Closed ciskakemper closed 3 years ago
I am currently working on branch iss27 on this issue, not on the master branch, as my previous changes on branch iss27 have not been merged with the master yet. @pscicluna, could you let me know once the branches are merged, so that I can work in the master branch again? In principle the issue 27 should be solved, but we cannot tell as the fit is done only to the photometry and the result is essentially the same.
Hi @ciskakemper, this issue was by design in early testing - I needed to limit it to the photometry to make the model more tractable in the short term. Simply uncommenting lines 35-38 should put the spectra back into dataset and make it work. We should probably add a more explanatory if statement to control this and make it easier to switch between different cases for testing.
Dear @pscicluna, I tried uncommenting lines 35-38, but it didn't work, so I thought they were commented out for a reason (namely that it didn't work) without documentation. It took me quite some time to figure out that the spectroscopy was not included in the fitting, as the fit results looked the same for the relative and absolute abundances, and in both cases looked quite off with the data (given that the spectroscopy was plotted).
After consultation with Peter this has been fixed.
I have found that the reason is that a good fit is not produced by pgQuasarSample.py is because the spectroscopy data is read in but never appended to the data set to be fitted. So the fit is only done to the photometry data. See lines 35-41 in pgQuasarSample.py