sahi v0.11.18 #58
sahi v0.11.18 #58
Conversation
…nda-forge-pinning 2024.07.10.08.41.05
|
Hi! This is the friendly automated conda-forge-linting service. I just wanted to let you know that I linted all conda-recipes in your PR ( |
| @@ -22,6 +22,8 @@ requirements: | |||
| - pip | |||
| - python >=3.6 | |||
There was a problem hiding this comment.
| - python >=3.6 | |
| - python >=3.7 |
There was a problem hiding this comment.
The python_requires is still >3.6 at https://github.com/obss/sahi/blob/0.11.18/setup.py#L36. Any reason to change it here?
There was a problem hiding this comment.
@weiji14 You are right, my mistake, we should revert this back. Thank you for noting 🙏🏻
|
WDYT @weiji14? 🤔 |
recipe/meta.yaml
Outdated
| run: | ||
| - py-opencv <=4.9 | ||
| - numpy <2.0.0 |
There was a problem hiding this comment.
Could you elaborate on the numpy<2 upper pin here? I saw obss/sahi#1057, but it doesn't mention any errors or what is incompatible.
The issue is that users can still conda install numpy=2 with sahi=0.11.16 or older, since the previous versions did not set this pin. There's a way to prevent that by submitting a repodata-patch to https://github.com/conda-forge/conda-forge-repodata-patches-feedstock, but want to get more context before doing all that work 🙂
There was a problem hiding this comment.
There are breaking changes in the recently released numpy version that collides with precompiled pytorch binaries. That's why I had to add that constrain @weiji14
There was a problem hiding this comment.
Please check this action run for the error I mention: https://github.com/obss/sahi/actions/runs/9866647136/job/27245723778
There was a problem hiding this comment.
Hmm, I don't think it's necessarily an issue on sahi, but rather on opencv. The error seems to be on the import cv2 line:
Traceback (most recent call last): File "/opt/hostedtoolcache/Python/3.9.19/x64/lib/python3.9/runpy.py", line 197, in _run_module_as_main
return _run_code(code, main_globals, None,
...
File "/home/runner/work/sahi/sahi/sahi/__init__.py", line 3, in <module>
from sahi.annotation import BoundingBox, Category, Mask
File "/home/runner/work/sahi/sahi/sahi/annotation.py", line 10, in <module>
from sahi.utils.cv import (
File "/home/runner/work/sahi/sahi/sahi/utils/cv.py", line 10, in <module>
import cv2
AttributeError: _ARRAY_API not found
E..
A module that was compiled using NumPy 1.x cannot be run in
NumPy 2.0.0 as it may crash. To support both 1.x and 2.x
versions of NumPy, modules must be compiled with NumPy 2.0.
Some module may need to rebuild instead e.g. with 'pybind11>=2.12'.
If you are a user of the module, the easiest solution will be to
downgrade to 'numpy<2' or try to upgrade the affected module.
We expect that some modules will need time to support NumPy 2.So we need a version of opencv compiled with NumPy 2.0, but that is only available with OpenCV 4.10.0 (opencv/opencv-python#943, conda-forge/opencv-feedstock#417), and sahi uses opencv<=4.9, hence why numpy 2.0 doesn't work.
In this case, the repodata-patch should probably be applied on opencv instead of sahi. I.e., we could remove the numpy<2.0 pin here, and have the numpy<2.0 be set in https://github.com/conda-forge/opencv-feedstock instead. Give me a few hours to open PRs for that.
There was a problem hiding this comment.
Hmm actually, py-opencv=4.9.0 already has a version constraint for numpy<2.0a0 (set in conda-forge/conda-forge-repodata-patches-feedstock#728, so we could probably just remove the numpy<2.0 pin here and things should work (specifically, sahi on conda-forge won't be able to install numpy=2.0 because py-opencv=4.8 can't):
| - numpy <2.0.0 | |
| - numpy |
Of course, the situation for pip install sahi because PyPI doesn't have repodata-patches, so you'll need to get sahi working with opencv=4.10.0 which supports NumPy 2.0 at some point.
There was a problem hiding this comment.
Yep, CI at https://dev.azure.com/conda-forge/feedstock-builds/_build/results?buildId=976316&view=logs&j=656edd35-690f-5c53-9ba3-09c10d0bea97&t=986b1512-c876-5f92-0d81-ba851554a0a3&l=1159 installs numpy=1.26.4 as expected instead of numpy=2.0, should be all good now!
Constraint should be set on py-opencv already in conda-forge/conda-forge-repodata-patches-feedstock#728, see comment at conda-forge#58 (comment) for context.
Also sorting dependency list alphabetically
It is very likely that the current package version for this feedstock is out of date.
Checklist before merging this PR:
license_fileis packagedInformation about this PR:
@conda-forge-admin,please add bot automergein the title and merge the resulting PR. This command will add our bot automerge feature to your feedstock.bot-rerunlabel to this PR. The bot will close this PR and schedule another one. If you do not have permissions to add this label, you can use the phrase@conda-forge-admin, please rerun botin a PR comment to have theconda-forge-adminadd it for you.Pending Dependency Version Updates
Here is a list of all the pending dependency version updates for this repo. Please double check all dependencies before merging.
Dependency Analysis
Please note that this analysis is highly experimental. The aim here is to make maintenance easier by inspecting the package's dependencies. Importantly this analysis does not support optional dependencies, please double check those before making changes. If you do not want hinting of this kind ever please add
bot: inspection: disabledto yourconda-forge.yml. If you encounter issues with this feature please ping the bot teamconda-forge/bot.Analysis by grayskull shows a discrepancy between it and the the package's stated requirements in the meta.yaml.
Packages found by grayskull but not in the meta.yaml:
Packages found in the meta.yaml but not found by grayskull:
This PR was created by the regro-cf-autotick-bot. The regro-cf-autotick-bot is a service to automatically track the dependency graph, migrate packages, and propose package version updates for conda-forge. Feel free to drop us a line if there are any issues! This PR was generated by - please use this URL for debugging.