Add data_analysis_script/src/data_analysis_script/core.py

data_analysis_script/src/data_analysis_script/core.py

@@ -0,0 +1,64 @@
+from typing import List, Dict, Any
+import numpy as np
+from scipy import stats
+
+
+def analyze_data(data: List[Dict[str, Any]]) -> Dict[str, float]:
+    """Analysiert Messdaten und berechnet Median-Amplitude, Ausreißerrate und statistische Signifikanz.
+
+    Args:
+        data: Liste von Messdatensätzen mit Schlüsseln 'condition_name' und 'amplitude'.
+
+    Returns:
+        Dictionary mit Feldern 'median_amplitude', 'outlier_rate', 'statistical_significance'.
+    """
+    if not data:
+        raise ValueError("data darf nicht leer sein")
+
+    amplitudes = []
+    conditions = {}
+
+    for record in data:
+        if not all(k in record for k in ("condition_name", "amplitude")):
+            raise ValueError("Jeder Datensatz muss 'condition_name' und 'amplitude' enthalten")
+        try:
+            amp = float(record["amplitude"])
+        except (TypeError, ValueError):
+            raise ValueError("Amplitude muss numerisch sein")
+        amplitudes.append(amp)
+        cond = record["condition_name"]
+        conditions.setdefault(cond, []).append(amp)
+
+    arr = np.array(amplitudes)
+    median_amplitude = float(np.median(arr))
+
+    q1, q3 = np.percentile(arr, [25, 75])
+    iqr = q3 - q1
+    lower_bound = q1 - 1.5 * iqr
+    upper_bound = q3 + 1.5 * iqr
+    outliers = ((arr < lower_bound) | (arr > upper_bound)).sum()
+    outlier_rate = float(outliers) / len(arr) if len(arr) > 0 else 0.0
+
+    p_value = 1.0
+    if len(conditions) == 2:
+        groups = list(conditions.values())
+        if all(len(g) > 0 for g in groups):
+            try:
+                _, p_value = stats.mannwhitneyu(groups[0], groups[1], alternative='two-sided')
+            except Exception:
+                p_value = 1.0
+    elif len(conditions) > 2:
+        groups = [g for g in conditions.values() if len(g) > 0]
+        if len(groups) > 1:
+            try:
+                _, p_value = stats.kruskal(*groups)
+            except Exception:
+                p_value = 1.0
+
+    result = {
+        "median_amplitude": median_amplitude,
+        "outlier_rate": outlier_rate,
+        "statistical_significance": float(p_value)
+    }
+
+    return result