Zonal stats with `method='exactextract'` not working on flat `xarray.Dataarray`

[maawoo]

Here is a reproducible example:

import xarray as xr
import numpy as np
import geopandas as gpd
from shapely.geometry import Point

x_coords = np.arange(500000, 500000 + 100 * 20, 20)
y_coords = np.arange(5500000, 5500000 + 100 * 20, 20)

# dummy Dataarray
test_da = xr.DataArray(
    np.random.random((100, 100)),
    coords={
        'y': y_coords,
        'x': x_coords
    },
    dims=['y', 'x'],
)
# dummy Dataset
test_ds = test_da.to_dataset(name='test_var')

test_gdf = gpd.GeoDataFrame({
    'geometry': [Point(x, y) for x, y in zip(np.random.uniform(x_coords.min(), x_coords.max(), 5), 
                                             np.random.uniform(y_coords.min(), y_coords.max(), 5))]}, 
                                             crs='EPSG:32632')
test_gdf['geometry_buffer'] = test_gdf.geometry.buffer(50)

Using the xarray.Datasetworks:

test_ds.xvec.zonal_stats(
    test_gdf.geometry_buffer, x_coords="x", y_coords="y",
    method='exactextract'
)

Using the xarray.Dataarray doesn't work and raises ValueError: Must pass non-zero number of levels/codes:

test_da.xvec.zonal_stats(
    test_gdf.geometry_buffer, x_coords="x", y_coords="y",
    method='exactextract'
)

arr_dims in the following part is empty when a Dataarray is passed and that results in stack to fail:

xvec/xvec/zonal.py

Lines 699 to 702 in 3ba5d0c

	# Stack the other dimensions into one dimension called "location"
	arr_dims = tuple(dim for dim in acc._obj.dims if dim not in [x_coords, y_coords])
	data = acc._obj.stack(location=arr_dims)
	locs = data.location.size

[martinfleis]

Thanks for the report! Would you be willing to fix it?

[maawoo]

Hi @martinfleis, sure I'll have a look in the next days 👍🏻