Merge branch 'issue-62' into 'main'

Fix db path issues when reusing DataLoaders from other Datamodules

Closes #62

See merge request biosignal/software/mednet!22

src/mednet/config/data/montgomery_shenzhen_indian/datamodule.py

@@ -4,6 +4,7 @@
 """Aggregated DataModule composed of Montgomery, Shenzhen and Indian datasets."""
 
 from ....data.datamodule import ConcatDataModule
+from ..indian.datamodule import CONFIGURATION_KEY_DATADIR as INDIAN_KEY_DATADIR
 from ..indian.datamodule import RawDataLoader as IndianLoader
 from ..indian.datamodule import make_split as make_indian_split
 from ..montgomery.datamodule import RawDataLoader as MontgomeryLoader
@@ -26,7 +27,7 @@ class DataModule(ConcatDataModule):
         montgomery_split = make_montgomery_split(split_filename)
         shenzhen_loader = ShenzhenLoader()
         shenzhen_split = make_shenzhen_split(split_filename)
-        indian_loader = IndianLoader()
+        indian_loader = IndianLoader(INDIAN_KEY_DATADIR)
         indian_split = make_indian_split(split_filename)
 
         super().__init__(

src/mednet/config/data/montgomery_shenzhen_indian_padchest/datamodule.py

@@ -4,6 +4,7 @@
 """Aggregated DataModule composed of Montgomery, Shenzhen, Indian, and PadChest datasets."""
 
 from ....data.datamodule import ConcatDataModule
+from ..indian.datamodule import CONFIGURATION_KEY_DATADIR as INDIAN_KEY_DATADIR
 from ..indian.datamodule import RawDataLoader as IndianLoader
 from ..indian.datamodule import make_split as make_indian_split
 from ..montgomery.datamodule import RawDataLoader as MontgomeryLoader
@@ -31,7 +32,7 @@ class DataModule(ConcatDataModule):
         montgomery_split = make_montgomery_split(split_filename)
         shenzhen_loader = ShenzhenLoader()
         shenzhen_split = make_shenzhen_split(split_filename)
-        indian_loader = IndianLoader()
+        indian_loader = IndianLoader(INDIAN_KEY_DATADIR)
         indian_split = make_indian_split(split_filename)
         padchest_loader = PadchestLoader()
         padchest_split = make_padchest_split(padchest_split_filename)

src/mednet/config/data/montgomery_shenzhen_indian_tbx11k/datamodule.py

@@ -4,6 +4,7 @@
 """Aggregated DataModule composed of Montgomery, Shenzhen, Indian, and TBX11k datasets."""
 
 from ....data.datamodule import ConcatDataModule
+from ..indian.datamodule import CONFIGURATION_KEY_DATADIR as INDIAN_KEY_DATADIR
 from ..indian.datamodule import RawDataLoader as IndianLoader
 from ..indian.datamodule import make_split as make_indian_split
 from ..montgomery.datamodule import RawDataLoader as MontgomeryLoader
@@ -31,7 +32,7 @@ class DataModule(ConcatDataModule):
         montgomery_split = make_montgomery_split(split_filename)
         shenzhen_loader = ShenzhenLoader()
         shenzhen_split = make_shenzhen_split(split_filename)
-        indian_loader = IndianLoader()
+        indian_loader = IndianLoader(INDIAN_KEY_DATADIR)
         indian_split = make_indian_split(split_filename)
         tbx11k_loader = TBX11kLoader()
         tbx11k_split = make_tbx11k_split(tbx11k_split_filename)

tests/test_hivtb.py

@@ -7,6 +7,8 @@ import importlib
 
 import pytest
 
+from click.testing import CliRunner
+
 
 def id_function(val):
     if isinstance(val, dict):
@@ -43,6 +45,17 @@ def test_protocol_consistency(
     )
 
 
+@pytest.mark.skip_if_rc_var_not_set("datadir.hivtb")
+def test_database_check():
+    from mednet.scripts.database import check
+
+    runner = CliRunner()
+    result = runner.invoke(check, ["--limit=10", "hivtb-f0"])
+    assert (
+        result.exit_code == 0
+    ), f"Exit code {result.exit_code} != 0 -- Output:\n{result.output}"
+
+
 @pytest.mark.skip_if_rc_var_not_set("datadir.hivtb")
 @pytest.mark.parametrize(
     "dataset",

tests/test_indian.py

@@ -10,6 +10,8 @@ import importlib
 
 import pytest
 
+from click.testing import CliRunner
+
 
 def id_function(val):
     if isinstance(val, dict):
@@ -47,6 +49,17 @@ def test_protocol_consistency(
     )
 
 
+@pytest.mark.skip_if_rc_var_not_set("datadir.indian")
+def test_database_check():
+    from mednet.scripts.database import check
+
+    runner = CliRunner()
+    result = runner.invoke(check, ["indian"])
+    assert (
+        result.exit_code == 0
+    ), f"Exit code {result.exit_code} != 0 -- Output:\n{result.output}"
+
+
 @pytest.mark.skip_if_rc_var_not_set("datadir.indian")
 @pytest.mark.parametrize(
     "dataset",

tests/test_montgomery.py

@@ -7,6 +7,8 @@ import importlib
 
 import pytest
 
+from click.testing import CliRunner
+
 
 def id_function(val):
     if isinstance(val, dict):
@@ -44,6 +46,17 @@ def test_protocol_consistency(
     )
 
 
+@pytest.mark.skip_if_rc_var_not_set("datadir.montgomery")
+def test_database_check():
+    from mednet.scripts.database import check
+
+    runner = CliRunner()
+    result = runner.invoke(check, ["montgomery"])
+    assert (
+        result.exit_code == 0
+    ), f"Exit code {result.exit_code} != 0 -- Output:\n{result.output}"
+
+
 @pytest.mark.skip_if_rc_var_not_set("datadir.montgomery")
 @pytest.mark.parametrize(
     "dataset",

tests/test_montgomery_shenzhen.py

@@ -7,6 +7,8 @@ import importlib
 
 import pytest
 
+from click.testing import CliRunner
+
 
 @pytest.mark.parametrize(
     "name",
@@ -53,3 +55,15 @@ def test_split_consistency(name: str):
         assert shenzhen.splits[split][0][0] == combined.splits[split][1][0]
         assert isinstance(shenzhen.splits[split][0][1], ShenzhenLoader)
         assert isinstance(combined.splits[split][1][1], ShenzhenLoader)
+
+
+@pytest.mark.skip_if_rc_var_not_set("datadir.montgomery")
+@pytest.mark.skip_if_rc_var_not_set("datadir.shenzhen")
+def test_database_check():
+    from mednet.scripts.database import check
+
+    runner = CliRunner()
+    result = runner.invoke(check, ["montgomery-shenzhen"])
+    assert (
+        result.exit_code == 0
+    ), f"Exit code {result.exit_code} != 0 -- Output:\n{result.output}"

tests/test_montgomery_shenzhen_indian.py

@@ -7,6 +7,8 @@ import importlib
 
 import pytest
 
+from click.testing import CliRunner
+
 
 @pytest.mark.parametrize(
     "name",
@@ -65,3 +67,16 @@ def test_split_consistency(name: str):
         assert indian.splits[split][0][0] == combined.splits[split][2][0]
         assert isinstance(indian.splits[split][0][1], IndianLoader)
         assert isinstance(combined.splits[split][2][1], IndianLoader)
+
+
+@pytest.mark.skip_if_rc_var_not_set("datadir.montgomery")
+@pytest.mark.skip_if_rc_var_not_set("datadir.indian")
+@pytest.mark.skip_if_rc_var_not_set("datadir.shenzhen")
+def test_database_check():
+    from mednet.scripts.database import check
+
+    runner = CliRunner()
+    result = runner.invoke(check, ["montgomery-shenzhen-indian"])
+    assert (
+        result.exit_code == 0
+    ), f"Exit code {result.exit_code} != 0 -- Output:\n{result.output}"

tests/test_montgomery_shenzhen_indian_padchest.py

@@ -7,6 +7,8 @@ import importlib
 
 import pytest
 
+from click.testing import CliRunner
+
 
 @pytest.mark.parametrize(
     "name,padchest_name",
@@ -69,3 +71,17 @@ def test_split_consistency(name: str, padchest_name: str):
             assert padchest.splits[split][0][0] == combined.splits[split][3][0]
             assert isinstance(padchest.splits[split][0][1], PadChestLoader)
             assert isinstance(combined.splits[split][3][1], PadChestLoader)
+
+
+@pytest.mark.skip_if_rc_var_not_set("datadir.montgomery")
+@pytest.mark.skip_if_rc_var_not_set("datadir.shenzhen")
+@pytest.mark.skip_if_rc_var_not_set("datadir.indian")
+@pytest.mark.skip_if_rc_var_not_set("datadir.padchest")
+def test_database_check():
+    from mednet.scripts.database import check
+
+    runner = CliRunner()
+    result = runner.invoke(check, ["montgomery-shenzhen-indian-padchest"])
+    assert (
+        result.exit_code == 0
+    ), f"Exit code {result.exit_code} != 0 -- Output:\n{result.output}"

tests/test_montgomery_shenzhen_indian_tbx11k.py

@@ -7,6 +7,8 @@ import importlib
 
 import pytest
 
+from click.testing import CliRunner
+
 
 @pytest.mark.parametrize(
     "name,tbx11k_name",
@@ -89,3 +91,17 @@ def test_split_consistency(name: str, tbx11k_name: str):
         assert tbx11k.splits[split][0][0] == combined.splits[split][3][0]
         assert isinstance(tbx11k.splits[split][0][1], TBX11kLoader)
         assert isinstance(combined.splits[split][3][1], TBX11kLoader)
+
+
+@pytest.mark.skip_if_rc_var_not_set("datadir.montgomery")
+@pytest.mark.skip_if_rc_var_not_set("datadir.shenzhen")
+@pytest.mark.skip_if_rc_var_not_set("datadir.indian")
+@pytest.mark.skip_if_rc_var_not_set("datadir.tbx11k")
+def test_database_check():
+    from mednet.scripts.database import check
+
+    runner = CliRunner()
+    result = runner.invoke(check, ["montgomery-shenzhen-indian-tbx11k-v1"])
+    assert (
+        result.exit_code == 0
+    ), f"Exit code {result.exit_code} != 0 -- Output:\n{result.output}"

tests/test_nih_cxr14.py

@@ -7,6 +7,8 @@ import importlib
 
 import pytest
 
+from click.testing import CliRunner
+
 
 def id_function(val):
     if isinstance(val, dict):
@@ -44,6 +46,17 @@ testdata = [
 ]
 
 
+@pytest.mark.skip_if_rc_var_not_set("datadir.nih_cxr14")
+def test_database_check():
+    from mednet.scripts.database import check
+
+    runner = CliRunner()
+    result = runner.invoke(check, ["--limit=10", "nih-cxr14"])
+    assert (
+        result.exit_code == 0
+    ), f"Exit code {result.exit_code} != 0 -- Output:\n{result.output}"
+
+
 @pytest.mark.skip_if_rc_var_not_set("datadir.nih_cxr14")
 @pytest.mark.parametrize("name,dataset,num_labels", testdata)
 def test_loading(database_checkers, name: str, dataset: str, num_labels: int):

tests/test_nih_cxr14_padchest.py

@@ -7,6 +7,8 @@ import importlib
 
 import pytest
 
+from click.testing import CliRunner
+
 
 @pytest.mark.parametrize(
     "name,padchest_name,combined_name",
@@ -45,3 +47,15 @@ def test_split_consistency(name: str, padchest_name: str, combined_name: str):
             assert padchest.splits[split][0][0] == combined.splits[split][1][0]
             assert isinstance(padchest.splits[split][0][1], PadChestLoader)
             assert isinstance(combined.splits[split][1][1], PadChestLoader)
+
+
+@pytest.mark.skip_if_rc_var_not_set("datadir.nih_cxr14")
+@pytest.mark.skip_if_rc_var_not_set("datadir.padchest")
+def test_database_check():
+    from mednet.scripts.database import check
+
+    runner = CliRunner()
+    result = runner.invoke(check, ["--limit=10", "nih-cxr14-padchest"])
+    assert (
+        result.exit_code == 0
+    ), f"Exit code {result.exit_code} != 0 -- Output:\n{result.output}"

tests/test_padchest.py

@@ -7,6 +7,8 @@ import importlib
 
 import pytest
 
+from click.testing import CliRunner
+
 
 def id_function(val):
     if isinstance(val, dict):
@@ -40,6 +42,17 @@ def test_protocol_consistency(
     )
 
 
+@pytest.mark.skip_if_rc_var_not_set("datadir.padchest")
+def test_database_check():
+    from mednet.scripts.database import check
+
+    runner = CliRunner()
+    result = runner.invoke(check, ["--limit=10", "padchest-idiap"])
+    assert (
+        result.exit_code == 0
+    ), f"Exit code {result.exit_code} != 0 -- Output:\n{result.output}"
+
+
 testdata = [
     ("idiap", "train", 193),
     ("idiap", "test", 1),

tests/test_shenzhen.py

@@ -7,6 +7,8 @@ import importlib
 
 import pytest
 
+from click.testing import CliRunner
+
 
 def id_function(val):
     if isinstance(val, dict):
@@ -44,6 +46,17 @@ def test_protocol_consistency(
     )
 
 
+@pytest.mark.skip_if_rc_var_not_set("datadir.shenzhen")
+def test_database_check():
+    from mednet.scripts.database import check
+
+    runner = CliRunner()
+    result = runner.invoke(check, ["shenzhen"])
+    assert (
+        result.exit_code == 0
+    ), f"Exit code {result.exit_code} != 0 -- Output:\n{result.output}"
+
+
 @pytest.mark.skip_if_rc_var_not_set("datadir.shenzhen")
 @pytest.mark.parametrize(
     "dataset",

tests/test_tbpoc.py

@@ -7,6 +7,8 @@ import importlib
 
 import pytest
 
+from click.testing import CliRunner
+
 
 def id_function(val):
     if isinstance(val, dict):
@@ -46,6 +48,17 @@ def test_protocol_consistency(
     )
 
 
+@pytest.mark.skip_if_rc_var_not_set("datadir.tbpoc")
+def test_database_check():
+    from mednet.scripts.database import check
+
+    runner = CliRunner()
+    result = runner.invoke(check, ["tbpoc-f0"])
+    assert (
+        result.exit_code == 0
+    ), f"Exit code {result.exit_code} != 0 -- Output:\n{result.output}"
+
+
 @pytest.mark.skip_if_rc_var_not_set("datadir.tbpoc")
 @pytest.mark.parametrize(
     "dataset",

tests/test_tbx11k.py

@@ -9,6 +9,8 @@ import typing
 import pytest
 import torch
 
+from click.testing import CliRunner
+
 
 def id_function(val):
     if isinstance(val, (dict, tuple)):
@@ -231,6 +233,22 @@ def check_loaded_batch(
     # __import__("pdb").set_trace()
 
 
+@pytest.mark.skip_if_rc_var_not_set("datadir.tbx11k")
+def test_database_check():
+    from mednet.scripts.database import check
+
+    runner = CliRunner()
+    result = runner.invoke(check, ["--limit=10", "tbx11k-v1-f0"])
+    assert (
+        result.exit_code == 0
+    ), f"Exit code {result.exit_code} != 0 -- Output:\n{result.output}"
+
+    result = runner.invoke(check, ["--limit=10", "tbx11k-v2-f0"])
+    assert (
+        result.exit_code == 0
+    ), f"Exit code {result.exit_code} != 0 -- Output:\n{result.output}"
+
+
 @pytest.mark.skip_if_rc_var_not_set("datadir.tbx11k")
 @pytest.mark.parametrize(
     "dataset",