Unable to load mas data with Psipy

[vreville]

Hi,

Since quite recently, I have been unable to load MAS data (from https://www.predsci.com/mhdweb/data_access.php) using pispy. The error message that I get is the following :

 File "/Users/vreville/Projets/MAS/mascoro_test.py", line 63, in <module>
    d=mas_as_pluto(directory)
  File "/Users/vreville/Projets/MAS/mascoro_test.py", line 35, in mas_as_pluto
    F=mas_output[ff]
      ~~~~~~~~~~^^^^
  File "/Users/vreville/env3.13/lib/python3.13/site-packages/psipy/model/base.py", line 64, in __getitem__
    data = self.load_file(var)
  File "/Users/vreville/env3.13/lib/python3.13/site-packages/psipy/model/mas.py", line 77, in load_file
    return read_mas_file(self.path, var)
  File "/Users/vreville/env3.13/lib/python3.13/site-packages/psipy/io/mas.py", line 57, in read_mas_file
    data = [_read_mas(f, var) for f in files]
            ~~~~~~~~~^^^^^^^^
  File "/Users/vreville/env3.13/lib/python3.13/site-packages/psipy/io/mas.py", line 77, in _read_mas
    data = xr.Dataset({var: xr.DataArray(data=data, coords=coords, dims=dims)})
                            ~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/vreville/env3.13/lib/python3.13/site-packages/xarray/core/dataarray.py", line 487, in __init__
    coords, dims = _infer_coords_and_dims(data.shape, coords, dims)
                   ~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/vreville/env3.13/lib/python3.13/site-packages/xarray/core/dataarray.py", line 192, in _infer_coords_and_dims
    raise ValueError(
    ...<2 lines>...
    )
ValueError: different number of dimensions on data and dims: 3 vs 4

However, the code is still working for previously downloaded MAS cube, so I suspect the product itself has changed recently.

Any help appreciated !

Victor

[thepeteriley]

Hi Victor,

That's a strange error. It does appear that there's something different about the datasets. Could you do a test, and print:

print(f"data.shape: {data.shape}, dims: {dims}")

for each of the datasets (one that worked and one that doesn't)?

Or, if you can point me to the relevant files on the data access page (e.g., different CRs), I'll take a look.

Cheers, Pete

[vreville]

Hi Pete,

thanks, I think I found the problem. When calling :

mas_output=MASOutput(directory)
mas_output["br"]

it tries to open both br002.hdf, and br_r0.hdf for which the dimensions are not matching, as this is the inner boundary condition. The easy fix is to remove br_r0.hdf from the directory, but there is surely a better way to handle this.

This file looks indeed like a recent addition to the downloaded zip folder.

Cheers,
Victor

[thepeteriley]

Hi Victor,
(Sorry, I thought I posted a response to this yesterday, but I must not have pushed it.) Yes, you're right. The older CRs didn't have the br_r0.hdf file. Your solution is definitely a good 'quick fix', but we should check for its presence and mask it from the search for the 3-D files.
Cheers, Pete