Work on impacts

[davidwalter2]

• Moving impact definitions into a separate folder under rabbit/impacts/ with pure tf functions that get called by the fitter.
• Adding back old global impacts definition based on the gaussian approximation as in https://arxiv.org/abs/2307.04007
  • Implementing this for observables as well
• fixing freezing parameter option

Adapting plotting scripts

• Steering impacts in the plotting scripts now with "--impactType "

[cippy]

I need more time to inspect the whole set of changes, but I was thinking about two things:

1. In the python files where impacts are defined, in the top comments you might also add a reference to the papers where traditional or global impacts are described (as we cite them in the W boson mass paper), so that future people will be able to understand where those come from.
2. I noticed that now we have some options defined in multiple scripts, such as the one to specify what impacts to use. Is it worth creating a common rabbit parser from where to pick the common options so to avoid repetitions?

[davidwalter2]

2. I noticed that now we have some options defined in multiple scripts, such as the one to specify what impacts to use. Is it worth creating a common rabbit parser from where to pick the common options so to avoid repetitions?

I've decided to open a new PR on this: #125

[davidwalter2]

1. In the python files where impacts are defined, in the top comments you might also add a reference to the papers where traditional or global impacts are described (as we cite them in the W boson mass paper), so that future people will be able to understand where those come from.

done