avoid and silence warnings

[wholmgren]

• Closes issue #xxxx
• I am familiar with the contributing guidelines.
• Fully tested. Added and/or modified tests to ensure correct behavior for all reasonable inputs. Tests (usually) must pass on the TravisCI and Appveyor testing services.
• Updates entries to docs/sphinx/source/api.rst for API changes.
• Adds description and name entries in the appropriate docs/sphinx/source/whatsnew file for all changes.
• Code quality and style is sufficient. Passes LGTM and SticklerCI checks.
• New code is fully documented. Includes sphinx/numpydoc compliant docstrings and comments in the code where necessary.
• Pull request is nearly complete and ready for detailed review.

Avoid or silence a lot of test warnings, plus a few that users might encounter (klucher and calcparams_desoto divide by 0).

Also some pep8 along the way. Waiting on #697.

[cwhanse]

Lots of format fixes !

[wholmgren]

Warnings status:

• py27-min: 2 old numpy warnings. let's overlook this.
• py27: forecast.py import warning (fix this). forecast's NDFD and HRRR_ESRL models produce warnings (maybe silence). pvfactors uses deprecated pvlib api -- maybe pvfactors 1.0.0 fixes this, but it's not compatible with our wrapper. cc @anomam
• py34: all of py27 plus invalid warning in test_ashraeiam_scalar. maybe does not warning invalid values in np.minimum and np.maximum because it's using very old pandas and numpy and this was a more recent problem. We should drop this build soon anyways.
• py35: all of py27 plus warnings about invalid values in np.minimum and np.maximum because conda-forge supplies only numpy 1.15 for python 3.5.
• py36: same as py27
• py37: only pvfactors warnings (this build doesn't test forecast.py).

To do:

• silence forecast.py import warnings
• look into NDFD and HRRR_ESRL warnings
• overlook everything else

Lots of format fixes !

Yeah :( Trying to strike a compromise with my IDE's linter. The stickler objections here are wrong. Maybe a flake8 bug.

[mikofski]

since you're making formatting changes, possibly consider:

result = irradiance.get_extra_radiation(
    times, method='nrel', how='numba', numthreads=4)

this is blacker code, tho not quite completely Black

[wholmgren]

I agree this better formatting in this case. Overall, I prefer the grey zone.

[mikofski]

Everything lgtm so far, thanks!

[wholmgren]

@cwhanse ready to merge if you approve of the last few commits. I'd like to keep the pvfactors and forecast.HRRR_ESRL warnings as a reminder of work to do.