rjnw
2019-5-12 01:51:44

jags for naive bayes takes a looong time to run, I had to cancel it cause it ran out of memory in the virtual machine.


ccshan
2019-5-12 03:35:38

Our paper says “a sweep by JAGS takes more than 500 seconds”.


ccshan
2019-5-12 03:35:53

Is that consistent with what you see? Does the VM need more memory?


rjnw
2019-5-12 03:36:27

jags has a startup of a couple hours for naive bayes


rjnw
2019-5-12 03:36:44

I did increase the memory for the VM but I haven’t tried running it again


rjnw
2019-5-12 03:37:22

but good news is I tested everything else and all the other run fine


ccshan
2019-5-12 04:07:47

So it sounds like you don’t know that the JAGS NB result in the VM doesn’t match the paper?


rjnw
2019-5-12 04:09:17

I just started jags again. jags result matched for gmm so I am sure the NB results will match too