删除 2.py

2.py

@@ -1,145 +0,0 @@
-import numpy as np
-from scipy import stats
-from collections import defaultdict
-
-def detect_outliers_3sigma(series, threshold=3):
-    """3倍标准差法检测离群点"""
-    mean = np.mean(series)
-    std = np.std(series)
-    
-    upper_bound = mean + threshold * std
-    lower_bound = mean - threshold * std
-    
-    outliers = (series > upper_bound) | (series < lower_bound)
-    outlier_indices = np.where(outliers)[0]
-    
-    return {
-        'series': series,
-        'series_name': series.name if series.name else '序列',
-        'mean': mean,
-        'std': std,
-        'upper_bound': upper_bound,
-        'lower_bound': lower_bound,
-        'outliers': series[outliers],
-        'outlier_indices': outlier_indices,
-        'threshold': threshold
-    }
-
-def detect_outliers_iqr(series, k=1.5):
-    """四分位数法检测离群点"""
-    q1 = series.quantile(0.25)
-    q3 = series.quantile(0.75)
-    iqr = q3 - q1
-    
-    upper_bound = q3 + k * iqr
-    lower_bound = q1 - k * iqr
-    
-    outliers = (series > upper_bound) | (series < lower_bound)
-    outlier_indices = np.where(outliers)[0]
-    
-    return {
-        'series': series,
-        'series_name': series.name if series.name else '序列',
-        'q1': q1,
-        'q3': q3,
-        'iqr': iqr,
-        'upper_bound': upper_bound,
-        'lower_bound': lower_bound,
-        'outliers': series[outliers],
-        'outlier_indices': outlier_indices,
-        'k': k
-    }
-
-def detect_outliers_grubbs(series, alpha=0.05):
-    """Grubbs法检测离群点"""
-    values = series.values
-    n = len(values)
-    outlier_indices = []
-    
-    while True:
-        if n <= 2:
-            break
-            
-        mean = np.mean(values)
-        std = np.std(values)
-        abs_dev = np.abs(values - mean)
-        max_idx = np.argmax(abs_dev)
-        g = abs_dev[max_idx] / std
-        
-        t = stats.t.ppf(1 - alpha / (2 * n), n - 2)
-        critical = (n - 1) / np.sqrt(n) * np.sqrt(t**2 / (n - 2 + t**2))
-        
-        if g > critical:
-            outlier_indices.append(series.index.get_loc(series.index[max_idx]))
-            values = np.delete(values, max_idx)
-            n -= 1
-        else:
-            break
-    
-    upper_bound = mean + critical * std
-    lower_bound = mean - critical * std
-    
-    return {
-        'series': series,
-        'series_name': series.name if series.name else '序列',
-        'mean': mean,
-        'std': std,
-        'upper_bound': upper_bound,
-        'lower_bound': lower_bound,
-        'outliers': series[outlier_indices],
-        'outlier_indices': outlier_indices,
-        'alpha': alpha
-    }
-
-def detect_outliers_gesd(series, alpha=0.05, max_outliers=None):
-    """GESD (广义极端学生化偏差) 方法检测离群点"""
-    values = series.copy()
-    n = len(values)
-    if max_outliers is None:
-        max_outliers = n // 10  # 默认最多检测10%的数据点为离群点
-        
-    outlier_indices = []
-    r_values = []
-    lambda_values = []
-    
-    for i in range(1, max_outliers + 1):
-        mean = np.mean(values)
-        std = np.std(values)
-        abs_dev = np.abs(values - mean)
-        max_idx = np.argmax(abs_dev)
-        r = abs_dev[max_idx] / std
-        r_values.append(r)
-        
-        p = 1 - alpha / (2 * (n - i + 1))
-        t = stats.t.ppf(p, n - i - 1)
-        lambda_val = (n - i) * t / np.sqrt((n - i - 1 + t**2) * (n - i + 1))
-        lambda_values.append(lambda_val)
-        
-        if r > lambda_val:
-            original_idx = series.index.get_loc(values.index[max_idx])
-            outlier_indices.append(original_idx)
-            values = values.drop(values.index[max_idx])
-        else:
-            break
-    
-    if len(outlier_indices) > 0:
-        upper_bound = series.iloc[outlier_indices].max() + 0.1 * series.std()
-        lower_bound = series.iloc[outlier_indices].min() - 0.1 * series.std()
-    else:
-        upper_bound = series.mean() + 3 * series.std()
-        lower_bound = series.mean() - 3 * series.std()
-    
-    return {
-        'series': series,
-        'series_name': series.name if series.name else '序列',
-        'mean': np.mean(series),
-        'std': np.std(series),
-        'upper_bound': upper_bound,
-        'lower_bound': lower_bound,
-        'outliers': series[outlier_indices],
-        'outlier_indices': outlier_indices,
-        'alpha': alpha,
-        'max_outliers': max_outliers,
-        'r_values': r_values,
-        'lambda_values': lambda_values
-    }