CLN: Fix dataset name -> CMACBench (#805)

* CLN: Fix dataset name -> CMACBench, add docstrings

* TST: Fix test for renamed CMACBench dataset

* Fix bug in CMACBench found with tests

* TST: Fix tests/test_datasets/test_get.py

* TST/CLN: Rename -> test_datasets/test_cmacbench.py

Author: NickleDave

src/vak/datasets/__init__.py

@@ -1,13 +1,13 @@
-from . import biosoundsegbench
-from .biosoundsegbench import BioSoundSegBench, SplitsMetadata
+from . import cmacbench
+from .cmacbench import CMACBench, SplitsMetadata
 from .get import get
 
 __all__ = [
-    "biosoundsegbench",
-    "BioSoundSegBench",
+    "cmacbench",
+    "CMACBench",
     "get",
     "SplitsMetadata",
 ]
 
 # TODO: make this a proper registry
-DATASETS = {"BioSoundSegBench": BioSoundSegBench}
+DATASETS = {"CMACBench": CMACBench}

src/vak/datasets/biosoundsegbench.py src/vak/datasets/cmacbench.pysrc/vak/datasets/biosoundsegbench.py renamed to src/vak/datasets/cmacbench.py

@@ -1,4 +1,4 @@
-"""Class representing BioSoundSegBench dataset."""
+"""Class representing CMACBench dataset."""
 
 from __future__ import annotations
 
@@ -35,13 +35,21 @@
 
 @define
 class SampleIDVectorPaths:
+    """Class that represents paths 
+    to npy files containing sample ID vectors.
+    """
     train: pathlib.Path
     val: pathlib.Path
     test: pathlib.Path
 
 
 @define
 class IndsInSampleVectorPaths:
+    """Class that represents paths 
+    to npy files containing vectors of 
+    indices within each sample/item
+    from the dataset.
+    """
     train: pathlib.Path
     val: pathlib.Path
     test: pathlib.Path
@@ -50,7 +58,7 @@ class IndsInSampleVectorPaths:
 @define
 class SplitsMetadata:
     """Class that represents metadata about dataset splits
-    in the BioSoundSegBench dataset, loaded from a json file"""
+    in the CMACBench dataset, loaded from a json file"""
 
     splits_csv_path: pathlib.Path
     sample_id_vector_paths: SampleIDVectorPaths
@@ -111,7 +119,7 @@ def from_paths(cls, json_path, dataset_path):
 class TrainingReplicateMetadata:
     """Class representing metadata for a
     pre-defined training replicate
-    in the BioSoundSegBench dataset.
+    in the CMACBench dataset.
     """
 
     biosound_group: str
@@ -124,7 +132,7 @@ class TrainingReplicateMetadata:
 
 
 def metadata_from_splits_json_path(
-    splits_json_path: pathlib.Path, datset_path: pathlib.Path
+    splits_json_path: pathlib.Path
 ) -> TrainingReplicateMetadata:
     name = splits_json_path.name
     try:
@@ -184,7 +192,7 @@ def metadata_from_splits_json_path(
 
 class TrainItemTransform:
     """Default transform used when training frame classification models
-    with :class:`BioSoundSegBench` dataset."""
+    with :class:`CMACBench` dataset."""
 
     def __init__(
         self,
@@ -247,7 +255,7 @@ def __call__(
 
 class InferItemTransform:
     """Default transform used when running inference on classification models
-    with :class:`BioSoundSegBench` dataset, for evaluation or to generate new predictions.
+    with :class:`CMACBench` dataset, for evaluation or to generate new predictions.
 
     Returned item includes frames reshaped into a stack of windows,
     with padded added to make reshaping possible.
@@ -369,8 +377,11 @@ def __call__(
         return item
 
 
-class BioSoundSegBench:
-    """Class representing BioSoundSegBench dataset."""
+class CMACBench:
+    """Class representing CMACBench dataset.
+    
+    https://github.com/vocalpy/CMACBench
+    """
 
     def __init__(
         self,
@@ -388,7 +399,7 @@ def __init__(
         return_frames_path: bool = False,
         item_transform: Callable | None = None,
     ):
-        """BioSoundSegBench dataset."""
+        """CMACBench dataset."""
         # ---- validate args, roughly in order
         dataset_path = pathlib.Path(dataset_path)
         if not dataset_path.exists() or not dataset_path.is_dir():
@@ -459,7 +470,7 @@ def __init__(
         # this is a bit convoluted: we are setting metadata, to set frame dur,
         # to be able to compute duration in property below
         self.training_replicate_metadata = metadata_from_splits_json_path(
-            self.splits_path, self.dataset_path
+            self.splits_path
         )
         self.frame_dur = (
             self.training_replicate_metadata.frame_dur * 1e-3

src/vak/predict/frame_classification.py

@@ -200,7 +200,7 @@ def predict_with_frame_classification_model(
         # we need "target_type" below when converting predictions to annotations,
         # but fail early here if we don't have it
         if "target_type" not in dataset_config["params"]:
-            from ..datasets.biosoundsegbench import VALID_TARGET_TYPES
+            from ..datasets.cmacbench import VALID_TARGET_TYPES
 
             raise ValueError(
                 "The dataset table in the configuration file requires a 'target_type' "

tests/test_datasets/conftest.py

@@ -340,7 +340,7 @@
 
 
 @pytest.fixture
-def mock_biosoundsegbench_dataset(tmp_path):
+def mock_cmacbench_dataset(tmp_path):
     dataset_path = tmp_path / "BioSoundSegBench"
     dataset_path.mkdir()
     splits_dir = dataset_path / "splits"

…s/test_datasets/test_biosoundsegbench.py tests/test_datasets/test_cmacbench.pytests/test_datasets/test_biosoundsegbench.py renamed to tests/test_datasets/test_cmacbench.py tests/test_datasets/test_biosoundsegbench.py renamed to tests/test_datasets/test_cmacbench.py

@@ -4,7 +4,7 @@
 import vak.datasets
 
 
-class TestBioSoundSegBench:
+class TestCMACBench:
     @pytest.mark.parametrize(
             'split, window_size, target_type',
             [
@@ -30,16 +30,16 @@ class TestBioSoundSegBench:
                 ),
             ]
     )
-    def test_init(self, split, window_size, target_type, mock_biosoundsegbench_dataset):
-        dataset_path, splits_path = mock_biosoundsegbench_dataset
-        dataset = vak.datasets.BioSoundSegBench(
+    def test_init(self, split, window_size, target_type, mock_cmacbench_dataset):
+        dataset_path, splits_path = mock_cmacbench_dataset
+        dataset = vak.datasets.CMACBench(
             dataset_path,
             splits_path,
             split,
             window_size,
             target_type,
         )
-        assert isinstance(dataset, vak.datasets.BioSoundSegBench)
+        assert isinstance(dataset, vak.datasets.CMACBench)
 
     @pytest.mark.parametrize(
             'dataset_path, splits_path, split, window_size, target_type, expected_exception',
@@ -110,17 +110,17 @@ def test_init(self, split, window_size, target_type, mock_biosoundsegbench_datas
             ]
     )
     def test_init_raises(
-            self, dataset_path, splits_path, split, window_size, target_type, expected_exception, mock_biosoundsegbench_dataset
+            self, dataset_path, splits_path, split, window_size, target_type, expected_exception, mock_cmacbench_dataset
         ):
         if dataset_path is None and splits_path is not None:
-            dataset_path, _ = mock_biosoundsegbench_dataset
+            dataset_path, _ = mock_cmacbench_dataset
         elif dataset_path is not None and splits_path is None:
-            _, splits_path = mock_biosoundsegbench_dataset
+            _, splits_path = mock_cmacbench_dataset
         elif dataset_path is None and splits_path is None:
-            dataset_path, splits_path = mock_biosoundsegbench_dataset
+            dataset_path, splits_path = mock_cmacbench_dataset
 
         with pytest.raises(expected_exception):
-            dataset = vak.datasets.BioSoundSegBench(
+            dataset = vak.datasets.CMACBench(
                 dataset_path,
                 splits_path,
                 split,

tests/test_datasets/test_get.py

@@ -10,7 +10,7 @@
         [
             (
                 {
-                    'name': 'BioSoundSegBench',
+                    'name': 'CMACBench',
                     'params': {
                         'window_size': 2000,
                         'target_type': 'multi_frame_labels',
@@ -20,7 +20,7 @@
             ),
             (
                 {
-                    'name': 'BioSoundSegBench',
+                    'name': 'CMACBench',
                     'params': {
                         'window_size': 2000,
                         'target_type': 'binary_frame_labels',
@@ -30,7 +30,7 @@
             ),
             (
                 {
-                    'name': 'BioSoundSegBench',
+                    'name': 'CMACBench',
                     'params': {
                         'window_size': 2000,
                         'target_type': 'boundary_frame_labels',
@@ -40,7 +40,7 @@
             ),
             (
                 {
-                    'name': 'BioSoundSegBench',
+                    'name': 'CMACBench',
                     'params': {
                         'window_size': 2000,
                         'target_type': 'boundary_frame_labels',
@@ -50,7 +50,7 @@
             ),
             (
                 {
-                    'name': 'BioSoundSegBench',
+                    'name': 'CMACBench',
                     'params': {
                         'window_size': 2000,
                         'target_type': 'boundary_frame_labels',
@@ -60,7 +60,7 @@
             ),
             (
                 {
-                    'name': 'BioSoundSegBench',
+                    'name': 'CMACBench',
                     'params': {
                         'window_size': 2000,
                         'target_type': 'boundary_frame_labels',
@@ -71,14 +71,14 @@
 
         ]
 )
-def test_get_biosoundsegbench(dataset_config, split, mock_biosoundsegbench_dataset):
-    dataset_path, splits_path = mock_biosoundsegbench_dataset
+def test_get_cmacbench(dataset_config, split, mock_cmacbench_dataset):
+    dataset_path, splits_path = mock_cmacbench_dataset
     dataset_config["path"] = dataset_path
     dataset_config["splits_path"] = splits_path
 
     dataset = vak.datasets.get(dataset_config, split)
 
-    assert isinstance(dataset, vak.datasets.BioSoundSegBench)
+    assert isinstance(dataset, vak.datasets.CMACBench)
     assert dataset.dataset_path == pathlib.Path(dataset_config["path"])
     assert dataset.splits_path == pathlib.Path(dataset_config["splits_path"])
     assert dataset.window_size == dataset_config["params"]["window_size"]