Fix inadvertent mixup of variable names in build_mapping_file()

[trhille]

Fix bug introduced by commit e51fcd1 that switched the variable names ds and ds_out in several places.

Checklist

• Document (in a comment titled Testing in this PR) any testing that was used to verify the changes

[trhille]

Testing

Without these changes, the landice/ismip6_forcing/atmosphere test case errors out immediately:

Calling the remapping function...
Creating a mapping file. Mapping method used: bilinear
Mapping file does not exist. Building one based on the input/output meshes
Creating temporary scrip files for source and destination grids...
      Failed
Exception raised while running the steps of the test case
Traceback (most recent call last):
  File "/global/cfs/cdirs/m4288/users/trhille/compass/compass/run/serial.py", line 322, in _log_and_run_test
    _run_test(test_case, available_resources)
    ~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/global/cfs/cdirs/m4288/users/trhille/compass/compass/run/serial.py", line 419, in _run_test
    _run_step(test_case, step, test_case.new_step_log_file,
    ~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
              available_resources)
              ^^^^^^^^^^^^^^^^^^^^
  File "/global/cfs/cdirs/m4288/users/trhille/compass/compass/run/serial.py", line 470, in _run_step
    step.run()
    ~~~~~~~~^^
  File "/global/cfs/cdirs/m4288/users/trhille/compass/compass/landice/tests/ismip6_forcing/atmosphere/process_smb_racmo.py", line 113, in run
    self.remap_source_smb_to_mali(racmo_file_temp1,
    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^
                                  remapped_file_temp,
                                  ^^^^^^^^^^^^^^^^^^^
                                  mali_mesh_name,
                                  ^^^^^^^^^^^^^^^
                                  mali_mesh_file,
                                  ^^^^^^^^^^^^^^^
                                  method_remap)
                                  ^^^^^^^^^^^^^
  File "/global/cfs/cdirs/m4288/users/trhille/compass/compass/landice/tests/ismip6_forcing/atmosphere/process_smb_racmo.py", line 190, in remap_source_smb_to_mali
    build_mapping_file(self.config, self.ntasks, self.logger,
    ~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                       input_file, mapping_file,
                       ^^^^^^^^^^^^^^^^^^^^^^^^^
                       scrip_from_latlon=True,
                       ^^^^^^^^^^^^^^^^^^^^^^^
                       mali_mesh_file=mali_mesh_file,
                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                       method_remap=method_remap)
                       ^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/global/cfs/cdirs/m4288/users/trhille/compass/compass/landice/tests/ismip6_forcing/create_mapfile.py", line 71, in build_mapping_file
    create_scrip_from_latlon(ismip6_grid_file, source_grid_scripfile)
    ~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/global/cfs/cdirs/m4288/users/trhille/compass/compass/landice/tests/ismip6_forcing/create_mapfile.py", line 166, in create_scrip_from_latlon
    np.ones(ds.sizes["grid_size"], dtype="int32"), dims=("grid_size",)
            ~~~~~~~~^^^^^^^^^^^^^
  File "/global/cfs/cdirs/fanssie/users/trhille/miniforge3/envs/dev_compass_1.8.0-alpha.2/lib/python3.13/site-packages/xarray/core/utils.py", line 466, in __getitem__
    return self.mapping[key]
           ~~~~~~~~~~~~^^^^^
KeyError: 'grid_size'

The other ismip6_forcing test cases passed, since they do not call create_scrip_from_latlon().

After this fix, the process_smb_racmo step no longer crashes. However, the process_smb step crashes after a few minutes (when using 8km input data) with this error:

  File "/global/cfs/cdirs/m4288/users/trhille/compass/compass/landice/tests/ismip6_forcing/atmosphere/process_smb.py", line 153, in run
    write_netcdf(input_file_combined, combined_file_temp)
<lots of other traceback pointing to xarray and scipy libraries, ending with>
OverflowError: Python integer 22653214272 out of bounds for int32

[xylar]

@trhille, sorry about that! Looks right to me now.

[xylar]

The process_smb error looks like a problem -- a bad value that's bigger than the maximum allowed integer -- and seemingly not related to this PR.

[trhille]

The process_smb error looks like a problem -- a bad value that's bigger than the maximum allowed integer -- and seemingly not related to this PR.

Yes, I'm working on tracking that down now, but I agree that it's not related to the bug in build_mapping_file().