meeting 2025 03 26 n57 - JacobPilawa/TriaxSchwarzschild_wiki_5 GitHub Wiki
Context
- Cleaning up results from the last meeting, lots of random testing and tracking down small discrepencies.
Diagnostics
- The first thing I started to look into was which bins seemed to be bounding around the most between the preliminary fits and the final MC fits. As a quick reminder, we were looking at these plots from last time:
- The left set of plots show the original (currently used in the minimization) kinematics plotted against the newest set of kinematics run with the "correct" settings.
- The right set of plots compare the preliminary fits of the newest set of kinematics to my attempt at reproducing the original kinematics with the new pPXF code.
| Comparison to "Published" | Comparison to Reproduction (Prelim) |
|---|---|
| [images/250326/original_kinematics_vs_new_kinematics.png]]](/JacobPilawa/TriaxSchwarzschild_wiki_5/wiki/[[images/250326/prelim_reproduced_original_kinematics_vs_new_kinematics.png) |
A few specific comments
- We noticed that two of the bins using the new code had quite large h4 relative to everything else. The full kinematics for these two bins are:
| Bin | v | v_err | sigma | sigma_err | h3 | h3_err | h4 | h4_err | h5 | h5_err | h6 | h6_err | h7 | h7_err | h8 | h8_err |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 49.79863 | 8.85107 | 327.60508 | 8.42134 | -0.00201 | 0.02020 | 0.10338 | 0.02153 | 0.01728 | 0.02326 | -0.01379 | 0.02523 | 0.08345 | 0.02331 | 0.11009 | 0.02719 |
| 19 | 56.81734 | 7.18033 | 278.63727 | 8.27980 | 0.06473 | 0.01635 | 0.09912 | 0.01800 | 0.05684 | 0.01953 | -0.01938 | 0.02214 | 0.04483 | 0.01763 | 0.01100 | 0.01915 |
- And here are the spectra and the accompanying fits for these two bins. It seems like me (looking at these and now looking back at the full spectal fits on this page), there appears to be a few features that should probably be masked out from this fit. There appears to be some very large "absorption"-like or "emission"-like features on the red side of some of the spectra which were not included in the initial spectral mask. It's likely that there is some variation in the kinematics coming from these weird features as you can almost start to see in the spectral fits below.
| Bin 1 | Bin 19 |
|---|---|
| [images/250326/pPXF_plot_00001-1.png]]](/JacobPilawa/TriaxSchwarzschild_wiki_5/wiki/[[images/250326/pPXF_plot_00019-1.png) |
- We also wanted to see which bins corresponded to the largest differences between the new and old spectra. The sigma point which is most discrepent between the original and my reproduction is the same Bin 1 shown above (with the weird h4 point near ~0.1).
- One thing that's a bit strange is that I think this weirdness is only present in trying to reproduce the original kinematics with the new code. To test this, I've plotted a few sets of kinematics below, showing the MC vs. prelim fits using the new code which appear to be in quite decent agreement. I can't exactly figure out how to get a perfect replication of the original code, still working on this.
- The plots below show:
- First column: comparison of the preliminary fits vs. the median of the MC realizations for the "correct" set of kinematics using the newest pPXF code.
- Second column: comparison of the preliminary fits trying to reproduce the original, "incorrect" kinematics using the newest pPXF code.
- Third column: comparison of the original kinematics vs. my attempt to reproduce them in the new code. Note that in this case the largest sigma discrepency is, as I had mentioned, the Bin 1 from above (which also has a very large h8, so more of a sign that this bin is just a bit strange).
- Fourt column: Repeated from above/last time, but this compares the new code's "correct" kinematics vs. original published kinematics currently used in the minimization. I'd like to figure out the reproduction issues, but this is just to show that we still maintain decent agreement between what is currently being used in the minimization vs. what I propose to use in future minimizations.
- Again, I think the key takeaway here is that the lagest sigma discrepency is associated with the highest h4/h8 values, and this lead me to look at the spectra a bit more closely. It looks like we should add a few more places in the spectral mask since this will hopefully clean up the agremeent between sets of data.
| Prelim vs. MC For "Correct" Settings (New Code) | Prelim vs. MC For "Incorrect" Settings (New Code) | Old Code Results vs. Attempt New Code Reprorduction | "Correct" Settings with New Code vs. Old Cude Results |
|---|---|---|---|
| [images/250326/prelim_new_kinematics_vs_mc_new_kinematics.png]]](/JacobPilawa/TriaxSchwarzschild_wiki_5/wiki/[[images/250326/prelim_reproduced_old_kinematics_vs_mc_reproduced_old_kinematics.png) | [images/250326/original_kinematics_vs_reproduced_old_kinematics.png]]](/JacobPilawa/TriaxSchwarzschild_wiki_5/wiki/[[images/250319/caseZ_vs_old_old_old_kinematics.png) |
- And for complete clarity, here's a comparison of the newest kinematics against my attempts to reproduce the original kinematics. Both of these use the newest code, and I've included the case where I compare the fiducial values as well as the MC results.
- I also included distributions of the moments and the errors for these two sets, which are in quite nice agreement:
| Fiducial | MC Results |
|---|---|
| [images/250326/fid_comparison.png]]](/JacobPilawa/TriaxSchwarzschild_wiki_5/wiki/[[images/250326/MC_comparison.png) | |
| [images/250326/mom_comp.png]]](/JacobPilawa/TriaxSchwarzschild_wiki_5/wiki/[[images/250326/error_comp.png) |