Code about getting mcmc sampling results

[0batsy0]

I hope to extract the data sampled by mcmc, so I used the following two different codes to try to obtain it, but the obtained results seem to be somewhat different. I don’t quite understand the difference between the functions of these two lines of code.Here are the distribution diagrams obtained by using these two writing methods for parameter q.

[kecnry]

This is a good question and probably something we should add to the tutorials somewhere!

By passing None to all the keyword arguments, phoebe will use the default proposed values in the solution for burning, thin, etc. Can you see if process_mcmc_chains_from_solution(tes, 'emcee_sol', burnin=0, thin=1, lnprob_cutoff=-np.inf) gives you what you'd expect?

To see the default values being applied when passing None: print(tes.filter(qualifier=['burnin', 'thin', 'lnprob_cutoff'], solution='emcee_sol'))

[0batsy0]

Thanks for your reply, I have successfully solved it!!!But I still don’t quite understand why when burnin exists by default (for example, I checked that the value is 226), the two methods will get the same number of data (my understanding is that the relative number of burnin should be subtracted)

[kecnry]

I would expect the shape/length to change as well. If you call .shape on each of the returned arrays, before plotting, what do they say?

[0batsy0]

Thank you!I found the problem!