Refactor Seismic Dataset Templates Documentation and Tests

- Updated docstrings for Seismic3DOffsetTilesTemplate and Seismic3DReceiverGathersTemplate to provide concise descriptions.
- Added a new test for chunk size calculation in both templates to ensure proper memory management.
- Adjusted existing tests to reflect changes in dataset naming conventions and template attributes.

These changes enhance clarity and maintainability of the codebase.

Author: BrianMichell

src/mdio/builder/templates/seismic_3d_offset_tiles.py

@@ -10,32 +10,7 @@
 
 
 class Seismic3DOffsetTilesTemplate(AbstractDatasetTemplate):
-    """Seismic Offset Tiles pre-stack 3D Dataset template.
-
-    A 5D template for wide-azimuth seismic data organized by offset vector tiles.
-    Data is binned by CDP location (inline, crossline) with offset vector
-    decomposition into inline and crossline offset tile components.
-
-    Dimensions:
-        - inline: Inline bin position
-        - crossline: Crossline bin position
-        - inline_offset_tile: Inline component of the offset vector tile
-        - crossline_offset_tile: Crossline component of the offset vector tile
-        - time/depth: Sample dimension
-
-    This organization is optimal for:
-        - Wide-azimuth data preservation
-        - 5D interpolation and regularization
-        - Azimuthal analysis and processing
-        - Offset vector filtering
-
-    The offset vector tiles partition the offset-azimuth space into a regular grid,
-    preserving both offset magnitude and azimuth information in a format suitable
-    for modern wide-azimuth processing workflows.
-
-    Args:
-        data_domain: The domain of the dataset ('time' or 'depth').
-    """
+    """Seismic 3D template for offset vector tile (OVT) binned gathers."""
 
     def __init__(self, data_domain: SeismicDataDomain = "time"):
         super().__init__(data_domain=data_domain)

src/mdio/builder/templates/seismic_3d_receiver_gathers.py

@@ -9,27 +9,7 @@
 
 
 class Seismic3DReceiverGathersTemplate(AbstractDatasetTemplate):
-    """Seismic receiver gather pre-stack 3D Dataset template.
-
-    A template for surveys with fixed receiver positions (OBN, OBC, land fixed-spread)
-    where data is organized by receiver position for receiver-side processing.
-    Shots are organized by shot lines with a calculated shot index.
-
-    This template is time-domain only as receiver gathers are typically used
-    for early-stage processing before depth conversion.
-
-    Dimensions:
-        - receiver: Index of the receiver node/station
-        - shot_line: Shot line or swath identifier
-        - shot_index: Sequential index of shots within the line (calculated, 0-N)
-        - time: Sample dimension
-
-    This organization is optimal for:
-        - Receiver-side wavefield separation (up/down)
-        - Receiver-consistent deconvolution
-        - Multi-component processing
-        - Mirror imaging from OBN
-    """
+    """Seismic 3D receiver gathers template with calculated shot index."""
 
     def __init__(self) -> None:
         super().__init__(data_domain="time")

tests/unit/v1/templates/test_seismic_3d_offset_tiles.py

@@ -110,7 +110,7 @@ class TestSeismic3DOffsetTilesTemplate:
     """Unit tests for Seismic3DOffsetTilesTemplate."""
 
     def test_configuration(self, data_domain: SeismicDataDomain) -> None:
-        """Unit tests for Seismic3DOffsetTilesTemplate configuration."""
+        """Test template configuration and attributes."""
         t = Seismic3DOffsetTilesTemplate(data_domain=data_domain)
 
         # Template attributes
@@ -128,19 +128,38 @@ def test_configuration(self, data_domain: SeismicDataDomain) -> None:
         assert attrs == {"surveyType": "3D", "gatherType": "offset_tiles"}
         assert t.default_variable_name == "amplitude"
 
+    def test_chunk_size_calculation(self, data_domain: SeismicDataDomain) -> None:
+        """Test that chunk shape produces approximately 9 MiB chunks.
+
+        The chunk shape (4, 4, 6, 6, 4096) produces:
+        4 * 4 * 6 * 6 * 4096 = 2,359,296 samples.
+        With float32 (4 bytes): 2,359,296 * 4 = 9,437,184 bytes = 9 MiB.
+        """
+        t = Seismic3DOffsetTilesTemplate(data_domain=data_domain)
+
+        chunk_shape = t.full_chunk_shape
+        assert chunk_shape == (4, 4, 6, 6, 4096)
+
+        samples_per_chunk = 1
+        for dim_size in chunk_shape:
+            samples_per_chunk *= dim_size
+
+        bytes_per_chunk = samples_per_chunk * 4
+        assert bytes_per_chunk == 9 * 1024 * 1024  # 9 MiB
+
     def test_build_dataset(self, data_domain: SeismicDataDomain, structured_headers: StructuredType) -> None:
-        """Unit tests for Seismic3DOffsetTilesTemplate build."""
+        """Test building a complete dataset with the template."""
         t = Seismic3DOffsetTilesTemplate(data_domain=data_domain)
         t.add_units({"cdp_x": UNITS_METER, "cdp_y": UNITS_METER})
         t.add_units({"time": UNITS_SECOND, "depth": UNITS_METER})
 
         dataset = t.build_dataset(
-            "Wide Azimuth Offset Tiles",
+            "OVT Gathers",
             sizes=(256, 256, 12, 12, 2048),
             header_dtype=structured_headers,
         )
 
-        assert dataset.metadata.name == "Wide Azimuth Offset Tiles"
+        assert dataset.metadata.name == "OVT Gathers"
         assert dataset.metadata.attributes["surveyType"] == "3D"
         assert dataset.metadata.attributes["gatherType"] == "offset_tiles"
 

tests/unit/v1/templates/test_seismic_3d_receiver_gathers.py

@@ -117,7 +117,7 @@ class TestSeismic3DReceiverGathersTemplate:
     """Unit tests for Seismic3DReceiverGathersTemplate."""
 
     def test_configuration(self) -> None:
-        """Unit tests for Seismic3DReceiverGathersTemplate configuration."""
+        """Test template configuration and attributes."""
         t = Seismic3DReceiverGathersTemplate()
 
         # Template attributes
@@ -137,8 +137,27 @@ def test_configuration(self) -> None:
         assert attrs == {"surveyType": "3D", "gatherType": "receiver_gathers"}
         assert t.default_variable_name == "amplitude"
 
+    def test_chunk_size_calculation(self) -> None:
+        """Test that chunk shape produces approximately 8 MiB chunks.
+
+        The chunk shape (1, 1, 512, 4096) produces:
+        1 * 1 * 512 * 4096 = 2,097,152 samples.
+        With float32 (4 bytes): 2,097,152 * 4 = 8,388,608 bytes = 8 MiB.
+        """
+        t = Seismic3DReceiverGathersTemplate()
+
+        chunk_shape = t.full_chunk_shape
+        assert chunk_shape == (1, 1, 512, 4096)
+
+        samples_per_chunk = 1
+        for dim_size in chunk_shape:
+            samples_per_chunk *= dim_size
+
+        bytes_per_chunk = samples_per_chunk * 4
+        assert bytes_per_chunk == 8 * 1024 * 1024  # 8 MiB
+
     def test_build_dataset(self, structured_headers: StructuredType) -> None:
-        """Unit tests for Seismic3DReceiverGathersTemplate build."""
+        """Test building a complete dataset with the template."""
         t = Seismic3DReceiverGathersTemplate()
         t.add_units({"receiver_x": UNITS_METER, "receiver_y": UNITS_METER})
         t.add_units({"source_coord_x": UNITS_METER, "source_coord_y": UNITS_METER})