fixed code for eicu dataset and added tests for it.

test-resources/core/eicudemo/SHA256SUMS.txt

@@ -0,0 +1,33 @@
+bcdfcd4f31c790e30fb645dc36414907172405897adff329ad379eed85d12017 LICENSE.txt
+607020180f1c7fbbc994a23552d6a4fdc70b86d4abc8e1b36f3ee28e96f15c42 admissionDx.csv.gz
+fcf0af0de0efe18ec932199d2fcb308143886df6d50daea869bea48464ea6d68 admissiondrug.csv.gz
+be890a4cfaf7a194596583f882f491ced515f1714e3d07fc59413a83af5dad31 allergy.csv.gz
+2dc6d63f9807c7e663ffa253998691605d527b702ac1db8cc5dda92308061e78 apacheApsVar.csv.gz
+ce799692737f7a712c59aa8fbb9a302cee7c719887f0638caed3ae144b166821 apachePatientResult.csv.gz
+99d5a4ab2c5cd07b6ee0a540c86a576e5975ed3b93df8525e3d287d5babf3eae apachePredVar.csv.gz
+af0dff697752a0d98200abdbf054a935a577319b26969a27a5287b18790d633e carePlanCareProvider.csv.gz
+dfb5eafc8cb6d0c96f6ac289dd49998939df353d919ca0451d74542fd13a2f7f carePlanEOL.csv.gz
+f095258743a05717dda031b9a4f00061cffe82e06ae793c0e51f2890e1dcd44e carePlanGeneral.csv.gz
+b56d3443683b1b6a29d64da1ffd1de2662aac87b1f49560af720c1bd1325558e carePlanGoal.csv.gz
+413ff454d00a3a5b295f4bbd1d8a697a5657c5ac828cbe757a43cd5e664082a8 carePlanInfectiousDisease.csv.gz
+58e32f4c0620b22323a46336f9e406d90ce128b4005349b8e40f181f41672617 customLab.csv.gz
+c99770c25fdeaee1eb34e490f982bd17459bc8902f951a2ae23adef65e7b1b52 diagnosis.csv.gz
+4ed356f9c4e7c8ce80e89241697e94b77f21c2f325cd6ad126794b1e92805ef4 hospital.csv.gz
+3b9f024d28682884a820763ac516ae68b247ebf17903fee4f38240c04886d527 infusiondrug.csv.gz
+45db8ee0e64a4e3745284bde5828f740175770c45d00ea0c27daff789d18106e intakeOutput.csv.gz
+e06852850abeb65211bc22ac43f17121ba41b14656fac862396449af84d8273a lab.csv.gz
+480a251570a27932110452f66f248061f712c306663c274f78f90875648cb720 medication.csv.gz
+0250a23d70033f27e1954ced50935cefedaaf3fddd5a5cf5bda2d5a6927eb94d microLab.csv.gz
+ec67b0cd294a555dc39849a15eea2760e4f8696abbb293094b2003662c997ffa note.csv.gz
+1664d3a7a0f9f535a15e2d032c270cfd64bab2d5a08b6bec1b31133893934d09 nurseAssessment.csv.gz
+3cfde5dc14cc324e68c9af0807c5572e71fd7e167c5315e6b68cc241e0ff8d47 nurseCare.csv.gz
+475000d63d4ef883b2e891d7197997201300bedc13fcf23e67add1abee11dafe nurseCharting.csv.gz
+3af6e3101c037347365d81daf73ea487ba12e14a6a10f0b2ed53beed2e30c7d5 pastHistory.csv.gz
+15e1eb52169828e3cfff3b67132a087298de4d0365bbb31bb294f6043bffd474 patient.csv.gz
+7c11cae91ce2828964b9212bb49e601a0e48911ef7df5480b547621845c9d73f physicalExam.csv.gz
+9e260512866498995b49d0f0676795b62115fc1045de3f7daf2e356769259de7 respiratoryCare.csv.gz
+21bfe647fea0a229d09ff04c775ec4270d20ae084131a987943da836df534235 respiratoryCharting.csv.gz
+0c9beffc86dcdd7c16b02268266a85510d122c4cbca69915245478a567840215 sqlite/eicu_v2_0_1.sqlite3.gz
+daefee030fa1ad29593dce880ad86bdfbdbb1e33fcb3e78fa21d71cb01a08628 treatment.csv.gz
+c294f46647990cb648b3ebfa589e83f330bdc0c2c922b5e8f23aa0ef99954635 vitalAperiodic.csv.gz
+04dc093a2e77cd4c52ac7f2a47a8356fbb7456c7f15deb20d5abb1ca4303231d vitalPeriodic.csv.gz

tests/core/test_eicu.py

@@ -0,0 +1,102 @@
+import unittest
+import os
+from pathlib import Path
+
+from pyhealth.datasets import eICUDataset
+
+
+class TesteICUDemo(unittest.TestCase):
+    """Test eICU dataset with demo data from local test resources."""
+
+    def setUp(self):
+        """Set up demo dataset path for each test."""
+        self._setup_dataset_path()
+        self._load_dataset()
+
+    def _setup_dataset_path(self):
+        """Get path to local eICU demo dataset in test resources."""
+        # Get the path to the test-resources/core/mimic3demo directory
+        test_dir = Path(__file__).parent.parent.parent
+        self.demo_dataset_path = str(test_dir / "test-resources" / "core" / "eicudemo")
+
+        print(f"\n{'='*60}")
+        print(f"Setting up eICU demo dataset")
+        print(f"Dataset path: {self.demo_dataset_path}")
+
+        # Verify the dataset exists
+        if not os.path.exists(self.demo_dataset_path):
+            raise unittest.SkipTest(
+                f"eICU demo dataset not found at {self.demo_dataset_path}"
+            )
+
+        # List files in the dataset directory
+        files = os.listdir(self.demo_dataset_path)
+        print(f"Found {len(files)} files in dataset directory:")
+        for f in sorted(files):
+            file_path = os.path.join(self.demo_dataset_path, f)
+            size = os.path.getsize(file_path) / 1024  # 
+            print(f"  - {f} ({size:.1f} KB)")
+        print(f"{'='*60}\n")
+
+    def _load_dataset(self):
+        """Load the dataset for testing."""
+        tables = ["hospital", "admissiondx", "diagnosis", "medication", "lab", "treatment", "physicalExam", "admissiondrug", "allergy", "apacheapsvar", 
+                  "apachepatientresult", "apachepredvar", "careplancareprovider", "careplaneol", "careplangeneral","careplangoal",
+                  "careplaninfectiousdisease","customlab","infusiondrug", "intakeoutput","microlab","note",
+                  "nurseassessment","nursecare","nursecharting", "pasthistory",
+                  "respiratorycare", "respiratorycharting",
+                  "vitalPeriodic","vitalAperiodic"]
+        print(f"Loading eICUDataset with tables: {tables}")
+        self.dataset = eICUDataset(root=self.demo_dataset_path, tables=tables)
+        print(f"✓ Dataset loaded successfully")
+        print(self.dataset)
+        #print(f"  Total patients: {len(self.dataset.patient)}")
+        print()
+
+    def test_stats(self):
+        """Test .stats() method execution."""
+        print(f"\n{'='*60}")
+        print("TEST: test_stats()")
+        print(f"{'='*60}")
+        try:
+            print("Calling dataset.stats()...")
+            self.dataset.stats()
+            print("✓ dataset.stats() executed successfully")
+        except Exception as e:
+            print(f"✗ dataset.stats() failed with error: {e}")
+            self.fail(f"dataset.stats() failed: {e}")
+
+    def test_get_events(self):
+        """Test get_patient and get_events methods with patient 141764."""
+        print(f"\n{'='*60}")
+        print("TEST: test_get_events()")
+        print(f"{'='*60}")
+
+        # Test get_patient method
+        print("Getting patient 141764...")
+        patient = self.dataset.get_patient("141764")
+        self.assertIsNotNone(patient, msg="Patient 141764 should exist in demo dataset")
+        print(f"✓ Patient 141764 found: {patient}")
+
+        # Test get_events method
+        print("Getting events for patient 141764...")
+        events = patient.get_events()
+        self.assertIsNotNone(events, msg="get_events() should not return None")
+        self.assertIsInstance(events, list, msg="get_events() should return a list")
+        self.assertGreater(
+            len(events), 0, msg="get_events() should not return an empty list"
+         )
+        print(f"✓ Retrieved {len(events)} events")
+      #  print(f"  Event types: {set(e[0] for e in events)}")
+
+        # Show sample events
+        print(f"\nSample events (first 3):")
+        for i, event in enumerate(events[:3]):
+            print(f"  {i+1}. Type: {event['event_type']}, Time: {event['timestamp']}")
+#            print(f"  {i+1}. Type: {event[0]}, Time: {event[1]}, Data: {event[2]}")
+
+        print(f"✓ test_get_events() passed successfully")
+
+
+if __name__ == "__main__":
+    unittest.main()