Changelog

v0.8.0 (2026-05-27)

Contributors to this version: Gabriel Rondeau-Genesse (@RondeauG).

New features and enhancements

  • All hydrological model outputs will now have standard dimensions and coordinates. (PR/417).

    • For distributed models, the spatial dimension will either be subbasin_id or unit_id depending resolution.

    • All outputs will now include ID, elevation, drainage area, and longitude and latitude coordinates when available, at the relevant scale(s).

  • An aggregate_outputs method has been added to the hydrological model classes to facilitate the aggregation of outputs to different spatial scales. (PR/417).

Breaking changes

  • The get_streamflow method of the hydrological model classes has been renamed to get_outputs. (PR/417).

Bug fixes

  • Fixed a misconfigured envlist setting in tox.toml. (PR/426).

Internal changes

  • A few notebooks have been moved to the pavics_notebooks directory to allow testing on PAVICS. (PR/414).

  • raven-hydro has been added as an explicit dependency in conda installation configurations. (PR/423).

  • Updated the cookiecutter template to the latest commit. (PR/426):

    • Updated the zizmor` workflow security checks to ignore some superfluous actions, addresses some minor workflow security issues.

    • Updated several development dependency versions.

    • Updated pre-commit hooks.

  • A _model_utils.py module has been added to the xhydro.modelling subpackage to hold utility functions common to all hydrological models. (PR/417).

    • standardize_outputs and aggregate_outputs functions have been added to this module.

  • The HYDROTEL notebook now uses the same mechanism as the tests to load the executable and demo project paths from environment variables. (PR/417).

  • Temporarily disabled the fail_on_warning setting in the Read the Docs configuration, until xscen is updated to >0.15.1. (PR/432).

v0.7.1 (2026-03-03)

Contributors to this version: Gabriel Rondeau-Genesse (@RondeauG), Trevor James Smith (@Zeitsperre).

New features and enhancements

  • Added support for Python 3.14. (PR/410).

  • Added support for the latest version of xclim, xscen, xarray, and pandas. (PR/410).

Breaking changes

  • Due to the added support for pandas v3.0.0, the leafmap dependency has been made optional when installing xhydro through pip. (PR/410).

    • The dependency can be installed separately with pip install xhydro[leafmap].

    • This does not affect the conda installation, which still includes leafmap as a dependency.

  • Development dependencies are now installed via pip “dependency-group” entries (see: PEP-735). Requires pip >=25.2. (PR/411).

    • New “optional-dependency” lists are now based on optional features (julia, leafmap, raven, speedups, all).

  • Depepndencies that are implicitly installed and should have been listed in the core package have been added (clisops, matplotlib, packaging, scikit-learn, shapely). (PR/411).

Bug fixes

  • No changes.

Internal changes

  • Updated the cookiecutter template to the latest commit. (PR/411):

    • pre-commit tool has been replaced by prek.

    • The project repository now requires Developer Certificate of Origin (DCO) in commit messages.

    • sphinx has been pinned below v9.0 due to a false-positive warning raised in the French documentation build.

    • The project now utilizes deptry to ensure that all necessary dependencies are listed in the core package.

  • ruff has been configured to temporarily ignore rule D420 due to an error in its implementation in v0.15.4. (PR/411.)

v0.7.0 (2026-02-24)

Contributors to this version: Trevor James Smith (@Zeitsperre), Gabriel Rondeau-Genesse (@RondeauG), Julián Ospina (@ospinajulian), Ève Larose (@e-larose), Thomas-Charles Fortier Filion (@TC-FF).

New features and enhancements

  • The xh.indicators.pmp.major_precipitation_events function now supports specifying a minimum precipitation threshold to consider an event. (PR/370).

  • New objective functions have been added to the calibration module. (GH/365, PR/366).

  • Added a new submodule xhydro.indicators.signatures to compute hydrological signatures. (GH/365, PR/366).

Breaking changes

  • The migration from stackstac to odc-stac in xhydro.gis has led to changes in the results of xhydro.gis.land_use_classification and xhydro.gis.surface_properties due to differences in projection handling between the two libraries. (PR/403).

    • Our tests show that the differences are generally minor, but users should verify that their results remain consistent after the update.

  • The Upstream Area (sq. km). column in the output of xh.gis.watershed_properties has been renamed to Upstream Area (sq. km). (PR/403).

  • Features that were deprecated and planned to be removed in v0.7.0 have now been removed. This includes: (PR/408).

    • The map argument in xh.gis.watershed_delineation (now m).

    • The use_defaults and check_missing arguments in the Hydrotel class.

    • Most of the arguments in the RavenpyModel.create_rv() method, which are now expected to be given to the class itself or to the update_data() method.

    • Streamflow is now always q.

    • return_period is now the only accepted argument for return periods in the xhfa module, and the old t, rp, and return_periods arguments have been removed.

    • xhfa.regional.get_group_from_fit has been renamed to xhfa.regional.get_clusters and the old function has been removed.

    • xhfa.regional.calculate_rp_from_afr has been renamed to xhfa.regional.calculate_return_period and the old function has been removed.

    • xhfa.uncertainties.calc_q_iter has been renamed to xhfa.uncertainties.calculate_quantiles_over_boostraped_groups and the old function has been removed.

Bug fixes

  • The plotting positions calculated by xhfa.local._plotting_positions are now assigned as coordinates to ensure compatibility with hvplot when combined to xarray >=2025.11.0. (PR/373).

  • Fixed a bug in xh.indicators.pmp.spatial_average_storm_configurations where certain instances were incorrectly assigned the string “x” and “y” instead of the expected value x and y. (PR/370).

  • Corrected the titles in the documentation of the Inputs for the Probable Maximum flood (PMF). (PR/370)

  • Added specific import for exactextract to prevent issues in the GIS modules if exactextract is prensent in the environment. (PR/381)

  • The version detection logic in xhydro.modelling._hydrotel should now be more robust. (PR/388).

  • Added a temporary workaround in xhydro.modelling._hydrotel to avoid OverflowError when chunks are automatically estimated as -1 by xscen.estimate_chunks. (PR/388).

    • Automated chunk estimation is also now skipped if the dataset is smaller than 100 MB.

  • xh.modelling.format_input now correctly handles subdaily meteorological data with Hydrotel. (GH/404, PR/405).

  • xh.modelling.format_input now correctly handles station-based meteorological data with Hydrotel. (PR/405).

Internal changes

  • Updated the cookiecutter template to the latest commit. (PR/371):

    • Migrated from tox.ini to tox.toml

    • pre-commit hooks have been updated

    • Python 3.13 has been set in CI workflows (replacing “3.x”)

    • tox builds no longer require python-coveralls (abandoned) and CI workflows now exclusively use coverallsapp/github-action

  • Addressed multiple FutureWarnings coming from xarray and pandas. (PR/403).

  • The backend used to load STAC data in xhydro.gis has been changed from stackstac to odc-stac. (PR/403).

    • This change addresses compatibility issues with new versions of rasterio.

  • Added an additional xfail condition to the tests calling planetary_computer to account for occasional server issues. (PR/403).

  • Updated test_hydrotel.py to enable testing with the Hydrotel executable and demo project. (PR/405).

    • You need to create a .env file in the tests directory and add the three required environment variables (HYDROTEL_EXECUTABLE, HYDROTEL_DEMO, and HYDROTEL_VERSION).

v0.6.1 (2025-10-22)

Contributors to this version: Gabriel Rondeau-Genesse (@RondeauG), Trevor James Smith (@Zeitsperre).

Bug fixes

  • Fixed a compatibility issue with xarray >=2025.9.1` in xhfa.regional.fit_pca. (PR/355).

  • Fixed a bug where the process would hang indefinitely on Windows systems when calling hydrological models through command line arguments. (PR/358, PR/363).

  • The long_name and description attributes of the q variable in Hydrotel outputs have been slightly modified to match RavenPy outputs. (PR/358).

  • Fixed a problem to get a regional tau4 instead of being local in order to compute sigma4 used in the computation of the Z score (PR/372).

Internal changes

  • Updated minimum numpy to v1.25+ and adjusted tox configuration to recompile the h5py library in order to deal with compatibility problems from pre-compiled wheels. (PR/357).

  • Updated the cookiecutter template to the latest commit. (PR/358):

    • black, blackdoc and isort have been dropped for modern ruff.

    • Some development dependencies have been updated to their latest versions.

    • Contributor Covenant Agreement has been updated to v3.0.

    • Added a CITATION.cff file to the repository.

    • License metadata now follows PEP 639 conventions.

    • Linting hooks are slightly more strict.

v0.6.0 (2025-09-22)

Contributors to this version: Louise Arnal (@lou-a), Thomas-Charles Fortier Filion (@TC-FF), Gabriel Rondeau-Genesse (@RondeauG), Julián Ospina (@ospinajulian).

New features and enhancements

  • Added a configuration for the Raven model in xhydro.modelling.format_input. (PR/257).

  • Generalization of xhydro.frequency_analysis.uncertainties.calc_q_iter. (PR/282).

  • New function xh.gis.watershed_to_raven_hru to extract HRU information from a watershed. (PR/303).

  • The RavenpyModel class now has a hru argument, to either provide the old HRU arguments (but now under a dictionary) or a GeoDataFrame with the HRU information. (GH/266, PR/303, PR/339).

  • The RavenpyModel class no longer writes new *.rv* files if they already exist. Additionally, a .create_rv() method has been added to the class to write the files. (PR/303).

  • The RavenpyModel class now accepts meteorological data in the form of a single station, multiple stations, or a 2D grid. (PR/303).

  • The RavenpyModel class now supports distributed HBVEC models. (PR/339).

  • Two new functions, update_data and update_config, have been added to the RavenpyModel class to facilitate updating the model with new data and configuration options. (PR/341, PR/345).

  • A executable argument has been added to the RavenpyModel class to specify a custom path to the Raven executable. (PR/341).

  • A gridweights argument has been added to the RavenpyModel class to specify a custom path to a grid weights file when using gridded meteorological data. (PR/344).

  • The qobs and alt_name_flow arguments in the RavenpyModel class have been re-added, but are currently only used to control the output_subbasins argument. (PR/339).

  • The xhydro.extreme_value_analysis module now uses Extremes.jl = “1.0.5” and Optim = “1.13.2”. (GH/292, PR/315).

  • Additional options can now be passed to the Hydrotel executable via the run_options argument in the run method. (PR/331).

  • A new function xhydro.indicators.pmp.pw_snowfall has been added to compute the Probable Maximum Snowfall Accumulation (PMSA). Additionally, the existing functions xhydro.indicators.pmp.precipitable_water_100y and xhydro.indicators.pmp.major_precipitation_events have been adapted accordingly. (PR/328).

Bug fixes

  • A warning will now appear if the calculated area in xh.gis.watershed_properties differs from the theoretical area from HydroBASINS. (PR/303).

  • If returning a GeoDataFrame in xh.gis.watershed_properties, column names have been changed to include the units. (GH/266, PR/303).

  • Multiple corrections to the xh.modelling.format_input function to ensure that the results are correctly formatted for Raven. (PR/303).

  • Importation will no longer fail if the ravenpy package is installed, but cannot find the Raven executable. (GH/305, PR/306).

  • Fixed a bug in Hydrotel modelling where the simulation file’s name was not correctly set to the ‘SIMULATION_COURANTE’. (PR/331).

  • Multiple changes to regional frequency analysis to have functions and variable names more explicit. (PR/332).

  • Fixed a bug in distributed Raven models where the order of the dimensions in the input data was not consistent with the expected order. (PR/344).

Breaking changes

  • The map argument in xh.gis.watershed_delineation has been renamed to m to avoid confusion with the built-in function. (PR/303).

  • The default CRS in xh.gis.watershed_properties and surface_properties has been changed to a call to geopandas.estimate_utm_crs instead of an hardcoded value. (PR/303).

  • The RavenpyModel class has abandoned the longitude, latitude, drainage_area and elevation arguments in favor of a hru argument. (PR/303).

  • The RavenpyModel class has abandoned the explicit evaporation and rain_snow_fraction, but they can still be passed as kwargs. (PR/303).

  • The basin_name coordinate produced by Raven is now renamed to subbasin_id. (PR/339).

  • The variables t from xhfa.local.parametric_quantiles, rp from xhfa.regional.calculate_rp_from_afr and return_periods from xhfa.uncertainties.calc_q_iter all renamed return_period. (GH/269, PR/317).

  • The function xhfa.regional.calculate_rp_from_afr was renamed xhfa.regional.calculate_return_period_from_afr. (PR/317).

  • The use_defaults argument in the Hydrotel class has been removed. (PR/331).

  • The internal xh.modelling._hydrotel._basic_checks function has been removed, as Hydrotel itself performs most of these checks. Checks that are still relevant have been moved to the run function. (PR/331).

  • The station_id dimension in the output of Hydrotel has been renamed to subbasin_id. (PR/331).

Internal changes

  • Running the docs translation steps from Makefile or Batchfile no longer executes the notebooks. (PR/330).

  • Added a security policy (SECURITY.md) to the repository. (PR/340).

  • The RavenpyModel class has been heavily refactored to improve code readability and maintainability. (PR/339, PR/341).

v0.5.0 (2025-04-24)

Contributors to this version: Thomas-Charles Fortier Filion (@TC-FF), Gabriel Rondeau-Genesse (@RondeauG), Trevor James Smith (@Zeitsperre), Julián Ospina (@ospinajulian), Essi Parent (@essicolo).

New features and enhancements

  • Added a new module xhydro.extreme_value_analysis that wraps the Extremes.jl package for Julia. (PR/175).

    • The module provides functions to fit extreme value distributions to data, compute return levels, and get uncertainty estimates.

    • The module is designed to be optional and requires the user to have Julia installed with the Extremes.jl package, along with the PyJuliaCall package for Python.

    • You can use pip install xhydro[julia] to install the required dependencies.

  • Multiple improvements to the documentation. (PR/274, PR/279, PR/293).

Bug fixes

  • Patched the outputs of xh.optimal_interpolation.execute to remove a superfluous station dimension and to ensure that the time dimension has coordinates. (PR/274).

    • Note that this change does not fix the underlying issue with the code, which will be addressed in a future release.

  • Added attributes to variables instead of global attributes in xh.extreme_value_analysis. Modified dimension names and introduced a new dimension, return_period, to the results of xh.extreme_value_analysis.return_level(). (PR/283).

Breaking changes

  • The xh.cc.sampled_indicators function has been separated into two functions: xh.cc.weighted_random_sampling and xh.cc.sampled_indicators. (PR/240).

    • Many of the arguments and outputs have been renamed or reorganized. Since no one was using this function yet AFAIK, no backward compatibility has been maintained.

  • The default use_defaults argument of Hydrotel has been changed from ‘True’ to ‘False’. (PR/233).

  • Multiple functions in the xhydro.frequency_analysis and xhydro.indicators.pmp modules have been updated to require literal arguments instead of positional arguments. (PR/274).

Internal changes

  • “land_use_classification” default collection has been changed to “io-lulc-annual-v02”, as the previous one will be deprecated in December 2024. (PR/227).

  • Added some collection, year, resolution and history attributes to xarray output of “land_use_classification”. (PR/227).

  • Added a “User-Agent” to fix an issue related to pooch calls in the notebooks for recent ReadTheDocs builds. (PR/231).

  • Patched the xhydro.testing.helpers.devereaux() function to add a “User-Agent” by default. (PR/234).

  • Fixed the URL joining logic of the load_registry() and devereaux() functions in the xhydro.testing.helpers module. (PR/234).

  • Updated uncertainties.calc_q_iter() to allow for ungaged catchments. (PR/235).

  • Updated the cookiecutter template. (PR/252):

    • GitHub Actions versions and Python dependencies updated.

    • Added pre-commit hooks for vulture (find dead code), codespell (spelling mistakes), gitleaks (token commit prevention), and zizmor (workflow security).

    • Added an advanced CodeQL workflow for security scanning.

  • Adjusted a handful of dependency pins to handle the latest version of sphinx. (PR/258).

  • The internal xh.cc._weighted_sampling function has been almost entirely rewritten to hopefully be more efficient. Results should be the same as before. (PR/240).

  • Reduced the number of tests performed on Notebooks. (PR/267).

  • Removed _fix_dates from _hydrotel module since it’s not relevant and likely to generate errors. (PR/233).

  • Updated and fixed many dependencies. (PR/295).

  • Added a workflow to test pip installations of xhydro. (PR/295).

  • Removed steps involving ESMF installation from the Makefile and the tox configuration. (PR/295).

v0.4.1 (2024-11-07)

Contributors to this version: Gabriel Rondeau-Genesse (@RondeauG), Trevor James Smith (@Zeitsperre), Louise Arnal (@lou-a).

New features and enhancements

  • A periods parameter was added to frequency_analysis.local.fit to compute multiple separate periods with a single call. (PR/219).

  • In xhydro.cc.sampled_indicators, the delta_type argument can now be a dictionary or None, in which case the attribute delta_kind is used. (PR/220).

  • In xhydro.cc.sampled_indicators, weights along a time or horizon dimension will no longer reduce that dimension. (PR/220).

Bug fixes

  • Fixed a bug in xhydro.modelling.format_input where the function would fail if the input data was a dask array. (PR/214).

  • The executable parameter in the Hydrotel model class is now always required. (PR/214).

Breaking changes

  • The xhydro testing utilities have been rewritten to use pooch for downloading and caching testing datasets from hydrologie/xhydro-testdata. (PR/212).

  • The xhydro testing utilities now require pytest-xdist as a development dependency. (PR/212).

  • Many core dependencies have been updated to more modern versions. (PR/218, PR/228).

  • The delta_type argument in xhydro.cc.sampled_indicators has been renamed to delta_kind and is no longer positional. (PR/220).

  • The xhydro.pmp module has been moved to xhydro.indicators.pmp. (PR/226).

Internal changes

  • Updated the notebooks to reduce the load on ReadTheDocs. (PR/211).

  • Tests can now be run using the pytest-xdist plugin for distributed testing. See the pytest-xdist documentation for more information. (PR/212).

  • Several tests reliant on online servers and services have been marked as online to prevent them from running in contexts where internet access is limited. (PR/212).

  • Many function docstrings and type hints have been updated for accuracy and precision. (PR/212).

  • The xHydro project now has a set of logos for use in documentation, the readme, and other materials. (PR/217).

  • xhydro.indicators is now a package with submodules for different types of indicators. Previous functions have been moved to a generic submodule and are imported at the top level for backwards compatibility. (PR/226).

  • xh.testing.utils.publish_release_notes now has a latest parameter to print only the latest release notes. (PR/228).

v0.4.0 (2024-10-04)

Contributors to this version: Trevor James Smith (@Zeitsperre), Gabriel Rondeau-Genesse (@RondeauG), Thomas-Charles Fortier Filion (@TC-FF), Julián Ospina (@ospinajulian).

New features and enhancements

  • xhydro now supports RavenPy v0.15.0 (RavenHydroFramework v3.8.1). (PR/161).

  • Regional frequency analysis functions as well as Resampling function for uncertainties have been added to the xhydro.frequency_analysis module. (PR/186).

  • New function xhydro.modelling.format_input to format CF-compliant input data for hydrological models (currently only supports Hydrotel). (PR/185).

  • xhydro now has a pmp module to compute the Probable Maximum Precipitation from climate change scenarios. (PR/176).

Internal changes

  • numpy has been pinned below v2.0.0 until xclim and other dependencies are updated to support it. (PR/161).

  • A helper script has been added in the CI directory to facilitate the translation of the xhydro documentation. (GH/63, PR/163).

  • The conda environment now relies on the newly created xdatasets package. (PR/164).

  • The cookiecutter has been updated to the latest commit. Changes include workflow fixes, stricter coding standards, and many small adjustments to the documentation. (PR/164).

  • A previously uncaught YAML formatting issue has been addressed. Stricter style conventions are now enforced. (PR/174).

  • Chunking was adjusted in a few functions to work with the new requirements of apply_ufunc. (PR/180).

  • Updated the cookiecutter template to the latest commit. (PR/177):

    • Actions have been updated and synchronized.

    • Warnings in Pull Requests from forks are now less buggy.

    • A new pre-commit hook and linting step for validating numpy docstrings has been added (numpydoc).

    • All pip-based dependencies used to run in CI are now managed by a CI/requirements_ci.txt that uses hashes of packages for security.

  • Added two new Batch (.bat) files to help facilitate the translation of and the generation of the xhydro documentation in Windows environments. (PR/196).

  • The bumpversion workflow now uses the Hydrologie Helper Bot to make signed commits. (PR/199).

  • Updated the cookiecutter template to the latest commit. (PR/199):

    • Updated development dependencies to the latest versions.

    • Staged support for Python3.13.

    • Added environment caching to existing workflows.

Breaking changes

  • xhydro now requires python >= 3.10. (PR/195).

v0.3.6 (2024-06-10)

Contributors to this version: Gabriel Rondeau-Genesse (@RondeauG), Richard Arsenault (@richardarsenault), Sébastien Langlois (@sebastienlanglois).

New features and enhancements

  • Added support for the Hydrotel hydrological model. (PR/18).

  • Added support for various hydrological models emulated through the Raven hydrological framework. (PR/128).

  • Added optimal interpolation functions for time-series and streamflow indicators. (PR/88, PR/129).

  • Added optimal interpolation notebooks. (PR/123).

  • Added surface properties (elevation, slope, aspect ratio) to the gis module. (PR/151).

Breaking changes

Internal changes

  • Tests using the gamma distribution were changed to the gumbel_r to avoid changes in xclim v0.49.0. (PR/145).

  • The cookiecutter template has been updated to the latest commit. Changes include the addition of a CODE_OF_CONDUCT.rst file, the renaming of CHANGES.rst to CHANGELOG.rst, and many small adjustments to the documentation. (PR/147).

  • Added a CODE_OF_CONDUCT.rst file with Contributor Covenant guidelines. (PR/147).

v0.3.5 (2024-03-20)

Contributors to this version: Trevor James Smith (@Zeitsperre), Thomas-Charles Fortier Filion (@TC-FF), Sébastien Langlois (@sebastienlanglois), Gabriel Rondeau-Genesse (@RondeauG).

New features and enhancements

  • xhydro has implemented a gis module that facilitates geospatial tasks needed for gathering hydrological inputs. (GH/60, PR/61).

Internal changes

  • Added a workflow based on actions/labeler to automatically label Pull Requests based on files changed. (PR/68).

  • Added a conditional trigger to the test-notebooks job to run in advance of pull request approval in the event that the notebooks found within docs/notebooks have been modified (labeled “notebooks”). (PR/68).

  • Significant changes to the Continuous Integration (CI) setup. (PR/65):

    • Added a workflow configuration using label_on_approval.yml and modifications of main.yml so that fewer tests are run on Pull Requests before they are fully approved.

    • Added some pre-commit configurations to both clean up the code within notebooks (NbQA) and strip their outputs (nbstripout).

    • tox is now fully v4.0-compliant.

    • Added a Makefile recipe to facilitate installation of esmpy when esmf is installed and visible on the $PATH.

    • Added a Makefile recipe for running tests over Jupyter notebooks.

    • Synchronized dependencies between pyproject.toml and conda configuration files.

  • Moved the notebooks under a Usage section in the documentation. (GH/114, PR/118).

v0.3.4 (2024-02-29)

Contributors to this version: Trevor James Smith (@Zeitsperre), Thomas-Charles Fortier Filion (@TC-FF), Gabriel Rondeau-Genesse (@RondeauG).

New features and enhancements

  • Added French language support to the documentation. (GH/53, PR/55).

  • Added a new set of functions to support creating and updating pooch registries, caching testing datasets from hydrologie/xhydro-testdata, and ensuring that testing datasets can be loaded into temporary directories. (PR/62).

  • xhydro is now configured to use pooch to download and cache testing datasets from hydrologie/xhydro-testdata. (PR/62).

  • xhydro is now Semantic Versioning v2.0.0 compliant. (PR/70).

  • Added new functions to xhydro.frequency_analysis.local to calculate plotting positions and to prepare plots. (PR/87).

  • xscen now supports Python3.12. (PR/99).

  • xscen now supports pandas >= 2.2.0, xarray >= 2023.11.0, and xclim >= 0.47.0. (PR/99).

  • Added xh.cc.sampled_indicators to compute future indicators using a perturbation approach and random sampling. (PR/54).

Breaking changes

  • Added pooch as an installation dependency. (PR/62).

  • xhydro now requires xarray>=2023.11.0, xclim>=0.48.2, xscen>=0.8.3, and, indirectly, pandas>=2.2.0. The main breaking change is in how yearly frequencies are called (‘YS-’ instead of ‘AS-‘). (PR/54).

  • Functions that output a dict with keys as xrfreq (namely, xh.indicators.compute_indicators) will now return the new nomenclature (e.g. “YS-JAN” instead of “AS-JAN”). (PR/54).

Internal changes

  • Added a new module for testing purposes: xhydro.testing.helpers with some new functions. (PR/62):

    • generate_registry: Parses data found in package (xhydro.testing.data), and adds it to the registry.txt

    • load_registry: Loads installed (or custom) registry and returns dictionary

    • populate_testing_data: Fetches the registry and optionally caches files at a different location (helpful for pytest-xdist).

  • Added a pre-commit hook (numpydoc) to ensure that numpy docstrings are formatted correctly. (PR/62).

  • The cookiecutter has been updated to the latest commit (PR/70, PR/106):

    • Added some workflows (Change file labelling, Cache cleaning, Dependency scans, OpenSSF Scorecard).

    • The README has been updated to organize badges in a table, including a badge for the OpenSSF Scorecard.

    • Updated pre-commit hook versions to the latest available.

    • Formatting tools are now pinned to their pre-commit equivalents.

    • actions-version-updater.yml has been replaced by dependabot.

    • Addressed a handful of misconfigurations in the workflows.

    • Updated ruff to v0.2.0 and black to v24.2.0.

  • Added a few functions missing from the API to their respective modules via __all__. (PR/99).

v0.3.0 (2023-12-01)

Contributors to this version: Gabriel Rondeau-Genesse (@RondeauG), Trevor James Smith (@Zeitsperre).

New features and enhancements

  • The xhydro planification was added to the documentation. (GH/39, PR/49).

Breaking changes

  • xhydro now adheres to PEPs 517/518/621 using the flit backend for building and packaging. (PR/50).

Bug fixes

  • The return_level dimension in xh.frequency_analysis.local.parametric_quantiles() is now the actual return level, not the quantile. (GH/41, PR/43).

Internal changes

  • Added xhydro.testing.utils.publish_release_notes() to help with the release process. (PR/37).

  • xh.frequency_analysis.local.parametric_quantiles() and xh.frequency_analysis.local.criteria() are now lazier. (GH/41, PR/43).

  • The cookiecutter template has been updated to the latest commit via cruft. (PR/50):

    • Manifest.in and setup.py have been removed.

    • pyproject.toml has been added, with most package configurations migrated into it.

    • HISTORY.rst has been renamed to CHANGES.rst.

    • actions-version-updater.yml has been added to automate the versioning of the package.

    • bump-version.yml has been added to automate patch versioning of the package.

    • pre-commit hooks have been updated to the latest versions; check-toml and toml-sort have been added to cleanup the pyproject.toml file.

    • ruff has been added to the linting tools to replace most flake8 and pydocstyle verifications.

v0.2.0 (2023-10-10)

Contributors to this version: Trevor James Smith (@Zeitsperre), Gabriel Rondeau-Genesse (@RondeauG), Thomas-Charles Fortier Filion (@TC-FF), Sébastien Langlois (@sebastienlanglois)

Announcements

  • Support for Python3.8 and lower has been dropped. (PR/11).

  • xHydro now hosts its documentation on Read the Docs. (GH/22, PR/26).

  • Local frequency analysis functions have been added under a new module xhydro.frequency_analysis. (PR/20, PR/27).

New features and enhancements

  • GitHub Workflows for automated testing using tox have been added. (PR/11).

  • Support for various xscen functions has been added to compute indicators and various climate change metrics. (PR/21).

  • New function xh.indicators.compute_volume to convert streamflow data to volumes. (PR/20, PR/27).

  • New function xh.indicators.get_yearly_op to compute block operation (e.g. block maxima, minima, etc.). (PR/20, PR/27).

Breaking changes

  • xHydro repository has renamed its primary development branch from master to main. (PR/13).

  • xHydro now requires a conda environment to be installed. (PR/21).

Bug fixes

  • N/A

Internal changes

  • Added a Pull Request template. (PR/14).

  • Various updates to the autogenerated boilerplate (Ouranosinc/cookiecutter-pypackage) via cruft. (PR/11, PR/12, PR/13):

    • General updates to pre-commit hooks, development dependencies, documentation.

    • Added configurations for Pull Request and Issues templates, Zenodo.

    • Documentation now makes use of sphinx directives for usernames, issues, and pull request hyperlinks (via sphinx.ext.extlinks). (GH/15).

    • GitHub Workflows have been added for automated testing, and publishing.

    • Some sphinx extensions have been added/enabled (sphinx-codeautolink, sphinx-copybutton).

    • Automated testing with tox now updated to use v4.0+ conventions.

    • Removed all references to travis.ci.

  • Deployments to TestPyPI and PyPI are now run using GitHub Workflow Environments as a safeguarding mechanism. (PR/28).

  • Various cleanups of the environment files. (GH/23, PR/30).

  • xhydro now uses the trusted publishing mechanism for PyPI and TestPyPI deployment. (PR/32).

  • Added tests. (PR/27).

0.1.2 (2023-05-10)

  • First release on PyPI.