Open colinorourke opened 1 year ago
Colin @colinorourke, thanks for this and apologies for the delay. This is a good catch. The natural fix is probably to replace any(...)
with any(..., na.rm = TRUE)
. The same fix should then be applied to other distributions where any(...)
is used.
Mikis @mstasinopoulos can you have a look? Do you agree?
I've found an issue with the BCCG distribution functions. They don't gracefully handle missing values due to the use of
any(q < 0)
, which will returnNA
if any of its inputs are missing. Here's a reprex, and a comparison to the behaviour of pnorm, which behaves as expected:Created on 2023-09-14 with reprex v2.0.2