Simplified DM Models

[ChrisJChang]

This is the merge containing all of the model code for the Simplified DM models project (Scalar, Dirac and Majorana). The vector DM model will come separately at a later date (it is in the second paper).

Main Changes:
-- Addition of 3 models run from GUM. These are named DMsimpVectorMed
-- Additional SpecBit module functions for Unitarity bounds (Majorana and Dirac)
-- Significant changes to the ColliderBit_InterpolatedYields code to include MonoJets and DiJets for these models.
-- 3 New interpolators in Utils/src/interp_collection. These are linear interpolators in 2, 4 and 5 dimensions. I wrote the 2D linear interpolator as an alternative to the existing gsl one because I found that it was very often unstable and would give vastly incorrect results. I wrote the 4D and 5D interpolators so that they could handle incomplete interpolation grids (we imposed additional cuts on which grid points to simulate in this study). This has led to them being a tad more complicated than they perhaps could be. They are initialised with an extra option specifying to let them have missing points or throw an error.

Note: One of the MonoJet data files was too large to push to GitHub (it was ~114Mb when the limit is 100Mb). I pushed the compressed file and added a note in the example YAML file for this model that this needs to be extracted first. If this is a problem, let me know.

[patscott]

Given that sort of size, it might be that we want to put some (or all?) of these interpolation grids into a downloadable resource rather than shipping them with the code. This could be a sensible use of @anderkve's idea about specifying downloadable datasets in #324.

[ChrisJChang]

I've updated this with a backend "DMsimp_data_1.0" that downloads these interpolation grids from Zenodo.

[anderkve]

Great, thanks for adding the backend. Once the work on #324 is done we can change it to be a downloadable dataset rather than backend in the cmake system.

[tegonzalo]

Not a proper review, just a few comments on thinks I picked up after a quick glance. @anderkve should do the proper review :)

[anderkve]

Not a proper review, just a few comments on thinks I picked up after a quick glance. @anderkve should do the proper review :)

Thanks! A couple of these overlap with comments I already have noted in my ongoing review (not published), but I'll just remove them from my review then to avoid duplicates.

[anderkve]

Great work, @ChrisJChang!

Most of my comments and requests should be very quick, I think. There are a couple of suggestions for larger changes in ColliderBit_InterpolatedYields.cpp to reduce code duplication.

[anderkve]

Hi @ChrisJChang -- sorry for the delay. This all looks good to me, so I'm happy for this to be merged once the current CI checks complete. (I merged in the latest master and fixed a small merge conflict, hence the new tests.)

@tegonzalo, since you are also listed as reviewer, did you want to take a final look before we merge?

PS: I don't have a Mathematica installed on my current laptop yet, so I can't actually do a final test of the full chain from running GUM --> running a GAMBIT scan with DMsimp.yaml. But I'm sure that works as expected given how many times you must have run that chain by now. :)

[tegonzalo]

I'll trust your judgement, if you say it can be merged, then by all means. @ChrisJChang are you happy to make the release of 2.3 yourself or do you prefer that either Anders or I do it?

[ChrisJChang]

I'm happy to make the release, although I've never done it before. I can see that there is a part in GitHub to do this with a certain branch (i.e. master). Is this all that there is to it? Also, I know that an email must be sent to the whole community with a list of changes. Do you have a list of the other changes since the last release?

…

________________________________ From: Tomas Gonzalo ***@***.***> Sent: Wednesday, October 5, 2022 6:02 PM To: GambitBSM/gambit ***@***.***> Cc: Christopher Chang ***@***.***>; Mention ***@***.***> Subject: Re: [GambitBSM/gambit] Simplified DM Models (PR #360) I'll trust your judgement, if you say it can be merged, then by all means. @ChrisJChang<https://github.com/ChrisJChang> are you happy to make the release of 2.3 yourself or do you prefer that either Anders or I do it? — Reply to this email directly, view it on GitHub<#360 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AOJAZHYHDJBRHWA73XQNH7DWBUYY3ANCNFSM5YWAA7IQ>. You are receiving this because you were mentioned.Message ID: ***@***.***>

[tegonzalo]

Check the code policies in the repo, there are instructions there on how to do it. The only thing you might not be able to do just yet is sending the announce email. Write to Ankit so that he adds you to the list of admins for gambit-announce list.