bug in get_relative_airmass(model='youngirvine1967')

[wholmgren]

Describe the bug
pvlib python implements the Young and Irvine 1967 model as:

    elif 'youngirvine1967' == model:
        am = ((1.0 / np.cos(zenith_rad)) *
              (1 - 0.0012*((1.0 / np.cos(zenith_rad)) ** 2) - 1))

The published formula is

am = sec(z) * (1 - 0.0012 * (sec(z)^2 - 1))

See http://adsbit.harvard.edu/cgi-bin/nph-iarticle_query?bibcode=1967AJ.....72..945Y&db_key=AST&page_ind=3&plate_select=NO&data_type=GIF&type=SCREEN_GIF&classic=YES

If you squint very hard you might notice that pvlib python's implementation is equivalent to

am = sec(z) * (1 - 0.0012 * sec(z)^2 - 1)

Versions:

• pvlib.__version__: 0.5.2. I probably broke it when cleaning up code in the 0.1-0.2 era.

Additional context
Discovered when implementing #158. Write good unit tests! I'll make a pr with a fix soon.

[cwhanse]

Nice squinting.

[wholmgren]

closed by #546