Bad solar flux data corrupts MSIS

[greglucas]

Plotting with bad solar flux values causes all subsequent values to produce "nan" output (not just that single day).

>>> from pymsis import msis
>>> msis.run("2005-09-10", 0, 0, 150)
array([[          nan,           nan,           nan,           nan,
                  nan,           nan,           nan,           nan,
        3.0979247e-17,           nan, 9.8023016e+02]], dtype=float32)

Specifically, the date 2005/9/10 causes issues if you try to plot over that day.

• Should we put limits on valid/realistic F10.7 values?
• Should we automatically interpolate over those values if there is a radio burst or something else going on that day?

Related to: #28

[greglucas]

This was fixed by #46

UserWarning: There is data that was either interpolated or predicted (not observed), use at your own risk.
  warnings.warn(
array([[1.5667554e-09, 2.2181889e+16, 2.3551942e+15, 1.5285736e+16,
        1.6321905e+13, 1.1822177e+12, 4.9340983e+13, 1.2026543e+13,
        3.0979247e-17, 4.1002158e+12, 6.0943359e+02]], dtype=float32)