2026-rff_mp/groshevava/docs/data/experiments.py

#проведение экспериментов и замер производительности

import time
import random
import csv
from linked_list import ll_insert, ll_find, ll_delete, ll_list_all
from hash_table import ht_create, ht_insert, ht_find, ht_delete, ht_list_all
from bst import bst_insert, bst_find, bst_delete, bst_list_all

#генерирует записи справочника. shuffled- случайный порядок, sorted-отсортированный по имени
def generate_test_data(n=10000, seed=42):
    random.seed(seed)
    
    names = [f"User_{i:05d}" for i in range(n)]
    phones = [f"+7-999-{i:07d}" for i in range(n)]
    
    records = list(zip(names, phones))
    
    records_shuffled = records.copy()
    random.shuffle(records_shuffled)
    
    records_sorted = sorted(records, key=lambda x: x[0])
    
    return records_shuffled, records_sorted

#замеряем время
def measure_insert(ll_structure, ht_structure, bst_structure, records, mode_name):
    results = []
    
    # Связный список
    start = time.perf_counter()
    head = None
    for name, phone in records:
        head = ll_insert(head, name, phone)
    end = time.perf_counter()
    results.append(["LinkedList", mode_name, "вставка", end - start])
    
    # Хеш-таблица
    start = time.perf_counter()
    buckets = ht_create(256)
    for name, phone in records:
        ht_insert(buckets, name, phone)
    end = time.perf_counter()
    results.append(["HashTable", mode_name, "вставка", end - start])
    
    # BST
    start = time.perf_counter()
    root = None
    for name, phone in records:
        root = bst_insert(root, name, phone)
    end = time.perf_counter()
    results.append(["BST", mode_name, "вставка", end - start])
    
    return results, head, buckets, root


def measure_find(head, buckets, root, all_names, mode_name):
    results = []
    
    # Выбираем 100 случайных существующих и 10 несуществующих имён
    existing_names = random.sample(all_names, 100)
    non_existing_names = [f"None_{i}" for i in range(10)]
    search_names = existing_names + non_existing_names
    
    # Связный список
    start = time.perf_counter()
    for name in search_names:
        ll_find(head, name)
    end = time.perf_counter()
    results.append(["LinkedList", mode_name, "поиск", end - start])
    
    # Хеш-таблица
    start = time.perf_counter()
    for name in search_names:
        ht_find(buckets, name)
    end = time.perf_counter()
    results.append(["HashTable", mode_name, "поиск", end - start])
    
    # BST
    start = time.perf_counter()
    for name in search_names:
        bst_find(root, name)
    end = time.perf_counter()
    results.append(["BST", mode_name, "поиск", end - start])
    
    return results


def measure_delete(head, buckets, root, all_names, mode_name):
    results = []
    
    # Выбираем 50 случайных имён для удаления
    delete_names = random.sample(all_names, 50)
    
    # Связный список
    start = time.perf_counter()
    for name in delete_names:
        head = ll_delete(head, name)
    end = time.perf_counter()
    results.append(["LinkedList", mode_name, "удаление", end - start])
    
    # Хеш-таблица
    start = time.perf_counter()
    for name in delete_names:
        ht_delete(buckets, name)
    end = time.perf_counter()
    results.append(["HashTable", mode_name, "удаление", end - start])
    
    # BST
    start = time.perf_counter()
    for name in delete_names:
        root = bst_delete(root, name)
    end = time.perf_counter()
    results.append(["BST", mode_name, "удаление", end - start])
    
    return results

#проводим эксперименты
def run_experiments(records_shuffled, records_sorted, repetitions=5):

    all_results = [
        ["Структура", "Режим", "Операция", "Время (сек)"]
    ]
    
    all_names = [record[0] for record in records_shuffled]
    
    for rep in range(repetitions):
        print(f"Повторение {rep + 1}/{repetitions}")
        
        # Шаффлированные данные
        results, head, buckets, root = measure_insert(
            None, None, None, records_shuffled, "случайный"
        )
        all_results.extend(results)
        
        results = measure_find(head, buckets, root, all_names, "случайный")
        all_results.extend(results)
        
        results = measure_delete(head, buckets, root, all_names, "случайный")
        all_results.extend(results)
        
        # Отсортированные данные
        results, head, buckets, root = measure_insert(
            None, None, None, records_sorted, "отсортированный"
        )
        all_results.extend(results)
        
        results = measure_find(head, buckets, root, all_names, "отсортированный")
        all_results.extend(results)
        
        results = measure_delete(head, buckets, root, all_names, "отсортированный")
        all_results.extend(results)
    
    return all_results


def save_results(results, filename="results.csv"):
    with open(filename, 'w', newline='', encoding='utf-8') as f:
        writer = csv.writer(f)
        writer.writerows(results)
    print(f"Результаты сохранены в {filename}")


def analyze_results(filename="results.csv"):
    from collections import defaultdict
    
    with open(filename, 'r', encoding='utf-8') as f:
        reader = csv.reader(f)
        next(reader)  # пропускаем заголовок
        data = list(reader)
    
    # Группируем для вычисления средних
    stats = defaultdict(list)
    for row in data:
        structure, mode, operation, time_str = row
        key = (structure, mode, operation)
        stats[key].append(float(time_str))
    
    print("\nСредние времена выполнения (сек):")
    print("-" * 60)
    print(f"{'Структура':<15} {'Режим':<20} {'Операция':<10} {'Время':<10}")
    print("-" * 60)
    
    for (structure, mode, operation), times in sorted(stats.items()):
        avg_time = sum(times) / len(times)
        print(f"{structure:<15} {mode:<20} {operation:<10} {avg_time:<10.6f}")


if __name__ == "__main__":
    print("Генерация тестовых данных...")
    records_shuffled, records_sorted = generate_test_data(10000)
    
    print("Запуск экспериментов...")
    results = run_experiments(records_shuffled, records_sorted, repetitions=5)
    
    save_results(results)
    
    analyze_results()