Retrieve value labels of labelled data

Description

This function retrieves the value labels of labelled data, which was created with the labelled or haven package, or imported from SPSS, SAS or STATA files (via read_spss, read_sas or read_stata) and

  • if x is a data frame or list of variables, returns all variables' value labels as list

  • or, if x is a vector, returns the value labels as string.

Usage

1
2
get_labels(x, attr.only = FALSE, include.values = NULL,
  include.non.labelled = FALSE, drop.na = TRUE)

Arguments

x

data.frame with variables that have value label attributes (e.g. from an imported SPSS, SAS or STATA data set, via read_spss, read_sas or read_stata); a variable (vector) with value label attributes; or a list of variables with values label attributes. If x has no label attributes, factor levels are returned. See 'Examples'.

attr.only

Logical, if TRUE, labels are only searched for in the the vector's attributes; else, if attr.only = FALSE and x has no label attributes, factor levels or string values are returned. See 'Examples'.

include.values

String, indicating whether the values associated with the value labels are returned as well. If include.values = "as.name" (or include.values = "n"), values are set as names attribute of the returned object. If include.values = "as.prefix" (or include.values = "p"), values are included as prefix to each label. See 'Examples'.

include.non.labelled

Logical, if TRUE, values without labels will also be included in the returned labels (see fill_labels).

drop.na

Logical, whether labels of tagged NA values (see tagged_na) should be included in the return value or not. By default, labelled (tagged) missing values are not returned. See get_na for more details on tagged NA values.

Details

This package can add (and read) value and variable labels either in foreign package style (attributes are named value.labels and variable.label) or in haven package style (attributes are named labels and label). By default, the haven package style is used.

Working with labelled data is a key feature of the sjPlot package, which accesses these attributes to automatically read label attributes for labelling axis categories and titles or table rows and columns in graphical or tabular outputs.

When working with labelled data, you can, e.g., use get_label or get_labels to get a vector of value and variable labels, which can then be used with other functions like barplot etc. See 'Examples'. Furthermore, value and variable labels are used when saving data, e.g. to SPSS (see write_spss), which means that the written SPSS file contains proper labels for each variable.

You can set a default label style (i.e. the names of the label attributes, see above) via options(value_labels = "haven") or options(value_labels = "foreign").

Value

Either a list with all value labels from all variables if x is a data.frame or list; a string with the value labels, if x is a variable; or NULL if no value label attribute was found.

Note

This function works with vectors that have value and variable label attributes (as provided, for instance, by labelled objects). Adding label attributes is automatically done when importing data sets with the read_spss, read_sas or read_stata functions. Labels can also manually be added using the set_labels and set_label functions. If vectors do not have label attributes, either factor-levels or the numeric values of the vector are returned as labels.

Most functions of the sjPlot package access value and variable label attributes to automatically detect labels in order to set them as axis, legend or title labels in plots (sjp.-functions) respectively as column or row headers in table outputs (sjt.-functions). See this and this online manuals for more details.

See Also

See package vignettes or online documentation for more details; set_labels to manually set value labels, get_label to get variable labels and get_values to retrieve value label associated values.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
# import SPSS data set
# mydat <- read_spss("my_spss_data.sav")

# retrieve variable labels
# mydat.var <- get_label(mydat)

# retrieve value labels
# mydat.val <- get_labels(mydat)

data(efc)
get_labels(efc$e42dep)

# simple barplot
barplot(table(efc$e42dep))
# get value labels to annotate barplot
barplot(table(efc$e42dep),
        names.arg = get_labels(efc$e42dep),
        main = get_label(efc$e42dep))

# include associated values
get_labels(efc$e42dep, include.values = "as.name")

# include associated values
get_labels(efc$e42dep, include.values = "as.prefix")

# get labels from multiple variables
get_labels(list(efc$e42dep, efc$e16sex, efc$e15relat))


# create a dummy factor
f1 <- factor(c("hi", "low", "mid"))
# search for label attributes only
get_labels(f1, attr.only = TRUE)
# search for factor levels as well
get_labels(f1)

# same for character vectors
c1 <- c("higher", "lower", "mid")
# search for label attributes only
get_labels(c1, attr.only = TRUE)
# search for string values as well
get_labels(c1)


# create vector
x <- c(1, 2, 3, 2, 4, NA)
# add less labels than values
x <- set_labels(x, c("yes", "maybe", "no"), force.values = FALSE)
# get labels for labelled values only
get_labels(x)
# get labels for all values
get_labels(x, include.non.labelled = TRUE)


# get labels, including tagged NA values
library(haven)
x <- labelled(c(1:3, tagged_na("a", "c", "z"), 4:1),
              c("Agreement" = 1, "Disagreement" = 4, "First" = tagged_na("c"),
                "Refused" = tagged_na("a"), "Not home" = tagged_na("z")))
# get current NA values
x
get_labels(x, include.values = "n", drop.na = FALSE)

Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.