ValueError: variable has no dimension scale associated with axis 0

[forestbat]

I want to read h5 data from my minio server, however it crashed again:

h5ds = xr.open_dataset(conf.FS.open('s3://watermodel-pub/smap/2019/01/14/SMAP_L4_SM_gph_20190114T073000_Vv7032_001.h5'), engine='h5netcdf')
Traceback (most recent call last):
  File "C:\Users\Administrator\AppData\Roaming\JetBrains\IntelliJIdea2023.3\plugins\python\helpers\pydev\pydevconsole.py", line 364, in runcode
    coro = func()
           ^^^^^^
  File "<input>", line 1, in <module>
  File "C:\Users\Administrator\IdeaProjects\hydrodata\venv\Lib\site-packages\xarray\backends\api.py", line 572, in open_dataset
    backend_ds = backend.open_dataset(
                 ^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\Administrator\IdeaProjects\hydrodata\venv\Lib\site-packages\xarray\backends\h5netcdf_.py", line 414, in open_dataset
    ds = store_entrypoint.open_dataset(
         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\Administrator\IdeaProjects\hydrodata\venv\Lib\site-packages\xarray\backends\store.py", line 43, in open_dataset
    vars, attrs = filename_or_obj.load()
                  ^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\Administrator\IdeaProjects\hydrodata\venv\Lib\site-packages\xarray\backends\common.py", line 210, in load
    (_decode_variable_name(k), v) for k, v in self.get_variables().items()
                                              ^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\Administrator\IdeaProjects\hydrodata\venv\Lib\site-packages\xarray\backends\h5netcdf_.py", line 233, in get_variables
    return FrozenDict(
           ^^^^^^^^^^^
  File "C:\Users\Administrator\IdeaProjects\hydrodata\venv\Lib\site-packages\xarray\core\utils.py", line 477, in FrozenDict
    return Frozen(dict(*args, **kwargs))
                  ^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\Administrator\IdeaProjects\hydrodata\venv\Lib\site-packages\xarray\backends\h5netcdf_.py", line 234, in <genexpr>
    (k, self.open_store_variable(k, v)) for k, v in self.ds.variables.items()
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\Administrator\IdeaProjects\hydrodata\venv\Lib\site-packages\xarray\backends\h5netcdf_.py", line 196, in open_store_variable
    dimensions = var.dimensions
                 ^^^^^^^^^^^^^^
  File "C:\Users\Administrator\IdeaProjects\hydrodata\venv\Lib\site-packages\h5netcdf\core.py", line 260, in dimensions
    self._dimensions = self._lookup_dimensions()
                       ^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\Administrator\IdeaProjects\hydrodata\venv\Lib\site-packages\h5netcdf\core.py", line 175, in _lookup_dimensions
    raise ValueError(
ValueError: variable '/EASE2_global_projection' has no dimension scale associated with axis 0. 
Use phony_dims='sort' for sorted naming or phony_dims='access' for per access naming.

My version of xarray is 2024.1.1.

So why it will happen and how to solve it?

[kmuehlbauer]

@forestbat You would need to add the suggested kwarg to the call to open_dataset. I'm not really sure if this has to be done via backend_kwargs.

ds = xr.open_dataset(file, engine="h5netcdf", phony_dims="access")

Reason is, that the provided h5-file does not declare dimension_scale for one or more variables ( that's a netcdf4 feature). The netcdf4 backend creates so called phony dimensions when reading the data. In h5netcdf we decided to let the user decide if this creation is done on variable access or in sorted manner when accessing the file.

[forestbat]

h5ds = xr.open_dataset(r'D:\Downloads\SMAP_L4_SM_gph_20190114T013000_Vv7032_001.h5', engine='h5netcdf', phony_dim='access')

Traceback (most recent call last):
  File "C:\Users\Administrator\AppData\Roaming\JetBrains\IntelliJIdea2023.3\plugins\python\helpers\pydev\pydevconsole.py", line 364, in runcode
    coro = func()
           ^^^^^^
  File "<input>", line 1, in <module>
  File "C:\Users\Administrator\IdeaProjects\hydrodata\venv\Lib\site-packages\xarray\backends\api.py", line 572, in open_dataset
    backend_ds = backend.open_dataset(
                 ^^^^^^^^^^^^^^^^^^^^^
TypeError: H5netcdfBackendEntrypoint.open_dataset() got an unexpected keyword argument 'phony_dim'

h5ds = xr.open_dataset(r'D:\Downloads\SMAP_L4_SM_gph_20190114T013000_Vv7032_001.h5', engine='h5netcdf', backend_kwargs={'phony_dim':'access'})
Traceback (most recent call last):
  File "C:\Users\Administrator\AppData\Roaming\JetBrains\IntelliJIdea2023.3\plugins\python\helpers\pydev\pydevconsole.py", line 364, in runcode
    coro = func()
           ^^^^^^
  File "<input>", line 1, in <module>
  File "C:\Users\Administrator\IdeaProjects\hydrodata\venv\Lib\site-packages\xarray\backends\api.py", line 572, in open_dataset
    backend_ds = backend.open_dataset(
                 ^^^^^^^^^^^^^^^^^^^^^
TypeError: H5netcdfBackendEntrypoint.open_dataset() got an unexpected keyword argument 'phony_dim'

Not only use phony as an argument directly but also use backend_kwargs failed.

[kmuehlbauer]

Sorry, the kwarg name is phony_dims not phony_dim. I've changed the above code.