meeting 2025 02 11 n315 - JacobPilawa/TriaxSchwarzschild_wiki_5 GitHub Wiki
Context
- I had forgotten that Jonelle asked for some scalings to test the dependence of the halo on the MGE. To do this, I've added two scalings (0.98 and 1.02) to the models run in this bullet point: https://github.com/JacobPilawa/TriaxSchwarzschild_wiki_5/wiki/meeting-2024-12-12-n315. Happy to add more but wanted to add a bit on both sides for this meeting so we had something to discuss.
- As a reminder, the initial test here was ~145 models over BH and ML. At the time we were only concerned with the MGE dependence on the halo, and potentially fixing the halo and shape could be influencing the model.
- The results at this stage currently look very nice! If I quote the K=80, nu=1.5 results, the results with MGE B3 are consistent essentially at 1sigma with our current results. Specifically the MBh = 2.95 +/- (0.17) billion Msun, and ML = 2.56 +/- 0.03. The halo additionally seems to be roughly where our model point was at, around 4.6e9 +/- (0.08e9) in code units. The halo doesnt seem't to want to move greatly from this location.
- Two really quick notes: *
Diagnostic Plots
- First, here are the 1d panels and the 2D scatter showing our current best fit with MGE B1 compared to the results from MGE B3 + scaled models.
| 1d panels | 2d Panel |
|---|---|
| [images/250211/MGEC_grid_250211.png]]](/JacobPilawa/TriaxSchwarzschild_wiki_5/wiki/[[images/250211/2d_scatter_240211.png) |
- And the results from running GPR/dynesty on the models. I've chosen a few different K and nu here, and have vertical plot-style comparison just below. Note that we haven't really been considering the K=60 models has having a high enough cutoff K to be reasonable, so there is a bit of wonkiness in these points.
| K=60 | K=80 | K=100 | |
|---|---|---|---|
| nu=0.5 | [images/250211/250211_MGEC_grid_alpha_K60_nu0.5-1.png]]](/JacobPilawa/TriaxSchwarzschild_wiki_5/wiki/[[images/250211/250211_MGEC_grid_alpha_K80_nu0.5-1.png) | images/250211/250211_MGEC_grid_alpha_K100_nu0.5-1.png | |
| nu=1.5 | [images/250211/250211_MGEC_grid_alpha_K60_nu1.5-1.png]]](/JacobPilawa/TriaxSchwarzschild_wiki_5/wiki/[[images/250211/250211_MGEC_grid_alpha_K80_nu1.5-1.png) | images/250211/250211_MGEC_grid_alpha_K100_nu1.5-1.png |
- And here's a vertical-style plot comparing these results. As usual, bottom-to-top increases the K we assume, and I switch back and forth between nu = 0.5 and nu = 1.5.
| Plot |
|---|
| images/250211/250211_MGEC_test.png |
