search

corteva / rioxarray

geospatial xarray extension powered by rasterio
https://corteva.github.io/rioxarray
Other
502 stars 80 forks source link

`reproject_match` raises `MissingSpatialDimensionError` with spatial dims set #768

Closed DahnJ closed 2 months ago

DahnJ commented 2 months ago

Code Sample, a copy-pastable example if possible

import numpy as np
import xarray as xr
import rioxarray

lat = [0.1, 0.15, 0.2]
lon = [0.1, 0.15, 0.2]
data = np.arange(1, 10).reshape(3, 3)
ds = xr.Dataset(
        data_vars={
            "foo": (["lat", "lon"], data),
            "bar": (["lat", "lon"], data),
        },
        coords={"lat": lat, "lon": lon},
    )

ds = ds.rio.set_spatial_dims(x_dim="lon", y_dim="lat")
ds = ds.rio.write_crs(4326)

ds = ds.rio.reproject_match(ds)

Thanks @bocalml for the MRE.

Problem description

This results in MissingSpatialDimensionError: x dimension (lon) not found.

This is likely related to the renaming mentioned in https://github.com/corteva/rioxarray/issues/751 but I haven't been able to make sense of this fully and the result is surprising to me.

Full traceback ```python --------------------------------------------------------------------------- MissingSpatialDimensionError Traceback (most recent call last) Cell In[31], line 1 ----> 1 ds = ds.rio.reproject_match(ds) File ~/miniconda3/envs/myenv/lib/python3.12/site-packages/rioxarray/raster_dataset.py:191, in RasterDataset.reproject_match(self, match_data_array, resampling, **reproject_kwargs) 189 raise 190 resampled_dataset[var] = self._obj[var].copy() --> 191 return resampled_dataset.rio.set_spatial_dims( 192 x_dim=self.x_dim, y_dim=self.y_dim, inplace=True 193 ) File ~/miniconda3/envs/myenv/lib/python3.12/site-packages/rioxarray/rioxarray.py:853, in XRasterBase.set_spatial_dims(self, x_dim, y_dim, inplace) 851 data_obj.rio._x_dim = x_dim 852 else: --> 853 raise MissingSpatialDimensionError( 854 f”x dimension ({x_dim}) not found.{_get_data_var_message(data_obj)}” 855 ) 856 if y_dim in data_obj.dims: 857 data_obj.rio._y_dim = y_dim MissingSpatialDimensionError: x dimension (lon) not found. ```

Expected Output

The dataset itself with no error.

Environment Information

rioxarray (0.15.3) deps:
  rasterio: 1.3.9
    xarray: 2024.3.0
      GDAL: 3.8.4
      GEOS: 3.12.1
      PROJ: 9.3.1
 PROJ DATA: [/Users/danieljahn/miniconda3/envs/myenv/share/proj](https://file+.vscode-resource.vscode-cdn.net/Users/danieljahn/miniconda3/envs/myenv/share/proj)
 GDAL DATA: [/Users/danieljahn/miniconda3/envs/myenv/share/gdal](https://file+.vscode-resource.vscode-cdn.net/Users/danieljahn/miniconda3/envs/myenv/share/gdal)

Other python deps:
     scipy: 1.12.0
    pyproj: 3.6.1

System:
    python: 3.12.2 | packaged by conda-forge | (main, Feb 16 2024, 20:54:21) [Clang 16.0.6 ]
executable: [/Users/danieljahn/miniconda3/envs/myenv/bin/python](https://file+.vscode-resource.vscode-cdn.net/Users/danieljahn/miniconda3/envs/myenv/bin/python)
   machine: macOS-14.2.1-arm64-arm-64bit

Installation method

Conda

Conda environment information (if you installed with conda):


Environment (conda list):

``` $ conda list | grep -E "rasterio|xarray|gdal" cf-xarray 0.9.0 pypi_0 pypi gdal 3.8.4 py312h56161e1_3 conda-forge libgdal 3.8.4 hd76467a_3 conda-forge rasterio 1.3.9 py312hd404879_2 conda-forge rioxarray 0.15.3 pyhd8ed1ab_0 conda-forge xarray 2024.3.0 pyhd8ed1ab_0 conda-forge ```


Details about conda and system ( conda info ):

``` $ conda info active environment : myenv active env location : /Users/danieljahn/miniconda3/envs/myenv shell level : 4 user config file : /Users/danieljahn/.condarc populated config files : /Users/danieljahn/.condarc conda version : 23.3.1 conda-build version : 3.23.3 python version : 3.11.0.final.0 virtual packages : __archspec=1=arm64 __osx=14.2.1=0 __unix=0=0 base environment : /Users/danieljahn/miniconda3 (writable) conda av data dir : /Users/danieljahn/miniconda3/etc/conda conda av metadata url : None channel URLs : https://conda.anaconda.org/t//conda-forge/osx-arm64 https://conda.anaconda.org/t//conda-forge/noarch package cache : /Users/danieljahn/miniconda3/pkgs /Users/danieljahn/.conda/pkgs envs directories : /Users/danieljahn/miniconda3/envs /Users/danieljahn/.conda/envs platform : osx-arm64 user-agent : conda/23.3.1 requests/2.31.0 CPython/3.11.0 Darwin/23.2.0 OSX/14.2.1 UID:GID : 501:20 netrc file : /Users/danieljahn/.netrc offline mode : False ```
snowman2 commented 2 months ago

Bug confirmed - thank you for your report!

Related #372

Even though this doesn't fix the issue, I recommend changing your code like as shown in the crs management example as it will help your spatial dimension persist better and follow the CF conventions for metadata:

import numpy
import xarray
import rioxarray

lat = [0.1, 0.15, 0.2]
lon = [0.1, 0.15, 0.2]
data = numpy.arange(1, 10).reshape(3, 3)
ds = xarray.Dataset(
        data_vars={
            "foo": (["lat", "lon"], data),
            "bar": (["lat", "lon"], data),
        },
        coords={"lat": lat, "lon": lon},
    )

ds = ds.rio.write_crs(
    4326, inplace=True
).rio.set_spatial_dims(
    x_dim="lon", y_dim="lat", inplace=True
).rio.write_coordinate_system(inplace=True)
ds = ds.rio.reproject_match(ds)