Conversation
| * @param dst output array that has the given shape | ||
| * @param shape target shape | ||
| */ | ||
| CV_EXPORTS_W void broadcast_to(InputArray src, const std::vector<int>& shape, OutputArray dst); |
There was a problem hiding this comment.
Please use OpenCV naming conventions. Should be broadcastTo
There was a problem hiding this comment.
Please use OpenCV naming conventions. Should be broadcastTo
Done.
|
Please take a look on failures of 32-bit builds: http://pullrequest.opencv.org/buildbot/builders/precommit_linux32/builds/100233
4GiB is too much for any OpenCV test. |
Target scale is down to {10, 100, 1000} now, which is about 3.8MB. |
|
@opencv-alalek friendly reminder. |
|
Hello, any reviewer updates for this pull request? |
* add broadcast_to with tests * change name * fix test * fix implicit type conversion * replace type of shape with InputArray * add perf test * add perf tests which takes care of axis * v2 from ficus expand * rename to broadcast * use randu in place of declare * doc improvement; smaller scale in perf * capture get_index by reference
* add broadcast_to with tests * change name * fix test * fix implicit type conversion * replace type of shape with InputArray * add perf test * add perf tests which takes care of axis * v2 from ficus expand * rename to broadcast * use randu in place of declare * doc improvement; smaller scale in perf * capture get_index by reference
Should work as
np.broadcast_to.Benchmarks
Results on M1:
Pull Request Readiness Checklist
See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request
Patch to opencv_extra has the same branch name.