Add data_analysis/tests/test_core.py

data_analysis/tests/test_core.py

@@ -0,0 +1,58 @@
+import pytest
+import pandas as pd
+import numpy as np
+from data_analysis.core import analyze_data, AnalysisResult
+from data_analysis.io_utils import LogEntry
+
+
+def _make_log_entries():
+    data = [
+        {"t": "2024-03-01T00:00:00Z", "Lx": 100, "dB": 40.5, "Temp": 20.0, "Inference": 0.2},
+        {"t": "2024-03-01T00:01:00Z", "Lx": 102, "dB": 42.0, "Temp": 20.1, "Inference": 0.25},
+        {"t": "2024-03-01T00:02:00Z", "Lx": 105, "dB": 80.0, "Temp": 20.2, "Inference": 0.9},  # anomaly
+        {"t": "2024-03-01T00:03:00Z", "Lx": 99,  "dB": 41.0, "Temp": 20.0, "Inference": 0.3},
+    ]
+    return [LogEntry(d["t"], d["Lx"], d["dB"], d["Temp"], d["Inference"]) for d in data]
+
+
+def test_analyze_data_returns_analysis_result():
+    logs = _make_log_entries()
+    result = analyze_data(logs)
+    assert isinstance(result, AnalysisResult)
+    assert hasattr(result, 'significant_patterns')
+    assert hasattr(result, 'anomaly_events')
+    assert isinstance(result.significant_patterns, list)
+    assert isinstance(result.anomaly_events, list)
+
+
+def test_analyze_data_detects_anomalies():
+    logs = _make_log_entries()
+    result = analyze_data(logs)
+    # At least one anomaly expected
+    assert any('anomaly' in str(ev).lower() or '80' in str(ev) for ev in result.anomaly_events)
+
+
+def test_analyze_data_pattern_length_consistency():
+    logs = _make_log_entries()
+    result1 = analyze_data(logs)
+    np.random.shuffle(logs)
+    result2 = analyze_data(logs)
+    assert len(result1.significant_patterns) == len(result2.significant_patterns)
+
+
+def test_analyze_data_empty_input():
+    result = analyze_data([])
+    assert isinstance(result, AnalysisResult)
+    assert result.significant_patterns == []
+    assert result.anomaly_events == []
+
+
+def test_invalid_input_type_raises():
+    with pytest.raises((TypeError, ValueError)):
+        analyze_data([{'wrong': 'structure'}])
+
+
+def test_analysis_result_repr_str():
+    result = AnalysisResult(['pattern'], ['event'])
+    s = str(result)
+    assert 'pattern' in s or 'event' in s