janno::process_age currently only produces the derived columns if Date_BC_AD_Prob is in the requested output columns, otherwise foregoing calibration altogether.
Example:
janno::process_age(
input_janno,
choices = c("Date_BC_AD_Start_Derived", "Date_BC_AD_Median_Derived", "Date_BC_AD_Stop_Derived"),
cal_curve = "intcal20"
)
This produces an unaltered input_janno, but adding "Date_BC_AD_Prob" in choices produces the desired output columns.
The real problem is that this behaviour is undocumented. Here's the explanation of choices in the docs:
| choices |
character vector. Which output columns should be added? By default all possible columns are added. |