Montepy is_atomic_fraction does not work as expected

[AlphonseM91]

Describe the bug

When montepy.data_inputs.material.is_atomic_fraction is set to False and I add a new material component and then write the output file, the fraction is positive (thus set to atomic fraction).

To Reproduce

A short code snippet of what you have ran.

zaid='1001.70c'
iso = montepy.data_inputs.isotope.Isotope(zaid)
mat = montepy.data_inputs.material_component.MaterialComponent(iso, 0.2)
# my MCNP_problem
print(pb.materials[99].is_atomic_fraction) --> True
# Then add a new material_component
pb.materials[99].material_components[iso]=mat
pb.write_to_file(output_path)

Then my output file has previous material components with negative fraction (i.e weight fraction) and the one I just added is positive. I find no way to write it with a minus sign before.

MCNP input file snippet

My M99 Material with the hydrogen i just added.

M99     12024.70c -0.4818 12025.70c -0.0610 12026.70c -0.0672 &
        8016.60C -0.39   $ MgO
1001.70c 0.2 

Version

• MontePy 0.5.2
• MCNP 6.1.0

Additional context

I know this should be fixed in MontePy 1.0 but I'd like to know if there is a quick fix I could do because I really need to be able to configure my Materials cards. Thank you!

[MicahGale]

Just to clarify did your material definition start out as?

M99     12024.70c -0.4818 12025.70c -0.0610 12026.70c -0.0672 &
        8016.60C -0.39   $ MgO

So good news: this is tested for and is currently working for the 1.0.0a1 candidate branch: mat_redesign. So a quick fix you could do right now is to switch to that branch, but that's very unstable as its feature development branch. I'm hoping to get an alpha release out within a month.

Bad news this is a bad bug that wouldn't have a quick and easy work around for the user. I can get a patch for 0.5.3 maybe in a week or longer that would be a stop gap measure.

So a few options:

1. expedite a 1.0.0 alpha release where the whole material interface changes (for the better)
2. expedite a patch for this which would require a lot of script rewriting for you to adapt to 1.0.0 in a month or so.
3. I write you a hacky script that manually modifies the syntax tree and circumvents the material object mostly.

Do you have a preference?

[MicahGale]

Also I believe you are saying is_atom_fraction is incorrect? Re-reading 5.6.1 of the 6.3 manual positive is atom fraction, and weight fraction is negative. The double negatives are hard to track.

[MicahGale]

As an update. This problem seemed actually pretty easy to solve. See the PR, next week we might be able to get this reviewed and merged.

[AlphonseM91]

Hi, thank you for your work. I have a technical issue that make me unable to install the code, so I couldn't test it. Maybe I'll wait the merge and it would work.

Also I believe you are saying is_atom_fraction is incorrect? Re-reading 5.6.1 of the 6.3 manual positive is atom fraction, and weight fraction is negative. The double negatives are hard to track.

Actually I was not saying that is_atom_fraction is incorrect : it is correct reading the input, but when it comes to writing it's wrong.

[MicahGale]

@AlphonseM91 this bug fix has been shipped to pypi: https://pypi.org/project/montepy/. You can get it with pip install --upgrade montepy.