Handle SMEFTsim topU3l <-> dim6top rotations

[bryates]

This is an attempt to handle rotations between SMEFTsim topU3l and dim6top in IM. The changes to the datacard maker could use some refactoring, but everything works.

[bryates]

I should add that the text2workspace step crashes on glados when using expr for WCs, but it runs on lxplus. We will need to debug this in the future.

[codecov]

Codecov Report

Attention: Patch coverage is 95.45455% with 1 line in your changes missing coverage. Please review.

Project coverage is 25.60%. Comparing base (3471b74) to head (a4c9cf1).

Files with missing lines	Patch %	Lines
topeft/modules/datacard_tools.py	95.45%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #458      +/-   ##
==========================================
+ Coverage   25.29%   25.60%   +0.30%     
==========================================
  Files          31       31              
  Lines        4977     4999      +22     
==========================================
+ Hits         1259     1280      +21     
- Misses       3718     3719       +1     

Flag	Coverage Δ
unittests	25.60% <95.45%> (+0.30%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[Andrew42]

Interesting, how does this deal with the decomposition step? i.e. where we break up the parameterization into separate "sub-"processes in order to encode the WC dependence.

[bryates]

Interesting, how does this deal with the decomposition step? i.e. where we break up the parameterization into separate "sub-"processes in order to encode the WC dependence.

We don't do that anymore with IM, everything is just stored in the scaling file. If we wanted to do something similar in AAC, it would probably still be casting the "new" WCs as functions of the "old".

[Andrew42]

Interesting, how does this deal with the decomposition step? i.e. where we break up the parameterization into separate "sub-"processes in order to encode the WC dependence.

We don't do that anymore with IM, everything is just stored in the scaling file. If we wanted to do something similar in AAC, it would probably still be casting the "new" WCs as functions of the "old".

Ah that's right, sorry! That does make things a lot simpler.

Looking at the code, I'm trying to understand what's going on in the block where you open/load the yaml file. My naive understanding is that the yaml file is mean to basically define a set of equations/functions whose arguments are purely WCs of the other basis, so when combine does the minimization in terms of those WCs the effect gets translated into the appropriate scaling in terms of the new/other WCs. What I don't get is all this manipulation you then need to do after you've loaded the yaml file. I also don't understand the use of eval in this block of code.

[bryates]

Interesting, how does this deal with the decomposition step? i.e. where we break up the parameterization into separate "sub-"processes in order to encode the WC dependence.

We don't do that anymore with IM, everything is just stored in the scaling file. If we wanted to do something similar in AAC, it would probably still be casting the "new" WCs as functions of the "old".

Ah that's right, sorry! That does make things a lot simpler.

Looking at the code, I'm trying to understand what's going on in the block where you open/load the yaml file. My naive understanding is that the yaml file is mean to basically define a set of equations/functions whose arguments are purely WCs of the other basis, so when combine does the minimization in terms of those WCs the effect gets translated into the appropriate scaling in terms of the new/other WCs. What I don't get is all this manipulation you then need to do after you've loaded the yaml file. I also don't understand the use of eval in this block of code.

The manipulations were my (probably inefficient) way to keep things general. I saved the cosine of the Weinberg angle as CW instead of hard coding it. The eval call is to compute SW.