/labelled documentation built on Oct. 11, 2024, 6:25 p.m.