Error when using MPI

[sshres]

I'm only seeing this error at the end when using MPIPool in the emcee sampler, it works fine with a multiprocessing pool.

> Traceback (most recent call last):
>   File "mpiemcee.py", line 121, in <module>
>     pos, prob, state = sampler.run_mcmc(p0, 10)
>   File "/home/sash7676/anaconda2/lib/python2.7/site-packages/emcee/sampler.py", line 172, in run_mcmc
>     **kwargs):
>   File "/home/sash7676/anaconda2/lib/python2.7/site-packages/emcee/ensemble.py", line 198, in sample
>     lnprob, blobs = self._get_lnprob(p)
>   File "/home/sash7676/anaconda2/lib/python2.7/site-packages/emcee/ensemble.py", line 382, in _get_lnprob
>     results = list(M(self.lnprobfn, [p[i] for i in range(len(p))]))
> TypeError: 'NoneType' object is not iterable

Has anybody had this problem?

[dfm]

What version of emcee are you using? How did you install it? Please make a minimal working example that demonstrates this issue.

[sshres]

I used 'pip install emcee', version emcee-2.2.1. The code worked fine when I used the ProcessingPool from pathos.multiprocessing but since I wanted to run my code on a cluster, I thought I'd use MPI which when I get the above posted error. Here's the full code: https://github.com/sshres/CoupledModel/blob/master/mpiemcee_coupled.py since the log posterior was a bit complicated to post here.

[dfm]

Can you please write the minimal working code snippet? It shouldn't depend on your data or your specific model. That way we can nail down where the issue is coming from. You should also say exactly what command line argument you're executing to run your script with MPI.

Also: what does emcee.version say? And how did you install mpi4py? What version of mpi4py is it? On Thu, Nov 9, 2017 at 11:40 PM sshres notifications@github.com wrote:

I used 'pip install emcee' today. The code worked fine when I used the ProcessingPool from pathos.multiprocessing but since I wanted to run my code on a cluster, I thought I'd use MPI which when I get the above posted error. Here's the full code: https://github.com/sshres/CoupledModel/blob/master/mpiemcee_coupled.py since the log posterior was a bit complicated to post here.

— You are receiving this because you commented.

Reply to this email directly, view it on GitHub https://github.com/dfm/emcee/issues/244#issuecomment-343373247, or mute the thread https://github.com/notifications/unsubscribe-auth/AAVYSme64_8coO6KFoMwvdw2JFp823GGks5s09PNgaJpZM4QY77w .

[dfm]

Also: you might want to double check the emcee MPI docs: http://emcee.readthedocs.io/en/stable/user/advanced.html#using-mpi-to-distribute-the-computations On Fri, Nov 10, 2017 at 6:31 AM Dan Foreman-Mackey foreman.mackey@gmail.com wrote:

Can you please write the minimal working code snippet? It shouldn't depend on your data or your specific model. That way we can nail down where the issue is coming from. You should also say exactly what command line argument you're executing to run your script with MPI.

Also: what does emcee.version say? And how did you install mpi4py? What version of mpi4py is it? On Thu, Nov 9, 2017 at 11:40 PM sshres notifications@github.com wrote:

I used 'pip install emcee' today. The code worked fine when I used the ProcessingPool from pathos.multiprocessing but since I wanted to run my code on a cluster, I thought I'd use MPI which when I get the above posted error. Here's the full code: https://github.com/sshres/CoupledModel/blob/master/mpiemcee_coupled.py since the log posterior was a bit complicated to post here.

— You are receiving this because you commented.

Reply to this email directly, view it on GitHub https://github.com/dfm/emcee/issues/244#issuecomment-343373247, or mute the thread https://github.com/notifications/unsubscribe-auth/AAVYSme64_8coO6KFoMwvdw2JFp823GGks5s09PNgaJpZM4QY77w .