Has siera (pharmaverse) make the ARD the reporting event describes:
writes the ADaM it reads as CSV files (siera's programmes read
<DATASET>.csv), the event as ARS JSON, has siera::readARS() write
one programme per output, and runs them. With tfl_build_ard() on the
same spec this is a round trip: the ARS says the same analyses as the
spec when the two ARDs hold the same numbers.
Usage
tfl_ars_ard(ars, adam, dir = tempfile("tfl_ars_"), keep = FALSE)Arguments
- ars
A
tfl_ars()written withprofile = "siera".- adam
The ADaM: a named list of data frames, or a folder (
tfl_read_adam()).- dir
A folder for the CSV files, the JSON and siera's programmes.
- keep
Keep
dir(by default it is removed on exit).
Value
The ARD (one data frame, all outputs), with siera's OutputId
and AnalysisId given back as the spec's output_id and the ARS
analysis id. With keep = TRUE, attribute dir.
Details
CSV keeps text and numbers; a date becomes text and a code with leading zeros may become a number. A variable of the event that is a date, or text of digits with a leading zero, is warned about before it is written.