search

tom-hc-park / MSc-RA-Bayesian-evidence-synthesis

Research Project at M.Sc. Statistics at University of British Columbia
4 stars 0 forks source link

JAGS error #5

Open tom-hc-park opened 4 years ago

tom-hc-park commented 4 years ago

Dear Professor Paul Gustafson (@paulgstf ),

Hope you are having a great weekend.

I've encountered to some error messages for the JAGS part. I wrote an Rmarkdown file which demonstrates the questions.

Since GitHub issue does not seem to allow me to upload the Rmarkdown file here, I would like to send you the file via e mail. I believe this way you don't have to clone the whole git repo to your local machine...

Thanks, Hyeongcheol

paulgstf commented 4 years ago

Sorry to be slow @tom-hc-park , please do send by e-mail.

paulgstf commented 4 years ago

Sorry, scratch that - see now that you sent it earlier.

paulgstf commented 4 years ago

@tom-hc-park , I took a look. Try changing the syntax so that ~ is for distributional assumptions, <- is for deterministic functions, e.g., blah ~ dunif... but thing.sq <- pow(thing, 2).

Keep me posted as to whether this fixed things or not.

tom-hc-park commented 4 years ago

Thank you ! I will try that and go further !

oxyaks commented 4 years ago

Hello Professor Paul Gustafson,

I am facing a peculiar problem. I have written a JAGS code for logistic regression and it works fine when i send the data from row1 but if take a different subset for example, x[500:1000] it gives error beta index out of bounds. But when i pass the data from row 1 as x[1:1000] or x[1:2000] it works just fine. Is there something I am not seeing here.

Many thanks, Aakash