Operating System: Linux Ubuntu 18.04 LTS, kernel v5.7.1-050701-generic
Description
After an Ultranest run, which is the chain file than can be imported outside the Python environment for external processing of the Monte Carlo samples over Xspec directly? Please note that I am using UltraNest over threeML, not BXA.
There are three text (not FITS) files available with parameter samples inside the /chains sub-directory (one of the 5 auto-generated to store the results in):
equal_weighted_post.txt
weighted_post.txt
weighted_post_untransformed.txt.
The last file is the run.txt, which contains the statistical bits.
The /info sub-directory has the posterior summary file (post_summary.csv) and a results.json file.
The sub-directories /extra and /plots are empty.
Another is the points.hdf5 file within the /results sub-directory. Is that the file I should convert to FITS and then use over Xspec?
Requirement
Knowing which file is the correct one is essential to me, since I could then try to sample on Xspec to compute the equivalent width of my line model component directly on Xspec. The threeML developers suggested I should ask this question here.
Description
After an Ultranest run, which is the chain file than can be imported outside the Python environment for external processing of the Monte Carlo samples over Xspec directly? Please note that I am using UltraNest over threeML, not BXA.
There are three text (not FITS) files available with parameter samples inside the
/chains
sub-directory (one of the 5 auto-generated to store the results in):equal_weighted_post.txt
weighted_post.txt
weighted_post_untransformed.txt
.The last file is the
run.txt
, which contains the statistical bits.The
/info
sub-directory has the posterior summary file (post_summary.csv
) and aresults.json
file.The sub-directories
/extra
and/plots
are empty.Another is the
points.hdf5
file within the/results
sub-directory. Is that the file I should convert to FITS and then use over Xspec?Requirement Knowing which file is the correct one is essential to me, since I could then try to sample on Xspec to compute the equivalent width of my line model component directly on Xspec. The threeML developers suggested I should ask this question here.