This is the first time I do this - the format might change.
This week has been a bit more messy than I would have liked. It has been quiet and nice, but cold in the office. I felt rather fatigued in the beginning of the week, but it has gotten less bad yesterday and today.
asterism-wise, I have been working on untanglin and debugging why the MLM model produces non-physical or just plain odd luminosities and radii.
Focusing just on the prior predictive distributions has helped quite a bit.
It is not solved but I do feel like I know what to debug and what to try this afternoon.
I met with Bill Monday regarding what we would like to investigate and do for post docs at the School this semester. Ideas were exchanged - I need to send some emails and get some help from other reps.
PLATO-wise I need to write a small summary of the progress in my work package and priorities for the remaining work.
I plan to send it later today. Maybe I should just get the unit tests done so that I am just waiting for the final version of SAS_tools.
I spent my Wednesday deep diving into the literature. Mostly on reaction rates in CNO cycles and how that affect the luminosities in the red giant phase.
This was not intentional - I had other plans. But I did learn something so I guess that is nice.
I also tested that predictions from the NN align with my expectations using the averaged HR-diagram and the GUI I made a couple of weeks ago.
Everything but fracage looked alright. I ran tests including Teff in the output along with photometric radii and luminosities (outlook_retainer_alias).
I also ran tests tweaking the weights in the loss function (cyclist_fable_going).
Higher weight/smaller scale weights on especially radius seems to be what we should do when the issue with fracage is resolved. Teff can stay where it was before.
I met with Jens Thursday regarding Rogue and Hennes. The paper is basically done, but he though that the frequency matching algorithm was acting up for some of the l=2 modes and thus not working properly for these evolved stars. It does, we investigated, and his tests did bias the result. We discussed what surface correction to use. We looked at the corner plots and most things looks promising and rather nice.
I promised I would give the dynamics of the stars a look - the targets are extremely interesting both from a seismic and Galactic point of view. They are 2 kpc away. I’ll look into their Galactic origin next week and send Jens something before he returns from vacation.
asterismfracage bug