added new type of plots; cloud point & overall view

.gitignore

@@ -2,3 +2,4 @@
 .py
 compile_commands.json
 build
+__pycache__

confidence.py

@@ -0,0 +1,68 @@
+import pandas as pd
+import numpy as np
+from scipy import stats
+
+def calculate_confidence_interval(data, confidence=0.95):
+    """
+    Вычисление доверительного интервала для среднего значения
+
+    Parameters:
+    data: массив данных
+    confidence: уровень доверия (по умолчанию 95%)
+
+    Returns:
+    tuple: (среднее, нижняя граница, верхняя граница, стандартная ошибка)
+    """
+    n = len(data)
+    mean = np.mean(data)
+    std_error = stats.sem(data)  # Стандартная ошибка среднего
+
+    # Вычисление доверительного интервала с использованием t-распределения
+    ci = stats.t.interval(confidence, n-1, loc=mean, scale=std_error)
+
+    return mean, ci[0], ci[1], std_error
+
+
+
+def compare_confidence_intervals(csv_file_path, confidence_levels=[0.90, 0.95, 0.99], delimiter=','):
+    """
+    Сравнение доверительных интервалов для разных уровней доверия
+    """
+    try:
+        df = pd.read_csv(csv_file_path, delimiter=delimiter, header=None,
+                         names=['index', 'delay_ms'])
+        delays = np.array(df['delay_ms'].values)
+
+        print("=" * 70)
+        print("СРАВНЕНИЕ ДОВЕРИТЕЛЬНЫХ ИНТЕРВАЛОВ ДЛЯ РАЗНЫХ УРОВНЕЙ ДОВЕРИЯ")
+        print("=" * 70)
+        print(f"Количество измерений: {len(delays)}")
+        print(f"Среднее значение: {np.mean(delays):.2f} мс")
+        print(f"Стандартная ошибка: {stats.sem(delays):.4f} мс\n")
+
+        results = []
+        for conf in confidence_levels:
+            mean, ci_lower, ci_upper, std_error = calculate_confidence_interval(
+                delays, conf)
+            width = ci_upper - ci_lower
+            margin = width / 2
+
+            results.append({
+                'confidence': conf,
+                'mean': mean,
+                'ci_lower': ci_lower,
+                'ci_upper': ci_upper,
+                'width': width,
+                'margin': margin
+            })
+
+            print(f"Доверительный интервал {conf*100:.0f}%:")
+            print(f"  [{ci_lower:.2f}, {ci_upper:.2f}] мс")
+            print(f"  Ширина: {width:.2f} мс, Погрешность: ±{margin:.2f} мс")
+            print()
+
+        return results
+
+    except Exception as e:
+        print(f"Ошибка при сравнении интервалов: {e}")
+        return None

delays.csv

@@ -1,527 +0,0 @@
-0,40.04
-1,45.29
-2,42.51
-3,47.70
-4,41.81
-5,38.94
-6,47.21
-7,44.49
-8,47.85
-9,37.17
-10,35.34
-11,38.51
-12,45.75
-13,42.85
-14,42.02
-15,45.20
-16,42.42
-17,48.65
-18,39.86
-19,44.28
-20,43.61
-21,36.93
-22,49.29
-23,44.65
-24,38.64
-25,36.87
-26,45.01
-27,36.96
-28,44.16
-29,44.49
-30,48.80
-31,41.11
-32,47.27
-33,48.46
-34,38.73
-35,41.96
-36,43.09
-37,36.38
-38,36.53
-39,38.64
-40,43.95
-41,40.16
-42,43.24
-43,39.37
-44,40.53
-45,39.79
-46,49.99
-47,44.13
-48,40.31
-49,42.57
-50,42.79
-51,39.92
-52,38.02
-53,36.32
-54,38.48
-55,38.57
-56,38.88
-57,40.04
-58,44.16
-59,45.56
-60,46.75
-61,44.95
-62,37.17
-63,46.30
-64,42.42
-65,37.72
-66,44.01
-67,48.19
-68,42.42
-69,38.54
-70,37.69
-71,41.90
-72,39.00
-73,43.37
-74,41.60
-75,39.61
-76,33.87
-77,39.98
-78,42.27
-79,38.27
-80,42.36
-81,40.62
-82,41.81
-83,39.83
-84,41.99
-85,45.07
-86,35.16
-87,49.29
-88,41.38
-89,41.60
-90,43.91
-91,40.99
-92,44.19
-93,42.48
-94,40.59
-95,34.79
-96,40.74
-97,43.76
-98,39.89
-99,40.01
-100,41.11
-101,40.22
-102,46.42
-103,41.69
-104,38.91
-105,34.91
-106,47.21
-107,41.38
-108,42.63
-109,35.64
-110,41.78
-111,35.77
-112,41.84
-113,46.02
-114,41.29
-115,36.35
-116,45.59
-117,42.79
-118,36.83
-119,41.11
-120,39.28
-121,47.58
-122,43.98
-123,44.22
-124,38.42
-125,36.62
-126,41.63
-127,43.82
-128,49.10
-129,36.35
-130,39.52
-131,41.60
-132,40.86
-133,36.07
-134,40.16
-135,43.37
-136,40.56
-137,38.54
-138,35.77
-139,41.05
-140,45.38
-141,43.46
-142,46.78
-143,42.02
-144,45.26
-145,39.49
-146,38.67
-147,39.92
-148,43.95
-149,40.04
-150,43.12
-151,38.09
-152,40.41
-153,35.55
-154,39.70
-155,39.83
-156,46.17
-157,39.22
-158,39.52
-159,37.60
-160,46.94
-161,42.94
-162,41.08
-163,39.09
-164,46.33
-165,41.38
-166,35.68
-167,38.85
-168,36.07
-169,41.11
-170,48.40
-171,41.56
-172,42.63
-173,41.87
-174,40.92
-175,38.24
-176,36.29
-177,47.52
-178,39.73
-179,38.76
-180,36.99
-181,37.11
-182,42.14
-183,40.31
-184,41.38
-185,36.56
-186,45.93
-187,40.13
-188,38.48
-189,39.55
-190,38.70
-191,38.94
-192,39.18
-193,39.28
-194,40.50
-195,41.60
-196,39.67
-197,41.72
-198,39.95
-199,43.06
-200,36.10
-201,35.34
-202,42.66
-203,36.99
-204,36.10
-205,35.10
-206,43.15
-207,35.22
-208,40.50
-209,40.71
-210,37.69
-211,40.89
-212,35.86
-213,42.97
-214,39.06
-215,42.33
-216,41.60
-217,42.69
-218,36.71
-219,42.79
-220,37.99
-221,38.06
-222,51.39
-223,39.55
-224,41.60
-225,35.80
-226,40.95
-227,43.30
-228,44.49
-229,41.66
-230,36.71
-231,48.98
-232,45.17
-233,41.44
-234,34.67
-235,41.81
-236,35.83
-237,43.03
-238,38.09
-239,35.34
-240,42.51
-241,42.69
-242,42.76
-243,50.02
-244,41.08
-245,42.45
-246,41.50
-247,47.67
-248,44.95
-249,41.14
-250,36.16
-251,35.28
-252,43.40
-253,48.58
-254,41.72
-255,40.80
-256,49.13
-257,41.38
-258,35.34
-259,38.42
-260,37.41
-261,37.60
-262,40.80
-263,44.89
-264,38.91
-265,36.90
-266,40.13
-267,43.27
-268,36.47
-269,43.55
-270,38.57
-271,42.63
-272,41.69
-273,48.00
-274,40.22
-275,39.43
-276,40.71
-277,40.74
-278,35.86
-279,38.06
-280,38.12
-281,46.26
-282,39.34
-283,41.50
-284,37.66
-285,37.84
-286,34.09
-287,43.15
-288,43.40
-289,45.53
-290,41.78
-291,42.91
-292,35.98
-293,48.07
-294,38.33
-295,40.50
-296,44.56
-297,42.91
-298,36.04
-299,37.23
-300,40.44
-301,41.66
-302,38.85
-303,42.14
-304,38.39
-305,36.62
-306,41.72
-307,39.00
-308,38.18
-309,43.21
-310,42.45
-311,37.69
-312,46.02
-313,42.11
-314,45.20
-315,43.58
-316,46.78
-317,39.03
-318,38.24
-319,45.59
-320,45.84
-321,45.20
-322,45.29
-323,40.41
-324,44.77
-325,42.97
-326,38.06
-327,39.09
-328,38.33
-329,34.58
-330,39.70
-331,43.88
-332,43.09
-333,38.24
-334,44.34
-335,41.56
-336,48.98
-337,39.18
-338,40.37
-339,39.58
-340,42.72
-341,44.10
-342,44.34
-343,40.47
-344,47.70
-345,37.96
-346,40.01
-347,36.19
-348,38.21
-349,39.40
-350,42.45
-351,45.62
-352,38.73
-353,35.95
-354,37.11
-355,45.47
-356,38.73
-357,47.06
-358,40.34
-359,35.58
-360,40.89
-361,34.94
-362,47.24
-363,42.30
-364,41.38
-365,37.63
-366,47.03
-367,39.18
-368,43.55
-369,37.63
-370,42.97
-371,42.24
-372,45.32
-373,38.39
-374,49.50
-375,39.67
-376,37.87
-377,39.06
-378,44.46
-379,38.70
-380,34.88
-381,38.02
-382,42.18
-383,40.47
-384,52.83
-385,45.01
-386,41.08
-387,50.51
-388,42.51
-389,53.71
-390,44.92
-391,39.06
-392,39.37
-393,44.74
-394,41.99
-395,36.16
-396,41.47
-397,38.67
-398,35.71
-399,46.08
-400,44.28
-401,36.25
-402,39.43
-403,45.59
-404,36.59
-405,39.79
-406,44.98
-407,49.13
-408,41.38
-409,40.41
-410,35.71
-411,44.89
-412,39.00
-413,46.11
-414,43.33
-415,48.65
-416,38.73
-417,38.88
-418,41.02
-419,44.25
-420,51.51
-421,45.84
-422,40.07
-423,44.22
-424,43.40
-425,46.75
-426,37.87
-427,42.97
-428,41.17
-429,38.21
-430,39.34
-431,34.33
-432,43.46
-433,44.59
-434,35.68
-435,39.76
-436,46.05
-437,40.34
-438,46.48
-439,43.67
-440,41.90
-441,53.31
-442,43.43
-443,39.46
-444,36.47
-445,46.69
-446,34.88
-447,41.96
-448,38.12
-449,36.38
-450,40.50
-451,45.84
-452,35.98
-453,38.18
-454,44.28
-455,36.29
-456,37.41
-457,42.72
-458,42.91
-459,43.21
-460,44.53
-461,37.51
-462,39.73
-463,44.89
-464,41.11
-465,43.52
-466,42.69
-467,37.90
-468,38.15
-469,35.40
-470,37.57
-471,44.83
-472,44.10
-473,40.44
-474,44.56
-475,42.69
-476,36.80
-477,39.00
-478,40.22
-479,37.32
-480,40.44
-481,40.71
-482,39.92
-483,45.29
-484,41.60
-485,41.66
-486,42.82
-487,47.21
-488,43.49
-489,47.76
-490,44.04
-491,41.29
-492,37.26
-493,39.49
-494,44.80
-495,39.03
-496,42.21
-497,45.59
-498,46.72
-499,41.05
-500,38.24
-501,40.44
-502,42.57
-503,53.38
-504,67.84
-505,44.98
-506,43.06
-507,43.18
-508,45.50
-509,47.88
-510,36.04
-511,40.22
-512,44.62
-513,43.76
-514,40.95
-515,39.15
-516,39.43
-517,39.55
-518,41.75
-519,48.00
-520,35.16
-521,43.52
-522,39.61
-523,36.68
-524,39.79
-525,51.21
-526,40.37

latency.py

@@ -1,225 +1,50 @@
-import pandas as pd
-import matplotlib.pyplot as plt
-import numpy as np
-from scipy import stats
+from latencyPlots import plot_delay_histogram_with_ci, plot_horizontal_hist_with_ci, plot_point_cloud
+from latencyData import LatencyData
+from confidence import compare_confidence_intervals
 import argparse
+import glob, logging, yaml, os
+import numpy as np
 
+logging.basicConfig(level=logging.INFO,
+                    format='%(asctime)s - %(levelname)s - %(message)s')
 
 def parse_args():
     parser = argparse.ArgumentParser()
-    parser.add_argument('-f', '--file', type=str, help='Path to the file')
+    parser.add_argument('-f', '--file', 
+                        type=str, 
+                        help='Path to the csv or yml file, depends on the mode', 
+                        required=True)
+    parser.add_argument('-m', '--mode', 
+                        type=int, 
+                        default=0, 
+                        help='Mode of the script; 0 - one file, 1 - dir')
     return parser.parse_args()
 
-
-def calculate_confidence_interval(data, confidence=0.95):
-    """
-    Вычисление доверительного интервала для среднего значения
-
-    Parameters:
-    data: массив данных
-    confidence: уровень доверия (по умолчанию 95%)
-
-    Returns:
-    tuple: (среднее, нижняя граница, верхняя граница, стандартная ошибка)
-    """
-    n = len(data)
-    mean = np.mean(data)
-    std_error = stats.sem(data)  # Стандартная ошибка среднего
-
-    # Вычисление доверительного интервала с использованием t-распределения
-    ci = stats.t.interval(confidence, n-1, loc=mean, scale=std_error)
-
-    return mean, ci[0], ci[1], std_error
-
-
-def plot_delay_histogram_with_ci(csv_file_path, save_to_file=True, show_plot=False,
-                                 confidence_level=0.99, title="default", delimiter=','):
-    """
-    Универсальный скрипт для построения гистограммы с доверительным интервалом
-    """
-
-    try:
-        # Чтение данных
-        df = pd.read_csv(csv_file_path, delimiter=delimiter, header=None,
-                         names=['index', 'delay_ms'])
-        delays = np.array(df['delay_ms'].values)
-        plt.rcParams.update({'font.size': 18})
-
-        # Вычисление базовой статистики
-        mean_delay = np.mean(delays)
-        median_delay = np.median(delays)
-        std_delay = np.std(delays)
-        min_delay = np.min(delays)
-        max_delay = np.max(delays)
-
-        # Вычисление доверительного интервала
-        mean_ci, ci_lower, ci_upper, std_error = calculate_confidence_interval(
-            delays, confidence_level)
-
-        # Вывод статистики
-        print("=" * 60)
-        print("АНАЛИЗ ЗАДЕРЖЕК С ДОВЕРИТЕЛЬНЫМ ИНТЕРВАЛОМ")
-        print("=" * 60)
-        print(f"Количество измерений (n): {len(delays)}")
-        print(f"Уровень доверия: {confidence_level*100}%")
-        print("\n--- ОСНОВНАЯ СТАТИСТИКА ---")
-        print(f"Минимальная задержка: {min_delay:.2f} мс")
-        print(f"Максимальная задержка: {max_delay:.2f} мс")
-        print(f"Средняя задержка: {mean_delay:.2f} мс")
-        print(f"Медианная задержка: {median_delay:.2f} мс")
-        print(f"Стандартное отклонение: {std_delay:.2f} мс")
-        print(f"Стандартная ошибка: {std_error:.4f} мс")
-
-        print(f"\n--- ДОВЕРИТЕЛЬНЫЙ ИНТЕРВАЛ {confidence_level*100}% ---")
-        print(f"Точечная оценка среднего: {mean_ci:.2f} мс")
-        print(f"Доверительный интервал: [{ci_lower:.2f}, {ci_upper:.2f}] мс")
-        print(f"Ширина интервала: {ci_upper - ci_lower:.2f} мс")
-        print(f"Погрешность: ±{(ci_upper - ci_lower)/2:.2f} мс")
-
-        # Дополнительные метрики
-        cv = (std_delay / mean_delay) * 100  # Коэффициент вариации
-        print(f"\n--- ДОПОЛНИТЕЛЬНЫЕ МЕТРИКИ ---")
-        print(f"Коэффициент вариации: {cv:.1f}%")
-        print(f"Q1 (25-й перцентиль): {np.percentile(delays, 25):.2f} мс")
-        print(f"Q3 (75-й перцентиль): {np.percentile(delays, 75):.2f} мс")
-        print(
-            f"IQR (межквартильный размах): {np.percentile(delays, 75) - np.percentile(delays, 25):.2f} мс")
-
-        # Создание гистограммы
-        plt.figure(figsize=(14, 8))
-        n_bins = min(30, len(delays) // 5)
-
-        # Гистограмма
-        n, bins, patches = plt.hist(delays, bins=n_bins, alpha=0.7, color='steelblue',
-                                    edgecolor='black', linewidth=0.5, density=False)
-
-        plt.xlabel('Задержка (мс)')
-        plt.ylabel('Частота')
-        plt.title(f'Распределение задержек на обработку нажатия мыши {title} (n={len(delays)}, CI={confidence_level*100}%)',
-                  fontsize=14, fontweight='bold')
-        plt.grid(True, alpha=0.3)
-
-        mean_delay = float(mean_delay)
-        median_delay = float(median_delay)
-        # Добавление статистических линий
-        plt.axvline(mean_delay, color='red', linestyle='-', linewidth=2,
-                    label=f'Среднее ({mean_delay:.1f} мс)')
-        plt.axvline(median_delay, color='green', linestyle='-', linewidth=2,
-                    label=f'Медиана ({median_delay:.1f} мс)')
-
-        # Добавление доверительного интервала
-        plt.axvspan(ci_lower, ci_upper, alpha=0.2, color='yellow',
-                    label=f'ДИ {confidence_level*100}%: [{ci_lower:.1f}, {ci_upper:.1f}] мс')
-
-        # Вертикальные линии для границ доверительного интервала
-        plt.axvline(ci_lower, color='orange',
-                    linestyle='--', linewidth=1, alpha=0.7)
-        plt.axvline(ci_upper, color='orange',
-                    linestyle='--', linewidth=1, alpha=0.7)
-
-        plt.legend(loc='upper right')
-        plt.tight_layout()
-
-        # Сохранение в файл
-        if save_to_file:
-            output_file = f'delay_distribution_{title}_ci_{int(confidence_level*100)}.png'
-            plt.savefig(output_file, dpi=300, bbox_inches='tight')
-            print(f"\n✓ Гистограмма сохранена как: {output_file}")
-
-        # Показ графика
-        if show_plot:
-            plt.show()
-        else:
-            plt.close()
-
-        print("\n✓ Анализ завершен успешно")
-
-        return {
-            'data': delays,
-            'mean': mean_delay,
-            'median': median_delay,
-            'std': std_delay,
-            'ci_lower': ci_lower,
-            'ci_upper': ci_upper,
-            'confidence_level': confidence_level
-        }
-
-    except FileNotFoundError:
-        print(f"✗ Ошибка: Файл '{csv_file_path}' не найден")
-        return None
-    except Exception as e:
-        print(f"✗ Ошибка: {e}")
-        return None
-
-
-def compare_confidence_intervals(csv_file_path, confidence_levels=[0.90, 0.95, 0.99], delimiter=','):
-    """
-    Сравнение доверительных интервалов для разных уровней доверия
-    """
-    try:
-        df = pd.read_csv(csv_file_path, delimiter=delimiter, header=None,
-                         names=['index', 'delay_ms'])
-        delays = np.array(df['delay_ms'].values)
-
-        print("=" * 70)
-        print("СРАВНЕНИЕ ДОВЕРИТЕЛЬНЫХ ИНТЕРВАЛОВ ДЛЯ РАЗНЫХ УРОВНЕЙ ДОВЕРИЯ")
-        print("=" * 70)
-        print(f"Количество измерений: {len(delays)}")
-        print(f"Среднее значение: {np.mean(delays):.2f} мс")
-        print(f"Стандартная ошибка: {stats.sem(delays):.4f} мс\n")
-
-        results = []
-        for conf in confidence_levels:
-            mean, ci_lower, ci_upper, std_error = calculate_confidence_interval(
-                delays, conf)
-            width = ci_upper - ci_lower
-            margin = width / 2
-
-            results.append({
-                'confidence': conf,
-                'mean': mean,
-                'ci_lower': ci_lower,
-                'ci_upper': ci_upper,
-                'width': width,
-                'margin': margin
-            })
-
-            print(f"Доверительный интервал {conf*100:.0f}%:")
-            print(f"  [{ci_lower:.2f}, {ci_upper:.2f}] мс")
-            print(f"  Ширина: {width:.2f} мс, Погрешность: ±{margin:.2f} мс")
-            print()
-
-        return results
-
-    except Exception as e:
-        print(f"Ошибка при сравнении интервалов: {e}")
-        return None
-
-
-# Пример использования
-if __name__ == "__main__":
-    # Укажите путь к вашему файлу
+def one_file_run():
     csv_file = parse_args().file
-    title = csv_file.split('/')[1].split('.')[0]
+    path_splitted = csv_file.split('/')
+    title = path_splitted[len(path_splitted) - 1].split('.')[0]
     print("АНАЛИЗ ДОВЕРИТЕЛЬНЫХ ИНТЕРВАЛОВ ДЛЯ ЗАДЕРЖЕК")
     print("=" * 50)
+    l = None
+    try:
+        l = LatencyData(csv_file, confidence_level=0.99)
+    except FileNotFoundError:
+        print(f"✗ Ошибка: Файл '{csv_file}' не найден")
+        exit(1)
 
-    # Основной анализ с 95% доверительным интервалом
     results = plot_delay_histogram_with_ci(
-        csv_file_path=csv_file,
         save_to_file=True,
         show_plot=False,
-        confidence_level=0.99,
-        title=title
+        title=title,
+        l=l
     )
 
     print("\n" + "=" * 50)
 
-    # Сравнение разных уровней доверия
     compare_results = compare_confidence_intervals(
         csv_file, [0.90, 0.95, 0.99])
 
-    # Интерпретация результатов
     if results:
         print("\n--- ИНТЕРПРЕТАЦИЯ РЕЗУЛЬТАТОВ ---")
         print(f"С вероятностью 95% истинное среднее значение задержки ")
@@ -227,3 +52,45 @@ if __name__ == "__main__":
             f"находится в интервале от {results['ci_lower']:.2f} до {results['ci_upper']:.2f} мс.")
         print(f"Это означает, что если бы мы повторили эксперимент много раз,")
         print(f"95% вычисленных таким образом интервалов содержали бы истинное среднее значение.")
+
+def validate_csv_list(csv_arr: list, dirConfig: dict) -> bool:
+    is_colors_valid = all(file in dirConfig['colors'].keys() for file in csv_arr)
+    is_labels_valid = all(file in dirConfig['labels'].keys() for file in csv_arr)
+    if not is_labels_valid:
+        logging.error("Labels in yaml is not fully valid")
+    if not is_colors_valid:
+        logging.error("Colors in yaml is not fully valid")
+    return is_colors_valid & is_labels_valid
+
+def dir_run():
+    dirConfig = None
+    with open(parse_args().file + '/dir.yml') as f:
+        dirConfig = yaml.safe_load(f)
+    if dirConfig is None: 
+        raise Exception("File not found or not loaded")
+
+    title = parse_args().file.split('/')[1]
+    all_csv_in_dir = glob.glob(parse_args().file + '/*.csv')
+    logging.info(f"Found {len(all_csv_in_dir)} csv files in directory {parse_args().file}")
+    lDataArr = {os.path.basename(file): LatencyData(file) for file in all_csv_in_dir}
+
+    validate_csv_list([os.path.basename(file) for file in all_csv_in_dir], dirConfig)
+    plot_point_cloud(lDataArr, 
+                     dirConfig['colors'], 
+                     dirConfig['labels'], 
+                     float(dirConfig['point_cloud']['dT']),
+                     title=title)
+
+    plot_horizontal_hist_with_ci(lDataArr, 
+                                 dirConfig['colors'], 
+                                 dirConfig['labels'],
+                                 title=title)
+    
+
+if __name__ == "__main__":
+    args = parse_args()
+    mode = args.mode
+    if mode == 0:
+        one_file_run()
+    elif mode == 1:
+        dir_run()

latencyData.py

@@ -0,0 +1,59 @@
+import pandas as pd
+import numpy as np
+from confidence import calculate_confidence_interval
+
+class LatencyData:
+    def __init__(self, filepath, delimiter=',', confidence_level=0.99):
+        self.df = pd.read_csv(filepath, delimiter=delimiter, header=None,
+                    names=['index', 'delay_ms'])
+        self.__parse(confidence_level)
+        
+    def __parse(self, confidence_level):
+        self.delays = np.array(self.df['delay_ms'].values)
+        self.confidence_level = confidence_level
+        # Вычисление базовой статистики
+        self.mean_delay = np.mean(self.delays)
+        self.median_delay = np.median(self.delays)
+        self.std_delay = np.std(self.delays)
+        self.min_delay = np.min(self.delays)
+        self.max_delay = np.max(self.delays)
+
+        # Вычисление доверительного интервала
+        self.mean_ci, self.ci_lower, self.ci_upper, self.std_error = calculate_confidence_interval(
+            self.delays, confidence_level)
+        
+        self.__output_data()
+
+    @property
+    def err(self):
+        return (self.ci_upper - self.ci_lower)/2
+
+    def __output_data(self):
+        # Вывод статистики
+        print("=" * 60)
+        print("АНАЛИЗ ЗАДЕРЖЕК С ДОВЕРИТЕЛЬНЫМ ИНТЕРВАЛОМ")
+        print("=" * 60)
+        print(f"Количество измерений (n): {len(self.delays)}")
+        print(f"Уровень доверия: {self.confidence_level*100}%")
+        print("\n--- ОСНОВНАЯ СТАТИСТИКА ---")
+        print(f"Минимальная задержка: {self.min_delay:.2f} мс")
+        print(f"Максимальная задержка: {self.max_delay:.2f} мс")
+        print(f"Средняя задержка: {self.mean_delay:.2f} мс")
+        print(f"Медианная задержка: {self.median_delay:.2f} мс")
+        print(f"Стандартное отклонение: {self.std_delay:.2f} мс")
+        print(f"Стандартная ошибка: {self.std_error:.4f} мс")
+
+        print(f"\n--- ДОВЕРИТЕЛЬНЫЙ ИНТЕРВАЛ {self.confidence_level*100}% ---")
+        print(f"Точечная оценка среднего: {self.mean_ci:.2f} мс")
+        print(f"Доверительный интервал: [{self.ci_lower:.2f}, {self.ci_upper:.2f}] мс")
+        print(f"Ширина интервала: {self.ci_upper - self.ci_lower:.2f} мс")
+        print(f"Погрешность: ±{(self.ci_upper - self.ci_lower)/2:.2f} мс")
+
+        # Дополнительные метрики
+        cv = (self.std_delay / self.mean_delay) * 100  # Коэффициент вариации
+        print(f"\n--- ДОПОЛНИТЕЛЬНЫЕ МЕТРИКИ ---")
+        print(f"Коэффициент вариации: {cv:.1f}%")
+        print(f"Q1 (25-й перцентиль): {np.percentile(self.delays, 25):.2f} мс")
+        print(f"Q3 (75-й перцентиль): {np.percentile(self.delays, 75):.2f} мс")
+        print(
+            f"IQR (межквартильный размах): {np.percentile(self.delays, 75) - np.percentile(self.delays, 25):.2f} мс")

latencyPlots.py

@@ -0,0 +1,182 @@
+import matplotlib.pyplot as plt
+from latencyData import LatencyData
+import logging
+
+def plot_delay_histogram_with_ci(l : LatencyData, save_to_file=True, show_plot=False, title="default"):
+    """
+    Универсальный скрипт для построения гистограммы с доверительным интервалом
+    """
+    try:
+        plt.rcParams.update({'font.size': 18})
+        # Создание гистограммы
+        plt.figure(figsize=(14, 8))
+        n_bins = min(30, len(l.delays) // 5)
+
+        n, bins, patches = plt.hist(l.delays, bins=n_bins, alpha=0.7, color='steelblue',
+                                    edgecolor='black', linewidth=0.5, density=False)
+
+        plt.xlabel('Задержка (мс)')
+        plt.ylabel('Частота')
+        plt.title(f'Распределение задержек на обработку нажатия мыши {title} (n={len(l.delays)}, CI={l.confidence_level*100}%)',
+                  fontsize=14, fontweight='bold')
+        plt.grid(True, alpha=0.3)
+
+        l.mean_delay = float(l.mean_delay)
+        l.median_delay = float(l.median_delay)
+        # Добавление статистических линий
+        plt.axvline(l.mean_delay, color='red', linestyle='-', linewidth=2,
+                    label=f'Среднее ({l.mean_delay:.1f} мс)')
+        plt.axvline(l.median_delay, color='green', linestyle='-', linewidth=2,
+                    label=f'Медиана ({l.median_delay:.1f} мс)')
+
+        # Добавление доверительного интервала
+        plt.axvspan(l.ci_lower, l.ci_upper, alpha=0.2, color='yellow',
+                    label=f'ДИ {l.confidence_level*100}%: [{l.ci_lower:.1f}, {l.ci_upper:.1f}] мс')
+
+        # Вертикальные линии для границ доверительного интервала
+        plt.axvline(l.ci_lower, color='orange',
+                    linestyle='--', linewidth=1, alpha=0.7)
+        plt.axvline(l.ci_upper, color='orange',
+                    linestyle='--', linewidth=1, alpha=0.7)
+
+        plt.legend(loc='upper right')
+        plt.tight_layout()
+
+        # Сохранение в файл
+        if save_to_file:
+            output_file = f'delay_distribution_{title}_ci_{int(l.confidence_level*100)}.png'
+            plt.savefig(output_file, dpi=300, bbox_inches='tight')
+            print(f"\n✓ Гистограмма сохранена как: {output_file}")
+
+        # Показ графика
+        if show_plot:
+            plt.show()
+        else:
+            plt.close()
+
+        print("\n✓ Анализ завершен успешно")
+
+        return {
+            'data': l.delays,
+            'mean': l.mean_delay,
+            'median': l.median_delay,
+            'std': l.std_delay,
+            'ci_lower': l.ci_lower,
+            'ci_upper': l.ci_upper,
+            'confidence_level': l.confidence_level
+        }
+    except Exception as e:
+        print(f"✗ Ошибка: {e}")
+        return None
+    
+def plot_horizontal_hist_with_ci(latencies : dict, colors: dict, labels: dict, save_to_file=True, show_plot=True, title="default"):
+    """
+    Parameters
+    ----------
+    latencies: dict
+        key is a filename, value is a latencyData object
+    
+    colors: dict
+        key is a filename, value is string in format tab:<color>
+    
+    labels: dict
+        key is a filename, value is a string label
+
+    """
+    plt.rcParams.update({'font.size': 14})
+    # Создание гистограммы
+    fig, ax = plt.subplots(figsize=(12,6))
+    
+    keys = latencies.keys()
+    values = latencies.values()
+
+    medians = [i.mean_delay for i in values]
+    errors = [i.err for i in values]
+    y_idx = range(len(keys))
+    logging.info(f"medians={medians}; errors={errors}")
+    # sorting of all arrays to have same order of elements
+    bar_colors = [colors[key] for key in keys]
+    bar_labels = [labels[key] for key in keys]
+
+    ax.barh(y_idx, medians, xerr=errors, color=bar_colors, capsize=15, align='center')
+
+    logging.info(f"y_idx={y_idx}; bar_labels={bar_labels}")
+    ax.set_yticks(y_idx, bar_labels)
+    ax.set_xlabel('Задержка (мс)')
+    fig.tight_layout()
+
+    if show_plot:
+        plt.show()
+    else:
+        plt.close()
+    
+    if save_to_file:
+        fig.savefig(f'delay_dist_overall_{title}')
+
+    
+def plot_point_cloud(latencies : dict, 
+                     colors: dict, 
+                     labels: dict, 
+                     dT : float, 
+                     hlines = dict,
+                     save_to_file=True, 
+                     show_plot=True, 
+                     title="default"):
+    """
+    Parameters
+    ----------
+    latencies: dict
+        key is a filename, value is a latencyData object
+    
+    colors: dict
+        key is a filename, value is string in format tab:<color>
+    
+    labels: dict
+        key is a filename, value is a string label
+
+    """
+
+    plt.rcParams.update({'font.size': 14})
+    # Создание гистограммы
+    fig, ax = plt.subplots(figsize=(10,8))
+    
+    keys = latencies.keys()
+    values = latencies.values()
+
+    medians = [i.mean_delay for i in values]
+
+    # sorting of all arrays to have same order of elements
+    point_labels = [labels[key] for key in keys]
+    point_colors = [colors[key] for key in keys]
+    point_xminall = min(len(value.delays) for value in values)
+    point_values = [value.delays[:point_xminall] for value in values]
+    point_x = list(range(point_xminall)) 
+    point_x = [float(x) * dT for x in point_x] 
+    point_ld = [value for value in values]
+    for i in range(len(point_values)):
+        ax.plot(point_x, 
+                point_values[i], 
+                color=point_colors[i], 
+                label=point_labels[i], 
+                marker='o',
+                linestyle='None')
+        ax.axhline(y=medians[i],
+                   color=point_colors[i],
+                   linestyle='--')
+        ax.axhspan(point_ld[i].ci_lower, 
+                   point_ld[i].ci_upper, 
+                   alpha=0.2, 
+                   color=point_colors[i])
+    ax.legend(ncol=3,bbox_to_anchor=(0.75, 1.1), fancybox=True)
+    ax.set_xlabel('Время (секунды)')
+    ax.set_ylabel('Задержка (мс)')
+    
+    fig.tight_layout()
+
+    if show_plot:
+        plt.show()
+    else:
+        plt.close()
+    
+    if save_to_file:
+        fig.savefig(f'delay_dist_cloud_{title}')

shit.py

@@ -10,6 +10,25 @@ import matplotlib.pyplot as plt
 logging.basicConfig(level=logging.INFO,
                     format='%(asctime)s - %(levelname)s - %(message)s')
 
+def mix_csv_column_pandas(filename, index_col=0, value_col=1):
+    """
+    Read a CSV file, mix up values in the specified column, and overwrite the file using pandas.
+    
+    Args:
+        filename (str): Path to the CSV file
+        index_col (int): Column index for identifiers (default: 0)
+        value_col (int): Column index for values to mix (default: 1)
+    """
+    # Read CSV
+    df = pd.read_csv(filename, header=None)
+    
+    # Mix values in the specified column
+    mixed_values = df[value_col].sample(frac=1).reset_index(drop=True)
+    df[value_col] = mixed_values
+    
+    # Save back to same file
+    df.to_csv(filename, index=False, header=False)
+
 class Interpol:
     """
     Huhu, interpol