v0.9.1

What's Changed

• Fix bug with ResultLocations.quantities by @ryan-kipawa in #136

Full Changelog: v0.9.0...v0.9.1

v0.9.0

What's Changed

The focus of this release was preparing for version 1.0. A few new features were added, but most changes were related to cleaning up the scripting API. Expect some deprecation warnings. There will likely be one more minor version before version 1.0 is released. If you have feedback on the API, now is the time to voice your opinion 😄

Added

• Create Res1D/Xns11 objects using mikeio1d.open().
• Filter dynamic data loaded in time now with mikeio1d.Res1d(..., time=(start,end)).
• Added some new notebook examples for Res1D.
• Add additional linting rules for documentation, numpy, and pandas.

Fixed

• Autocompletion was flaky when accessing objects with many dots.
• Improved docstring consistency throughout codebase.

Changed

• Updated notebooks to latest scripting API and reorganized.
• Only load header by default (performance improvement).
• Reduce calls to pythonnet (performance improvement).
• Speed up CI tests.
• Clean Res1D scripting API, adding deprecation warnings.
• Use Ruff instead of black for formatting and linting.

Full Changelog: v0.8.2...v0.9.0

MIKE IO 1D v0.8.2

What's Changed

Fixed

• Derived quantities were missing from object html representations.

Full Changelog: v0.8.1...v0.8.2

MIKE IO 1D v0.8.1

What's Changed

This is the same release as v0.8.0, but includes necessary dependency files.

Fixed

• Wheel and source builds did not include all necessary binary dependencies.

Full Changelog: v0.8.0...v0.8.1

MIKE IO 1D v0.8.0

What's Changed

The main edition with this version is the concept of derived quantities, along with a set of default derived quantities.

Added

• Derived quantity concept introduced with an API the same as regular quantities.
• Nine default derived quantities (e.g. 'Node Flooding', 'Reach Filling', etc.).
• Ability to extend MIKE IO 1D with custom derived quantities.
• Quantity units are now more consistently visible in object representations.

Fixed

• Fixed bug where rounding to milliseconds sometimes failed.

Changed

• Updated documentation and README examples.

Full Changelog: v0.7.0...v0.8.0

MIKE IO 1D v0.7.0

What's Changed

The main highlight of this release is new functionality related to xns11 files.

Added

• New API for reading and writing xns11 files (see new notebook examples).
• Access to both raw and processed data in xns11 files.
• Export xns11 sections and markers to GeoPandas.
• Gridpoint indexing from ResultReach by either chainage or number.
• Extra gridpoint static attributes: chainage, reach name, and x/y coordinates.

Fixed

• Various warning fixes related to new Pandas and GeoPandas versions.

Changed

• Removed support for Python 3.8 (to be compatible with Pandas >= 2.1).
• Iterating over IRes1DReach objects must now be done via ResultReach.reaches.

Full Changelog: v0.6.1...v0.7.0

MIKE IO 1D v0.6.1

MIKE IO 1D v0.6.1 is a bugfix release.

Fixed

• Loading MIKE IO 1D together with MIKE+Py
• Fixed override_name parameter that was not working in ResultFrameAggregator
• Fixed converting res11 to res1d
• Fixed calling ResultQuantityCollection.plot with kwargs

Changelog: v0.6...v0.6.1

MIKE IO 1D v0.6

MIKE IO 1D v0.6 is a substantial upgrade to the package relative to previous releases. A fundamental change was the introduction of TimeSeriesId. For more details, see below:

Added

• Introduced TimeSeriesId to uniquely identify results.
• Read methods now include 'column_mode' parameter that enables multiindex reading (e.g. column_mode='compact').
• Added more type hints to improve IDE auto-completion and docstring peeking.
• Merging of regular and LTS extreme/periodic res1d files.
• Convert reaches to GeoPandas in two modes: 'segmented' and 'combined'.
• Export to GeoPandas with quantities aggregated in time.

Changed

• Result reading/writing fundamentally uses TimeSeriesId now instead of QueryData
• DataFrames previously including duplicates are now resolved by TimeSeriesId (especially for reach segments, the 'tag' level is used)
• Following are now abstract base classes: ResultReader, QueryData, ResultLocation
• GeoPandas conversion now includes extra columns matching some TimeSeriesId fields.

Changelog: v0.5...v0.6

MIKE IO 1D v0.5

MIKE IO 1D v0.5 includes the following changes:

Added

• Support for Python 3.12
• Linux support (experimental).
• Initial support for GeoPandas (ability to export static network)
• Geometry package for converting IRes1DLocation objects to corresponding Shapely objects
• Updated documentation hosted on GitHub Pages.

Changed

• Consistent and pythonic test file structure

MIKE IO 1D v0.4.1

MIKE IO 1D v0.4.1 includes the following changes:

• Fixes filtering of reaches that was in some cases producing the wrong results.
• Use MIKE 1D NuGet packages v22.0.3 and v22.0.4 for DHI.Mike1D.ResultDataAccess (related to above)
• A new 'mikenet' module that makes DHI .NET libraries easier to load.