Releases: DHI/mikeio1d
Releases · DHI/mikeio1d
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.
MIKE IO 1D v0.4
MIKE IO 1D v0.4 has the following functionality:
- Faster reading of large result files.
- Changes to ResultQuantity.plot to be more similar to the familiar Matplotlib style.
- A few bug fixes (see changelog for details).
Note: Support for Python 3.6 is removed in this version.
MIKE IO 1D v0.3
MIKE IO 1D v0.3 has the following functionality added:
- Ability to add queries using auto-completion
- Ability to modify res1d file contents using a data frame
- Ability to extract time series to csv, dfs0, and txt files
- Support for querying structures and global data items
MIKE IO 1D v0.2
MIKE IO 1D v0.2 has the following functionality added:
- Ability to read result files in a filtered way
- More detailed information about result files in repr
- Dictionaries containing catchment, node, reach, and global result item classes
- Support for reading SWMM and EPANET result files by upgrading to MIKE 1D v20.1.0
- Support for reading LTS result files
Initial release
This package was previously part of MIKE IO