feat: add PTBXLDataset and PTBXLMultilabelClassification task

.gitignore

@@ -139,4 +139,13 @@ data/physionet.org/
 .vscode/
 
 # Model weight files (large binaries, distributed separately)
-weightfiles/
+weightfiles/
+# CS-598 project data (downloaded separately — not library code)
+cs598_project/ptbxl_database.csv
+cs598_project/scp_statements.csv
+cs598_project/ptbxl-records-pyhealth.csv
+cs598_project/ptbxl_v103_pyhealth.yaml
+cs598_project/WFDB
+cs598_project/output/
+cs598_project/**/*.pkl
+cs598_project/**/*.ckpt

docs/api/tasks.rst

@@ -229,3 +229,4 @@ Available Tasks
     Mutation Pathogenicity (COSMIC) <tasks/pyhealth.tasks.MutationPathogenicityPrediction>
     Cancer Survival Prediction (TCGA) <tasks/pyhealth.tasks.CancerSurvivalPrediction>
     Cancer Mutation Burden (TCGA) <tasks/pyhealth.tasks.CancerMutationBurden>
+    PTB-XL Multi-Label ECG Classification <tasks/pyhealth.tasks.PTBXLMultilabelClassification>

docs/api/tasks/pyhealth.tasks.PTBXLMultilabelClassification.rst

@@ -0,0 +1,18 @@
+pyhealth.tasks.PTBXLMultilabelClassification
+============================================
+
+PTB-XL is a large publicly available 12-lead ECG dataset annotated with SNOMED-CT codes.
+This task turns a :class:`~pyhealth.datasets.PTBXLDataset` into a **multi-label classification** problem.
+
+Two label spaces are supported via the ``label_type`` argument:
+
+- ``"superdiagnostic"`` — 5 coarse diagnostic classes (NORM, MI, STTC, CD, HYP)
+- ``"diagnostic"`` — 27 SNOMED-CT classes scored in the PhysioNet / CinC Challenge 2020
+
+The ``sampling_rate`` argument (100 or 500 Hz) controls temporal resolution, enabling
+an ablation study across both axes.
+
+.. autoclass:: pyhealth.tasks.PTBXLMultilabelClassification
+    :members:
+    :undoc-members:
+    :show-inheritance: