Moved padchest configs to data

pyproject.toml

@@ -476,11 +476,11 @@ nih_cxr14_cm = "ptbench.configs.datasets.nih_cxr14_re.cardiomegaly"
 # NIH CXR14 / PadChest aggregated dataset
 nih_cxr14_pc_idiap = "ptbench.configs.datasets.nih_cxr14_re_pc.idiap"
 # PadChest
-padchest_idiap = "ptbench.configs.datasets.padchest.idiap"
-padchest_tb_idiap = "ptbench.configs.datasets.padchest.tb_idiap"
-padchest_no_tb_idiap = "ptbench.configs.datasets.padchest.no_tb_idiap"
-padchest_tb_idiap_rgb = "ptbench.configs.datasets.padchest.tb_idiap_rgb"
-padchest_cm_idiap = "ptbench.configs.datasets.padchest.cardiomegaly_idiap"
+padchest_idiap = "ptbench.data.padchest.idiap"
+padchest_tb_idiap = "ptbench.data.padchest.tb_idiap"
+padchest_no_tb_idiap = "ptbench.data.padchest.no_tb_idiap"
+padchest_tb_idiap_rgb = "ptbench.data.padchest.tb_idiap_rgb"
+padchest_cm_idiap = "ptbench.data.padchest.cardiomegaly_idiap"
 # extended PadChestTB dataset (with radiological signs)
 padchest_tb_idiap_rs = "ptbench.configs.datasets.padchest_RS.tb_idiap"
 

src/ptbench/configs/datasets/padchest/__init__.py

-# SPDX-FileCopyrightText: Copyright © 2023 Idiap Research Institute <contact@idiap.ch>
-#
-# SPDX-License-Identifier: GPL-3.0-or-later
-
-
-def _maker(protocol, resize_size=512, cc_size=512, RGB=True):
-    import torchvision.transforms as transforms
-
-    from ....data.padchest import dataset as raw
-    from ....data.transforms import SingleAutoLevel16to8
-    from .. import make_dataset as mk
-
-    post_transforms = []
-    if not RGB:
-        post_transforms = [transforms.Lambda(lambda x: x.convert("L"))]
-
-    return mk(
-        [raw.subsets(protocol)],
-        [
-            SingleAutoLevel16to8(),
-            transforms.Resize(resize_size),
-            transforms.CenterCrop(cc_size),
-        ],
-        [transforms.RandomHorizontalFlip()],
-        post_transforms,
-    )

src/ptbench/configs/datasets/padchest/idiap.py

-# SPDX-FileCopyrightText: Copyright © 2023 Idiap Research Institute <contact@idiap.ch>
-#
-# SPDX-License-Identifier: GPL-3.0-or-later
-
-"""Padchest (idiap protocol) dataset for computer-aided diagnosis.
-
-* See :py:mod:`ptbench.data.padchest` for dataset details
-* This configuration resolution: 512 x 512 (default)
-"""
-
-from . import _maker
-
-dataset = _maker("idiap")

src/ptbench/data/padchest/__init__.py

@@ -226,6 +226,7 @@ import importlib.resources
 import os
 
 from ...utils.rc import load_rc
+from .. import make_dataset
 from ..dataset import JSONDataset
 from ..loader import load_pil, make_delayed
 
@@ -252,9 +253,30 @@ def _loader(context, sample):
     return make_delayed(sample, _raw_data_loader)
 
 
-dataset = JSONDataset(
+json_dataset = JSONDataset(
     protocols=_protocols,
     fieldnames=("data", "label"),
     loader=_loader,
 )
 """Padchest dataset object."""
+
+
+def _maker(protocol, resize_size=512, cc_size=512, RGB=True):
+    import torchvision.transforms as transforms
+
+    from ..transforms import SingleAutoLevel16to8
+
+    post_transforms = []
+    if not RGB:
+        post_transforms = [transforms.Lambda(lambda x: x.convert("L"))]
+
+    return make_dataset(
+        [json_dataset.subsets(protocol)],
+        [
+            SingleAutoLevel16to8(),
+            transforms.Resize(resize_size),
+            transforms.CenterCrop(cc_size),
+        ],
+        [transforms.RandomHorizontalFlip()],
+        post_transforms,
+    )

src/ptbench/configs/datasets/padchest/cardiomegaly_idiap.py → src/ptbench/data/padchest/cardiomegaly_idiap.py

@@ -12,6 +12,38 @@ parameters: Label = "Normal", MethodLabel = "Physician", Projection = "PA"
 * This configuration resolution: 512 x 512 (default)
 """
 
+from clapper.logging import setup
+
+from .. import return_subsets
+from ..base_datamodule import BaseDataModule
 from . import _maker
 
-dataset = _maker("cardiomegaly_idiap", RGB=False)
+logger = setup(__name__.split(".")[0], format="%(levelname)s: %(message)s")
+
+
+class DefaultModule(BaseDataModule):
+    def __init__(
+        self,
+        train_batch_size=1,
+        predict_batch_size=1,
+        drop_incomplete_batch=False,
+        multiproc_kwargs=None,
+    ):
+        super().__init__(
+            train_batch_size=train_batch_size,
+            predict_batch_size=predict_batch_size,
+            drop_incomplete_batch=drop_incomplete_batch,
+            multiproc_kwargs=multiproc_kwargs,
+        )
+
+    def setup(self, stage: str):
+        self.dataset = _maker("cardiomegaly_idiap", RGB=False)
+        (
+            self.train_dataset,
+            self.validation_dataset,
+            self.extra_validation_datasets,
+            self.predict_dataset,
+        ) = return_subsets(self.dataset)
+
+
+datamodule = DefaultModule

src/ptbench/data/padchest/idiap.py

+# SPDX-FileCopyrightText: Copyright © 2023 Idiap Research Institute <contact@idiap.ch>
+#
+# SPDX-License-Identifier: GPL-3.0-or-later
+
+"""Padchest (idiap protocol) dataset for computer-aided diagnosis.
+
+* See :py:mod:`ptbench.data.padchest` for dataset details
+* This configuration resolution: 512 x 512 (default)
+"""
+
+from clapper.logging import setup
+
+from .. import return_subsets
+from ..base_datamodule import BaseDataModule
+from . import _maker
+
+logger = setup(__name__.split(".")[0], format="%(levelname)s: %(message)s")
+
+
+class DefaultModule(BaseDataModule):
+    def __init__(
+        self,
+        train_batch_size=1,
+        predict_batch_size=1,
+        drop_incomplete_batch=False,
+        multiproc_kwargs=None,
+    ):
+        super().__init__(
+            train_batch_size=train_batch_size,
+            predict_batch_size=predict_batch_size,
+            drop_incomplete_batch=drop_incomplete_batch,
+            multiproc_kwargs=multiproc_kwargs,
+        )
+
+    def setup(self, stage: str):
+        self.dataset = _maker("idiap")
+        (
+            self.train_dataset,
+            self.validation_dataset,
+            self.extra_validation_datasets,
+            self.predict_dataset,
+        ) = return_subsets(self.dataset)
+
+
+datamodule = DefaultModule

src/ptbench/configs/datasets/padchest/no_tb_idiap.py → src/ptbench/data/padchest/no_tb_idiap.py

@@ -19,6 +19,38 @@ diagnosis.
   fibrosis, edema and consolidation
 """
 
+from clapper.logging import setup
+
+from .. import return_subsets
+from ..base_datamodule import BaseDataModule
 from . import _maker
 
-dataset = _maker("no_tb_idiap")
+logger = setup(__name__.split(".")[0], format="%(levelname)s: %(message)s")
+
+
+class DefaultModule(BaseDataModule):
+    def __init__(
+        self,
+        train_batch_size=1,
+        predict_batch_size=1,
+        drop_incomplete_batch=False,
+        multiproc_kwargs=None,
+    ):
+        super().__init__(
+            train_batch_size=train_batch_size,
+            predict_batch_size=predict_batch_size,
+            drop_incomplete_batch=drop_incomplete_batch,
+            multiproc_kwargs=multiproc_kwargs,
+        )
+
+    def setup(self, stage: str):
+        self.dataset = _maker("no_tb_idiap")
+        (
+            self.train_dataset,
+            self.validation_dataset,
+            self.extra_validation_datasets,
+            self.predict_dataset,
+        ) = return_subsets(self.dataset)
+
+
+datamodule = DefaultModule

src/ptbench/configs/datasets/padchest/tb_idiap.py → src/ptbench/data/padchest/tb_idiap.py

@@ -12,6 +12,38 @@ parameters: Label = "Normal", MethodLabel = "Physician", Projection = "PA"
 * This configuration resolution: 512 x 512 (default)
 """
 
+from clapper.logging import setup
+
+from .. import return_subsets
+from ..base_datamodule import BaseDataModule
 from . import _maker
 
-dataset = _maker("tb_idiap", RGB=False)
+logger = setup(__name__.split(".")[0], format="%(levelname)s: %(message)s")
+
+
+class DefaultModule(BaseDataModule):
+    def __init__(
+        self,
+        train_batch_size=1,
+        predict_batch_size=1,
+        drop_incomplete_batch=False,
+        multiproc_kwargs=None,
+    ):
+        super().__init__(
+            train_batch_size=train_batch_size,
+            predict_batch_size=predict_batch_size,
+            drop_incomplete_batch=drop_incomplete_batch,
+            multiproc_kwargs=multiproc_kwargs,
+        )
+
+    def setup(self, stage: str):
+        self.dataset = _maker("tb_idiap", RGB=False)
+        (
+            self.train_dataset,
+            self.validation_dataset,
+            self.extra_validation_datasets,
+            self.predict_dataset,
+        ) = return_subsets(self.dataset)
+
+
+datamodule = DefaultModule

src/ptbench/configs/datasets/padchest/tb_idiap_rgb.py → src/ptbench/data/padchest/tb_idiap_rgb.py

@@ -13,6 +13,38 @@ parameters: Label = "Normal", MethodLabel = "Physician", Projection = "PA"
 * This configuration resolution: 224 x 224 (default)
 """
 
+from clapper.logging import setup
+
+from .. import return_subsets
+from ..base_datamodule import BaseDataModule
 from . import _maker
 
-dataset = _maker("tb_idiap", resize_size=256, cc_size=224)
+logger = setup(__name__.split(".")[0], format="%(levelname)s: %(message)s")
+
+
+class DefaultModule(BaseDataModule):
+    def __init__(
+        self,
+        train_batch_size=1,
+        predict_batch_size=1,
+        drop_incomplete_batch=False,
+        multiproc_kwargs=None,
+    ):
+        super().__init__(
+            train_batch_size=train_batch_size,
+            predict_batch_size=predict_batch_size,
+            drop_incomplete_batch=drop_incomplete_batch,
+            multiproc_kwargs=multiproc_kwargs,
+        )
+
+    def setup(self, stage: str):
+        self.dataset = _maker("tb_idiap", resize_size=256, cc_size=224)
+        (
+            self.train_dataset,
+            self.validation_dataset,
+            self.extra_validation_datasets,
+            self.predict_dataset,
+        ) = return_subsets(self.dataset)
+
+
+datamodule = DefaultModule