Implement orient3d + insphere #23
Conversation
|
Irrespective of reviews (I am doubtful that a human can meaningfully review the robust code in general without immense effort) we should port over the tests from https://github.com/mourner/robust-predicates/blob/main/test/test.js and https://github.com/mourner/robust-predicates/tree/main/test/fixtures. The latter is…some work and doesn't have to happen in this PR, but we should do it. |
|
Thanks for the contribution @dmyyy ; thanks @urschrei for also porting the tests and ensuring correctness of this. At some point, we could add these predicates into the demo, so we can also get a visual feedback for the robustness. It's built for 2-input predicate because the output is 2-d image, but I guess we can fix one input to just get a sense. |
|
Would anyone else like to review this? I'm happy to merge otherwise. |
|
It looks plausible, but as you said, it's tough to review the robust math meaningfully. Provided we also merge #24, this LGTM. |
Agreed with this. +1 to merge. |
24: Use test fixtures from Nanevski et al (2001) r=rmanoka a=urschrei Ported over from https://github.com/mourner/robust-predicates/tree/main/test/fixtures This PR is based on #23, which should be merged first. Co-authored-by: Stephan Hügel <[email protected]>
24: Use test fixtures from Nanevski et al (2001) r=rmanoka a=urschrei Ported over from https://github.com/mourner/robust-predicates/tree/main/test/fixtures This PR is based on #23, which should be merged first. Co-authored-by: Stephan Hügel <[email protected]>
Added tests + tests pass.
Would appreciate another pair of eyes on two_one_product implementation. (+ general changes since there's a lot and it's possible I missed something)