webknossos-libs
webknossos-libs copied to clipboard
Python API and CLI tools for working with WEBKNOSSOS datasets, annotations and server interactions. Includes converter to OME-Zarr.
It should be possible to create, read, edit and delete tasks and their adjacent types from the Python library. It would also be great to have downloads of the annotations...
Issue #325 list many files, which still use `wkw` directly. This issue is used to track the progress. - Integrate the Dataset API into the CLI: - [ ] `recubing.py`...
PR #683 added an empty test.sh, which should contain all tests from CI, and be used in CI instead.
After PR #683, initial typing is introduced to cluster_tools. However, we should enforce complete typing by uncommenting `# disallow_untyped_defs = true` in the pyproject.toml and adding the respective type-hints.
At the moment, the docs reflect the current master state. However, users usually want to see the version that is currently available via PyPI. Since they might not use the...
The webknossos-libs feature the fantastic `Dataset`-API upon all of the wkcuber features and others build. Since, we want to support Ome-Zarr in the webKnossos ecosystem, I think it would be...
Once sharding is implemented in [zarr-python](https://github.com/zarr-developers/zarr-python), we should enable it for Zarr datasets in this library. See https://github.com/zarr-developers/zarr-python/issues/877
The PR #627 brings initial support for Zarr in the Dataset API. One major limitation is that only Blosc+Zstd compression is enabled. It would be desirable to enable a larger...
The `downsample` method should accept a `bounding_boxes: Optional[List[BoundingBox]]` parameter. When specified, only these bounding boxes should be subject to downsampling.
We should respect the layer's bounding box when handling wkw files explicitly in the (mag)view code, e.g. * `MagView.get_bounding_boxes_on_disk()` and `MagView.get_views_on_disk()` should get a `respect_layer_bounding_box=True` argument, which by default only...