Parallel II Estimator#1842
Conversation
|
@dcoeurjo My last commit features the ordered (yet parallel version) of the estimator. I realized that because we rely on the input range for surfel, we have no need to build locally the surface and we can simply extract the required surfels from the range. Not to blow my own trumpet, but I think the final code looks pretty good :) (And in my computer, even with only 2000 surfels and 4 threads it does run faster than the sequential one) |
|
I will have a look, thanks. |
dcoeurjo
left a comment
There was a problem hiding this comment.
Looks good to me. Just the doc is missing and some LGPL headers. I've added another domain splitter and a unit test for that.
Code should be in "kernel/domains/" subfolder by the way
|
/builddoc |
|
There was an error while building the doc. Check the GitHub actions for debugging. |
|
/builddoc |
|
The documentation is built. It will be available, after a few minutes, here: https://dgtal-team.github.io/doc-pr/pr1842/index.html |
|
I've updated the geometry shortcuts to enable multithreaded curvature estimation. |
|
/builddoc |
|
The documentation is built. It will be available, after a few minutes, here: https://dgtal-team.github.io/doc-pr/pr1842/index.html |
PR Description
A new WIP branch for parallel Integral Invariant estimators.
Checklist