gamma angular correlation validation

[gipert]

Here's some work I did today on the topic, I won't have time to finish this for a while so if anyone wants to help I'd be happy :)

• add gamma correlation tests / plots
• think about how to boost statistics
• add comparison with literature

Message from @jasondet:

My code of course computes them on the fly from the nuclear polarity and the multipolarities assigned to each transition so that all gammas in a cascade can be properly correlated with each other, I'm attaching my summary of what the code actually does for your info (I sent it to the G4 guys a decade ago for a paper that's still "in progress"). The math can also produce a compact expression for the angular correlation between any two gammas. When I did validation, I calculated the coefficients myself using expressions in the attached paper from the 60's, see eq 3.46 and the text following it. Also useful for me was the attached tabulation by Taylor et al, and the attached Korean article from 2002 that summarizes the calculation for 2 consecutive gammas in modern notation with numerical values provided for 60Co. To perform the calcs, you might need the functions I implemented in G4Clebsch and G4PolarizationTransition.

G4GammaCorrDoc.pdf
1967RoseBrink_RevModPhys.39.306.pdf
1971Taylor_GGCorrCoeffs_NuclDataTablesA9_1.pdf
2002Kim_JKorNuclSoc34_22.pdf

General note

I think we should adopt this strategy to run tests: CTest only calls pytest, and all tests are written in python.

[gipert]

[ManuelHu]

we will also have to turn off this test for G4 11.2. I get quite a lot of segfaults with enabled gamma correlation (which is a known problem with G4 11.2 when you hit negative-parity states...)

[tdixon97]

Then I think we should switch to requiring 11.3 (or was the fix in some 11.2 patch?)

[ManuelHu]

no, 11.2 never got a fix for this bug (and, I guess, 11.3.0 also had it)

[gipert]

Then I think we should switch to requiring 11.3 (or was the fix in some 11.2 patch?)

though radioactive decays is only one application of remage. Maybe we can just add a warning to the docs, rather than requiring 11.3.whatever.