Add Global Spline Fit model

[The-Ludwig]

Hi, first off: Sorry again for pushing to main.

I implemented the GlobalSplineFit primary model by Dembinski et. al. Proceeding for the model can be found here and an overview of the model as well as the data can be found here.
This model is, as far as I understand it, the state-of-the-art model right now and is actually fited with real data.

Unfortunately, the model is not yet publicly available, yet, since the developers are still working on the paper. Once the paper is released, they will also release a python package. The best thing to do right now (according to Hans Dembinski) is to download the tabulated data and interpolate between the values.

Of course, this raises questions for this pull request, as it is unclear how to distribute these tabulated values and how the copyright on that will work (that's also why the Pre-Commit tests will fail).
I think shipping the .txt file with the source code is okay for now. Sadly, the website does not seem to have an api, which would permit downloading the table in an install script. The other option I can think of is to upload the table in an release as an asset and downloading that asset in the install script.

To interpolate I use scipy.interpolate.CubicSpline. Since I did not want to bloat the nice and short dependency list, I added it as an optional dependency and raise an error if scipy is not found.

Anyway, please let me know what you think about this and how I can improve this PR.

Context

For our prompt-muon analysis this model will be very interesting to test, since it does seem to produce the highest prompt-muon fluxes compared to other models.

[codecov-commenter]

Codecov Report

Patch coverage: 100.00% and no project coverage change.

Comparison is base (8813a86) 100.00% compared to head (47d73a1) 100.00%.

📣 This organization is not using Codecov’s GitHub App Integration. We recommend you install it so Codecov can continue to function properly for your repositories. Learn more

Additional details and impacted files

@@            Coverage Diff            @@
##              main       #10   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files           12        12           
  Lines          669       693   +24     
=========================================
+ Hits           669       693   +24     

Impacted Files	Coverage Δ
src/simweights/_fluxes.py	100.00% <100.00%> (ø)
src/simweights/_pdgcode.py	100.00% <100.00%> (ø)

☔ View full report in Codecov by Sentry.
📢 Do you have feedback about the report comment? Let us know in this issue.

[The-Ludwig]

@kjmeagher From my side this is ready for review. As said, the pre-commit is failing, since I do not know how to give the right license to the GSF data table.

[kjmeagher]

this looks good aside from the comments above. For the data file license you can ask @ @HDembinski if he can license it under BSD-2-Clause

[HDembinski]

Hi Ludwig, btw IceCube internally should have the GSF model code already.

For public code, it is an acceptable compromise to use the public tables, as you do. The tables themselves are not under any license, they are not source code. Perhaps I should nevertheless clarify under what permissions one can use the tables.

I hereby declare that the tables are available under CC BY 4.0 https://creativecommons.org/licenses/by/4.0/. Please add a license file to the package and indicate that the tables fall under that license. Scipy is a good example on how to handle code/data licensed under compatible licenses, see e.g. https://github.com/scipy/scipy/blob/main/LICENSES_bundled.txt

[The-Ludwig]

@HDembinski Thanks for the reply! I added the license, although I'm not sure how to give details about compatibility.

@kjmeagher Pre-commit is now failing because of the large scipy dependency, what can we do about that?

[HDembinski]

CC BY 4.0 is compatible with BSD 2-clause, in the sense that both give you similar distribution and modification rights. I chose CC BY 4.0 mainly because it guarantees attribution.

[kjmeagher]

This all looks good. Thanks to @HDembinski for the relicense.