ax = fig.add_axes([0.10, 0.15, 0.85, 0.75])
plt.plot(chn[::nthin2, j + 1],
chn[::nthin2, i + 1],
'o',
markeredgecolor='blue',
markerfacecolor='b',
markersize=5)
ax.set_xlabel(vnames[j], fontsize=22)
ax.set_ylabel(vnames[i], fontsize=22)
plt.savefig('line_infer.chn_' + vnames[j] + '_' + vnames[i] +
'.pdf')
plt.clf()
## Plot posterior PDF 'triangle' ################################################
np_kde = 100
postp.plot_all_posteriors(chn[:, 1:], vnames, np_kde,
"line_infer.posteriors", debug, dense_plots)
if generate_stats:
postp.get_mcmc_stats(all_samples, vnames, "line_infer", debug)
# ## Find PC coefficients corresponding to the chain samples #######################
# samfile="chain_samples.dat"
# # save chain samples minus the step column
# np.savetxt(samfile,chn[:,1:])
# print("Running KDE-Rosenblatt transformation to build input PCE")
# cmd = pcequad+' -o' + str(pcord)+' -f ' + samfile + ' -x ' + pctype + ' -w' + str(bw) + ' > pcequad.log'
# print("Running",cmd)
# os.system(cmd)
# for res_file in [chainfile, samfile, 'quadpts.dat', 'quadpts_mapped.dat', 'PCcoeff.dat']:
# shutil.copyfile(res_file, 'prob5_'+res_file)
desired_stats)
print("\nMAP parameter set:")
for idx in range(len(map_params)):
print(v_names[idx], ":", map_params[idx])
# Save MAP parameter info to a file
MAP_info = np.stack((v_names, map_params), axis=-1)
np.savetxt(out_file_base + "_map_params.dat", MAP_info, fmt='%-8s')
# Plot posteriors
# first and last three columns do not contain variable samples
if post_plots:
uqtkinfpp.plot_all_posteriors(all_samples[:, 1:(1 + len(v_names))],
v_names,
np_kde,
out_file_base,
demo_verbose,
dense=True)
# Make graphs for each parameter
if parameter_graphs:
graph_tools.plot_parameter_graphs(out_file_base +
"_parameter_graphs",
all_samples,
v_names,
interactive=interactive,
jpeg=jpeg)
# Make graphs for the agreement between data and observations:
model.model_info["fit_params"] = map_params
plt.clf()
indx = indx + 10
else:
break
## Scatter plots from posterior ##################################################################
for i in range(n_all_vars):
for j in range(i):
fig = plt.figure(figsize=(10, 10))
ax = fig.add_axes([0.12, 0.12, 0.8, 0.8])
plt.plot(chn[:, j + 1],
chn[:, i + 1],
'o',
markeredgecolor='blue',
markerfacecolor='b',
markersize=5)
ax.set_xlabel(vnames[j], fontsize=22)
ax.set_ylabel(vnames[i], fontsize=22)
plt.savefig('tmcmc_bimodal.chn_' + vnames[j] + '_' + vnames[i] +
'.posterior.pdf')
plt.clf()
## Plot posterior PDF 'triangle' ################################################
np_kde = 100
postp.plot_all_posteriors(chn[:, 1:], vnames, np_kde,
"tmcmc_bimodal.posteriors", 0, dense_plots)
print(
"END: TMCMC example done: check out tmcmc_bimodal.*.pdf files for results."
)