Merge pull request #247 from scipp/transform-sizes

Add Transform.sizes property for nicer downstream behavior

src/scippnexus/nxtransformations.py

@@ -69,7 +69,12 @@ class Transform:
     value: sc.Variable | sc.DataArray | sc.DataGroup
     vector: sc.Variable
     depends_on: DependsOn
-    offset: sc.Variable | None
+    offset: sc.Variable | None = None
+
+    @property
+    def sizes(self) -> dict[str, int]:
+        """Convenience property to access sizes of the value."""
+        return self.value.sizes
 
     def __post_init__(self):
         if self.transformation_type not in ['translation', 'rotation']:

tests/transform_test.py

@@ -0,0 +1,38 @@
+import pytest
+import scipp as sc
+
+from scippnexus import DependsOn
+from scippnexus.nxtransformations import Transform, TransformationError
+
+
+@pytest.fixture()
+def depends_on() -> DependsOn:
+    return DependsOn(parent='/', value='.')
+
+
+@pytest.fixture()
+def z_vector() -> sc.Variable:
+    return sc.vector(value=[0, 0, 1], unit='m')
+
+
+def test_init_raises_if_transformation_type_is_invalid(depends_on, z_vector) -> None:
+    with pytest.raises(TransformationError, match='transformation_type'):
+        Transform(
+            name='t1',
+            transformation_type='trans',
+            value=sc.ones(dims=['x', 'y', 'z'], shape=(2, 3, 4)),
+            vector=z_vector,
+            depends_on=depends_on,
+        )
+
+
+def test_sizes_returns_sizes_of_value(depends_on, z_vector) -> None:
+    value = sc.ones(dims=['x', 'y', 'z'], shape=(2, 3, 4))
+    transform = Transform(
+        name='t1',
+        transformation_type='translation',
+        value=value,
+        vector=z_vector,
+        depends_on=depends_on,
+    )
+    assert transform.sizes == {'x': 2, 'y': 3, 'z': 4}