search

hippke / tls

Transit Least Squares: An optimized transit-fitting algorithm to search for periodic transits of small planets
MIT License
48 stars 25 forks source link

Please provide a [Rs] #75

Closed 3fon3fonov closed 5 years ago

3fon3fonov commented 5 years ago

Hello Michael!

Can you please provide also as an output a_rs (semi-major axis in units of stellar radii), as this is an input parameter in "batman", and is easy later to take TLS results and feed batman again.

I see you have "rp_rs", which is great, but the semi-major axis is missing.

hippke commented 5 years ago

That‘s unfortunately not possible. It requires fitting a full model, e.g. using the PyTransit, juliet, or exoplanet package. You can calculate a rough estimate using Kepler‘s law with some assumptions on the masses of the bodies.

3fon3fonov commented 5 years ago

I am sorry, I am not sure I understand. First, I can fit a full model myself using my “Exo-Striker” tool (which you should probably advertise more as it also uses the TLS). Juliet, The Exo-Striker, and correct me if i am wrong TLS, use the batman-package, which uses “a_rs” as a parameter. So why not? How do you “feed” Batman inside the TLS?

What usually i do with the Exo-Striker is to run TLS and when TLS is done adopt its parameters in order to star a full model, which so far relies on batman.

hippke commented 5 years ago

Yes, TLS uses batman to generate models. If you have a prior on any model parameter (perhaps from stability arguments, known planets, etc.), you can actual feed it into TLS and use that. There are also the model pre-sets default, grazing and box. If no model parameters are given, TLS uses the default model. This has been optimized with the average transit shape of small planets (Figure 2 in the paper). In this model, a_rs=23.1 (see source). As this parameter is a fixed value, one should run a full model fit afterwards to find a "better" a_rs. After all, you want all parameters to be fitted, including limb darkening etc. And thanks for the pointer to Exo-Striker - very cool package! I will advertise that in the future.

3fon3fonov commented 5 years ago

Ok that explains a lot. I just thought that you are also varying these parameters too. It was probably naive from my side to assume so. It must be very expensive to vary yet another dimension. Sorry if i missed this in the paper. About the Exo-Striker, i showed you a demo about 6 months ago, you must had forgotten.

It is still work in progress, but it gets more and more mature.

Trifon