Variables labels and packed columns

Joseph Larmarange

Source: vignettes/packed_columns.Rmd

The tidyr package allows to group several columns of a tibble into one single df-column, see tidyr::pack(). Such df-column is itself a tibble. It’s not currently clear why you would ever want to pack columns since few functions work with this sort of data.

library(tidyr)
d <- iris %>%
  as_tibble() %>%
  pack(
    Sepal = starts_with("Sepal"),
    Petal = starts_with("Petal"),
    .names_sep = "."
  )
str(d)
## tibble [150 × 3] (S3: tbl_df/tbl/data.frame)
##  $ Species: Factor w/ 3 levels "setosa","versicolor",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Sepal  : tibble [150 × 2] (S3: tbl_df/tbl/data.frame)
##   ..$ Length: num [1:150] 5.1 4.9 4.7 4.6 5 5.4 4.6 5 4.4 4.9 ...
##   ..$ Width : num [1:150] 3.5 3 3.2 3.1 3.6 3.9 3.4 3.4 2.9 3.1 ...
##  $ Petal  : tibble [150 × 2] (S3: tbl_df/tbl/data.frame)
##   ..$ Length: num [1:150] 1.4 1.4 1.3 1.5 1.4 1.7 1.4 1.5 1.4 1.5 ...
##   ..$ Width : num [1:150] 0.2 0.2 0.2 0.2 0.2 0.4 0.3 0.2 0.2 0.1 ...
class(d$Sepal)
## [1] "tbl_df"     "tbl"        "data.frame"

Regarding variable labels, you may want to define a label for one sub-column of a df-column, or eventually a label for the df-column itself.

For a sub-column, you could use easily var_label() to define your label.

library(labelled)
var_label(d$Sepal$Length) <- "Length of the sepal"
str(d)
## tibble [150 × 3] (S3: tbl_df/tbl/data.frame)
##  $ Species: Factor w/ 3 levels "setosa","versicolor",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Sepal  : tibble [150 × 2] (S3: tbl_df/tbl/data.frame)
##   ..$ Length: num [1:150] 5.1 4.9 4.7 4.6 5 5.4 4.6 5 4.4 4.9 ...
##   .. ..- attr(*, "label")= chr "Length of the sepal"
##   ..$ Width : num [1:150] 3.5 3 3.2 3.1 3.6 3.9 3.4 3.4 2.9 3.1 ...
##  $ Petal  : tibble [150 × 2] (S3: tbl_df/tbl/data.frame)
##   ..$ Length: num [1:150] 1.4 1.4 1.3 1.5 1.4 1.7 1.4 1.5 1.4 1.5 ...
##   ..$ Width : num [1:150] 0.2 0.2 0.2 0.2 0.2 0.4 0.3 0.2 0.2 0.1 ...

But you cannot use directly var_label() for the df-column.

var_label(d$Petal) <- "wrong label for Petal"
str(d)
## tibble [150 × 3] (S3: tbl_df/tbl/data.frame)
##  $ Species: Factor w/ 3 levels "setosa","versicolor",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Sepal  : tibble [150 × 2] (S3: tbl_df/tbl/data.frame)
##   ..$ Length: num [1:150] 5.1 4.9 4.7 4.6 5 5.4 4.6 5 4.4 4.9 ...
##   .. ..- attr(*, "label")= chr "Length of the sepal"
##   ..$ Width : num [1:150] 3.5 3 3.2 3.1 3.6 3.9 3.4 3.4 2.9 3.1 ...
##  $ Petal  : tibble [150 × 2] (S3: tbl_df/tbl/data.frame)
##   ..$ Length: num [1:150] 1.4 1.4 1.3 1.5 1.4 1.7 1.4 1.5 1.4 1.5 ...
##   .. ..- attr(*, "label")= chr "wrong label for Petal"
##   ..$ Width : num [1:150] 0.2 0.2 0.2 0.2 0.2 0.4 0.3 0.2 0.2 0.1 ...
##   .. ..- attr(*, "label")= chr "wrong label for Petal"

As d$Petal is itself a tibble, applying var_label() on it would have an effect on each sub-column. To change a variable label to the df-column itself, you could use label_attribute().

label_attribute(d$Petal) <- "correct label for Petal"
str(d)
## tibble [150 × 3] (S3: tbl_df/tbl/data.frame)
##  $ Species: Factor w/ 3 levels "setosa","versicolor",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Sepal  : tibble [150 × 2] (S3: tbl_df/tbl/data.frame)
##   ..$ Length: num [1:150] 5.1 4.9 4.7 4.6 5 5.4 4.6 5 4.4 4.9 ...
##   .. ..- attr(*, "label")= chr "Length of the sepal"
##   ..$ Width : num [1:150] 3.5 3 3.2 3.1 3.6 3.9 3.4 3.4 2.9 3.1 ...
##  $ Petal  : tibble [150 × 2] (S3: tbl_df/tbl/data.frame)
##   ..$ Length: num [1:150] 1.4 1.4 1.3 1.5 1.4 1.7 1.4 1.5 1.4 1.5 ...
##   .. ..- attr(*, "label")= chr "wrong label for Petal"
##   ..$ Width : num [1:150] 0.2 0.2 0.2 0.2 0.2 0.4 0.3 0.2 0.2 0.1 ...
##   .. ..- attr(*, "label")= chr "wrong label for Petal"
##   ..- attr(*, "label")= chr "correct label for Petal"

On the other hand, set_variable_labels() works differently, as the primary intention of this function is to work on the columns of a tibble.

d <- d %>% set_variable_labels(Sepal = "Label of the Sepal df-column")
str(d)
## tibble [150 × 3] (S3: tbl_df/tbl/data.frame)
##  $ Species: Factor w/ 3 levels "setosa","versicolor",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Sepal  : tibble [150 × 2] (S3: tbl_df/tbl/data.frame)
##   ..$ Length: num [1:150] 5.1 4.9 4.7 4.6 5 5.4 4.6 5 4.4 4.9 ...
##   .. ..- attr(*, "label")= chr "Length of the sepal"
##   ..$ Width : num [1:150] 3.5 3 3.2 3.1 3.6 3.9 3.4 3.4 2.9 3.1 ...
##   ..- attr(*, "label")= chr "Label of the Sepal df-column"
##  $ Petal  : tibble [150 × 2] (S3: tbl_df/tbl/data.frame)
##   ..$ Length: num [1:150] 1.4 1.4 1.3 1.5 1.4 1.7 1.4 1.5 1.4 1.5 ...
##   .. ..- attr(*, "label")= chr "wrong label for Petal"
##   ..$ Width : num [1:150] 0.2 0.2 0.2 0.2 0.2 0.4 0.3 0.2 0.2 0.1 ...
##   .. ..- attr(*, "label")= chr "wrong label for Petal"
##   ..- attr(*, "label")= chr "correct label for Petal"

This is equivalent to:

var_label(d) <- list(Sepal = "Label of the Sepal df-column")
str(d)
## tibble [150 × 3] (S3: tbl_df/tbl/data.frame)
##  $ Species: Factor w/ 3 levels "setosa","versicolor",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Sepal  : tibble [150 × 2] (S3: tbl_df/tbl/data.frame)
##   ..$ Length: num [1:150] 5.1 4.9 4.7 4.6 5 5.4 4.6 5 4.4 4.9 ...
##   .. ..- attr(*, "label")= chr "Length of the sepal"
##   ..$ Width : num [1:150] 3.5 3 3.2 3.1 3.6 3.9 3.4 3.4 2.9 3.1 ...
##   ..- attr(*, "label")= chr "Label of the Sepal df-column"
##  $ Petal  : tibble [150 × 2] (S3: tbl_df/tbl/data.frame)
##   ..$ Length: num [1:150] 1.4 1.4 1.3 1.5 1.4 1.7 1.4 1.5 1.4 1.5 ...
##   .. ..- attr(*, "label")= chr "wrong label for Petal"
##   ..$ Width : num [1:150] 0.2 0.2 0.2 0.2 0.2 0.4 0.3 0.2 0.2 0.1 ...
##   .. ..- attr(*, "label")= chr "wrong label for Petal"
##   ..- attr(*, "label")= chr "correct label for Petal"

To use set_variable_labels() on sub-columns, you should use this syntax:

d$Petal <- d$Petal %>%
  set_variable_labels(
    Length = "Petal length",
    Width = "Petal width"
  )
str(d)
## tibble [150 × 3] (S3: tbl_df/tbl/data.frame)
##  $ Species: Factor w/ 3 levels "setosa","versicolor",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Sepal  : tibble [150 × 2] (S3: tbl_df/tbl/data.frame)
##   ..$ Length: num [1:150] 5.1 4.9 4.7 4.6 5 5.4 4.6 5 4.4 4.9 ...
##   .. ..- attr(*, "label")= chr "Length of the sepal"
##   ..$ Width : num [1:150] 3.5 3 3.2 3.1 3.6 3.9 3.4 3.4 2.9 3.1 ...
##   ..- attr(*, "label")= chr "Label of the Sepal df-column"
##  $ Petal  : tibble [150 × 2] (S3: tbl_df/tbl/data.frame)
##   ..$ Length: num [1:150] 1.4 1.4 1.3 1.5 1.4 1.7 1.4 1.5 1.4 1.5 ...
##   .. ..- attr(*, "label")= chr "Petal length"
##   ..$ Width : num [1:150] 0.2 0.2 0.2 0.2 0.2 0.4 0.3 0.2 0.2 0.1 ...
##   .. ..- attr(*, "label")= chr "Petal width"
##   ..- attr(*, "label")= chr "correct label for Petal"

If you want to get the list of variable labels of a tibble, by default var_label() or get_variable_labels() will return the labels of the first level of columns.

## $Species
## NULL
## 
## $Sepal
## [1] "Label of the Sepal df-column"
## 
## $Petal
## [1] "correct label for Petal"

To obtain the list of variable labels for sub-columns, you could use recurse = TRUE:

d %>% get_variable_labels(recurse = TRUE)
## $Species
## NULL
## 
## $Sepal
## $Sepal$Length
## [1] "Length of the sepal"
## 
## $Sepal$Width
## NULL
## 
## 
## $Petal
## $Petal$Length
## [1] "Petal length"
## 
## $Petal$Width
## [1] "Petal width"
d %>%
  get_variable_labels(
    recurse = TRUE,
    null_action = "fill",
    unlist = TRUE
  )
##               Species          Sepal.Length           Sepal.Width 
##             "Species" "Length of the sepal"               "Width" 
##          Petal.Length           Petal.Width 
##        "Petal length"         "Petal width"