Issue181 solar irradiation orientation/tilt

[coveralls]

Coverage remained the same at 69.493% when pulling 5cdfcae8dcc4329d8ad55bec06871e98c1893939 on JrtPec:issue181-solar-irradiation-orientation/tilt into 3165cdab2d1d5e36317b2a86e5d4a9860ed920d4 on opengridcc:develop.

[JrtPec]

Yay! Solar Irradiance on a tilted and oriented plane! Just as discussed in #181.

To see how it works, check out the demo_forecast notebook. Basically, you supply a list of orientation/tilt pairs that you want included in your dataframes, and it just works.

There are some notes though (although I doubt anyone will notice):

• The days method no longer has these stupid "include this", "include that" flags. Everything is included now. (They were originally designed to speed up things when you didn't need certain calculated columns, but that increase in speed isn't really noticeable (if it even exists), and the flags made the code way to complicated and maintainable.
• Daytime averages have been deprecated (again I don't think anyone was using those), because I had to redesign (simplify) the aggregation process. We might want to use a Daytime Average Temperature again in the future, if that is the case just let me know.
• The "get your responses in a special timezone to your liking" has been removed, because it conflicted with the new aggregation methods and overcomplicated the code. You get all dataframes in the local timezone of the requested location. If you don't like that, just add tz_convert. Simple.