dyachenkoas/docs/data/1laba.py

@@ -0,0 +1,391 @@
+import time
+import random
+import csv
+import os
+import matplotlib.pyplot as plt
+import numpy as np
+
+# ===================== 1. Связный список =====================
+def ll_insert(head, name, phone):
+    """Вставка в конец (или обновление), возвращает голову."""
+    new_node = {'name': name, 'phone': phone, 'next': None}
+    if head is None:
+        return new_node
+    cur = head
+    while True:
+        if cur['name'] == name:
+            cur['phone'] = phone
+            return head
+        if cur['next'] is None:
+            break
+        cur = cur['next']
+    cur['next'] = new_node
+    return head
+
+def ll_find(head, name):
+    cur = head
+    while cur:
+        if cur['name'] == name:
+            return cur['phone']
+        cur = cur['next']
+    return None
+
+def ll_delete(head, name):
+    if head is None:
+        return None
+    if head['name'] == name:
+        return head['next']
+    cur = head
+    while cur['next']:
+        if cur['next']['name'] == name:
+            cur['next'] = cur['next']['next']
+            return head
+        cur = cur['next']
+    return head
+
+def ll_list_all(head):
+    result = []
+    cur = head
+    while cur:
+        result.append((cur['name'], cur['phone']))
+        cur = cur['next']
+    result.sort(key=lambda x: x[0])
+    return result
+
+# ===================== 2. Хеш-таблица =====================
+def ht_hash(name, size):
+    h = 0
+    for ch in name:
+        h = (h * 31 + ord(ch)) % size
+    return h
+
+def ht_insert(buckets, name, phone):
+    idx = ht_hash(name, len(buckets))
+    buckets[idx] = ll_insert(buckets[idx], name, phone)
+
+def ht_find(buckets, name):
+    idx = ht_hash(name, len(buckets))
+    return ll_find(buckets[idx], name)
+
+def ht_delete(buckets, name):
+    idx = ht_hash(name, len(buckets))
+    buckets[idx] = ll_delete(buckets[idx], name)
+
+def ht_list_all(buckets):
+    result = []
+    for head in buckets:
+        cur = head
+        while cur:
+            result.append((cur['name'], cur['phone']))
+            cur = cur['next']
+    result.sort(key=lambda x: x[0])
+    return result
+
+# ===================== 3. BST =====================
+def bst_insert(root, name, phone):
+    """Итеративная вставка, не вызывает переполнения стека."""
+    new_node = {'name': name, 'phone': phone, 'left': None, 'right': None}
+    if root is None:
+        return new_node
+    
+    cur = root
+    while True:
+        if name < cur['name']:
+            if cur['left'] is None:
+                cur['left'] = new_node
+                break
+            cur = cur['left']
+        elif name > cur['name']:
+            if cur['right'] is None:
+                cur['right'] = new_node
+                break
+            cur = cur['right']
+        else:
+            cur['phone'] = phone  # обновление
+            break
+    return root
+
+def bst_find(root, name):
+    cur = root
+    while cur:
+        if name == cur['name']:
+            return cur['phone']
+        elif name < cur['name']:
+            cur = cur['left']
+        else:
+            cur = cur['right']
+    return None
+
+def bst_delete(root, name):
+    # Ищем узел и его родителя
+    parent = None
+    cur = root
+    while cur and cur['name'] != name:
+        parent = cur
+        if name < cur['name']:
+            cur = cur['left']
+        else:
+            cur = cur['right']
+    if cur is None:  # не найден
+        return root
+
+    # Случай 1: нет левого потомка
+    if cur['left'] is None:
+        child = cur['right']
+    # Случай 2: нет правого потомка
+    elif cur['right'] is None:
+        child = cur['left']
+    else:
+        # Случай 3: два потомка — ищем минимальный в правом поддереве
+        succ_parent = cur
+        succ = cur['right']
+        while succ['left']:
+            succ_parent = succ
+            succ = succ['left']
+        # Копируем данные
+        cur['name'] = succ['name']
+        cur['phone'] = succ['phone']
+        # Удаляем succ (у него нет левого потомка)
+        if succ_parent['left'] == succ:
+            succ_parent['left'] = succ['right']
+        else:
+            succ_parent['right'] = succ['right']
+        return root
+
+    # Подключаем child вместо cur
+    if parent is None:
+        return child
+    if parent['left'] == cur:
+        parent['left'] = child
+    else:
+        parent['right'] = child
+    return root
+
+def bst_list_all(root):
+    result = []
+    stack = []
+    cur = root
+    while stack or cur:
+        while cur:
+            stack.append(cur)
+            cur = cur['left']
+        cur = stack.pop()
+        result.append((cur['name'], cur['phone']))
+        cur = cur['right']
+    return result
+
+# ===================== Генерация данных =====================
+def generate_data(n=10000):
+    records = [(f"User_{i:05d}", f"8800{i:07d}") for i in range(n)]
+    shuffled = records[:]
+    random.shuffle(shuffled)
+    sorted_rec = sorted(records, key=lambda x: x[0])
+    return shuffled, sorted_rec
+
+# ===================== Замеры =====================
+def run_experiment(struct_type, records, n_searches=100, n_missing=10, n_deletes=50, repeats=5):
+    """
+    struct_type: 'll', 'ht', 'bst'
+    Возвращает словарь с усреднёнными замерами.
+    """
+    all_insert_times = []
+    all_search_times = []
+    all_delete_times = []
+
+    for _ in range(repeats):
+        # --- инициализация структуры ---
+        if struct_type == 'll':
+            head = None
+        elif struct_type == 'ht':
+            buckets = [None] * 512  # размер хеш-таблицы
+        else:  # bst
+            root = None
+
+        # --- вставка ---
+        start = time.perf_counter()
+        if struct_type == 'll':
+            for name, phone in records:
+                head = ll_insert(head, name, phone)
+        elif struct_type == 'ht':
+            for name, phone in records:
+                ht_insert(buckets, name, phone)
+        else:
+            for name, phone in records:
+                root = bst_insert(root, name, phone)
+        insert_time = time.perf_counter() - start
+        all_insert_times.append(insert_time)
+
+        # --- поиск ---
+        existing = random.sample(records, min(n_searches, len(records)))
+        missing = [(f"Missing_{i}", "") for i in range(n_missing)]
+        test_keys = existing + missing
+        random.shuffle(test_keys)
+
+        start = time.perf_counter()
+        if struct_type == 'll':
+            for name, _ in test_keys:
+                ll_find(head, name)
+        elif struct_type == 'ht':
+            for name, _ in test_keys:
+                ht_find(buckets, name)
+        else:
+            for name, _ in test_keys:
+                bst_find(root, name)
+        search_time = time.perf_counter() - start
+        all_search_times.append(search_time)
+
+        # --- удаление ---
+        del_sample = random.sample(records, min(n_deletes, len(records)))
+        start = time.perf_counter()
+        if struct_type == 'll':
+            for name, _ in del_sample:
+                head = ll_delete(head, name)
+        elif struct_type == 'ht':
+            for name, _ in del_sample:
+                ht_delete(buckets, name)
+        else:
+            for name, _ in del_sample:
+                root = bst_delete(root, name)
+        delete_time = time.perf_counter() - start
+        all_delete_times.append(delete_time)
+
+    return {
+        'struct': struct_type,
+        'insert_avg': sum(all_insert_times) / repeats,
+        'search_avg': sum(all_search_times) / repeats,
+        'delete_avg': sum(all_delete_times) / repeats,
+        'insert_all': all_insert_times,
+        'search_all': all_search_times,
+        'delete_all': all_delete_times,
+    }
+
+def main():
+    random.seed(42)
+    N = 10000
+    shuffled, sorted_rec = generate_data(N)
+
+    results = []
+    for struct_name, label in [('ll', 'LinkedList'), ('ht', 'HashTable'), ('bst', 'BST')]:
+        for order_name, records in [('shuffled', shuffled), ('sorted', sorted_rec)]:
+            print(f"Тестирую {label} на {order_name} данных...")
+            res = run_experiment(struct_name, records)
+            res['order'] = order_name
+            res['label'] = label
+            results.append(res)
+            print(f"{label:15} | {order_name:10} | insert: {res['insert_avg']:.6f}s | "
+                  f"search: {res['search_avg']:.6f}s | delete: {res['delete_avg']:.6f}s")
+
+    # Сохраняем в CSV
+    os.makedirs('docs/data', exist_ok=True)
+    with open('docs/data/benchmark_results.csv', 'w', newline='') as f:
+        writer = csv.writer(f)
+        writer.writerow(['structure', 'order', 'run', 'insert', 'search', 'delete'])
+        for r in results:
+            for i in range(len(r['insert_all'])):
+                writer.writerow([r['label'], r['order'], i + 1,
+                                 r['insert_all'][i], r['search_all'][i], r['delete_all'][i]])
+    print("\nCSV сохранён в docs/data/benchmark_results.csv")
+
+    # ===================== ГРАФИКИ =====================
+    structures = ['LinkedList', 'HashTable', 'BST']
+    orders = ['shuffled', 'sorted']
+    metrics = ['insert', 'search', 'delete']
+    metric_names = {'insert': 'Вставка (сек)', 'search': 'Поиск (сек)', 'delete': 'Удаление (сек)'}
+    colors = {'shuffled': '#4CAF50', 'sorted': '#FF5722'}
+
+    fig, axes = plt.subplots(1, 3, figsize=(16, 5.5))
+    
+    for idx, metric in enumerate(metrics):
+        ax = axes[idx]
+        x = np.arange(len(structures))
+        width = 0.35
+        
+        # Собираем данные
+        shuffled_vals = []
+        sorted_vals = []
+        for struct in structures:
+            for res in results:
+                if res['label'] == struct and res['order'] == 'shuffled':
+                    shuffled_vals.append(res[f'{metric}_avg'])
+                elif res['label'] == struct and res['order'] == 'sorted':
+                    sorted_vals.append(res[f'{metric}_avg'])
+        
+        bars1 = ax.bar(x - width/2, shuffled_vals, width, label='Случайный порядок', 
+                       color=colors['shuffled'], edgecolor='black', linewidth=0.5)
+        bars2 = ax.bar(x + width/2, sorted_vals, width, label='Отсортированный порядок', 
+                       color=colors['sorted'], edgecolor='black', linewidth=0.5)
+        
+        # Подписи значений на столбцах
+        for bar in bars1:
+            height = bar.get_height()
+            ax.text(bar.get_x() + bar.get_width()/2., height + max(shuffled_vals)*0.01,
+                    f'{height:.4f}', ha='center', va='bottom', fontsize=7)
+        for bar in bars2:
+            height = bar.get_height()
+            ax.text(bar.get_x() + bar.get_width()/2., height + max(sorted_vals)*0.01,
+                    f'{height:.4f}', ha='center', va='bottom', fontsize=7)
+        
+        ax.set_title(metric_names[metric], fontsize=12, fontweight='bold')
+        ax.set_xticks(x)
+        ax.set_xticklabels(structures, fontsize=10)
+        ax.legend(fontsize=9)
+        ax.grid(axis='y', alpha=0.3, linestyle='--')
+        
+        # Для поиска — логарифмическая шкала (чтобы было видно разницу)
+        if metric == 'search':
+            ax.set_yscale('log')
+            ax.set_ylabel('Время (сек, лог. шкала)', fontsize=9)
+        else:
+            ax.set_ylabel('Время (сек)', fontsize=9)
+
+    plt.suptitle('Сравнение производительности структур данных (N = 10 000 записей)', 
+                 fontsize=14, fontweight='bold', y=1.02)
+    plt.tight_layout()
+    
+    # Сохраняем график
+    graph_path = 'docs/benchmark_graph.png'
+    os.makedirs('docs', exist_ok=True)
+    plt.savefig(graph_path, dpi=150, bbox_inches='tight')
+    plt.show()
+    print(f"График сохранён в {graph_path}")
+
+    print("АНАЛИЗ РЕЗУЛЬТАТОВ")
+    
+    print("\n1. Влияние порядка данных на BST:")
+    bst_shuffled_insert = next(r['insert_avg'] for r in results if r['label']=='BST' and r['order']=='shuffled')
+    bst_sorted_insert = next(r['insert_avg'] for r in results if r['label']=='BST' and r['order']=='sorted')
+    print(f"   - Случайные данные: {bst_shuffled_insert:.6f} сек")
+    print(f"   - Отсортированные данные: {bst_sorted_insert:.6f} сек")
+    print(f"   - Замедление в {bst_sorted_insert/bst_shuffled_insert:.1f} раз")
+    print("   Причина: на отсортированных данных BST вырождается в связный список (глубина = N)")
+    
+    print("\n2. Стабильность хеш-таблицы:")
+    ht_shuffled = next(r['insert_avg'] for r in results if r['label']=='HashTable' and r['order']=='shuffled')
+    ht_sorted = next(r['insert_avg'] for r in results if r['label']=='HashTable' and r['order']=='sorted')
+    print(f"   - Случайные: {ht_shuffled:.6f} сек")
+    print(f"   - Отсортированные: {ht_sorted:.6f} сек")
+    print("   Причина: хеш-функция равномерно распределяет ключи независимо от порядка")
+    
+    print("\n3. Медленный поиск в связном списке:")
+    ll_search = next(r['search_avg'] for r in results if r['label']=='LinkedList' and r['order']=='shuffled')
+    ht_search = next(r['search_avg'] for r in results if r['label']=='HashTable' and r['order']=='shuffled')
+    print(f"   - LinkedList: {ll_search:.6f} сек")
+    print(f"   - HashTable: {ht_search:.6f} сек")
+    print(f"   - Хеш-таблица быстрее в {ll_search/ht_search:.1f} раз")
+    print("   Причина: поиск в списке всегда O(n), в хеш-таблице ~O(1)")
+    
+    print("\n4. Удаление:")
+    for label in ['LinkedList', 'HashTable', 'BST']:
+        del_shuff = next(r['delete_avg'] for r in results if r['label']==label and r['order']=='shuffled')
+        del_sort = next(r['delete_avg'] for r in results if r['label']==label and r['order']=='sorted')
+        print(f"   - {label:15}: случ.={del_shuff:.6f} сек, отсорт.={del_sort:.6f} сек")
+    
+    print("\n5. Рекомендации:")
+    print("   - Частый поиск + вставки → Хеш-таблица")
+    print("   - Нужна сортировка «из коробки» → Сбалансированное BST (AVL/Красно-чёрное)")
+    print("   - Только добавление в конец → Связный список")
+    print("   - Обычный BST опасен на реальных частично упорядоченных данных!")
+    print("="*60)
+
+if __name__ == '__main__':
+    main()

dyachenkoas/docs/data/benchmark_results.csv

@@ -0,0 +1,31 @@
+structure,order,run,insert,search,delete
+LinkedList,shuffled,1,2.009218709077686,0.01879545859992504,0.015042624901980162
+LinkedList,shuffled,2,2.0021930830553174,0.019880667328834534,0.011847833171486855
+LinkedList,shuffled,3,2.0060967500321567,0.01650112494826317,0.014535124879330397
+LinkedList,shuffled,4,2.0117608746513724,0.01841795863583684,0.01226008404046297
+LinkedList,shuffled,5,2.0219967076554894,0.019554249942302704,0.013240499887615442
+LinkedList,sorted,1,1.9876009593717754,0.01887020794674754,0.011140415910631418
+LinkedList,sorted,2,1.9921909999102354,0.01734908390790224,0.012648874893784523
+LinkedList,sorted,3,2.005885625258088,0.016392583958804607,0.012753374874591827
+LinkedList,sorted,4,2.0059890002012253,0.018063416704535484,0.013081958051770926
+LinkedList,sorted,5,2.000846417155117,0.01971287466585636,0.012666041031479836
+HashTable,shuffled,1,0.016287750098854303,0.00015062512829899788,7.462501525878906e-05
+HashTable,shuffled,2,0.014905208256095648,0.00014308281242847443,9.108288213610649e-05
+HashTable,shuffled,3,0.014663124922662973,0.00014704186469316483,7.82911665737629e-05
+HashTable,shuffled,4,0.014399250037968159,0.00014016637578606606,8.183391764760017e-05
+HashTable,shuffled,5,0.014289166778326035,0.000143333338201046,8.44169408082962e-05
+HashTable,sorted,1,0.014408249873667955,0.0001459997147321701,7.950002327561378e-05
+HashTable,sorted,2,0.016188541892915964,0.00016799988225102425,7.862504571676254e-05
+HashTable,sorted,3,0.022037209011614323,0.00014124996960163116,8.16253013908863e-05
+HashTable,sorted,4,0.01406783377751708,0.0001532919704914093,8.27917829155922e-05
+HashTable,sorted,5,0.014112749602645636,0.0001559997908771038,9.04998742043972e-05
+BST,shuffled,1,0.012917417101562023,0.0001227916218340397,7.3291826993227e-05
+BST,shuffled,2,0.01313945883885026,0.000122124794870615,7.370905950665474e-05
+BST,shuffled,3,0.01313587510958314,0.00011783279478549957,7.433397695422173e-05
+BST,shuffled,4,0.012769625056535006,0.00012508314102888107,6.770854815840721e-05
+BST,shuffled,5,0.012868000194430351,0.0001216246746480465,7.262500002980232e-05
+BST,sorted,1,3.3953831251710653,0.023627332877367735,0.013505042064934969
+BST,sorted,2,3.3977634580805898,0.025384000036865473,0.015041666105389595
+BST,sorted,3,3.404989833943546,0.02827158337458968,0.012459500227123499
+BST,sorted,4,3.389576541259885,0.025892207864671946,0.015427417121827602
+BST,sorted,5,3.408438625279814,0.025629667099565268,0.013972874730825424