meeting 2024 01 17 n315 - JacobPilawa/TriaxSchwarzschild_wiki_5 GitHub Wiki
Context
I've been cleaning some of the results from last time, included some more direct comparisons between my best-fitting Gh6Mh6/Gh8Mh6 data/compared to what Irina has published in the previous kinematic papers.
Infil Plots for Various Cases
adeg = -1 (no additive polynomial case)
Plot
Gh6Mh6 non-point symm
Gh6Mh6 point symm
Gh8Mh6 non-point symm
Gh8Mh6 point symm
adeg = 0 (a constant polynomial)
Plot
Gh6Mh6 non-point symm
Gh6Mh6 point symm
Gh8Mh6 non-point symm
Gh8Mh6 point symm
Running pPXF with Exactly Irina's Settings
We discussed last time trying to reproduce Irina's exact results, and I've gotten started a bit on that front:
Wavelength range (rest): 8420-8770 A; But masking the GMOS chip gap cuts off fit at ~8750 A.
Template list: (15 stars from Barth et al. 2002 -- spectra are from MILES Calcium Triplet library)
Spectrum mask: /GMOS/galaxy_files/N315/spectra/N315_spectra_mask.fits (bright sky lines and chip gaps as assessed from inspection of GMOS spectra)
Initial velocity shift: 4950 km/s
Multiplicative polynomial: degree 3
Additive polynomial: degree 0 (constant offset)
Bias parameter: 0
100 MCMC iterations
Despite running what I believe to be the exact settings Irina used, I am unable to get perfect agreement with her results. I hadn't had a chance to dig into the diagnostics on the template weights yet which I certainly will do, but I wanted to put up some plots on here so that we can discuss how important this is in the meantime.
Direct Comparsion of Her Data vs. Mine
Spectral Fit Comparison:
|Bin #|Her Fits|My Gh6 Fits|
|-----|--------|-----------|
|0| ||
|19| ||
|21| ||
|163| ||
|175| ||
I was also curious what happens if I run Irina's exact settings, but fit for 8 moments instead of 6: