Asy_peakbag: add glitches to the model

[grd349]

Some stars are showing such strong glitch signatures that it makes sense to consider adding glitches into asy_peakbag.

We know how to do this, but will need to generate some prior data. We can probably use the Vrard results in the first instance to get things setup.

[ojhall94]

I'd love to help implement this, I'd like some glitch knowledge.

[grd349]

All the info is in https://arxiv.org/abs/1505.07280 :) . It includes some relations that we can use as a starting prior.

We would probably add an exponential decay on top of the glitch sinusoid.

The work plan would be something like:

Implement the extra parameters and the extra term in the model in asy_peakbag.
Add a prior based on Vrard on the new parameters.
Run on lots of stars.
Use the results of the last step to build a new epsilon prior that now includes the glitch properties.
Remove the Vrard prior.
Run lots of tests.
Merge.

We can discuss offline and then we can add a summary here.