Finish implementing the time-to-CCF-SNR feature

[danechever]

I've added a version of Dimitri's code which inverts the CCF SNR equation to get the exposure time required to get a specific SNR goal. However, Dimitri's version of the code assumes that he can compute the FPS based on the detector linearity and well depth. The current version of PSISIM has a user-provided n_exposures such that we can't dynamically compute the FPS for this function. Dimitri's version of this code also neglects the systematics so it only holds when we are not systematics limited. If possible, it'd be good to find a more general solution.

For now, the version of this function that I've implemented works well in the photon-noise-limited regime.

psisim/psisim/signal.py

Line 76 in b7d63a9

warnings.warn('This function is incomplete at the moment. Double check all results for accuracy.')

This raises a separate question: do we want to dynamically set the n_exposures as Dimitri does in his code or do we want to keep this as a user-provided value? As it is now, the user could set an exposure time, n_exoposures combination that would saturate the detector.

[danechever]

Related to #32 which will allow us to dynamically set the exposure time and frames.