Dfs3.read takes an area argument, documents it as "Read only data within the specified rectangular area (x0, x1, y0, y1)", and then raises unconditionally:
d = mikeio.open("tests/testdata/dissolved_oxygen.dfs3")
d.read(area=(600000, 6150000, 610000, 6160000))
# NotImplementedError: area subsetting is not yet implemented for Dfs3The parameter appears in the signature, in the docstring, and in the generated API reference, so it looks supported from every angle a user would check. Dfs2.read implements it, which makes the dfs3 signature look like a promise of parity.
This is the same shape of problem as #968 (isel(layer=...) exposed on FM 3D DataArrays where it can never work): an argument that exists only to reject you. Either implement it, or drop it from the signature and docstring so the gap is visible before runtime.
Area subsetting is arguably more valuable for dfs3 than for dfs2, since dfs3 files are larger and reading a sub-area is one of the few ways to avoid loading the whole thing. layers= already does this for the vertical direction.
Code sites:
src/mikeio/dfs/_dfs3.py (Dfs3.read)
Dfs3.readtakes anareaargument, documents it as "Read only data within the specified rectangular area (x0, x1, y0, y1)", and then raises unconditionally:The parameter appears in the signature, in the docstring, and in the generated API reference, so it looks supported from every angle a user would check.
Dfs2.readimplements it, which makes the dfs3 signature look like a promise of parity.This is the same shape of problem as
#968(isel(layer=...)exposed on FM 3D DataArrays where it can never work): an argument that exists only to reject you. Either implement it, or drop it from the signature and docstring so the gap is visible before runtime.Area subsetting is arguably more valuable for dfs3 than for dfs2, since dfs3 files are larger and reading a sub-area is one of the few ways to avoid loading the whole thing.
layers=already does this for the vertical direction.Code sites:
src/mikeio/dfs/_dfs3.py (Dfs3.read)