correctly handle nan bg_value

src/depiction/image/image_normalization.py

@@ -37,10 +37,10 @@ def _normalize_single_xarray(self, image: xarray.DataArray, variant: ImageNormal
         with xarray.set_options(keep_attrs=True):
             if variant == ImageNormalizationVariant.VEC_NORM:
                 norm = ((image**2).sum(["c"])) ** 0.5
-                return xarray.where(norm != 0, image / norm, 0)
+                return xarray.where(norm != 0, image / norm, image.attrs.get("bg_value", 0))
             elif variant == ImageNormalizationVariant.STD:
                 std = image.std("c")
-                return xarray.where(std != 0, (image - image.mean("c")) / std, 0)
+                return xarray.where(std != 0, (image - image.mean("c")) / std, image.attrs.get("bg_value", 0))
             else:
                 raise NotImplementedError(f"Unknown variant: {variant}")
 

src/depiction/tools/clustering.py

@@ -100,7 +100,10 @@ def clustering(
         full_features=image_features.data_flat.values.T,
     )
     label_image = MultiChannelImage.from_sparse(
-        values=full_labels[:, np.newaxis], coordinates=image_full_features.coordinates_flat, channel_names=["cluster"]
+        values=full_labels[:, np.newaxis],
+        coordinates=image_full_features.coordinates_flat,
+        channel_names=["cluster"],
+        bg_value=np.nan,
     )
 
     # write the result of the operation

tests/unit/image/test_image_normalization.py

@@ -1,3 +1,4 @@
+import numpy as np
 import pytest
 import xarray as xr
 
@@ -33,6 +34,22 @@ def test_normalize_xarray_single_vec_norm(image_normalizer, single_image):
     assert norm_vec.attrs["bg_value"] == single_image.attrs["bg_value"]
 
 
+def test_normalize_xarray_single_vec_norm_with_nans(image_normalizer):
+    image_with_nans = xr.DataArray(
+        data=[[[2, np.nan], [0, 2]], [[1, 1], [4, np.nan]], [[np.nan, 0], [0, 0]]],
+        dims=["y", "x", "c"],
+        attrs={"bg_value": np.nan},
+    )
+    norm_vec = image_normalizer.normalize_xarray(image_with_nans, variant=ImageNormalizationVariant.VEC_NORM)
+    expected = xr.DataArray(
+        data=[[[1, np.nan], [0, 1]], [[0.707107, 0.707107], [1, np.nan]], [[np.nan, np.nan], [np.nan, np.nan]]],
+        dims=["y", "x", "c"],
+        attrs={"bg_value": np.nan},
+    )
+    xr.testing.assert_allclose(expected, norm_vec)
+    assert np.isnan(norm_vec.attrs["bg_value"])
+
+
 def test_normalize_xarray_single_std(image_normalizer, single_image):
     norm_std = image_normalizer.normalize_xarray(single_image, variant=ImageNormalizationVariant.STD)
     expected = xr.DataArray(