Stan Installation Error or Coding Error


#1

Dear Stanians,
It’s time of the year when I need to update both R and accordingly RStudio as well as rstan :( I did it, but my r codes written two years ago don’t work any more. I am not sure if that’s related to how I installed these three or any coding error in the Stan coding section. I am trying to simply estimate an ordered logit model using a Bayesian approach. Attached please find the log file. Somehow, the log file (sink()) wasn’t able to capture the few warning messages which I don’t think led to the errors in the log file that I attached and
read as follows:
DIAGNOSTIC(S) FROM PARSER:
Warning (non-fatal): assignment operator <- deprecated in the Stan language; use = instead.
Warning (non-fatal): assignment operator <- deprecated in the Stan language; use = instead.
Warning (non-fatal): assignment operator <- deprecated in the Stan language; use = instead.
Warning (non-fatal): assignment operator <- deprecated in the Stan language; use = instead.
Warning (non-fatal): assignment operator <- deprecated in the Stan language; use = instead.
Warning (non-fatal): assignment operator <- deprecated in the Stan language; use = instead.
Warning (non-fatal): assignment operator <- deprecated in the Stan language; use = instead.
Warning (non-fatal): assignment operator <- deprecated in the Stan language; use = instead.
Warning (non-fatal): assignment operator <- deprecated in the Stan language; use = instead.
Warning: Deprecated function ‘categorical_log’; please replace suffix ‘_log’ with ‘_lpdf’ for density functions or ‘_lpmf’ for mass functions

bayesGoLogitMod01aStan2018-03-16.txt (40.8 KB)

Thanks a lot for any help!

Jun Xu, PhD
Professor
Department of Sociology
Ball State University
Muncie, IN47306


#2

I’m afraid I don’t understand what you’re asking here.

Just a warning, so you don’t need to fix it yet. It will break some time in the future when we finally remove <-. The fix to get rid of the arning is just what the message says: replace <- with =.

Same thing. Use categorical_lpmf rather than categorical_log.