search

DarkEnergySurvey / ugali

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

Update absolute magnitude calculations #60

Closed kadrlica closed 5 years ago

kadrlica commented 5 years ago

These changes generalize the absolute_magnitude and absolute_magnitude_martin calculations for all surveys and filter combinations. This is implemented by creating a copy of the isochrone with the same properties (age, metallicity, distance) in the SDSS g,r filter system. The absolute magnitude is then calculated directly from the transformation equations from Jester 2005.

This solution is not perfect! When calculating the Martin magnitude, we are now cutting on the SDSS g-band magnitude rather than the band_1 value for the data set we are analyzing. Given the already flawed nature of absolute_magnitude_martin, I'm ok with this change since it will have a minimal impact on DECam data (due to the similarity between the filter sets). However, users need to be aware that the detection cut will be placed on g-band regardless of the band_1 value for their analysis.

This PR is not ready to merge (I'll remove the absolute_magnitude2 and absolute_magnitude_martin2 methods); however, I would like to hear feed back on the design from @bechtol and @sidneymau.

sidneymau commented 5 years ago

the Mv and Mv_martin functions appear to now be inconsistent:

  Mv:
  - -0.536496493824758
  - [-0.21438180994432998, -0.7983019321085529]
  Mv_martin:
  - -0.02143419887591655
  - [0.703355944803804, -1.6784901363401887]