Испавил ген.алгоритм, надо только нормально по папочкам расскидать (мне сного шутить, что у меня нет папы?)

genetic_algorithm/genetic_algorithm.py

@@ -1,11 +1,15 @@
 import json
 from datetime import datetime, timedelta
 import random
-from typing import Dict, List
+from typing import Dict, List, Tuple, Any
 from fastapi import FastAPI, HTTPException, Request
 
-def load_graph_from_request(request) -> Dict[str, Dict[str, List]]:
+from schemas import TripRequest
+
+
+def load_graph_from_request(request: TripRequest) -> Tuple[Dict[str, Dict[str, List[Any]]], str, str, List[Dict]]:
     graph = {}
+    flights_data = []
 
     # Загрузить данные о полетах из запроса
     flights = request.flights
@@ -19,6 +23,15 @@ def load_graph_from_request(request) -> Dict[str, Dict[str, List]]:
         departure_time = datetime.fromisoformat(flight.departureTime)
         destination_time = datetime.fromisoformat(flight.destinationTime)
 
+        # Сохранение данных о полете для вывода
+        flights_data.append({
+            "id": flight.id,
+            "departurePoint": departure_point,
+            "destinationPoint": destination_point,
+            "departureTime": departure_time,
+            "destinationTime": destination_time
+        })
+
         # Проверка наличия вершин в графе и их создание при необходимости
         if departure_point not in graph:
             graph[departure_point] = {}
@@ -28,9 +41,9 @@ def load_graph_from_request(request) -> Dict[str, Dict[str, List]]:
         # Добавление ребра в граф
         graph[departure_point][destination_point] = [flight.distance, departure_time, destination_time]
 
-    return graph, start_point, end_point
+    return graph, start_point, end_point, flights_data
+
 
-# Функция для вычисления длины и времени пути с учетом минимального интервала времени
 def path_length_and_time(path, graph):
     length = 0
     if path[0] not in graph or path[1] not in graph[path[0]]:
@@ -49,8 +62,8 @@ def path_length_and_time(path, graph):
         end_time = graph[path[i]][path[i + 1]][2]
     return length, start_time, end_time
 
-# Функция для генерации начальной популяции
-def generate_population(size, start, end, graph, max_attempts=100):
+
+def generate_population(size, start, end, graph, max_attempts=1000):
     population = []
     for i in range(size):
         attempts = 0
@@ -74,13 +87,12 @@ def generate_population(size, start, end, graph, max_attempts=100):
             attempts += 1
     return population
 
-# Функция для селекции родителей
+
 def select_parents(population, graph):
     sorted_population = sorted(population, key=lambda p: path_length_and_time(p, graph)[0])
     return sorted_population[:len(sorted_population) // 2]
 
 
-# Функция для скрещивания (кроссинговера)
 def crossover(parent1, parent2):
     crossover_point = len(parent1) // 2
     child = parent1[:crossover_point]
@@ -90,7 +102,6 @@ def crossover(parent1, parent2):
     return child
 
 
-# Функция для мутации
 def mutate(child):
     if len(child) <= 2:
         return child  # Если путь состоит из одной вершины, мутация не требуется
@@ -100,9 +111,18 @@ def mutate(child):
     return child
 
 
-# Главная функция генетического алгоритма
-def genetic_algorithm(start, end, graph, population_size=100, generations=100):
+def update_best_paths(best_paths, path, graph, max_best_paths=3):
+    length, _, _ = path_length_and_time(path, graph)
+    if path not in best_paths:
+        best_paths[path] = length
+    if len(best_paths) > max_best_paths:
+        worst_path = max(best_paths, key=best_paths.get)
+        del best_paths[worst_path]
+
+
+def genetic_algorithm(start, end, graph, flights_data, population_size=100, generations=100):
     population = generate_population(population_size, start, end, graph)
+    best_paths = {}  # Словарь для хранения уникальных лучших маршрутов и их длин
     for generation in range(generations):
         parents = select_parents(population, graph)
         new_population = parents[:]
@@ -112,8 +132,35 @@ def genetic_algorithm(start, end, graph, population_size=100, generations=100):
             child = crossover(parent1, parent2)
             if random.random() < 0.1:  # Вероятность мутации
                 child = mutate(child)
-            new_population.append(child)
+            if child[-1] == end:
+                new_population.append(child)
+                update_best_paths(best_paths, tuple(child), graph)
         population = new_population
-    best_path = min(population, key=lambda p: path_length_and_time(p, graph)[0])
-    best_length, start_time, end_time = path_length_and_time(best_path, graph)
-    return best_path, best_length, start_time, end_time
+
+    best_paths = sorted(best_paths.items(), key=lambda item: item[1])
+    top_paths = [list(path) for path, _ in best_paths[:3]]
+
+    result = []
+    for path in top_paths:
+        path_data = []
+        for i in range(len(path) - 1):
+            for flight in flights_data:
+                if flight['departurePoint'] == path[i] and flight['destinationPoint'] == path[i + 1]:
+                    path_data.append({
+                        "id": flight['id'],
+                        "departurePoint": flight['departurePoint'],
+                        "destinationPoint": flight['destinationPoint']
+                    })
+                    break
+        length, start_time, end_time = path_length_and_time(path, graph)
+        # Проверка на корректность длины пути и времени
+        if length == float('inf') or start_time is None or end_time is None:
+            continue
+        result.append({
+            "to": path_data,
+            "start_time": start_time.isoformat() if start_time else None,
+            "end_time": end_time.isoformat() if end_time else None,
+            "best_length": length
+        })
+
+    return result

router_class.py

@@ -17,9 +17,11 @@ router = APIRouter(
 
 @router.post("/get_flight/")
 async def get_flight(request: TripRequest):
-    graph, start_point, end_point = load_graph_from_request(request)
-    best_path, best_length, start_time, end_time = genetic_algorithm(start_point, end_point, graph)
-    return best_path, best_length, start_time, end_time
+    graph, start_point, end_point, flights_data = load_graph_from_request(request)
+    result = genetic_algorithm(start_point, end_point, graph, flights_data)
+    if not result:
+        raise HTTPException(status_code=404, detail="No valid paths found")
+    return result
 
 @router.get("/negative")
 async def get_class_names() -> List[str]: