[Mesa-users] Regarding some clarifications

[Meridith Joyce]

Hi Nathatli,

Be wary of (pseudo) chi-sq approaches in situations where many models
intersect an uncertainty domain, as is likely to be the case when fitting
tracks or isochrones to single-star observations. Any models that produce
normalized (pseudo) chi-sq scores below 1 should be treated as equally
valid; it is not meaningful to rank their consistency beneath this
threshold---to do so would be an error of overfitting.

Forgive the shameless self-referencing, but these kinds of problems are one
of my areas of expertise. Maybe some of these papers would be useful:

(1) Determining the best fit to a single star with asteroseismic mode IDs
https://ui.adsabs.harvard.edu/abs/2021MNRAS.502.1633M/abstract

(2) Fully agnostic, maximum likelihood approach to isochrone fitting when
observational uncertainties are large (analogous to your situation in many
ways)
https://ui.adsabs.harvard.edu/abs/2023ApJ...946...28J/abstract

(3) Discussion of the importance of *indpendent* observational constraints
when doing fit statistics
https://ui.adsabs.harvard.edu/abs/2018ApJ...864...99J/abstract

(4) more fitting
https://ui.adsabs.harvard.edu/abs/2018ApJ...856...10J/abstract

Good luck,



On Fri, May 5, 2023 at 11:46 AM Nathali Perera via Mesa-users <
mesa-users at lists.mesastar.org> wrote:

> Hello everyone,
> I'm a beginner to MESA and I'm working with this software for my final
> year research project. I'm currently modeling detached binary star systems
> which have almost similar masses like 1.875 and 1.856 solar mass. Since the
> binary star system is detached and not assumed to be mass transferring,
> these two stars were modeled as single evolutionary stars. Then I tried to
> change their mass (starting from 1.75 to 1.9 with a step size of 0.05),
> chemical composition (changes 0.01, 0.014, 0.018), mixing length (change as
> 1.8 and 0.2), and overshooting (0.005, 0.01,0.015)to see the effect for
> their evolutionary tracks. My question is, if there is more than one
> evolutionary track that intersects the star's error box, how do we know the
> best suitable track for that star?
>
> Secondly, I would like to clarify some other questions regarding the code
> which I used. How can we select when to use 'kap_file_prefix' or
> 'use_type2_opaities'? Should we use 'kap_file_prefix' for stars that are
> rich in hydrogen and 'use_type2_opacities' for stars that have heavy
> metals? If I use both of these two conditions at the same time will MESA
> select one condition among them?
>
> And if anyone sees any mistake in my code, please let me know. I would
> like to learn and correct my mistakes.
>
> And I'm using mesa release r22.11.1.I have attached a figure that I
> obtained from my work and in that figure, I have marked the binary star
> system with error bars and several evolutionary tracks.
>
> Thank You
> Nathali Perera
> _______________________________________________
> mesa-users at lists.mesastar.org
> https://lists.mesastar.org/mailman/listinfo/mesa-users
>
>

-- 

Dr Meridith Joyce
Marie Sklodowska-Curie Actions Widening Fellow
CSFK Konkoly Observatory, Budapest, Hungary 1121
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20230508/0a78a48b/attachment.htm>