Added WG00 attenuation curves

[dcorre]

The attenuation curves of Witt & Gordon are now included. There are some tests for WG00 and C00, so far only for testing the inputs (wavelength and tau_V or Av).

The docstrings for both classes are up to date.

There are 2 examples for the documentation in fit_attenuation.rst, showing how to use these 2 classes to fit an extinction curve.

Inputs are now in microns.

[dcorre]

Hello Karl,

apparently during the building of the documentation scipy is not found when astropy.modeling is called... I do not understand what is the reason. May be you should add scipy to the required packages to install?

On my laptop I did python setup.py build_docs and it worked fine.

David

[karllark]

You can add in scipy as a requirement in setup.cfg. The line is near the bottom.

[dcorre]

I tried to add scipy to setup.cfg as install_requires = astropy, scipy but during the docs building scipy is still not found. It brought a new error with lapack library for one test so I went back to: install_requires = astropy

I can not figure out how to solve the problem for the documentation. It is building correctly on my laptop. May be need to specify to install scipy in the docs/conf.py ?

Bug seems similar to: https://github.com/rtfd/readthedocs.org/issues/1639

I'll look in more details in a few weeks

[karllark]

Not sure the scipy error. Have not seen this before and the dust_extinction package uses the same fitter and so should get the same error. This may be an issue with readthedocs that might be solved now or soon. Go ahead and fix doc "===..." issues and we'll see if new tests with travis work.

[karllark]

BTW, amazing amount of work you've done on this already! Very cool! Looking forward to playing around with the new C00 and WG00 models.

[karllark]

I found the issue. Please change the conda dependencies in the .travis.yml file to be:

    - CONDA_DEPENDENCIES='Cython scipy'

That should fix the scipy issue in the build_docs test.

Found by looking at the dust_extinction travis build. Must have "solved" this problem in the past for that package and forgotten the solution. Argh.

[dcorre]

I have corrected the '===' lines in the docs but the problem with scipy persists. I'll ask the people at my lab when I'll come back there at the end of the month.

If you want to play with the data, I think you could accept the pull request as the tests for the docs also failed for the first one you accepted. The docs should build correctly on your computer typing python setup.py build_docs I have not modified the global documentation so far, so we will work on it soon

[dcorre]

I just saw your message, this sounds to be the solution!!

[dcorre]

Thank you it worked! I just added scipy as we do not use cython so far.

[karllark]

Most excellent. Going to merge now.

[karllark]

Addresses issue #4.