get TRACTOR data for given galaxy

[Amarchuk]

Hi,

Not sure if this is a right place for this question. I try to get some properties of TRACTOR models in DR9 for specific galaxies, for example get B/T (bulge to total) for PGC146275 (coordinates 66.392840 -8.535853). Is there any simple way to do this?

I can find the proper brickid, which is 0662m085 and then find all models, which are lie within some radius of galaxy's center. But I am not sure how to attribute them to one or another object. In this particular example there are 2 EXP, 1 SER and 2 REX models within 10 arcsec. Is it a good approach to just take the closest EXP, SER and REX model (not sure about the last one)?

Thanks a lot

[dstndstn]

Hi,

If you look in the viewer with the DR9 catalog overlay on, you can just barely see that there are multiple sources right at the center of that galaxy. https://www.legacysurvey.org/viewer/?ra=66.3925&dec=-8.5358&layer=ls-dr9&zoom=16&sources-dr9

Our modeling can often go awry on large resolved galaxies like this; we'll end up "explaining" the galaxy using multiple galaxy entries in the catalog. Depending on your interest, you might want to sum their fluxes, or take the brightest.

If you look in the catalog at the REF_CAT column, you should see that one of them is "L3", which means it came from the Siena Galaxy Atlas. If you want to look at a single catalog entry, that's probably a good choice.

Also, we don't really have a B/T ratio in the catalogs. All the models we're fitting here are achromatic -- galaxies are assumed not to have color gradients - and we don't fit multi-component models (at least not on purpose!). The fitting is really geared toward smaller, unresolved galaxies.

If you're looking at larger galaxies, definitely check out the Siena Galaxy Atlas -- it has many more detailed measurements.

cheers, --dustin

On Mon, Mar 1, 2021 at 4:42 AM Amarchuk notifications@github.com wrote:

Hi,

Not sure if this is a right place for this question. I try to get some properties of TRACTOR models in DR9 for specific galaxies, for example get B/T (bulge to total) for PGC146275 (coordinates 66.392840 -8.535853). Is there any simple way to do this?

I can find the proper brickid, which is 0662m085 and then find all models, which are lie within some radius of galaxy's center. But I am not sure how to attribute them to one or another object. In this particular example there are 2 EXP, 1 SER and 2 REX models within 10 arcsec. Is it a good approach to just take the closest EXP, SER and REX model (not sure about the last one)?

Thanks a lot

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/legacysurvey/legacypipe/issues/679, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAIEH7PHE4BEBN3QMRWOT5LTBNOQRANCNFSM4YMHILCQ .

[Amarchuk]

Thank you, Dustin!

Don't get the "achromatic" part of your answer (can't we just attribute SER to bulge and EXP to disk in cases like PGC146275, when there are several models? Probably not), but overall it was very clear, thank you.

I check SGA before and just notice REF_CAT by myself, check it for 2000 galaxies I have and see most of the models with L3 are SER. And yes, I actually want to find all entries in the catalog, related for the given galaxy. Is it true, that this can't be done in any reasonably effective way besides just get all models close to the center and maybe filter them by flux/shape_r?

[dstndstn]

The models aren't fit simultaneously, so you definitely should not attribute one model to each galaxy component.

You could try checking the "maskbits" column for the GALAXY bit -- indicating the source is inside the ellipse of the SGA galaxy. But that will pick up stars and other objects inside that ellipse as well. https://github.com/legacysurvey/legacypipe/blob/master/py/legacypipe/bits.py#L60

In short, there is no magic that will let you find all the models that "belong" to a galaxy, so you just need to do something that makes sense for your science.

[Amarchuk]

@dstndstn Thanks for the explanations!