issues

What happened?

In recent xarray versions, da.interp() behaves differently when using floats as input, in comparison to using a DataArray with a new dimension as input. In the attached example, the variables interp_with_floats and interp_with_da return a different amount of float/nan values: [35.824154 35.810417 35.796684 nan nan] [35.824154 35.810417 nan nan nan] This is the case with xarray 2023.1.0/2023.2.0, but was not yet happening in xarray 2022.6.0/2022.12.1.dev15+gd6d24507. I cannot find .interp() related changes in https://docs.xarray.dev/en/stable/whats-new.html. A CMEMS use case is described in https://github.com/Deltares/dfm_tools/issues/287

What did you expect to happen?

Equal interpolation results with both methods since the input values are equal.

Minimal Complete Verifiable Example

```Python import numpy as np import xarray as xr

ds = xr.Dataset() so_np = np.array([[[35.819576, 35.82568 , 35.82873 ], [35.819576, 35.824154, 35.831783], [35.822628, 35.824154, 35.82873 ]],

              [[35.802788, 35.80584 , 35.815   ],
               [35.815   , 35.810417, 35.821102],
               [35.824154, 35.813473, 35.81805 ]],

              [[35.786003, 35.789055,       np.nan],
               [35.807365, 35.796684,       np.nan],
               [35.824154, 35.80584 ,       np.nan]],

              [[35.776848,       np.nan,       np.nan],
               [35.792107,       np.nan,       np.nan],
               [35.822628,       np.nan,       np.nan]],

              [[35.781425,       np.nan,       np.nan],
               [35.792107,       np.nan,       np.nan],
               [35.789055,       np.nan,       np.nan]]])

ds['so'] = xr.DataArray(so_np,dims=('depth','latitude','longitude')) ds['longitude'] = xr.DataArray([-9.6, -9.5, -9.4], dims=('longitude')) ds['latitude'] = xr.DataArray([42.9, 43.0, 43.1], dims=('latitude'))

x_xr = xr.DataArray([-9.5],dims=('plipoints')) y_xr = xr.DataArray([43],dims=('plipoints'))

interp_with_floats = ds.interp(longitude=x_xr[0], latitude=y_xr[0], method='linear').so #selecting one value from the da drops the new plipoints dimension interp_with_da_existing = ds.interp(longitude=x_xr.values, latitude=y_xr.values, method='linear').so.isel(longitude=0,latitude=0) #using the DataArray values keeps lat/lon dimenions, gives the same interp result interp_with_da_newdim = ds.interp(longitude=x_xr, latitude=y_xr, method='linear').so.isel(plipoints=0) #using the DataArray introduces a plipoints dimension, which gives different interp result print(interp_with_floats.to_numpy()) print(interp_with_da_existing.to_numpy()) print(interp_with_da_newdim.to_numpy()) print(xr.version)

assert (interp_with_floats.isnull()==interp_with_da_existing.isnull()).all() #success assert (interp_with_floats.isnull()==interp_with_da_newdim.isnull()).all() #fails with scipy>=1.10.0 ```

MVCE confirmation

• [X] Minimal example — the example is as focused as reasonably possible to demonstrate the underlying issue in xarray.
• [X] Complete example — the example is self-contained, including all data and the text of any traceback.
• [X] Verifiable example — the example copy & pastes into an IPython prompt or Binder notebook, returning the result.
• [X] New issue — a search of GitHub Issues suggests this is not a duplicate.

Relevant log output

No response

Anything else we need to know?

No response

Environment

• [x] Closes #4061
• [x] Closes #49
• [x] Closes #7014
• [x] Tests added
• [x] User visible changes (including notable bug fixes) are documented in whats-new.rst
• [ ] New functions/methods are listed in api.rst >> not applicable

Is your feature request related to a problem?

Yes, xarray not being able to open netcdf files with non-unique variable/dimension names is an issue for many people. Me personally have come across several files where this is an issue. As a not so nice workaround I use drop_variables for variables x/y before reading the file with xarray, and then adding them again under a different name with help of the netCDF4 library. It would be very convenient to have a rename_variables argument for xr.open_dataset() (and xr.open_mfdataset()).

This suggestion was also made by others in other issues but I cannot find a feature request for it: by @jthielen in https://github.com/pydata/xarray/issues/2233#issuecomment-782188310 by @markelg in https://github.com/pydata/xarray/issues/2368#issuecomment-415402600

Describe the solution you'd like

rename_variables argument for xr.open_dataset() (and xr.open_mfdataset()).

Describe alternatives you've considered

No response

Additional context

No response