MAINT: Merge duplicate implementations of *vander2d and *vander3d functions
#13079
Conversation
… functions Every implementation is the same right now, other than calling a different `*vander` function. Merging these into a single private function taking a callback results in significant deduplication. Found by LGTM.
*vander2d and *vander3d functons*vander2d and *vander3d functions
|
LGTM, might be interesting to see if the Thanks Eric. |
|
Let's see if the LGTM nightly today spots that the duplicates are gone... |
…der3d` functions One was missing from numpygh-13079
|
@jhelie: Any idea why lgtm would show 0 alerts fixed on this PR, yet still reduce the total number of project alerts? |
|
Hi Eric, I'm guessing the fixed alerts correspond to duplicate code flagged by rules such as this one. If so the automated code review does not show such non-attributable alerts, as documented here (point 3). The idea is to only show results which can be fully accounted for by the PR analysed - which is not the case for statistical queries. But I too find it confusing and agree the current situation is not ideal, and we're thinking on how to improve the display of our results. In the meantime you can check which queries have the non-attributable tag, see for instance here for the Python ones. Hope that helps! |
|
Thanks for the info. Kind of a shame that the fixes don't get attributed me, but I suppose they are only arbitrary internet points. It would be nice if there were at least some way to be sure the alerts had been addressed, even if you don't want to attach an owner to them. Perhaps "N possibly unrelated alerts fixed" or something in the PR feedback. |
|
Thanks for the feedback, I totally agree and we are discussing internally some ways to make this better. |
Every implementation is the same right now, other than calling a different
*vanderfunction.Merging these into a single private function taking a callback results in significant deduplication.
Found by LGTM.
Split from #13078, since one commit needed further discussion, and I made some typos in the commit message anyway