feat: add detect-anomalies CLI command for time series anomaly detection (closes #292)

Author: HeerakKashyap

igel/anomaly_detection.py

@@ -0,0 +1,59 @@
+"""
+Time Series Anomaly Detection Utilities
+
+- Isolation Forest for anomaly detection
+- Statistical methods for outlier detection
+- Time series specific anomaly detection
+"""
+
+import numpy as np
+from sklearn.ensemble import IsolationForest
+from sklearn.preprocessing import StandardScaler
+
+def detect_anomalies_isolation_forest(data, contamination=0.1):
+    """
+    Detect anomalies using Isolation Forest.
+    
+    Args:
+        data: Time series data (1D array)
+        contamination: Expected proportion of anomalies
+    
+    Returns:
+        Dictionary with anomaly scores and predictions
+    """
+    # Reshape data for sklearn
+    X = data.reshape(-1, 1)
+    
+    # Fit isolation forest
+    iso_forest = IsolationForest(contamination=contamination, random_state=42)
+    predictions = iso_forest.fit_predict(X)
+    scores = iso_forest.score_samples(X)
+    
+    return {
+        "predictions": predictions,  # -1 for anomalies, 1 for normal
+        "scores": scores,
+        "anomaly_indices": np.where(predictions == -1)[0]
+    }
+
+def detect_statistical_anomalies(data, threshold=3):
+    """
+    Detect anomalies using statistical methods (Z-score).
+    
+    Args:
+        data: Time series data
+        threshold: Number of standard deviations for anomaly detection
+    
+    Returns:
+        Dictionary with anomaly information
+    """
+    mean = np.mean(data)
+    std = np.std(data)
+    z_scores = np.abs((data - mean) / std)
+    
+    anomalies = z_scores > threshold
+    
+    return {
+        "anomaly_indices": np.where(anomalies)[0],
+        "z_scores": z_scores,
+        "threshold": threshold
+    }