ENH: Adding model for Dust extincted ellipticity projection

[jucordero]

Description

Implements an achromatic model for dust obscured ellipticity distributions, from Padilla & Strauss 2008. This would close #341

The model follows a similar structure to ryden04_ellipticity. The dust_extincted_ellipticity samples axis ratios and projects the resultant 3D ellipsoid using Binney formulation in triaxial_axis_ratio_extincted. Here, instead of sampling from a uniform cosine distribution for the inclination angle, extincted_angle_schechter is called to sample a random orientation assuming a Schechter luminosity function and the dust model.

Luminosity function parameters are assumed fixed (i.e. not arrays), which simplifies the broadcasting to get the sample size right, and also fits a more realistic use case.

Checklist

• [x] Follow the Contributor Guidelines
• [x] Write unit tests
• [x] Write documentation strings
• [x] Assign someone from your working team to review this pull request
• [x] Assign someone from the infrastructure team to review this pull request

[jucordero]

I've added a notebook with some detailed explanation of how the functions are built in this enhancement to the projected-ellipticity repository (private)

https://github.com/skypyproject/projected-ellipticities/blob/master/Dust-extincted%20ellipticity%20model.ipynb

I'm not sure the unit tests I added are completely appropriate. Suggestions appreciated.

[jucordero]

Just had a chat at the SkyPy meeting. In essencem, this PR models ellipticity as a function of extinction, while what we really want is a model for extinction as a function of ellipticity, which is a much simpler task. We've also discussed about the possibility that extinction as a function of ellipticity could interfere with the k-correction templates.