search

PMEAL / porespy

A set of tools for characterizing and analying 3D images of porous materials
https://porespy.org
MIT License
301 stars 98 forks source link

Implement parallel extraction #958

Closed Arenhart closed 3 months ago

Arenhart commented 4 months ago

We have been working on the regions_to_network function, improving it's performance with Numba. We got an improvement of up to x60 faster network extraction when running in parallel. We also modified it to allow for anisotropic volumes.

While we tried to keep the results as close as possible to the original function, some differences were included, listed as following:

Only valid for 3D volumes.
Throat and perimeter calculations modified to allow anisotropic volumes (in a very simplified explanation, marching squares is applied to a flattened throat image).
Distance local and global maxima is the average of all coordinates that have the highest distance transform value.
Changed the marching cubes algorithm to an implementation that outputs the area and volume without the need of creating the mesh, "standard" accuracy was removed due to similar calculation time.

There are some minor fixes and modifications that got mixed in the branch, they could be moved to other issues/PRs.

Also, the branch shouldn't work unless Pyedt is available as a package. I will try to add it as soon as possible.

ma-sadeghi commented 4 months ago

Thanks for the PR and the amazing speedup! To avoid git conflicts, please let us know once you're done with your changes, so we can review it.

jgostick commented 4 months ago

Hi @Arenhart, I am not quite sure how to deal with PRs from other people's repos...like how can I pull your branch and make changes locally, etc. So, I am going to create a new temp branch, merge your code in there, then make all my changes on temp, and finally merge into dev. I really appreciate your amazing contribution and I want it to wither on the vine, so I am going to forge ahead even if it's the wrong way to do it.

jgostick commented 4 months ago

OK, I think I have figured out how to proceed, and also why all the conflicts. Firstly, I have added the ltracegeo fork as an additional remote on my local PC, then pulled the branch. Secondly, there are conflicts because @Arenhart edited some files that @ma-sadeghi moved to the src folder. I will attempt to sort all this out, THEN push my updated branch to the porespy repo. I could probably try to push it to the ltracegeo repo, but I don't want to disturb their code.