I am working through: Lotka-Volterra multiple ways and ran into an issue when doing the SMC example (with likelihood). That section looks like this: sampler = "SMC with Likelihood" draws = 2000 with model: trace_SMC_like = pm.sample_smc(draws) trace = trace_SMC_like az.summary(trace) fails wit…

Should work if you define the Op in a python file and import it in your notebook. This kind of issue is common when combining multiprocessing and interactive python environments

Pm.sample_smc fails if chains > 1 on Mac OS

Questions v5

ricardoV94 March 12, 2024, 4:01pm 9

I am not sure why, may be related to stuff like this: python - Jupyter Notebook Multiprocessing code not working - Stack Overflow

1 Like

Topic		Replies	Views
Using multiple chains in SMC in PyMC3 Questions smc_abc	9	1580	December 3, 2021
Multiprocess sampling with SMC fails on linux cluster theano	0	788	December 3, 2021
Combining parameters into single tensor Questions	19	3538	October 30, 2017
Problem with multiprocessing in PyMC3 Questions	5	3722	August 20, 2018
When ever I tried running NUTS sampling my notebook gets disconnected from kernel Questions	4	1872	July 7, 2018

Pm.sample_smc fails if chains > 1 on Mac OS

Related topics