Dear all, Can you please suggest me how can I correct in the following code in order to avoid pickle error ? import pymc3 as pm import pymc_forward_model as PFM # create our Op sigma=1 mu = PFM.MyLikelihoodFunctionGrad(PFM.log_likelihood, data, sigma, XX) #mu = PFM.MyLikelihoodFunction(PFM.log_li…

[image] marty: NameError for chain, seed, start in zip(range(chains), seeds, start_points): name 'trace' is not defined File "C:\Users\Username\Anaconda2\envs\pymc\lib\site-packages\pymc3\parallel_sampling.py", line 313, in <listcomp> Is your underlying exception. It seems that you a…

Indeed I updated pymc3 with: pip install git+https://github.com/pymc-devs/pymc3, and I get this: Traceback (most recent call last): File "<string>", line 1, in <module> File "C:\Users\Username\Anaconda2\envs\pymc2\lib\multiprocessing\spawn.py", line 105, in spawn_main exitcode = _main(fd) …

Great, now the traceback is much cleaner and points the relevant line that tried to get the trace but failed.

maybe the solution is then to install ubuntu ?

[image] marty: I have the same probleme with the methode: “black box” likelihood function: https://docs.pymc.io/notebooks/blackbox_external_likelihood.html Hi! I am using the same example, but when I try to use HMC which needs to calculate the gradient, raise the UserWarning (from gradient f…

Try moving the trace plot into the model context used during sampling. Could you also paste the full batch script that you’re using here or upload it to a gist? The problem is hard to debug on windows but should be fixable

Please see this thread and let me know if this helps [image] When I am custom likelihood example I get pickle error when multiple chains are used Questions Hello Everyone, I am trying to use this example ([ https://docs.pymc.io/notebooks/blackbox_external_likelihood.html ]( h…

Here the full script, (I just replaced the my_model function with a simple example, because the real one is a large python code). I did not get what you mean by “Try moving the trace plot into the model context used during sampling” ? import pymc3 as pm import numpy as np import theano.tensor as t…

I think this refers to a different issue which has already been solved by replacing the lambda function by: def my_mu(v): return mu(v)

You have to add one indentation level to the following: [image] marty: _ = pm.traceplot(trace) print('COUNT : %s'%(COUNT)) print('COUNT/ITER : %s'%(COUNT/(ndraws+nburn)/chains)) print('ndraws : %s'%(ndraws)) print('nburn : %s'%(nburn)) print('chains : %s'%(chains)) print('cores : %s'%(cores)…

Pickle error while using lambda function

Questions

avi April 30, 2019, 12:53pm 18

Please see this thread and let me know if this helps

Topic		Replies	Views
When I am custom likelihood example I get pickle error when multiple chains are used Questions	2	553	April 5, 2019
I can't get pymc3 to work Questions bug	4	2628	January 15, 2021
Pymc3 getting stuck after initialization Questions	41	9396	February 9, 2022
Cannot pickle '_thread.lock' object type error when running pm.sample v5	3	673	May 4, 2023
Multiprocessing (windows 10) BrokenPipeError: [Errno 32] Broken pipe Questions	14	30091	November 27, 2018

Pickle error while using lambda function

Related topics