GH-38007: [C++] Add VariableShapeTensor implementation#38008
GH-38007: [C++] Add VariableShapeTensor implementation#38008rok merged 73 commits intoapache:mainfrom
Conversation
eec16ec to
6287402
Compare
dff5441 to
9c3b464
Compare
|
Hi @rok! It would be great to have this in 15.0.0 (added the milestone for visibility). Do you think you have enough bandwidth to work on it at the moment? |
|
@AlenkaF yes I'd very much like to continue working on this, I was waiting for the release before chasing reviewers :D |
|
@pitrou I addressed the review comments and would like to proceed and merge this tomorrow. Any further comments? |
| const std::vector<int64_t>& permutation = {}, | ||
| const std::vector<std::string>& dim_names = {}, | ||
| const std::vector<std::optional<int64_t>>& uniform_shape = {}); |
There was a problem hiding this comment.
Why not pass those vectors by value just like in the constructor?
| const std::vector<int64_t>& permutation = {}, | ||
| const std::vector<std::string>& dim_names = {}, | ||
| const std::vector<std::optional<int64_t>>& uniform_shape = {}); |
|
|
||
| #include "arrow/status.h" | ||
| #include "arrow/util/print_internal.h" | ||
| #include <span> |
There was a problem hiding this comment.
Please move these with the other stdlib includes above?
|
By the way, do you want to expose it in Python soon after? That would allow validating against NumPy. |
Yes, I'm jumping on #40354 right after we merge here :) |
|
Oh wait, rebases are cheap now #40354 |
|
Thank you for all the reviews @pitrou & others! |
|
Thanks for pushing this through @rok ! It's a good thing we finally have an implementation 🎉 |
|
After merging your PR, Conbench analyzed the 3 benchmarking runs that have been run so far on merge-commit 2fcc3ec. There were no benchmark performance regressions. 🎉 The full Conbench report has more details. It also includes information about 9 possible false positives for unstable benchmarks that are known to sometimes produce them. |
Rationale for this change
We want to add VariableShapeTensor extension type definition for arrays containing tensors with variable shapes.
What changes are included in this PR?
This adds a C++ implementation.
Are these changes tested?
Yes.
Are there any user-facing changes?
This adds a new extension type C++.