Categorical fields are exported as integers but are encoded with labels.
For example 20116 “Smoking status”:
| Coding | Meaning |
|---|---|
| -3 | Prefer not to answer |
| 0 | Never |
| 1 | Previous |
| 2 | Current |
This package includes two functions to label a single UK Biobank field or a data frame of them using the UK Biobank encoding schema. Examples:
# update the Smoking status field
ukb <- label_ukb_field(ukb, field="p20116_i0")
table(ukb$p20116_i0) # tabulates the values
#> -3 0 1 2
#> 2057 273405 172966 52949
table(haven::as_factor(ukb$p20116_i0)) # tabulates the labels
#> Prefer not to answer Never Previous Current
#> 2057 273405 172966 52949
haven::print_labels(ukb$p20116_i0) # show the value:label mapping for this variable
#> Labels:
#> value label
#> -3 Prefer not to answer
#> 0 Never
#> 1 Previous
#> 2 Current
#
# if you have a whole data frame of exported fields, you can use the wrapper function label_ukb_fields()
# say the `ukb` data frame contains 4 variables: `eid`, `p54_i0`, `p31` and `age_at_assessment`
# update the variables that looks like UK Biobank fields with titles and, where cateogrical, labels
# i.e., `p54_i0` and `p31` only -- `eid` and `age_at_assessment` are ignored
ukb <- label_ukb_fields(ukb)
table(ukb$p31) # tabulates the values
#> 0 1
#> 273238 229031
table(haven::as_factor(ukb$p31)) # tabulates the labels
#> Female Male
#> 273238 229031