Update basicsim.py

[marquezj]

This test checks for coherent scattering physics for NCrystal materials in OpenMC.

[tkittel]

Thanks @marquezj !

So since we didn't have the in-person meeting about the new development tools yet, here is how to quickly run some checks locally before pushing. First of all, you probably want to add ./devel/bin to your PATH so you get the new ncdevtool command (you can simply source ./devel/setup.sh which does this for you):

. ./devel/setup.sh

Now you can launch ncdevtool checks for some very fast code checks. However, since this is a development branch, I currently have a few fixme's left in the code which will cause the fixme check to fail. So try everything except that (the fixme check is only run in the CI workflow named fixme, so it does not cause other spurious failures):

ncdevtool check -n fixme

That is most likely all you need since you are just adding a python test script, but for reference here are a few other commands that one could in general use to check before pushing:

• ncdevtool cmake (builds with CMake and runs ~100 tests with CTest)
• ncdevtool sb -t (builds with simplebuild and launches tests)

In devel/reqs you can find environment files for conda and pip with all dependencies.

In this particular case, the ncdevtool check -n fixme complains that you have too many trailing empty lines in the file you added.

[tkittel]

Btw., I can tell that I have to fix something so that the access token we use to get the openmc data is also available in PRs. Otherwise the openmc workflow will always fail in PRs.

[tkittel]

Hmm... it is a bit complicated security wise. I think for the short term we might have to live without the openmc job succeeding in PRs, perhaps we can find a fix later.

[tkittel]

To save you some time, you should try to run ncdevtool check -n fixme before pushing, since it fails with:

Run ./devel/bin/ncdevtool check -n "fix""me"
Trailing spaces at end-of-line found in /home/runner/work/ncrystal/ncrystal/devel/openmc/basicsim.py

Additionally, could you rename TRUE_ to REF_ in your variable names? The latter seems a bit less omniscient ;-)

[tkittel]

The openmc CI jobs ran succesfully, which is nice. However, could you add a few more printouts when you are performing tests versus the reference, etc. Because the log file just ends with:

 ============================>     RESULTS     <============================

 Leakage Fraction            = 0.99334 +/- 0.00003

which makes one wonder if there were any checks at all (I guess there were, but without a printout it is hard to tell if the code by mistake aborted too early or what).