Baluev method plus an notebook example

[nhuntwalker]

Added Baluev 2008 method for calculating false alarm probabilities to lomb_scargle.py. Also added an ipynb for demonstrating the comparison between Baluev 2008 theoretical model and data from Gaussian white noise in the examples directory

[jakevdp]

Looks good. A couple things:

• this is only valid for single-term lomb-scarlge, yes? In that case, this implementation should probably be in LombScargle rather than the base PeriodicModeler, where it would be inherited by other modelers for which the assumptions don't hold. I would probably add a NotImplemented method in the base class, and specialize it with the Baluev method in the LombScargle class.
• we need some unit tests for the functionality. Probably just compute what the result should be for a known input, and confirm that the output matches (in case of future refactorings)

Thanks for putting this together!

[jakevdp]

Looks great, Nick! It looks like something has been changed in master, and there's now a merge conflict here. Can you merge your branch with master and fix the conflict?