Changelog

Note: Numbers like (#123) point to closed Pull Requests on the fractal-tasks-core repository.

1.3.2

• Tasks:
  • Add percentile-based rescaling to calculate registration task to make it more robust (#848)
• Dependencies:
• Relax pandas constraint to <2.
• Relax torch constraint to <=3.0.0.
• Relax numpy constraint to <2.1.0.
• Add python 3.12 to the CI matrix (#770).
• Chores:
• fix future warning when using Series.getitem with positional arguments.

1.3.1

• Testing
  • Use latest version of Zenodo tiny-ome-zarr dataset (#817).
  • Relax pip-version constraint in external-packages-manifest tests (#825).
  • Run pip-based CI also regularly and on-demand (#829).
  • Update GitHub actions for upload/download/coverage (#832).
• Dependencies:
  • Require pydantic<=2.8.2 (#836).
  • Require stackview<=0.9.0 (#829).
• Documentation:
  • Bump mkdocstrings and mkdocstrings-python to support griffe v1 (#818).

1.3.0

• Tasks:
  • image_glob_patterns are renamed to include_glob_patterns in Convert Cellvoyager to OME-Zarr (regular & multiplexing) (#812).
  • Convert Cellvoyager to OME-Zarr (regular & multiplexing) gain exclusion patterns to exclude specific patterns of images from being processed (#812).
  • Fix issue with arbitrary acquisition names in Convert Cellvoyager Multiplexing to OME-Zarr (#812).
  • In Convert Cellvoyager to OME-Zarr (regular & multiplexing), handle channels in the mrf metadata file that aren't present in the mlf metadata better (#812).
  • In Convert Cellvoyager to OME-Zarr, improve plate metadata for image list when multiple plates with the same plate name are processed (#812).
  • Catch errors for missing mlf & mrf files better in Convert Cellvoyager to OME-Zarr (regular & multiplexing) (#812).
  • Drop defusexml dependency for cellvoyager metadata conversion (#812).
  • Rename Maximum Intensity Projection HCS Plate task to Project Image (HCS Plate) (#814).
  • Expose selection of projection modes in Project Image (HCS Plate): MIP, MINIP, MEANIP & SUMIP (#814).
  • Rename task function from maximum_intensity_projection to projection and modified parameters in fractal_tasks_core.tasks.io_models.InitArgsMIP (#814).

1.2.1

• Core-library
  • Add create_roi_table_from_df_list library function in fractal_tasks_core.v1.roi: It combines a list of ROI table dataframes into an AnnData ROI table and handles repeating labels (#811).
  • Move create_well_acquisition_dict & _split_well_path_image_path from fractal_tasks_core.tasks._registration_utils.py & fractal_tasks_core.tasks._zarr_utils into fractal_tasks_core.utils (#811).
• Tasks:
  • Fixes issue 810 for Cellpose task: Avoids creating duplicated entries in ROI tables when masking ROI table input was used (#811).

1.2.0

• Core-library and tasks:
  • Switch all core models to Pydantic V2 (#793).
• JSON Schema generation tools:
  • Move JSON-Schema tools to Pydantic V2 (#793).
• Testing:
  • Remove dependency on pytest-pretty (#793).
  • Update manifest_external_packages.yml GitHub Action so that it installs the current fractal-tasks-core (#793).

1.1.1

• Tasks:
  • Fix issue with masked ROI & relabeling in Cellpose task (#786).
  • Fix issue with masking ROI label types in masked_loading_wrapper for Cellpose task (#786).
  • Enable workaround to support yx images in Cellpose task (#789).
  • Fix error handling in calculate_registration_image_based (#799).
  • Fix minor issues with call-signature and type hints in calculate_registration_image_based (#799).

1.1.0

NOTE: Starting from this release, fractal-tasks-core can coexist with Pydantic V2 but it still uses pydantic v1 under the hood for the time being. When working with Pydantic V1, the required version is >=1.10.16.

• Tasks:
  • Refactor Cellpose Task inputs: Combine Channel inputs & channel normalization parameters (#738).
  • Refactor Cellpose Task inputs: Group advanced Cellpose parameters into the CellposeModelParams model that's provided via advanced_cellpose_model_params (#738).
  • Refactor Cellpose Task inputs: Support independent normalization of 2 input channels in the Cellpose task (#738).
  • Rename task.cellpose_transforms into tasks.cellpose_utils (#738).
  • Fix wrong repeated overlap checks for bounding-boxes in Cellpose task (#778).
  • Fix minor MIP issues related to plate metadata and expecting acquisition metadata in all NGFF plates (#781).
  • Add chi2_shift option to Calculate Registration (image-based) task (#741).
• Development:
  • Switch to transitional pydantic.v1 imports, changes pydantic requirement to ==1.10.16 or >=2.6.3 (#760).
  • Support JSON-Schema generation for Enum task arguments (#749).
  • Make JSON-Schema generation tools more flexible, to simplify testing (#749).
• Documentation:
  • Update documentation (#751).
  • Improve/extend page showing tasks from other packages (#759, #777).
• JSON Schema generation:
  • Test manifest creation for three other tasks packages (#763).
• NGFF subpackage
  • Fix Plate model to correspond better to 0.4.0 NGFF spec: Now makes acquisition metadata optional (#781).
• Dependencies:
  • Add image_registration within fractal-tasks extra (#741).

1.0.2

• Fix bug in plate metadata in MIP task (in the copy_ome_zarr_hcs_plate init function) (#736).

1.0.1

• Add support for converting 1536 well plates in cellvoyager converters (#715).
• Testing:
  • Make validation of NGFF Zarr attributes more strict, in tests (#732).
• Development:
  • Update poetry to 1.8.2 (#734).

1.0.0

• Update all tasks to use the new Fractal API from Fractal server 2.0 (#671)
• Provide new dev tooling to create Fractal manifest for new task API (#671)
• Add Pydantic models for OME-NGFF HCS Plate validation (#671)
• Breaking changes in core library:
  • In get_acquisition_paths helper function of NgffWellMeta: The dictionary now contains a list of paths as values, not single paths. The NotImplementedError for multiple images with the same acquisition was removed.
  • The utils.get_table_path_dict helper function was made private & changed its input parameters: It's now _get_table_path_dict(zarr_url: str)
• Breaking changes in task sub-package:
  • Rename tasks for increase clarity (#671 & #706).
  • Changed registration tasks flow: Now 2 Compound tasks, 1 Parallel task (#671).
  • Changed parameter names in registration tasks: acquisition instead of cycle (#719).
  • Changed parameter names & defaults in illumination correction task: (#671).
    • Now uses illumination_profiles instead of dict_corr.
    • Changes default of background subtraction from 110 to 0.
• Dependencies:
  • Add filelock (#718).

0.14.3

• Make Cellpose task work for non HCS plate OME-Zarr images (#659)
• Add option to Illumination Correction to specify the ROI table name (#665)

0.14.2

• Add custom normalization options to the Cellpose task (#650)
• Add more options to the Cellpose task to control model behavior (#650)
• For Cellpose task, switch to using Enums for model_type (see issue #401)

0.14.1

• Fix bug in cellpose_segmentation upon using masked loading and setting channel2 (#639). Thanks @FranziskaMoos-FMI and @enricotagliavini.
• Improve handling of potential race condition in "Apply Registration to image" task (#638).

0.14.0

• Breaking changes in tasks:
  • Make NapariWorkflowsOutput.label_name attribute required, and use it to fill the region["path"] table attribute (#613).
• Breaking changes in core library:
  • ⚠️ Refactor the whole package structure, leading to breaking changes for most imports (#613); more details at this page.
  • In prepare_label_group helper function:
    • Make label_attrs function argument required (#613).
    • Validate label_attrs with NgffImageMeta model (#613).
    • Override multiscale name in label_attrs with label_name (#613).
  • In write_table helper function:
    • Drop logger function argument (#613).
    • Add table_name function argument, taking priority over table_attrs (#613).
    • Raise an error if no table type is not provided (#613).
    • Raise an error if table attributes do not comply with table specs (#613).
• Other internal changes:
  • Comply with table specs V1, by writing all required Zarr attributes (#613).
  • Remove has_args_schema obsolete property from manifest (#603).
  • Handle GroupNotFoundError in load_NgffImageMeta and load_NgffWellMeta (#622).
• Bug fixes:
  • Fix table selection in calculate registration image-based (#615).
• Documentation
  • Clarify table specs V1 (#613).
• Testing:
  • Use more recent Zenodo datasets, created with fractal-tasks-core>=0.12 (#623).
  • Use poetry 1.7.1 in GitHub actions (#620).
  • Align with new Zenodo API (#601).
  • Update test_valid_manifest (#606).
  • Use pooch to download test files (#610).
• Documentation:
  • Add list of tasks (#625).
• Dependencies:
  • Remove Pillow <10.1.0 constraint (#626).

0.13.1

• Always use write_table in tasks, rather than AnnData write_elem (#581).
• Remove assumptions on ROI-table columns from get_ROI_table_with_translation helper function of calculate_registration_image_based task (#591).
• Testing:
  • Cache Zenodo data, within GitHub actions (#585).
• Documentation:
  • Define V1 of table specs (#582).
  • Add mathjax support (#582).
  • Add cross-reference inventories to external APIs (#582).

0.13.0

• Tasks:
  • New task and helper functions:
    • Introduce import_ome_zarr task (#557, #579).
    • Introduce get_single_image_ROI and get_image_grid_ROIs (#557).
    • Introduce detect_ome_ngff_type (#557).
    • Introduce update_omero_channels (#579).
  • Make maximum_intensity_projection independent from ROI tables (#557).
  • Make Cellpose task work when input_ROI_table is empty (#566).
  • Fix bug of missing attributes in ROI-table Zarr group (#573).
• Dependencies:
  • Restrict Pillow version to <10.1 (#571).
  • Support AnnData 0.10 (#574).
• Testing:
  • Align with new Zenodo API (#568).
  • Use ubuntu-22 for GitHub CI (#576).

0.12.2

• Relax check_valid_ROI_indices to support search-first scenario (#555).
• Do not install docs dependencies in GitHub CI (#551).

0.12.1

• Make Channel.window attribute optional in lib_ngff.py (#548).
• Automate procedure for publishing package to PyPI (#545).

0.12.0

This release includes work on Pydantic models for NGFF specs and on ROI tables.

• NGFF Pydantic models:
  • Introduce Pydantic models for NGFF metadata in lib_ngff.py (#528).
  • Extract num_levels and coarsening_xy parameters from NGFF objects, rather than from metadata task input (#528).
  • Transform several lib_zattrs_utils.py functions (get_axes_names, extract_zyx_pixel_sizes and get_acquisition_paths) into lib_ngff.py methods (#528).
  • Load Zarr attributes from groups, rather than from .zattrs files (#528).
• Regions of interest:
  • Set FOV_ROI_table and well_ROI_table ZYX origin to zero (#524).
  • Remove heuristics to determine whether to reset origin, in cellpose_segmentation task (#524).
  • Remove obsolete reset_origin argument from convert_ROI_table_to_indices function (#524).
  • Remove redundant reset_origin call from apply_registration_to_ROI_tables task (#524).
  • Add check on non-negative ROI indices (#534).
  • Add check on ROI indices not starting at (0,0,0), to highlight v0.12/v0.11 incompatibility (#534).
  • Fix bug in creation of bounding-box ROIs when cellpose_segmentation loops of FOVs (#524).
  • Update type of metadata parameter of prepare_FOV_ROI_table and prepare_well_ROI_table functions (#524).
  • Fix reset_origin so that it returns an updated copy of its input (#524).
• Dependencies:
  • Relax fsspec<=2023.6 constraint into fsspec!=2023.9.0 (#539).

0.11.0

• Tasks:
  • (major) Introduce new tasks for registration of multiplexing cycles: calculate_registration_image_based, apply_registration_to_ROI_tables, apply_registration_to_image (#487).
  • (major) Introduce new overwrite argument for tasks create_ome_zarr, create_ome_zarr_multiplex, yokogawa_to_ome_zarr, copy_ome_zarr, maximum_intensity_projection, cellpose_segmentation, napari_workflows_wrapper (#499).
  • (major) Rename illumination_correction parameter from overwrite to overwrite_input (#499).
  • Fix plate-selection bug in copy_ome_zarr task (#513).
  • Fix bug in definition of metadata["plate"] in create_ome_zarr_multiplex task (#513).
  • Introduce new helper functions write_table, prepare_label_group and open_zarr_group_with_overwrite (#499).
  • Introduce new helper functions are_ROI_table_columns_valid, convert_indices_to_regions, reset_origin, is_standard_roi_table, get_acquisition_paths, get_table_path_dict, get_axes_names, add_zero_translation_columns, calculate_min_max_across_dfs, apply_registration_to_single_ROI_table, write_registered_zarr, calculate_physical_shifts, get_ROI_table_with_translation (#487).
• Testing:
  • Add tests for overwrite-related task behaviors (#499).
  • Introduce mock-up of napari_skimage_regionprops package, for testing of napari_workflows_wrapper task (#499).
• Dependencies:
  • Require fsspec version to be <=2023.6 (#509).

0.10.1

• Tasks:
  • Improve validation for OmeroChannel.color field (#488).
  • Include image-label/source/image OME-NGFF attribute when creating labels (#478).
  • Update default values for tolerance (tol) in lib_ROI_overlaps.py functions (#466).
• Development tools:
  • Include docs_info and docs_link attributes in manifest tasks (#486).
  • Rename and revamp scripts to update/check the manifest (#486).
  • Improve logging and error-handling in tools for args-schema creation (#469).
• Documentation:
  • Convert docstrings to Google style (#473, #479).
  • Switch from sphinx to mkdocs for documentation (#479).
  • Update generic type hints (#462, #479).
  • Align examples to recent package version, and mention them in the documentation (#470).
• Testing:
  • Improve coverage of core library (#459, #467, #468).
  • Update Zenodo datasets used in tests (#454).
  • Run tests both for the poetry-installed and pip-installed package (#455).
• Dependencies:
  • Relax numpy required version to <2 (#477).
  • Relax dask required version to >=2023.1.0 (#455).
  • Relax zarr required version to >=2.13.6,<3 (#455).
  • Relax pandas required version to >=1.2.0,<2 (#455).
  • Relax Pillow required version to >=9.1.1,<10.0.0 (#455).
  • Full update of poetry.lock file (mutiple PRs, e.g. #472).
  • Include requests and wget in the dev poetry dependency group (#455).

0.10.0

• Restructure the package and repository:
  • Move tasks to tasks subpackage (#390)
  • Create new dev subpackage (#384).
  • Make tasks-related dependencies optional, and installable via fractal-tasks extra (#390).
  • Remove tools package extra (#384), and split the subpackage content into lib_ROI_overlaps and examples (#390).
• (major) Modify task arguments
  • Add Pydantic model lib_channels.OmeroChannel (#410, #422);
  • Add Pydantic model tasks._input_models.Channel (#422);
  • Add Pydantic model tasks._input_models.NapariWorkflowsInput (#422);
  • Add Pydantic model tasks._input_models.NapariWorkflowsOutput (#422);
  • Move all Pydantic models to main package (#438).
  • Modify arguments of illumination_correction task (#431);
  • Modify arguments of create_ome_zarr and create_ome_zarr_multiplex (#433).
  • Modify argument default for ROI_table_names, in copy_ome_zarr (#449).
  • Remove the delete option from yokogawa to ome zarr (#443).
  • Reorder task inputs (#451).
• JSON Schemas for task arguments:
  • Add JSON Schemas for task arguments in the package manifest (#369, #384).
  • Add JSON Schemas for attributes of custom task-argument Pydantic models (#436).
  • Make schema-generation tools more general, when handling custom Pydantic models (#445).
  • Include titles for custom-model-typed arguments and argument attributes (#447).
  • Remove TaskArguments models and switch to Pydantic V1 validate_arguments (#369).
  • Make coercing&validating task arguments required, rather than optional (#408).
  • Remove default_args from manifest (#379, #393).
• Other:
  • Make pydantic dependency required for running tasks, and pin it to V1 (#408).
  • Remove legacy executor definitions from manifest (#361).
  • Add GitHub action for testing pip install with/without fractal-tasks extra (#390).
  • Remove sqlmodel from dev dependencies (#374).
  • Relax constraint on torch version, from ==1.12.1 to <=2.0.0 (#406).
  • Review task docstrings and improve documentation (#413, #416).
  • Update anndata dependency requirements (from ^0.8.0 to >=0.8.0,<=0.9.1), and replace anndata.experimental.write_elem with anndata._io.specs.write_elem (#428).

0.9.4

• Relax constraint on scikit-image version, by only requiring a version >=0.19 (#367).

0.9.3

• For labeling tasks (cellpose_segmentation or napari_worfklows_wrapper), allow empty ROI tables as input or output (#365).
• Relax constraint related to the presence of channels in create_zarr_structure_multiplex task (#365).

0.9.2

• Increase memory requirements for some tasks in manifest (#363).

0.9.1

• Add use_gpu argument for cellpose_segmentation task (#350).
• Add dummy return object to napari-workflows task (#359).
• Include memory/cpu/gpu requirements in manifest, in view of new fractal-server SLURM backend (#360).

0.9.0

• Introduce a module for masked loading of ROIs, and update the cellpose_segmentation task accordingly (#306).
• Rename task arguments: ROI_table_name->input_ROI_table and bounding_box_ROI_table_name->output_ROI_table (#306).
• Implement part of the proposed table support in OME-NGFF specs, both for the tables zarr group and then for each table subgroup (#306).
• Rename module: lib_remove_FOV_overlaps.py->lib_ROI_overlaps.py (#306).
• Add new functions to existing modules: lib_regions_of_interest.convert_region_to_low_res, lib_ROI_overlaps.find_overlaps_in_ROI_indices (#306).

0.8.1

• Disable bugged validation of model_type argument in cellpose_segmentation (#344).
• Raise an error if the user provides an unexpected argument to a task (#337); this applies to the case of running a task as a script, with a pydantic model for task-argument validation.

0.8.0

• (major) Update task interface: remove filename extension from input_paths and output_path for all tasks, and add new arguments (image_extension,image_glob_pattern) to create_ome_zarr task (#323).
• Implement logic for handling image_glob_patterns argument, both when globbing images and in Yokogawa metadata parsing (#326).
• Fix minor bugs in task arguments (#329).

0.7.5

• Update cellpose_segmentation defaults and parse additional parameters (#316).
• Add dual-channel input for cellpose_segmentation task (#315).

0.7.4

• Add tests for python 3.10 (#309).
• Drop support for python 3.8 (#319).
• Update task interface: use string arguments instead of pathlib.Path, and only set defaults in function call signatures (#303).

0.7.3

• Add reset_origin argument to convert_ROI_table_to_indices (#305).
• Do not overwrite existing labels in cellpose_segmentation task (#308).

0.7.2

• Remove pyqt5-related dependencies (#288).

0.7.1

Missing

0.7.0

• Replace dask.array.core.get_mapper() with zarr.storage.FSStore() (#282).
• Pin dask version to >=2023.1.0, <2023.2.
• Pin zarr version to >=2.13.6, <2.14.
• Pin numpy version to >=1.23.5,<1.24.
• Pin cellpose version to >=2.2,<2.3.

0.6.5

• Remove FOV overlaps with more flexibility (#265).

0.6.4

• Created tools submodule and installation extra (#262).

0.6.3

• Added napari dependency, pinned to 0.4.16 version.
• Fixed type-hinting bug in task to create multiplexing OME-Zarr structure (#258).

0.6.2

• Support passing a pre-made metadata table to tasks creating the OME-Zarr structure (#252).

0.6.1

• Add option for padding an array with zeros in upscale_array (#251).
• Simplified imagecodecs and PyQt5 dependencies (#248).

0.6.0

• (major) Refactor of how to address channels (#239).
• Fix bug in well ROI table (#245).

0.5.1

• Fix sorting of image files when number of Z planes passes 100 (#237).

0.5.0

• (major) Deprecate measurement task (#235).
• (major) Use more uniform names for tasks, both in python modules and manifest (#235).
• Remove deprecated manifest from __init__.py (#233).

0.4.6

• Skip image files if filename is not parsable (#219).
• Preserve order of input_paths for multiplexing subfolders (#222).
• Major refactor of replicate_zarr_structure, also enabling support for zarr files with multiple images (#223).

0.4.5

• Replace Cellpose wrapper with CellposeModel, to support pretrained_model argument (#218).
• Update cellpose version (it was pinned to 2.0, in previous versions) (#218).
• Pin torch dependency to version 1.12.1, to support CUDA version 10.2 (#218).

0.4.4

Missing due to releasing error.

0.4.3

• In create_zarr_structure_multiplex, always use/require strings for acquisition field (#217).

0.4.2

• Bugfixes

0.4.1

• Only use strings as keys of channel_parameters (in create_zarr_structure_multiplex).

0.4.0

• (major) Rename well to image (both in metadata list and in manifest) and add an actual well field (#210).
• Add create_ome_zarr_multiplexing, and adapt yokogawa_to_zarr (#210).
• Relax constraint about outputs in napari_worfklows_wrapper (#209).

0.3.4

• Always log START/END times for each task (#204).
• Add label_name argument to cellpose_segmentation (#207).
• Add pretrained_model argument to cellpose_segmentation (#207).

0.3.3

• Added napari_worfklows_wrapper to manifest.

0.3.2

• Compute bounding boxes of labels, in cellpose_segmentation (#192).
• Parse image filenames in a more robust way (#191).
• Update manifest, moving parallelization_level and executor to meta attribute.

0.3.1

• Fix executable fields in manifest.
• Remove graphviz dependency.

0.3.0

• Conform to Fractal v1, through new task manifest (#162) and standard input/output interface (#155, #157).
• Add several type hints (#148) and validate them in the standard task interface (#175).
• Update napari_worfklows_wrapper: pyramid level for labeling worfklows (#148), label-only inputs (#163, #171), relabeling (#167), 2D/3D handling (#166).
• Deprecate dummy and dummy_fail tasks.

0.2.6

• Setup sphinx docs, to be built and hosted on https://fractal-tasks-core.readthedocs.io; include some preliminary updates of docstrings (#143).
• Dependency cleanup via deptry (#144).

0.2.5

• Add napari_workflows_wrapper task (#141).
• Add lib_upscale_array.py module (#141).

0.2.4

• Major updates to metadata_parsing.py (#136).