search

DarkEnergySurvey / ugali

Ultra-faint galaxy likelihood toolkit
MIT License
15 stars 15 forks source link

Marigo2017 isochrones have incorrect values #29

Closed apace7 closed 6 years ago

apace7 commented 6 years ago

The Marigo2017 isochrone tarball has two points which are incorrect. One looks to be the last data point in the file which has a discontinuity to the other points.

test.pdf

kadrlica commented 6 years ago

I see from an off-line thread that these plots are generated with Marigo2017.mag_1 and Marigo2017.mag_2 (rather than the output of Marigo2017.sample):

"I checked that the plots were made with iso.mag_1 and iso.mag_2. It looks like these are directly pulled from the tar balls."

This means that it is probably an issue with the output of the Padova web interface (http://stev.oapd.inaf.it/cgi-bin/cmd_3.0).

kadrlica commented 6 years ago

Do we know if these points are actually unintended or are they an artifact of trying to include post-AGB objects?

apace7 commented 6 years ago

One of the data point is likely artifact (at M~30). The last point in every file has this problem. Looking at a couple of the files the last data point has logg = -9.999.

I think the other point that is disjoint is for the AGB stars.

kadrlica commented 6 years ago

Thanks, that's good to know.

kadrlica commented 6 years ago

I've looked at the affected file. If I'm counting columns correctly, I think it's logL = -9.999 (not logg). This is at least consistent with the mag ~ 30 values). I've also noticed that stage = 9, which I do not see defined as a valid stage here.

I'm going to reach out to the isochrone maintainers and update parsec.py to reject points with stage == 9.