Make the seed a global variable

pyMBE-dev / pyMBE

pyMBE provides tools to facilitate building up molecules with complex architectures in the Molecular Dynamics software ESPResSo. For an up-to-date API documention please check our website:

https://pymbe-dev.github.io/pyMBE/pyMBE.html

GNU General Public License v3.0

5 stars 7 forks source link

Make the seed a global variable #50

Closed davidbbeyer closed 2 months ago

davidbbeyer commented 2 months ago

Solves #44, #43

@pm-blanco I have made the SEED a mandatory attribute of the pmb class, which is now used to set up the various reaction methods. Furthermore, I have added an attribute rng that is initialized using the provided SEED and used in generate_random_points_in_a_sphere. This way it is avoided that the rng, which was before initialized within generate_random_points_in_a_sphere, is each time re-initialized with the same seed.

All samples, tests, etc. have been updated accordingly.

Moreover, I have also fixed the bug discovered by @kosovan regarding non-cubic simulation boxes (#43)

pm-blanco commented 2 months ago

thanks @davidbbeyer for taking care of this, we should make a couple of CI tests for this:

one that tests that center_molecule_in_simulation_box places the molecule in the center of the simulation box, for both cubic and non cubic simulation boxes. Probably could be part of testsuite/create_molecule_position_test.py
We should have a CI test that checks that pyMBE produces deterministic results for a given SEED. I am not completely sure what is the best way we could check this. A simple test could be to create a somewhat complex molecule twice into espresso with two different instances of pyMBE with the same SEED and check that one gets the same structure

davidbbeyer commented 2 months ago

@pm-blanco I have implemented your suggestions:

testsuite/create_molecule_position_test.py now contains two tests that check if the function center_molecule_in_simulation_box works correctly. Note that I had to slightly modify the function center_molecule_in_simulation_box, since it was performing a check that was not strictly necessary and in some cases might cause that the molecule is not centered in the box (depending on the order that objects with a given molecule id are added to the dataframe). I also adjusted the function create_molecule such that the returned dictionary now actually matches the description in the docstring.
I added a test testsuite/seed_test.py that checks if a molecule is created at exactly the same position if the same global seed is provided to pyMBE.