Longlist of new features

[jsmariegaard]

Bugs/inconsistencies

• Notebook. Simplify "Contour lines" in "Dfsu - Export to shapefile.ipynb"
• Notebook. Generic.ipynb needs updating to MIKE IO 1.0
• Fix: keepdims not implemented for Dfs3. Current behaviour is opposite to default argument of keepdims (=False). Fix dfs3 single layer #460
• Fix: renaming a DataArray is possible but is only partly reflected in Dataset (not in keys of _data_vars). Maybe _data_vars should be a sequence instead... Change Dataset to a Sequence instead of a Mapping #423
• Fix. Dfsu contour/contourf plots with nan data fails - Dfsu: contourf plot type results error with NaN values #343
• Fix: concat of Datasets/DataArrays without time axis is not working. dataset concatenation single timestep #441
• consistency in argument names: x,y, z or coords or both?
• projection vs projection_string: remove use of the latter.

New features

• improve/harmonize string representation of all objects (especially dfs objects)
• Fix: axis-name for Grid1D after selection of line in Grid2D
• Fix: try-catch when reading dfs files to handle partly corrupt files
• add generic to global namesspace - so that one can mikeio.concat() etc?
• All generic methods should be able to take either dfs files or Dataset/DataArray
• mikeio.concat() that takes both files, Dataset and DataArray
• Dfsu interp - pseudo-Laplacian interpolation
• Dfs2 interp - bilinear interp (be careful with 2dspectrum)
• ds.to_dfs() add deletevalue argument (and remove deletevalue property on Dataset/DataArray)
• EUMUnit nicer on plots: e.g. [m**2*s/deg] instead of [meter_pow_2_sec_per_deg]
• Dfsu plot: boundary_outline should only be "bold" for nodes with code>0
• Grid Geometry: named dims (e.g. 'z') for dfs1/dfs2 geometries such that we can do better plotting of e.g. vertical data
• Interp of directions (by eum)
• Dfsu3D should have volume property (or method if the dynamic volume needs to be calculated)
• Dfsu3D select transect by extracting a number of profiles along a line (with fixed spacing)
• Dfsu3D interp vertical column
• Dfsu aggregate should have support for taking area/volume into account where possible (e.g. average)
• Fix: isel() for layered dfsu is calculating node_ids twice (refactor)
• Geometry. __eq__ for comparison of two geometries - for large FM geometries, compare only shape and first and last coordinates.
• DataArray.sel() selecting point/line/area outside domain should only be possible with sel(..., method='nearest')
• Dfs2 proper plotting of rotated grids
• Dfs2 spectrum files to use same plotting as Dfsu-plotter (via specialized Grid2D object). Note: First and last row can be the same (0=360 remove from plotting).
• Grid2D to_geometryFM() to be used as basis for Dfs2->Dfsu conversion, Grid2D to GeometryFM #374
• Dfs2 Datasets -> save to dfsu
• Dfs2 handle the reading of files with static items
• Edit custom blocks (and maintain after read/write) especially for dfs2, Read/modify dfs custom blocks #202
• DataArray. Long/lat global datasets - handle circular x-axis (e.g. for interpolation)
• DataArray. interp_like() similar to xarray for more Geometry types.
• Dataset. plotting of vector fields like speed/direction (optionally with custom scalar field as colors)
• Dataset/DataArray: sel() with area argument as polygon or bbox to select all elements inside area
• Dataset. create new Dataset by providing a list of mixed Datasets and DataArrays: ds=mikeio.Dataset([ds1, ds2, da1, da2])
• Dataset/DataArray: aggregate, quantile etc over layers (i.e. allow "layer" as axis)
• Dataset/DataArray: aggregate, quantile for Dfs2/Dfs3 over single axis should give Dfs1/Dfs2
• Dataset/DataArray: argmin(), argmax() methods that return index integers, DataArray argmin/argmax #409
• Temporal aggregation, group by e.g. month, to easily obtain monthly statistics
• Dataset. to_xarray() for Dfsu
• Dataset. from_xarray() - would make it alot easier to convert grib data to dfs2! Add new method Dataset.from_xarray #369
• (Dfsu spectral: aggregating on spectral axis)

Notebooks/examples/documentation:

• Example showing how to plot data on top of cartopy