Improve error messaging for derive_param_tte()

[rossfarrugia]

What happened?

Any reason why derive_param_tte() argument source_datasets doesn't like underscores?

Session Information

Using {admiral} 1.1.1

Reproducible Example

This code works (assuming I have a valid adsl_death dataset and I've set rand_censor):

death <- event_source(
  dataset_name = "adsl",
  filter = DTHFL == "Y",
  date = DTHDT
)

test_os <- derive_param_tte(
  dataset_adsl = adsl_death,
  start_date = RANDDT,
  event_conditions = list(death),
  censor_conditions = list(rand_censor),
  source_datasets = list(adsl = adsl_death),
  set_values_to = exprs(
    PARAMCD = "OS"
  )
)

This code errors:

death <- event_source(
  dataset_name = "adsl_death",
  filter = DTHFL == "Y",
  date = DTHDT
)

test_os <- derive_param_tte(
  dataset_adsl = adsl_death,
  start_date = RANDDT,
  event_conditions = list(death),
  censor_conditions = list(rand_censor),
  source_datasets = list(adsl_death = adsl_death),
  set_values_to = exprs(
    PARAMCD = "OS"
  )
)

The error given is:

Error in `fun(..., .envir = .envir)`:
! Could not evaluate cli `{}` expression: `source_names`.
Caused by error in `eval(expr, envir = envir)`:
! object 'source_names' not found
Type .Last.error to see the more details.

[bundfussr]

@rossfarrugia , I think the issue is not related to underscores but an issue in the error message. The correct error message should be:

Error in `derive_param_tte()`:
! The dataset names must be included in the list specified for the `source_datasets` argument.
ℹ Following names were provided by `source_datasets`: "adsl_death"
ℹ But, `censor_conditions[[1]]$dataset_name = adsl`

I assume in your example rand_censor is referring to adsl instead of adsl_death. Could you check?

[rossfarrugia]

Yes! That's it @bundfussr - great spot. It was an error on my side but the error message could definitely be improved, as I'd have spent hours trying to figure that out.

[bundfussr]

Yes! That's it @bundfussr - great spot. It was an error on my side but the error message could definitely be improved, as I'd have spent hours trying to figure that out.

@rossfarrugia , I have fixed the bug in the error message. If you want to try it out, you need to call remotes::install_github("pharmaverse/admiraldev", ref = "469_fix_assert_list_element") to install the admiraldev version where the bug is fixed.

[rossfarrugia]

yes, works now by giving better error message when re-running the same code 👍

[bms63]

Yes! That's it @bundfussr - great spot. It was an error on my side but the error message could definitely be improved, as I'd have spent hours trying to figure that out.

hours!! oh no!!