recoding SPSS/.sav files in R while retaining metadata

Question

I would like to use the attributes of the dataset to recode my spss data while preserving the metadata/label attributes. The problem I am having is that anything I do removes the existing attributes while trying to recode the numeric values of my data into the text.

For example the variable AWR_A_5 below the 1's correspond to 'Kelley Blue Book'

TestFile Data

testfile <- haven::read_sav("TestFile.sav")

> savfile$AWR_A_5
<labelled<double>[10]>: Which of the following automotive resource sites, if any, 
have you heard of? Please select all that apply. Kelley Blue Book
 [1] NA NA  0 NA NA  1  1 NA  1  0

Labels:
 value            label
     1   Kelley Blue Book

So I would like to recode those 1's in the data to Kelley Blue Book while retaining the attributes. So it would look like the following. I understand this would change the column type from numeric to character.

> savfile$AWR_A_5
<labelled<character>[10]>: Which of the following automotive resource sites, if any,
 have you heard of? Please select all that apply. Kelley Blue Book
 [1] NA NA  0 NA NA  Kelley Blue Book  Kelley Blue Book NA  Kelley Blue Book  0

Labels:
 value            label
     1   Kelley Blue Book

Answer 1

One option to achieve your desired result would be to convert both the values or the data type and the labels to characters (as I deal regularly with such transformations I wrapped this code inside a function).

Using some fake random example data:

library(haven)

set.seed(123)

savfile <- data.frame(
  AWR_A_5 = sample(c(0, 1, NA), 10, replace = TRUE)
)

savfile$AWR_A_5 <- labelled(
  savfile$AWR_A_5,
  label = "Which of the following automotive resource sites, if any, have you heard of? Please select all that apply. Kelley Blue Book",
  labels = c("Kelley Blue Book" = 1)
)

savfile$AWR_A_5
#> <labelled<double>[10]>: Which of the following automotive resource sites, if any, have you heard of? Please select all that apply. Kelley Blue Book
#>  [1] NA NA NA  1 NA  1  1  1 NA  0
#> 
#> Labels:
#>  value            label
#>      1 Kelley Blue Book

x <- savfile$AWR_A_5
label <- attr(x, "label")
labels <- attr(x, "labels")
labels <- setNames(as.character(labels), names(labels))

labelled(
  as.character(haven::as_factor(x)),
  label = label,
  labels = labels
)
#> <labelled<character>[10]>: Which of the following automotive resource sites, if any, have you heard of? Please select all that apply. Kelley Blue Book
#>  [1] NA               NA               NA               Kelley Blue Book
#>  [5] NA               Kelley Blue Book Kelley Blue Book Kelley Blue Book
#>  [9] NA               0               
#> 
#> Labels:
#>  value            label
#>      1 Kelley Blue Book