improve function argument ordering in pvsystem

[wholmgren]

This PR improves some function signatures as discussed in #197.

Also addresses the input/output type issues described #201 for ashraeiam, physicaliam, and snlinverter.

Adds test_tools.py.

I'll merge tomorrow unless there are comments.

Closes #197.

[jforbess]

It looks good at a high level, but I noticed that you didn't do the same type of reordering for the sapm function. I see that you have just updated sapm to better align with Matlab, but the intention was only to align algorithmically, right, so we'd still want the parameters ordered with effective_irradiance and temperature first?

[wholmgren]

Good catch @jforbess. The functions introduced in #218 should be changed, too.

[wholmgren]

@jforbess I updated the sapm functions. I'll let this sit for another day.