Extract the labels attribute from “labeled” tibble columns from a haven import from Stata

Question

Hadley Wickham\'s haven package, applied to a Stata file, returns a tibble with many columns of type \"labeled\". You can see these with str(), e.g.:

Answer 1

Jumping off @omar-waslow answer above, but adding the use of attr_getter.

If the data (some_df) is imported using read_dta in the haven package, then each column in the tibble has an attr called "label". So we split up the dataframe, going column by column. This creates a two column dataframe which can be joined back (after pivot_longer, for example).

library(tidyverse)
label_lookup_map <- tibble(
   col_name = some_df %>% names(),
   labels = some_df %>% map_chr(attr_getter("label"))
)