v0.13.0

• Added support for Birdhouse Helper Bot (for bumping versions).
• CI Actions and Python dependencies are now pinned to commit hashes.
• Permissions have been set and restricted as needed for all workflows.
• Pinned cf-xarray below v0.9.3.
• Reformatted and renamed CHANGES.rst to CHANGELOG.rst.
• Dropped support for Python 3.8 and 3.9. The supported versions are now Python 3.10, 3.11 and 3.12.
• black has been updated to v2024.10.0, and coding conventions have been updated to Python3.10+.
• Several dependencies now are pinned to baseline versions, including anyascii, dask, ipython, matplotlib, nbsphinx numpy, and sphinxcontrib-bibtex.
• Added xscen dependency. Now used in spatial averaging of ensemble processes.
• Updated the build system to use flit v3.9.0 (PEP 517 and PEP 621). Several configuration files have been migrated to pyproject.toml and .flake8.
• finch now uses a src-based layout for the package. The finch package is now located in the src directory.
• The pre-commit configuration has been updated to use pre-commit v3.5.0. Several hooks have been added, including pygrep-hooks (security fixes), ruff (code formatting), and vulture (dead code).
• Documentation examples now build without warnings. ReadTheDocs is now configured to fail_on_warning.
• xclim has been updated to v0.52.2.
• Modernized the documentation to reflect recent changes.
• Added pylint to the linting checks and pre-commit steps.
• Added deployment workflows for PyPI and TestPyPI.
• Synchronized more dependencies.

v0.12.1

• Replaced unidecode with anyascii due to a licensing issue.
• Synchronized some dependencies across build systems.
• Added a workaround in wps_geoseries_to_netcdf to handle a pandas v2.0 behaviour change.

v0.12.0

• Renamed the installed package from finch to birdhouse-finch.
• First release of the birdhouse-finch package on PyPI.
• Versioning now adheres to SemVer v2.0.0.
• Added a Makefile recipe to the repository to evaluate notebooks while ignoring the output cells.
• Cleaned up documentation to facilitate easier navigation.
• Slightly reorganized the documentation for easier navigation.
• Fast-forwarded the cookiecutter.
• Fixed the Manifest.in to add all necessary files to wheel.
• Removed references to files that have never existed (apidoc).
• Cleaned up the setup code.
• Added more files to be ignored in the .gitignore and in the Manifest.in.

v0.11.4

• Fixed a bug that occurred when fixing a broken cftime-index with newer cftime versions.
• Placed pins on xarray and pandas to prevent future errors from changes to frequency codes.

v0.11.3

• Updated ReadTheDocs to use the new mambaforge version (2022.9).
• Addressed calls in GitHub Action usage that were emitting warnings.
• Updated MANIFEST.in to include and exclude the relevant files for the source distribution.
• Modified the setup.py to only include the files necessary in the wheel.
• Updated AUTHORS.rst to list more contributors.
• Removed namespace file (__init__.py) from tests to ensure that they aren't treated like an importable package.
• Updated pre-commit hooks.
• Sorted software requirements for legibility.
• Removed Travis-CI shell script.

v0.11.2

• Added a Docker-based testing suite to the GitHub Workflows.
• Added a wider range of Python versions to test against in the GitHub Workflows.
• Migrated conda-build action from mamba-org/provision-with-micromamba to mamba-org/setup-micromamba.
• Cleaned up the Dockerfile. Docker now pip-installs finch directly from the GitHub repository.
• Finch now explicitly supports Python3.11.
• Pinned Python below 3.12 on conda and removed pin on pint for ReadTheDocs builds.
• Added a GitHub Workflow to bump the version number and to create tags from the version bumping process on dispatch.
• Added a pre-commit hook for validating the ReadTheDocs configuration and GitHub Workflows.

v0.11.1

• Update to xclim 0.43.0.
• Added xclim yml module support
  • Added humidex days above calculation via yml module
  • Reimplmented streamflow indicators via yml module (adjust for xclim 0.41 breaking changes)
• Fixed bug for CanDCS-U6 ensemble "26models" list
• Passing an empty string to ensemble_percentiles in ensemble processes will return the merged un-reduced ensemble. The different members are listed along the realization coordinates through raw names allowing for basic distinction between the input members.

v0.10.0

• Generalize ensemble datasets configuration.
  • Datasets usable by ensemble processes are now specified through a YAML file which is pointed to in the configuration.
  • As a consequence processes are generated according to the available variables. Similarly for the allowed values of some inputs on these processes.
  • The output name now includes the dataset name (if a custom name was not specified).
  • finch.processes.xclim was removed, there is no static module of processes.
  • Input "rcp" has been renamed to "scenario".
  • Input "dataset_name" has been fixed and renamed to "dataset".
• Update to xclim 0.38.0.
• Improved subset_grid_point_dataset & subset_bbox_dataset performance when using local files.

v0.9.1

Bug fix release.

• Avoid using a broken version of libarchive in the Docker image.

v0.9.0

• Fix use of output_name, add output_format to xclim indicators.

• Change all outputs to use output as the main output field name (instead of output_netcdf).

• Updated to xclim 0.37:

  • Percentile inputs of xclim indicators have been renamed with generic names, excluding an explicit mention to the target percentile.
  • In ensemble processes, these percentiles can now be chosen through perc_[var] inputs. The default values are inherited from earlier versions of xclim.

• Average shape process downgraded to be single-threaded, as ESMF seems to have issues with multithreading.

• Removed deprecated processes subset_ensemble_bbox_BCCAQv2, subset_ensemble_BCCAQv2 and BCCAQv2_heat_wave_frequency_gridpoint.

• Added csv_precision to all processes allowing CSV output. When given, it controls the number of decimal places in the output.