Test various fields with clumerge()

tests/conftest.py

@@ -95,6 +95,7 @@ def pytest_addoption(parser):
 t_ds_ot_n: Sequence[int]
 t_ds_od_n: Sequence[int]
 t_no_clusters_field: Sequence[bool]
+t_ds_cgs_n: Sequence[int]
 
 
 def pytest_report_header(config):
@@ -128,6 +129,7 @@ def pytest_generate_tests(metafunc):
         t_ds_ot_n = [0]
         t_ds_od_n = [0, 1]
         t_no_clusters_field = (False,)
+        t_ds_cgs_n = [2]
     elif test_level == "ci":
         # CI test level
         seeds = [123]
@@ -151,6 +153,7 @@ def pytest_generate_tests(metafunc):
         t_ds_ot_n = [0, 1]
         t_ds_od_n = [0, 1]
         t_no_clusters_field = [False, True]
+        t_ds_cgs_n = [2, 3]
     elif test_level == "normal":
         seeds = [0, 123, 6789]
         t_ndims = [1, 2, 3, 10]
@@ -173,6 +176,7 @@ def pytest_generate_tests(metafunc):
         t_ds_ot_n = [0, 1]
         t_ds_od_n = [0, 1, 2]
         t_no_clusters_field = [False, True]
+        t_ds_cgs_n = [2, 3, 4]
     elif test_level == "full":
         seeds = [0, 123, 6789, 9876543]
         t_ndims = [1, 2, 3, 5, 10, 30]
@@ -195,6 +199,7 @@ def pytest_generate_tests(metafunc):
         t_ds_ot_n = [0, 1, 2]
         t_ds_od_n = [0, 1, 2]
         t_no_clusters_field = [False, True]
+        t_ds_cgs_n = [2, 3, 4, 5]
     else:
         raise ValueError(f"Unknown test level {test_level!r}")
 
@@ -224,6 +229,7 @@ def param_if(param: str, value: Sequence[Any]):
     param_if("ds_ot_n", t_ds_ot_n)
     param_if("ds_od_n", t_ds_od_n)
     param_if("no_clusters_field", t_no_clusters_field)
+    param_if("ds_cgs_n", t_ds_cgs_n)
 
 
 @pytest.fixture()

tests/test_main.py

@@ -928,3 +928,73 @@ def test_clumerge_general(
         assert mds["points"].shape == expect_shape
         assert max(mds["clusters"]) == tclu
         assert can_cast(mds["clusters"].dtype, int64)
+
+
+def test_clumerge_fields(
+    prng: Generator,
+    ndims,
+    ds_cgs_n,
+):
+    """# Test clumerge with data from clugen() and merging more fields."""
+    datasets: MutableSequence[NamedTuple | Mapping[str, ArrayLike]] = []
+    tclu: int = 0
+    tclu_i: int = 0
+    tpts: int = 0
+
+    # Create data sets with clugen()
+    for _ in range(ds_cgs_n):
+        # clugen() should run without problem
+        with warnings.catch_warnings():
+            # Check that the function runs without warnings
+            warnings.simplefilter("error")
+
+            ds_cgs = clugen(
+                ndims,
+                prng.integers(1, high=11),
+                prng.integers(1, high=101),
+                prng.random(size=ndims),
+                prng.random(),
+                prng.random(size=ndims),
+                prng.random(),
+                prng.random(),
+                prng.random(),
+                allow_empty=True,
+                rng=prng,
+            )
+
+            tclu += len(unique(ds_cgs.clusters))
+            tpts += len(ds_cgs.points)
+            tclu_i += len(ds_cgs.sizes)
+            datasets.append(ds_cgs)
+
+        # Check that clumerge() is able to merge data set fields related to points
+        # without warnings
+        with warnings.catch_warnings():
+            # Check that the function runs without warnings
+            warnings.simplefilter("error")
+            mds = clumerge(*datasets, fields=("points", "clusters", "projections"))
+
+        # Check that the number of clusters and points is correct
+        expect_shape = (tpts,) if ndims == 1 else (tpts, ndims)
+        assert mds["points"].shape == expect_shape
+        assert mds["projections"].shape == expect_shape
+        assert max(mds["clusters"]) == tclu
+        assert can_cast(mds["clusters"].dtype, int64)
+
+        # Check that clumerge() is able to merge data set fields related to clusters
+        # without warnings
+        with warnings.catch_warnings():
+            # Check that the function runs without warnings
+            warnings.simplefilter("error")
+            mds = clumerge(*datasets, fields=("sizes", "centers", "directions", "angles", "lengths"),
+            clusters_field=None,
+        )
+
+        # Check that the cluster-related fields have the correct sizes
+        expect_shape = (tclu_i,) if ndims == 1 else  (tclu_i, ndims)
+        assert len(mds["sizes"]) == tclu_i
+        assert can_cast(mds["sizes"], int64)
+        assert mds["centers"].shape == expect_shape
+        assert mds["directions"].shape == expect_shape
+        assert len(mds["angles"]) == tclu_i
+        assert len(mds["lengths"]) == tclu_i