Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix errors in continuous integration #788

Merged
merged 41 commits into from
May 17, 2023

Conversation

JoostJM
Copy link
Collaborator

@JoostJM JoostJM commented Sep 20, 2022

  • Move building for Mac to CircleCI, stop building on TravisCI
  • Update code to deal with recent changes

ToDo:

  • Fix Conda CI for windows

Use quotes to ensure correct installation of version-constrained pip installs.
Periods are not allowed in job name.
ensures it does not fail if the directory already exists
These python distributions fail to build on CircleCI (failing during setup of python environment).
When tag version specifies alfa, beta or rc release, push to test pypi server.
No longer supported since September 13th, 2020
Repo organization name has changed to AIM-Harvard, update the check to reflect this.
Ensure patchelf can be installed in linux
Fix tag check in AppVeyor
Twin upload is handled in powershell, requiring a different reference to environment variables.
When Label_channel column is present in the CSV input, it needs to be parsed as an int.
However, due to a bug, the value for label was used instead. Related to 54a3782, but now for voxel-based extraction
Force scikit-ci to also use the configured python by first correcting the path. Removes support of python 3.6 build, also remove this from the linux build.
@JoostJM JoostJM force-pushed the circle-ci-mac-os branch 2 times, most recently from 14c3fd5 to 0e52366 Compare December 13, 2022 10:13
- Call twine using python -m twine in Linux
- Correctly setup miniconda environment in Appveyor. See also conda/conda#8836
- Add compiler requirement conda meta.yaml
Allow for newer versions to be used. Consistency in results is covered by testing.
Both label and label_channel are used in SimpleITK, which is more sensitive to datatypes (i.e. will fail when either parameter is int64_t). Therefore, force these values to int.
Label_channel was added to support 3D slicer's Segmentation format. However, with newer versions, Segmentations created in 3D slicer may be stored as labelmaps (more space efficient). Add a note to point users to this in `getMask` function.
Fix new definition of long description to ensure it is rendered on PyPi
Update status badges in readme to reflect new CI integration.
@JoostJM JoostJM closed this May 17, 2023
@JoostJM JoostJM reopened this May 17, 2023
@JoostJM
Copy link
Collaborator Author

JoostJM commented May 17, 2023

Linked to #825 (this PR does not merge into master however):

  • MacOs build support was added to CircleCI, and free plans are no longer available on TravisCI. Therefore move the MacOs build to CircleCI
  • Add Python 3.8 and 3.9 support
  • Switch to use of pytest instead of nose (deprecated)
  • Fix bug in selecting the label_channel based on input csv in batchprocessing
  • Force label and label_channel to int for compatibility with SimpleITK

@JoostJM JoostJM merged commit 8017bdb into AIM-Harvard:master May 17, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant