Allow for MDCalc to be run with relax_cell=True during the initial relaxation

[Andrew-S-Rosen]

Closes #95 by making it possible to relax the unit cell before the MD run.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 96.29%. Comparing base (971197a) to head (945d676).

Additional details and impacted files

@@           Coverage Diff           @@
##             main      #96   +/-   ##
=======================================
  Coverage   96.29%   96.29%           
=======================================
  Files          22       22           
  Lines        1160     1161    +1     
=======================================
+ Hits         1117     1118    +1     
  Misses         43       43           

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[Andrew-S-Rosen]

@shyuep - This is ready for your review.

[shyuep]

@rul048 you are the maintainer. Pls review and merge.

[Andrew-S-Rosen]

Thanks, will ping @rul048 if there is anything going forward. 👍

[rul048]

Hello Prof. Rosen. Thank you for your contribution. The initial reason we set relax_cell=False while only allowing users to optimize the atomic positions was to achieve the similar functionality of the minimize command in LAMMPS within MDCalc.

This approach ensures that the volume of the input structure remains exactly the same as the one used for the subsequent MD simulation, especially for NVT/NVE ensembles. However, it does limit flexibility for users who wish to perform NPT simulations with a fully optimized structure.

Thank you again for pointing this out in issue #95. Your method offers a flexible solution, and the implementation looks good to me.

[Andrew-S-Rosen]

@rul048: Thank you for reviewing this PR and for your detailed explanation of the history, which certainly makes a lot of sense. I greatly appreciate it!