Charles Driver
Charles Driver
Any chance of this happening? Or can anyone suggest how to use an ensemble with a custom metric that uses additional data to compute the performance?
I'm just aware of this now, happy to try modify the stan code in ctsem but also not sure where to start. I'm available for any discussion / ideas @bgoodri...
so, the stanc() function seems? broken, in that it requires a file argument when, if you use the model_code argument, file should not be needed.
``` library(rstan) st
That works and everything compiles and runs, and looks like normal, but the log probability / parameter values are all wrong. Will take me some time to see why.
That was just a cheap mwe, not my actual test problem ;) but sure. I've narrowed down the problem to the gradient calculation of some limited set of functions, but...
Ok. I spent the morning diagnosing the 'incorrect gradient under certain conditions' thing, and it magically fixed itself :/ No idea how / why, maybe some remnant left in memory?...
https://www.dropbox.com/s/1mrvpossovz3sa0/ctsm.stan?dl=0
Not sure I follow. It was just a miscapitalisation (should have been JAx) .
yes that is a valid error. However, when I use stanc("", model_code=x) , it doesn't give the error, which is the problem.