sumo_DQN.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: pyvirtualdisplay in c:\\users\\asus\\appdata\\local\\programs\\python\\python310\\lib\\site-packages (3.0)\n",
      "Requirement already satisfied: Pillow in c:\\users\\asus\\appdata\\local\\programs\\python\\python310\\lib\\site-packages (10.3.0)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING: Ignoring invalid distribution -orch (c:\\users\\asus\\appdata\\local\\programs\\python\\python310\\lib\\site-packages)\n",
      "WARNING: Ignoring invalid distribution -orch (c:\\users\\asus\\appdata\\local\\programs\\python\\python310\\lib\\site-packages)\n",
      "WARNING: Ignoring invalid distribution -orch (c:\\users\\asus\\appdata\\local\\programs\\python\\python310\\lib\\site-packages)\n"
     ]
    }
   ],
   "source": [
    "!pip install pyvirtualdisplay Pillow\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# SUMO Setups"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SUMO_HOME is set to: C:\\Program Files (x86)\\Eclipse\\Sumo\n",
      "Does SUMO_HOME exist? True\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import torch\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import gymnasium as gym\n",
    "from tqdm import tqdm\n",
    "from dotenv import load_dotenv\n",
    "from gymnasium import spaces\n",
    "import matplotlib.patches as patches\n",
    "import shutil\n",
    "\n",
    "SUMO_HOME = 'C:\\\\Program Files (x86)\\\\Eclipse\\\\Sumo'\n",
    "\n",
    "os.environ['SUMO_HOME'] = SUMO_HOME\n",
    "\n",
    "# Print the SUMO_HOME environment variable to verify\n",
    "print(\"SUMO_HOME is set to:\", os.environ.get('SUMO_HOME'))\n",
    "\n",
    "# Check if the SUMO_HOME path exists\n",
    "print(\"Does SUMO_HOME exist?\", os.path.exists(os.environ.get('SUMO_HOME')))\n",
    "\n",
    "\n",
    "# Define file paths\n",
    "net_file = 'C:/Users/ASUS/Sumo/2024-11-03-18-23-47/osm.net.xml/osm.net.xml'\n",
    "route_file = 'C:/Users/ASUS/Sumo/2024-11-03-18-23-47/osm.passenger.rou.xml'\n",
    "out_csv_name = 'C:/Users/ASUS/Sumo/2024-11-03-18-23-47/osm.passenger.csv'\n",
    "config_file = 'C:/Users/ASUS/Sumo/2024-11-03-18-23-47/osm.sumocfg'\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Importing Required libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import gymnasium as gym\n",
    "from tqdm import tqdm\n",
    "import os\n",
    "from dotenv import load_dotenv\n",
    "import sumo_rl\n",
    "import traci"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Environment Setup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Environment:\n",
    "    def __init__(self, env_name, route_file, net_file, out_csv_name, render_mode='human', num_seconds=100000):\n",
    "        self.env_name = env_name\n",
    "        self.use_gui = True if render_mode == 'human' else False\n",
    "        self.route_file = route_file\n",
    "        self.net_file = net_file\n",
    "        self.out_csv_name = out_csv_name\n",
    "        self.env = gym.make(\n",
    "                        env_name,\n",
    "                        net_file=net_file,\n",
    "                        route_file=route_file,\n",
    "                        out_csv_name=out_csv_name,\n",
    "                        use_gui=self.use_gui,\n",
    "                        num_seconds=num_seconds\n",
    "                    )\n",
    "     \n",
    "        self.state, _ = self.env.reset()\n",
    "        self.done = False\n",
    "        self.observation_space = self.env.observation_space.shape[0]\n",
    "        self.action_space = self.env.action_space\n",
    "        self.traffic_signals = self.env.traffic_signals\n",
    "    \n",
    "    def reset(self):\n",
    "        self.state, _ = self.env.reset()\n",
    "        self.done = False\n",
    "        return self.state\n",
    "    \n",
    "    def custom_reward(self, traffic_signal, reward_type='average_speed', reward_method='simple'):\n",
    "        # print(\"Inside custom reward method\")\n",
    "        if reward_method == 'simple':\n",
    "            match reward_type:\n",
    "                case 'average_speed':\n",
    "                    return traffic_signal.get_avgerage_speed()\n",
    "                case 'congesion':\n",
    "                    return -1 * traffic_signal.get_pressure()\n",
    "                case 'emissions':\n",
    "                    return -1* traffic_signal.get_emission_co2()\n",
    "                case 'throughput':\n",
    "                    return traffic_signal.get_throughput()\n",
    "\n",
    "        else:\n",
    "            # Weighted sum of the metrics\n",
    "            reward = 0\n",
    "            if weights is None:\n",
    "                weights = {\n",
    "                    'average_speed': 0.4,\n",
    "                    'waiting_time': 0.3,\n",
    "                    'emissions': 0.2,\n",
    "                    'throughput': 0.1\n",
    "                }\n",
    "\n",
    "            # Calculate individual rewards\n",
    "            average_speed = traffic_signal.get_average_speed()\n",
    "            waiting_time = -1* traffic_signal._diff_waiting_time_reward()\n",
    "            total_queue = -1 * traffic_signal.get_total_queued()\n",
    "            congesion = traffic_signal.get_pressure()\n",
    "\n",
    "            print(average_speed, waiting_time, total_queue, congesion)\n",
    "            weighted_reward = (\n",
    "                weights['average_speed'] * average_speed +\n",
    "                weights['waiting_time'] * waiting_time +\n",
    "                weights['emissions'] * total_queue +\n",
    "                weights['throughput'] * congesion\n",
    "            )\n",
    "\n",
    "            return weighted_reward\n",
    "            \n",
    "            \n",
    "    def step(self, action):\n",
    "        # print(\"Inside step method\")\n",
    "        next_state, reward, terminated, truncated, info = self.env.step(action)\n",
    "        # print(\"Step taken\", next_state, terminated, truncated, info)\n",
    "\n",
    "        # print(\"Traffic signals:\", list(self.traffic_signals.values())[0])\n",
    "        # traffic_signal = list(self.traffic_signals.values())[0]\n",
    "        # print(\"Pressure:\", traffic_signal.get_pressure())\n",
    "        # # print(traffic_signal.get_average_speed(), traffic_signal.get_total_queued(), traffic_signal._diff_waiting_time_reward(), traffic_signal.get_pressure())\n",
    "        # reward = self.custom_reward(traffic_signal, reward_type='congesion', reward_method='simple')\n",
    "        # print(\"Reward:\", reward)\n",
    "\n",
    "        self.state = next_state\n",
    "        self.done = terminated\n",
    "        return next_state, reward, self.done or truncated\n",
    "    \n",
    "    def render(self):\n",
    "        self.env.render()\n",
    "    \n",
    "    def close(self):\n",
    "        try:\n",
    "            self.env.close()\n",
    "            if traci.isLoaded():\n",
    "                traci.close()\n",
    "            print(\"Env and Traci closed successfully.\")\n",
    "        except Exception as e:\n",
    "            print(\"Error while closing the environment:\", e)\n",
    "    \n",
    "    def get_state(self):\n",
    "        return self.state\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Random Agent Run Basic"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Running Environment (Simple Intersection)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\gymnasium\\core.py:311: UserWarning: \u001b[33mWARN: env.traffic_signals to get variables from other wrappers is deprecated and will be removed in v1.0, to get this variable you can do `env.unwrapped.traffic_signals` for environment variables or `env.get_wrapper_attr('traffic_signals')` that will search the reminding wrappers.\u001b[0m\n",
      "  logger.warn(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Observation Space: 11\n",
      "Action Space: 2\n",
      "Initial State: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "action space Discrete(2)\n",
      "Action  1  Reward  0.0  Done  False observation [1.         0.         0.         0.05048805 0.         0.05283551\n",
      " 0.05283551 0.         0.         0.         0.        ]\n",
      "Action  1  Reward  0.0  Done  False observation [1.         0.         1.         0.05048805 0.05048805 0.15850651\n",
      " 0.15850651 0.         0.         0.         0.        ]\n",
      "Action  1  Reward  0.0  Done  False observation [0.         1.         0.         0.1009761  0.05048805 0.21134202\n",
      " 0.21134202 0.         0.         0.         0.        ]\n",
      "Action  0  Reward  -0.03  Done  False observation [0.         1.         1.         0.1009761  0.05048805 0.10567101\n",
      " 0.15850651 0.05048805 0.         0.         0.        ]\n",
      "Action  0  Reward  -0.010000000000000002  Done  False observation [1.         0.         0.         0.05048805 0.         0.15850651\n",
      " 0.15850651 0.         0.         0.05283551 0.05283551]\n",
      "Action  1  Reward  -0.18  Done  False observation [1.         0.         1.         0.05048805 0.05048805 0.21134202\n",
      " 0.21134202 0.         0.         0.10567101 0.10567101]\n",
      "Action  0  Reward  -0.23  Done  False observation [1.         0.         1.         0.1009761  0.05048805 0.21134202\n",
      " 0.26417753 0.         0.         0.10567101 0.15850651]\n",
      "Action  0  Reward  -0.31  Done  False observation [1.         0.         1.         0.05048805 0.05048805 0.26417753\n",
      " 0.31701303 0.         0.         0.15850651 0.21134202]\n",
      "Action  0  Reward  -0.3999999999999999  Done  False observation [1.         0.         1.         0.         0.05048805 0.31701303\n",
      " 0.31701303 0.         0.         0.21134202 0.26417753]\n",
      "Action  1  Reward  0.10999999999999988  Done  False observation [0.         1.         0.         0.05048805 0.05048805 0.31701303\n",
      " 0.36984855 0.         0.         0.10567101 0.10567101]\n",
      "Action  1  Reward  0.6200000000000001  Done  False observation [0.         1.         1.         0.05048805 0.05048805 0.31701303\n",
      " 0.31701303 0.05048805 0.         0.         0.        ]\n",
      "Action  0  Reward  -0.03999999999999998  Done  False observation [1.         0.         0.         0.         0.05048805 0.36984855\n",
      " 0.36984855 0.         0.         0.10567101 0.05283551]\n",
      "Action  0  Reward  -0.33000000000000007  Done  False observation [1.         0.         1.         0.05048805 0.         0.42268404\n",
      " 0.42268404 0.         0.         0.26417753 0.21134202]\n",
      "Action  0  Reward  -0.6099999999999999  Done  False observation [1.         0.         1.         0.1009761  0.05048805 0.47551954\n",
      " 0.47551954 0.         0.         0.36984855 0.36984855]\n",
      "Action  0  Reward  -0.7500000000000002  Done  False observation [1.         0.         1.         0.05048805 0.1009761  0.52835506\n",
      " 0.5811906  0.         0.         0.42268404 0.42268404]\n",
      "Action  0  Reward  -0.8899999999999997  Done  False observation [1.         0.         1.         0.05048805 0.1009761  0.5811906\n",
      " 0.63402605 0.         0.         0.47551954 0.52835506]\n",
      "Action  0  Reward  -1.0200000000000005  Done  False observation [1.         0.         1.         0.05048805 0.05048805 0.5811906\n",
      " 0.63402605 0.         0.         0.5811906  0.63402605]\n",
      "Action  1  Reward  -0.1899999999999995  Done  False observation [0.         1.         0.         0.05048805 0.1009761  0.63402605\n",
      " 0.6868616  0.         0.         0.42268404 0.47551954]\n",
      "Action  0  Reward  0.9199999999999999  Done  False observation [0.         1.         1.         0.1009761  0.15146415 0.52835506\n",
      " 0.63402605 0.05048805 0.05048805 0.21134202 0.26417753]\n",
      "Action  1  Reward  0.9199999999999999  Done  False observation [0.         1.         1.         0.15146415 0.15146415 0.47551954\n",
      " 0.52835506 0.05048805 0.05048805 0.05283551 0.15850651]\n",
      "Action  0  Reward  0.08000000000000007  Done  False observation [1.         0.         0.         0.1009761  0.15146415 0.52835506\n",
      " 0.5811906  0.         0.         0.05283551 0.05283551]\n",
      "Action  1  Reward  -0.27  Done  False observation [1.         0.         1.         0.1009761  0.05048805 0.63402605\n",
      " 0.6868616  0.         0.         0.21134202 0.15850651]\n",
      "Action  0  Reward  -0.56  Done  False observation [1.         0.         1.         0.05048805 0.1009761  0.6868616\n",
      " 0.7396971  0.         0.         0.36984855 0.36984855]\n",
      "Action  0  Reward  -0.9000000000000004  Done  False observation [1.         0.         1.         0.05048805 0.05048805 0.7396971\n",
      " 0.79253256 0.         0.         0.5811906  0.52835506]\n",
      "Action  1  Reward  -0.27999999999999936  Done  False observation [0.         1.         0.         0.1009761  0.05048805 0.6868616\n",
      " 0.7396971  0.05048805 0.         0.5811906  0.52835506]\n",
      "Action  1  Reward  0.5799999999999996  Done  False observation [0.         1.         1.         0.1009761  0.05048805 0.63402605\n",
      " 0.6868616  0.05048805 0.05048805 0.31701303 0.36984855]\n",
      "Action  0  Reward  -0.3500000000000001  Done  False observation [1.         0.         0.         0.05048805 0.05048805 0.7396971\n",
      " 0.79253256 0.         0.         0.21134202 0.31701303]\n",
      "Action  0  Reward  -0.5099999999999998  Done  False observation [1.         0.         1.         0.05048805 0.05048805 0.79253256\n",
      " 0.8453681  0.         0.         0.21134202 0.36984855]\n",
      "Action  1  Reward  0.29000000000000004  Done  False observation [0.         1.         0.         0.1009761  0.1009761  0.79253256\n",
      " 0.79253256 0.         0.         0.21134202 0.26417753]\n",
      "Action  0  Reward  0.9899999999999998  Done  False observation [0.         1.         1.         0.1009761  0.1009761  0.79253256\n",
      " 0.79253256 0.05048805 0.05048805 0.15850651 0.15850651]\n",
      "Action  0  Reward  0.3799999999999999  Done  False observation [1.         0.         0.         0.05048805 0.05048805 0.7396971\n",
      " 0.7396971  0.         0.         0.10567101 0.05283551]\n",
      "Action  1  Reward  -0.3599999999999999  Done  False observation [1.         0.         1.         0.05048805 0.05048805 0.79253256\n",
      " 0.79253256 0.         0.         0.31701303 0.26417753]\n",
      "Action  1  Reward  0.1499999999999999  Done  False observation [0.         1.         0.         0.05048805 0.05048805 0.7396971\n",
      " 0.7396971  0.         0.         0.21134202 0.21134202]\n",
      "Action  1  Reward  0.71  Done  False observation [0.         1.         1.         0.1009761  0.1009761  0.7396971\n",
      " 0.7396971  0.05048805 0.05048805 0.10567101 0.10567101]\n",
      "Action  1  Reward  1.0400000000000003  Done  False observation [0.         1.         1.         0.1009761  0.15146415 0.63402605\n",
      " 0.63402605 0.05048805 0.05048805 0.         0.05283551]\n",
      "Action  0  Reward  0.08999999999999986  Done  False observation [1.         0.         0.         0.05048805 0.1009761  0.6868616\n",
      " 0.6868616  0.         0.         0.05283551 0.05283551]\n",
      "Action  0  Reward  -0.24  Done  False observation [1.         0.         1.         0.         0.         0.6868616\n",
      " 0.6868616  0.         0.         0.15850651 0.21134202]\n",
      "Action  1  Reward  0.1200000000000001  Done  False observation [0.         1.         0.         0.05048805 0.         0.7396971\n",
      " 0.7396971  0.         0.         0.26417753 0.21134202]\n",
      "Action  1  Reward  0.43999999999999995  Done  False observation [0.         1.         1.         0.05048805 0.         0.7396971\n",
      " 0.7396971  0.         0.         0.10567101 0.10567101]\n",
      "Action  1  Reward  0.37  Done  False observation [0.         1.         1.         0.05048805 0.         0.6868616\n",
      " 0.63402605 0.         0.         0.05283551 0.10567101]\n",
      "Action  1  Reward  0.33  Done  False observation [0.         1.         1.         0.1009761  0.05048805 0.5811906\n",
      " 0.52835506 0.         0.         0.05283551 0.        ]\n",
      "Action  1  Reward  0.09  Done  False observation [0.         1.         1.         0.1009761  0.1009761  0.42268404\n",
      " 0.42268404 0.05048805 0.         0.         0.        ]\n",
      "Action  0  Reward  0.07  Done  False observation [1.         0.         0.         0.05048805 0.05048805 0.47551954\n",
      " 0.47551954 0.         0.         0.         0.        ]\n",
      "Action  1  Reward  -0.18999999999999997  Done  False observation [1.         0.         1.         0.05048805 0.         0.5811906\n",
      " 0.52835506 0.         0.         0.15850651 0.15850651]\n",
      "Action  1  Reward  -0.14  Done  False observation [0.         1.         0.         0.1009761  0.1009761  0.6868616\n",
      " 0.5811906  0.         0.         0.10567101 0.15850651]\n",
      "Action  1  Reward  0.14999999999999997  Done  False observation [0.         1.         1.         0.1009761  0.15146415 0.6868616\n",
      " 0.63402605 0.         0.         0.         0.05283551]\n",
      "Action  0  Reward  -0.09999999999999998  Done  False observation [1.         0.         0.         0.1009761  0.1009761  0.79253256\n",
      " 0.7396971  0.         0.         0.10567101 0.05283551]\n",
      "Action  0  Reward  -0.34  Done  False observation [1.         0.         1.         0.05048805 0.         0.8453681\n",
      " 0.79253256 0.         0.         0.21134202 0.26417753]\n",
      "Action  0  Reward  -0.6699999999999999  Done  False observation [1.         0.         1.         0.05048805 0.         0.8453681\n",
      " 0.8453681  0.         0.         0.47551954 0.36984855]\n",
      "Action  1  Reward  -0.45999999999999996  Done  False observation [0.         1.         0.         0.05048805 0.05048805 0.8453681\n",
      " 0.8982036  0.         0.         0.42268404 0.42268404]\n",
      "Action  0  Reward  -0.040000000000000036  Done  False observation [0.         1.         1.         0.1009761  0.1009761  0.79253256\n",
      " 0.8453681  0.         0.         0.31701303 0.42268404]\n",
      "Action  0  Reward  -0.5700000000000001  Done  False observation [1.         0.         0.         0.1009761  0.1009761  0.79253256\n",
      " 0.8453681  0.         0.         0.36984855 0.42268404]\n",
      "Action  0  Reward  -0.7399999999999998  Done  False observation [1.         0.         1.         0.1009761  0.1009761  0.79253256\n",
      " 0.8453681  0.         0.         0.31701303 0.42268404]\n",
      "Action  1  Reward  -0.08000000000000007  Done  False observation [0.         1.         0.         0.05048805 0.15146415 0.7396971\n",
      " 0.79253256 0.         0.05048805 0.21134202 0.26417753]\n",
      "Action  0  Reward  0.44999999999999973  Done  False observation [0.         1.         1.         0.15146415 0.15146415 0.7396971\n",
      " 0.7396971  0.         0.05048805 0.05283551 0.15850651]\n",
      "Action  0  Reward  0.28000000000000025  Done  False observation [1.         0.         0.         0.15146415 0.15146415 0.7396971\n",
      " 0.7396971  0.         0.         0.         0.15850651]\n",
      "Action  0  Reward  -0.29000000000000004  Done  False observation [1.         0.         1.         0.1009761  0.15146415 0.79253256\n",
      " 0.79253256 0.         0.         0.21134202 0.21134202]\n",
      "Action  1  Reward  0.06000000000000005  Done  False observation [0.         1.         0.         0.1009761  0.15146415 0.7396971\n",
      " 0.7396971  0.         0.         0.21134202 0.15850651]\n",
      "Action  0  Reward  0.48  Done  False observation [0.         1.         1.         0.15146415 0.15146415 0.7396971\n",
      " 0.7396971  0.05048805 0.1009761  0.10567101 0.10567101]\n",
      "Action  1  Reward  0.69  Done  False observation [0.         1.         1.         0.2019522  0.2019522  0.63402605\n",
      " 0.63402605 0.1009761  0.15146415 0.         0.05283551]\n",
      "Action  0  Reward  0.010000000000000009  Done  False observation [1.         0.         0.         0.15146415 0.2019522  0.6868616\n",
      " 0.6868616  0.         0.         0.05283551 0.05283551]\n",
      "Action  1  Reward  -0.020000000000000018  Done  False observation [1.         0.         1.         0.1009761  0.1009761  0.6868616\n",
      " 0.6868616  0.         0.         0.21134202 0.21134202]\n",
      "Action  0  Reward  -0.55  Done  False observation [1.         0.         1.         0.05048805 0.05048805 0.79253256\n",
      " 0.79253256 0.         0.         0.36984855 0.31701303]\n",
      "Action  1  Reward  -0.2400000000000002  Done  False observation [0.         1.         0.         0.1009761  0.05048805 0.8453681\n",
      " 0.8453681  0.         0.         0.42268404 0.36984855]\n",
      "Action  0  Reward  0.3200000000000003  Done  False observation [0.         1.         1.         0.1009761  0.05048805 0.79253256\n",
      " 0.79253256 0.05048805 0.         0.36984855 0.26417753]\n",
      "Action  1  Reward  0.5499999999999998  Done  False observation [0.         1.         1.         0.15146415 0.1009761  0.6868616\n",
      " 0.6868616  0.1009761  0.05048805 0.21134202 0.21134202]\n",
      "Action  0  Reward  -0.30999999999999983  Done  False observation [1.         0.         0.         0.1009761  0.05048805 0.6868616\n",
      " 0.6868616  0.         0.         0.15850651 0.15850651]\n",
      "Action  1  Reward  -0.27000000000000024  Done  False observation [1.         0.         1.         0.         0.05048805 0.6868616\n",
      " 0.6868616  0.         0.         0.15850651 0.21134202]\n",
      "Action  1  Reward  -0.06999999999999984  Done  False observation [0.         1.         0.         0.05048805 0.1009761  0.7396971\n",
      " 0.7396971  0.         0.         0.21134202 0.26417753]\n",
      "Action  1  Reward  0.42999999999999994  Done  False observation [0.         1.         1.         0.1009761  0.1009761  0.7396971\n",
      " 0.7396971  0.         0.         0.05283551 0.10567101]\n",
      "Action  1  Reward  0.6100000000000001  Done  False observation [0.         1.         1.         0.1009761  0.1009761  0.63402605\n",
      " 0.63402605 0.         0.05048805 0.05283551 0.        ]\n",
      "Action  0  Reward  0.0  Done  False observation [1.         0.         0.         0.05048805 0.1009761  0.6868616\n",
      " 0.6868616  0.         0.         0.         0.10567101]\n",
      "Action  1  Reward  -0.27  Done  False observation [1.         0.         1.         0.05048805 0.05048805 0.6868616\n",
      " 0.6868616  0.         0.         0.21134202 0.21134202]\n",
      "Action  0  Reward  -0.6099999999999999  Done  False observation [1.         0.         1.         0.05048805 0.05048805 0.7396971\n",
      " 0.79253256 0.         0.         0.42268404 0.36984855]\n",
      "Action  1  Reward  -0.3600000000000003  Done  False observation [0.         1.         0.         0.1009761  0.1009761  0.79253256\n",
      " 0.8453681  0.         0.         0.36984855 0.42268404]\n",
      "Action  0  Reward  0.33000000000000007  Done  False observation [0.         1.         1.         0.1009761  0.1009761  0.7396971\n",
      " 0.79253256 0.05048805 0.         0.31701303 0.31701303]\n",
      "Action  1  Reward  0.33000000000000007  Done  False observation [0.         1.         1.         0.2019522  0.2019522  0.6868616\n",
      " 0.6868616  0.05048805 0.05048805 0.26417753 0.15850651]\n",
      "Action  0  Reward  -0.33999999999999986  Done  False observation [1.         0.         0.         0.15146415 0.15146415 0.6868616\n",
      " 0.6868616  0.         0.         0.21134202 0.21134202]\n",
      "Action  1  Reward  -0.33000000000000007  Done  False observation [1.         0.         1.         0.1009761  0.05048805 0.6868616\n",
      " 0.6868616  0.         0.         0.21134202 0.26417753]\n",
      "Action  0  Reward  -0.6200000000000001  Done  False observation [1.         0.         1.         0.1009761  0.05048805 0.79253256\n",
      " 0.79253256 0.         0.         0.36984855 0.42268404]\n",
      "Action  0  Reward  -0.8900000000000001  Done  False observation [1.         0.         1.         0.1009761  0.05048805 0.8982036\n",
      " 0.8982036  0.         0.         0.52835506 0.52835506]\n",
      "Action  1  Reward  -0.3900000000000001  Done  False observation [0.         1.         0.         0.1009761  0.05048805 0.8982036\n",
      " 0.8982036  0.         0.         0.5811906  0.5811906 ]\n",
      "Action  0  Reward  0.15000000000000036  Done  False observation [0.         1.         1.         0.15146415 0.1009761  0.8453681\n",
      " 0.8453681  0.05048805 0.         0.47551954 0.47551954]\n",
      "Action  1  Reward  0.52  Done  False observation [0.         1.         1.         0.15146415 0.1009761  0.6868616\n",
      " 0.6868616  0.1009761  0.05048805 0.36984855 0.36984855]\n",
      "Action  1  Reward  0.52  Done  False observation [0.         1.         1.         0.15146415 0.15146415 0.5811906\n",
      " 0.52835506 0.1009761  0.05048805 0.15850651 0.10567101]\n",
      "Action  1  Reward  0.69  Done  False observation [0.         1.         1.         0.15146415 0.15146415 0.42268404\n",
      " 0.42268404 0.15146415 0.1009761  0.         0.        ]\n",
      "Action  0  Reward  0.18000000000000016  Done  False observation [1.         0.         0.         0.15146415 0.15146415 0.52835506\n",
      " 0.52835506 0.         0.         0.05283551 0.05283551]\n",
      "Action  1  Reward  -0.010000000000000231  Done  False observation [1.         0.         1.         0.05048805 0.05048805 0.63402605\n",
      " 0.63402605 0.         0.         0.21134202 0.21134202]\n",
      "Action  0  Reward  -0.5899999999999999  Done  False observation [1.         0.         1.         0.05048805 0.05048805 0.7396971\n",
      " 0.79253256 0.         0.         0.36984855 0.36984855]\n",
      "Action  0  Reward  -0.8999999999999999  Done  False observation [1.         0.         1.         0.1009761  0.1009761  0.8982036\n",
      " 0.8982036  0.         0.         0.5811906  0.52835506]\n",
      "Action  0  Reward  -1.2700000000000005  Done  False observation [1.         0.         1.         0.05048805 0.1009761  0.9510391\n",
      " 0.9510391  0.         0.         0.7396971  0.7396971 ]\n",
      "Action  1  Reward  -0.6699999999999999  Done  False observation [0.         1.         0.         0.05048805 0.1009761  0.9510391\n",
      " 0.9510391  0.         0.05048805 0.7396971  0.7396971 ]\n",
      "Action  1  Reward  0.10000000000000053  Done  False observation [0.         1.         1.         0.1009761  0.15146415 0.8453681\n",
      " 0.8453681  0.05048805 0.1009761  0.5811906  0.5811906 ]\n",
      "Action  0  Reward  -0.8000000000000007  Done  False observation [1.         0.         0.         0.05048805 0.1009761  0.8453681\n",
      " 0.8453681  0.         0.         0.36984855 0.47551954]\n",
      "Action  0  Reward  -0.6899999999999995  Done  False observation [1.         0.         1.         0.         0.05048805 0.8453681\n",
      " 0.8453681  0.         0.         0.36984855 0.36984855]\n",
      "Action  0  Reward  -0.6399999999999997  Done  False observation [1.         0.         1.         0.         0.         0.8453681\n",
      " 0.8453681  0.         0.         0.36984855 0.42268404]\n",
      "Action  0  Reward  -0.8800000000000008  Done  False observation [1.         0.         1.         0.05048805 0.05048805 0.9510391\n",
      " 0.9510391  0.         0.         0.47551954 0.5811906 ]\n",
      "Action  0  Reward  -1.2199999999999989  Done  False observation [1.        0.        1.        0.1009761 0.1009761 0.9510391 0.9510391\n",
      " 0.        0.        0.6868616 0.7396971]\n",
      "Action  0  Reward  -1.6100000000000012  Done  False observation [1.        0.        1.        0.1009761 0.1009761 1.        1.\n",
      " 0.        0.        0.8982036 1.       ]\n",
      "Action  1  Reward  -0.5499999999999989  Done  False observation [0.         1.         0.         0.05048805 0.05048805 0.9510391\n",
      " 0.9510391  0.         0.         0.8453681  0.8453681 ]\n",
      "Env and Traci closed successfully.\n"
     ]
    }
   ],
   "source": [
    "# Define paths to the network and route files (Buffalo Road Intersection)\n",
    "\n",
    "# net_file = 'C:/Users/ASUS/Sumo/2024-11-03-18-23-47/osm.net.xml/osm.net.xml'\n",
    "# route_file = 'C:/Users/ASUS/Sumo/2024-11-03-18-23-47/osm.passenger.rou.xml'\n",
    "# out_csv_name = 'C:/Users/ASUS/Sumo/2024-11-03-18-23-47/osm.passenger.csv'\n",
    "\n",
    "\n",
    "# Using Custom simple Traffic Intersection\n",
    "nets_dir = 'nets'\n",
    "\n",
    "file_name = 'single_intersection_simple'\n",
    "out_csv_name = f'{file_name}.csv'\n",
    "\n",
    "nets_file = os.path.join(nets_dir, f'{file_name}.net.xml')\n",
    "routes_file = os.path.join(nets_dir, f'{file_name}.rou.xml')\n",
    "out_csv_name = os.path.join(nets_dir, f'{file_name}.csv')\n",
    "\n",
    "file_exists = lambda file_path: os.path.exists(file_path)\n",
    "\n",
    "if not file_exists(nets_file):\n",
    "    raise FileNotFoundError(f\"Net file not found: {net_file}\")\n",
    "if not file_exists(routes_file):\n",
    "    raise FileNotFoundError(f\"Route file not found: {route_file}\")\n",
    "\n",
    "\n",
    "# Instantiate the environment\n",
    "sumo_env = Environment('sumo-rl-v0', net_file=nets_file, route_file=routes_file, out_csv_name=out_csv_name, render_mode=None, num_seconds=100000)\n",
    "\n",
    "print(\"Observation Space:\", sumo_env.observation_space)\n",
    "print(\"Action Space:\", sumo_env.action_space.n)\n",
    "print(\"Initial State:\", sumo_env.state)\n",
    "\n",
    "max_steps = 100\n",
    "\n",
    "try:\n",
    "    print(\"action space\", sumo_env.action_space)\n",
    "    for step in range(max_steps):\n",
    "        # sumo_env.render()\n",
    "        action = sumo_env.env.action_space.sample()\n",
    "        next_state, reward, done = sumo_env.step(action)\n",
    "        print(\"Action \", action, \" Reward \", reward, \" Done \", done, \"observation\", next_state)\n",
    "        \n",
    "        if done:\n",
    "            sumo_env.reset()\n",
    "            break\n",
    "\n",
    "except KeyboardInterrupt:\n",
    "    print(\"Keyboard interrupt detected. Closing the environment.\")\n",
    "\n",
    "except Exception as e:\n",
    "    print(\"Error occurred:\", e)\n",
    "\n",
    "finally:\n",
    "    sumo_env.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training DQN"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## DQN Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Setting up basic Neural Network for DQN\n",
    "from torch import nn\n",
    "\n",
    "input_nodes = 128\n",
    "\n",
    "class DQN(nn.Module):\n",
    "    def __init__(self, observation_size, action_size, input_nodes=128):\n",
    "        super(DQN, self).__init__()\n",
    "        print(\"Observation size Network\", observation_size)\n",
    "        print(\"Action size Network\", action_size)\n",
    "\n",
    "        self.fc1 = nn.Linear(observation_size, input_nodes)\n",
    "        self.fc2 = nn.Linear(input_nodes, input_nodes)\n",
    "        self.fc3 = nn.Linear(input_nodes, action_size)\n",
    "        \n",
    "    \n",
    "    def forward(self, x):\n",
    "        x = nn.functional.relu(self.fc1(x))\n",
    "        x = nn.functional.relu(self.fc2(x))\n",
    "        x = self.fc3(x)\n",
    "        return x"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Replay Memory"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Class for Replay Memory\n",
    "from collections import namedtuple, deque\n",
    "import random\n",
    "\n",
    "Transition = namedtuple('Transition', ('state', 'action', 'next_state', 'reward', 'done'))\n",
    "\n",
    "class ReplayMemory:\n",
    "    def __init__(self, capacity):\n",
    "        self.memory = deque([], maxlen=capacity)\n",
    "\n",
    "    def push(self, *args):\n",
    "        self.memory.append(Transition(*args))\n",
    "\n",
    "    def sample(self, batch_size):\n",
    "        return random.sample(self.memory, batch_size)\n",
    "\n",
    "    def __len__(self):\n",
    "        return len(self.memory)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## DQN Agent"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Setting up the Agent Class for doing the DQN \n",
    "class Agent:\n",
    "    def __init__(self, env, epsilon, gamma, learning_rate, epsilon_decay, mem_size=5000):\n",
    "        self.env = env\n",
    "        self.state_size = env.observation_space\n",
    "        self.action_size = env.action_space.n\n",
    "        self.epsilon = epsilon\n",
    "        self.gamma = gamma\n",
    "        self.learning_rate = learning_rate\n",
    "        self.min_epsilon = 0.01\n",
    "        self.max_epsilon = 1\n",
    "        self.epsilon_decay = epsilon_decay\n",
    "\n",
    "        self.memory_size_max = mem_size\n",
    "        self.memory_buffer = ReplayMemory(self.memory_size_max)\n",
    "        self.q_network = DQN(self.state_size, self.action_size)\n",
    "        self.q_target_network = DQN(self.state_size, self.action_size)\n",
    "        self.q_target_network.load_state_dict(self.q_network.state_dict())\n",
    "        self.loss_function = nn.MSELoss()\n",
    "        self.optimizer = torch.optim.Adam(self.q_network.parameters(), lr=self.learning_rate)\n",
    "    \n",
    "    def choose_action(self, state):\n",
    "        if np.random.random() < self.epsilon:\n",
    "            return np.random.choice(self.action_size)\n",
    "        else:\n",
    "            state = torch.FloatTensor(state).unsqueeze(0)    \n",
    "            with torch.no_grad():\n",
    "                q_values = self.q_network(state)\n",
    "                return torch.argmax(q_values).item()\n",
    "    \n",
    "    def update_experience(self, state, action, reward, next_state, done):\n",
    "        state = torch.FloatTensor(state).unsqueeze(0)\n",
    "        next_state = torch.FloatTensor(next_state).unsqueeze(0)\n",
    "        action = torch.LongTensor([action]).view(-1, 1)\n",
    "        reward = torch.FloatTensor([reward]).to(torch.float32)\n",
    "        done = torch.FloatTensor([done]).to(torch.float32)\n",
    "        self.memory_buffer.push(state, action, next_state, reward, done)\n",
    "\n",
    "        if self.memory_buffer.__len__() > self.memory_size_max:\n",
    "            self.memory_buffer.pop(0)\n",
    "\n",
    "    def update_exploration_probability(self, episode):\n",
    "        self.epsilon = self.min_epsilon + (self.max_epsilon - self.min_epsilon) * np.exp(-self.epsilon_decay * episode)\n",
    "    \n",
    "    def optimize(self, batch_size, sample_threshold=0.1):\n",
    "        if len(self.memory_buffer) < batch_size:\n",
    "            return\n",
    "        \n",
    "        if len(self.memory_buffer) < sample_threshold * self.memory_size_max:\n",
    "            return\n",
    "\n",
    "        transitions = self.memory_buffer.sample(batch_size)\n",
    "        batch = Transition(*zip(*transitions))\n",
    "        \n",
    "        # Convert the parameters to torch tensors\n",
    "        # state_batch = torch.cat(batch.state)\n",
    "        # action_batch = torch.tensor(batch.action).view(-1, 1)\n",
    "        # reward_batch = torch.tensor(batch.reward).float()\n",
    "        # next_state_batch = torch.cat(batch.next_state)\n",
    "\n",
    "        state_batch = torch.cat(batch.state)\n",
    "        action_batch = torch.cat(batch.action)\n",
    "        reward_batch = torch.cat(batch.reward)\n",
    "        next_state_batch = torch.cat(batch.next_state)\n",
    "        done_batch = torch.cat(batch.done)\n",
    "\n",
    "        curr_q_vals = self.q_network(state_batch)\n",
    "        curr_q_vals = curr_q_vals.gather(1, action_batch)\n",
    "\n",
    "        # update target q value and current q values, backpropagate the loss\n",
    "        with torch.no_grad():\n",
    "            next_q_vals = self.q_target_network(next_state_batch)\n",
    "            max_next_q_vals = torch.max(next_q_vals, 1)[0]\n",
    "            target_q_vals = reward_batch + self.gamma * max_next_q_vals * (1 - done_batch)\n",
    "\n",
    "        loss = self.loss_function(curr_q_vals.squeeze(), target_q_vals)\n",
    "        self.optimizer.zero_grad()\n",
    "        loss.backward()\n",
    "        # Handling the exploding gradient problem\n",
    "        nn.utils.clip_grad_norm_(self.q_network.parameters(), 1)\n",
    "        self.optimizer.step()\n",
    "\n",
    "        return loss.item()\n",
    "    \n",
    "    def update_target_network(self):\n",
    "        self.q_target_network.load_state_dict(self.q_network.state_dict())\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Trainer DQN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "def trainer_dqn(env, agent, n_episodes, batch_size, epsilon, gamma, learning_rate, epsilon_decay, C, max_steps=100, tau=0.005):\n",
    "    max_steps = 500\n",
    "    use_wandb = os.environ.get('USE_WANDB', 'false').lower() == 'true'\n",
    "    epsilon_start = 1\n",
    "    epsilon_end = 0.01\n",
    "\n",
    "    if use_wandb:\n",
    "        wandb_config = {\n",
    "            'env': env.env_name,\n",
    "            'algorithm': 'DQN',\n",
    "            'batch_size': batch_size,\n",
    "            'gamma': gamma,\n",
    "            'eps_start': epsilon_start,\n",
    "            'eps_end': epsilon_end,\n",
    "            'eps_decay': epsilon_decay,\n",
    "            'target_update': C,\n",
    "            'lr': learning_rate,\n",
    "            'num_episodes': n_episodes,\n",
    "            'max_timesteps': max_steps,\n",
    "            'seed': 0\n",
    "        }\n",
    "\n",
    "        method = 'DQN' if not agent.use_double_dqn else 'DoubleDQN'\n",
    "        \n",
    "        cust_wandb = WanDB(wandb_config, f'assignment-2-{env.env_name}', env.env_name)\n",
    "       \n",
    "    \n",
    "    loss_list = []\n",
    "    reward_list = []\n",
    "    rewards_per_episode = {}\n",
    "    epsilon_values = []\n",
    "\n",
    "    p_bar = tqdm(range(n_episodes), colour='green', desc='Training progress', unit='Episode')\n",
    "    \n",
    "\n",
    "    tot_steps = 0\n",
    "\n",
    "    try:\n",
    "        for episode in p_bar:\n",
    "            state = env.reset()\n",
    "            done = False\n",
    "            total_reward = 0\n",
    "            total_loss = 0            \n",
    "    \n",
    "            while not done:\n",
    "                    tot_steps += 1\n",
    "                    action = agent.choose_action(state)\n",
    "                    # print(\"Action \", action)\n",
    "                    next_state, reward, done = env.step(action)\n",
    "\n",
    "                    # print(\"Action \", action, \" Reward \", reward)\n",
    "                    # Save the transition in the replay memory\n",
    "                    agent.update_experience(state, action, reward, next_state, done)\n",
    "\n",
    "                    # Run the optimization step\n",
    "                    loss = agent.optimize(batch_size)\n",
    "                    if loss:\n",
    "                        if use_wandb:\n",
    "                            cust_wandb.wandb_log({\n",
    "                                    'loss': loss, \n",
    "                                    'step_reward': reward, \n",
    "                                    'step': tot_steps\n",
    "                                })\n",
    "                        \n",
    "                        total_loss += loss\n",
    "                    total_reward += reward\n",
    "                    state = next_state\n",
    "\n",
    "                    if done:\n",
    "                        break\n",
    "\n",
    "            if episode % C == 0:\n",
    "                if agent.use_double_dqn:\n",
    "                    agent.update_target_network(tau)\n",
    "                else:\n",
    "                    agent.update_target_network()\n",
    "\n",
    "            agent.update_exploration_probability(episode)\n",
    "            epsilon_values.append(agent.epsilon)\n",
    "\n",
    "            loss_list.append(total_loss)\n",
    "            reward_list.append(total_reward)\n",
    "            rewards_per_episode[episode] = total_reward\n",
    "\n",
    "            if use_wandb:\n",
    "                cust_wandb.wandb_log({\n",
    "                            'epsilon': agent.epsilon,\n",
    "                            'total_reward': total_reward,\n",
    "                            'epsilon': agent.epsilon,\n",
    "                            'total_loss': total_loss,\n",
    "                            'episode': episode\n",
    "                })\n",
    "    except traci.exceptions.FatalTraCIError as e:\n",
    "        print(\"TraCI error:\", e)\n",
    "        \n",
    "    except KeyboardInterrupt:\n",
    "        print(\"Keyboard interrupt detected. Closing the environment.\")\n",
    "\n",
    "\n",
    "    except Exception as e:\n",
    "        print(f\"Error occurred in episode {episode}, step {step_}: {e}\")\n",
    "        raise\n",
    "\n",
    "        \n",
    "    finally:\n",
    "        if use_wandb:\n",
    "            cust_wandb.wandb_finish()\n",
    "            print(f'Episode: {episode}, Total Reward: {total_reward}, Loss: {total_loss}')\n",
    "    env.close()        \n",
    "    return reward_list, loss_list, rewards_per_episode, epsilon_values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training Agent"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Observation Space: 11\n",
      "Action Space: Discrete(2)\n",
      "Observation size Network 11\n",
      "Action size Network 2\n",
      "Observation size Network 11\n",
      "Action size Network 2\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Training progress:   0%|\u001b[32m          \u001b[0m| 2/1000 [00:15<2:10:33,  7.85s/Episode]Exception ignored in: <function SumoEnvironment.__del__ at 0x000001BD1142C820>\n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sumo_rl\\environment\\env.py\", line 460, in __del__\n",
      "    self.close()\n",
      "  File \"c:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sumo_rl\\environment\\env.py\", line 450, in close\n",
      "    traci.close()\n",
      "  File \"c:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\main.py\", line 262, in close\n",
      "    connection.check().close(wait)\n",
      "  File \"c:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\connection.py\", line 397, in close\n",
      "    self._sendCmd(tc.CMD_CLOSE, None, None)\n",
      "  File \"c:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\connection.py\", line 232, in _sendCmd\n",
      "    return self._sendExact()\n",
      "  File \"c:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\connection.py\", line 130, in _sendExact\n",
      "    self._socket.send(length + self._string)\n",
      "ConnectionResetError: [WinError 10054] An existing connection was forcibly closed by the remote host\n",
      "Exception ignored in: <function SumoEnvironment.__del__ at 0x000001BD1142C820>\n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sumo_rl\\environment\\env.py\", line 460, in __del__\n",
      "    self.close()\n",
      "  File \"c:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sumo_rl\\environment\\env.py\", line 450, in close\n",
      "    traci.close()\n",
      "  File \"c:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\main.py\", line 262, in close\n",
      "    connection.check().close(wait)\n",
      "  File \"c:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\connection.py\", line 397, in close\n",
      "    self._sendCmd(tc.CMD_CLOSE, None, None)\n",
      "  File \"c:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\connection.py\", line 232, in _sendCmd\n",
      "    return self._sendExact()\n",
      "  File \"c:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\connection.py\", line 130, in _sendExact\n",
      "    self._socket.send(length + self._string)\n",
      "ConnectionResetError: [WinError 10054] An existing connection was forcibly closed by the remote host\n",
      "Exception ignored in: <function SumoEnvironment.__del__ at 0x000001BD1142C820>\n",
      "Traceback (most recent call last):\n",
      "  File \"c:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sumo_rl\\environment\\env.py\", line 460, in __del__\n",
      "    self.close()\n",
      "  File \"c:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sumo_rl\\environment\\env.py\", line 450, in close\n",
      "    traci.close()\n",
      "  File \"c:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\main.py\", line 262, in close\n",
      "    connection.check().close(wait)\n",
      "  File \"c:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\connection.py\", line 397, in close\n",
      "    self._sendCmd(tc.CMD_CLOSE, None, None)\n",
      "  File \"c:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\connection.py\", line 232, in _sendCmd\n",
      "    return self._sendExact()\n",
      "  File \"c:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\connection.py\", line 130, in _sendExact\n",
      "    self._socket.send(length + self._string)\n",
      "ConnectionAbortedError: [WinError 10053] An established connection was aborted by the software in your host machine\n",
      "Training progress: 100%|\u001b[32m██████████\u001b[0m| 1000/1000 [1:40:10<00:00,  6.01s/Episode]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Env and Traci closed successfully.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "# Trainer method\n",
    "\n",
    "n_episodes = 1000\n",
    "batch_size = 64\n",
    "epsilon = 1\n",
    "gamma = 0.99\n",
    "learning_rate = 1e-3\n",
    "epsilon_decay = 0.995\n",
    "C = 5\n",
    "num_seconds = 500\n",
    "\n",
    "nets_dir = 'nets'\n",
    "\n",
    "file_name = 'single_intersection_simple'\n",
    "\n",
    "nets_file = os.path.join(nets_dir, f'{file_name}.net.xml')\n",
    "routes_file = os.path.join(nets_dir, f'{file_name}.rou.xml')\n",
    "out_csv_name = os.path.join(nets_dir+\"/results_csv\", f'{file_name}.passenger.csv')\n",
    "\n",
    "file_exists = lambda file_path: os.path.exists(file_path)\n",
    "\n",
    "results_dir = os.path.join(nets_dir, 'results_csv')\n",
    "if os.path.exists(results_dir):\n",
    "    shutil.rmtree(results_dir)\n",
    "os.makedirs(results_dir, exist_ok=True)\n",
    "\n",
    "if not file_exists(nets_file):\n",
    "    raise FileNotFoundError(f\"Net file not found: {net_file}\")\n",
    "if not file_exists(routes_file):\n",
    "    raise FileNotFoundError(f\"Route file not found: {route_file}\")\n",
    "\n",
    "sumo_env = Environment('sumo-rl-v0', net_file=nets_file, route_file=routes_file, out_csv_name=out_csv_name, render_mode=None, num_seconds=num_seconds)\n",
    "\n",
    "\n",
    "print(\"Observation Space:\", sumo_env.observation_space)\n",
    "print(\"Action Space:\", sumo_env.action_space)\n",
    "# print(\"Initial State:\", sumo_env.state)\n",
    "\n",
    "agent = Agent(sumo_env, epsilon, gamma, learning_rate, epsilon_decay)\n",
    "\n",
    "rewards, losses, rewards_per_episode, epsilon_values = trainer_dqn(sumo_env, agent, n_episodes, batch_size, epsilon, gamma, learning_rate, epsilon_decay, C)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "# Graph plotting methods\n",
    "def epsilon_decay_plot(epsilon_values, environment_type):\n",
    "  # plt.figure(figsize=(6, 4))\n",
    "  plt.title(f'Epsilon Decay {environment_type}')\n",
    "  plt.plot(epsilon_values)\n",
    "  plt.xlabel('Episode', fontsize=18)\n",
    "  plt.ylabel('Epsilon Values', fontsize=18)\n",
    "  plt.title('Epsilon Decay', fontsize=26)\n",
    "  plt.xticks(fontsize=10)\n",
    "  plt.yticks(fontsize=10)\n",
    "  # plt.ylim(ymin=0, ymax=1)\n",
    "  # plt.xlim(xmin=0, xmax=500)\n",
    "  plt.grid()\n",
    "  plt.show()\n",
    "\n",
    "def rewards_per_episode_plot(rewards_per_ep, environment_type, epsilons=[]):\n",
    "  episodes = np.array(list(rewards_per_ep.keys())).flatten()\n",
    "  rewards = np.array(list(rewards_per_ep.values())).flatten()\n",
    "\n",
    "  plt.figure(figsize=(10, 6))\n",
    "  plt.plot(episodes, rewards, marker='o', linestyle='-', markersize=4)\n",
    "  plt.title(f'Rewards per Episode {environment_type}')\n",
    "  plt.xlabel('Episode')\n",
    "  plt.ylabel('Reward')\n",
    "  plt.grid(True)\n",
    "  plt.xticks(range(0, max(episodes)+1, max(1, len(episodes)//10)))\n",
    "\n",
    "  plt.tight_layout()\n",
    "  plt.show()\n",
    "\n",
    "def rewards_per_episode_plot_2(rewards_per_ep, environment_type, epsilons=None, window_size=10):\n",
    "    \"\"\"\n",
    "    Plots rewards per episode with optional smoothing.\n",
    "    \n",
    "    Args:\n",
    "        rewards_per_ep (dict): A dictionary where keys are episodes and values are rewards.\n",
    "        environment_type (str): The type of environment (for plot title).\n",
    "        epsilons (list, optional): Epsilon values per episode, if you'd like to include them in the plot. Default is None.\n",
    "        window_size (int, optional): The window size for moving average smoothing. Default is 10.\n",
    "    \"\"\"\n",
    "    episodes = np.array(list(rewards_per_ep.keys())).flatten()\n",
    "    rewards = np.array(list(rewards_per_ep.values())).flatten()\n",
    "\n",
    "    plt.figure(figsize=(10, 6))\n",
    "    plt.plot(episodes, rewards, marker='o', linestyle='-', markersize=4, label='Rewards per Episode')\n",
    "\n",
    "    # Moving average for smoothing\n",
    "    if window_size > 1:\n",
    "        moving_avg = np.convolve(rewards, np.ones(window_size)/window_size, mode='valid')\n",
    "        plt.plot(episodes[:len(moving_avg)], moving_avg, color='r', label=f'Moving Avg (window={window_size})')\n",
    "\n",
    "    plt.title(f'Rewards per Episode {environment_type}', fontsize=20)\n",
    "    plt.xlabel('Episode', fontsize=14)\n",
    "    plt.ylabel('Reward', fontsize=14)\n",
    "    plt.grid(True)\n",
    "    \n",
    "    # Include epsilon plot if provided\n",
    "    if epsilons:\n",
    "        ax2 = plt.gca().twinx()\n",
    "        ax2.plot(episodes, epsilons, color='g', alpha=0.6, linestyle='--', label='Epsilon')\n",
    "        ax2.set_ylabel('Epsilon', fontsize=14)\n",
    "        ax2.tick_params(axis='y', labelsize=12)\n",
    "\n",
    "    plt.legend()\n",
    "    plt.tight_layout()\n",
    "    plt.show()\n",
    "    return plt\n",
    "\n",
    "\n",
    "def plot_results_csv_per_step(folder_path=\"nets/results_csv\", plot_steps=False, plot_episodes=True):\n",
    "  folder_path = \"nets/results_csv\"\n",
    "\n",
    "  global_steps = []\n",
    "  global_cumulative_waiting_times = []\n",
    "\n",
    "  current_step = 0\n",
    "\n",
    "  global_episodes = []\n",
    "  episode_numbers = []\n",
    "  cumulative_waiting_times = []\n",
    "\n",
    "  for episode_idx, file_name in enumerate(sorted(os.listdir(folder_path)), start=1):\n",
    "      if file_name.endswith(\".csv\"):\n",
    "          file_path = os.path.join(folder_path, file_name)\n",
    "          data = pd.read_csv(file_path)\n",
    "\n",
    "          if plot_steps:\n",
    "            steps = data['step'] + current_step\n",
    "            cumulative_waiting_time = data['system_total_waiting_time'].cumsum()\n",
    "            \n",
    "            global_steps.extend(steps)\n",
    "            global_cumulative_waiting_times.extend(cumulative_waiting_time)\n",
    "\n",
    "            current_step += data['step'].iloc[-1] + 1\n",
    "\n",
    "          else:        \n",
    "            # Compute the total cumulative waiting time for the episode\n",
    "            total_waiting_time = data['system_total_waiting_time'].sum()\n",
    "            \n",
    "            # Append episode number and total waiting time\n",
    "            episode_numbers.append(episode_idx)\n",
    "            cumulative_waiting_times.append(total_waiting_time)\n",
    "\n",
    "  # Plot the data\n",
    "  plt.figure(figsize=(12, 6))\n",
    "  if plot_episodes:\n",
    "    plt.plot(episode_numbers, cumulative_waiting_times, label=\"Cumulative Waiting Time\", color='red')\n",
    "  if plot_steps:\n",
    "    plt.plot(global_steps, global_cumulative_waiting_times, label=\"Cumulative Waiting Time\", color='blue')\n",
    "\n",
    "  # Add title, labels, grid, and legend\n",
    "  title_text = \"Cumulative Waiting Time vs Total Steps Across Episodes\" if plot_steps else \"Cumulative Waiting Time vs Total Episodes\"\n",
    "  plt.title(title_text)\n",
    "  plt.xlabel(f\"Total {'Steps' if plot_steps else 'Episodes'}\")\n",
    "  plt.ylabel(\"Cumulative Waiting Time\")\n",
    "  plt.grid(True)\n",
    "  plt.legend()\n",
    "  plt.show()\n",
    "\n",
    "def plot_results_csv_per_episode(folder_path=\"nets/results_csv\"):\n",
    "  episode_numbers = []\n",
    "  cumulative_waiting_times = []\n",
    "\n",
    "  # Iterate over all files in the folder, sorted by filename for consistency\n",
    "  for episode_idx, file_name in enumerate(sorted(os.listdir(folder_path)), start=1):\n",
    "      if file_name.endswith(\".csv\"):\n",
    "          # Load the CSV file\n",
    "          file_path = os.path.join(folder_path, file_name)\n",
    "          data = pd.read_csv(file_path)\n",
    "          \n",
    "          # Compute the total cumulative waiting time for the episode\n",
    "          total_waiting_time = data['system_total_waiting_time'].sum()\n",
    "          \n",
    "          # Append episode number and total waiting time\n",
    "          episode_numbers.append(episode_idx)\n",
    "          cumulative_waiting_times.append(total_waiting_time)\n",
    "\n",
    "  # Plot the data\n",
    "  plt.figure(figsize=(12, 6))\n",
    "  plt.plot(episode_numbers, cumulative_waiting_times, marker='o', linestyle='-', color='b', alpha=0.8)\n",
    "\n",
    "  # Add title, labels, grid, and legend\n",
    "  plt.title(\"Total Waiting Time vs Episodes\")\n",
    "  plt.xlabel(\"Episode\")\n",
    "  plt.ylabel(\"Total Waiting Time (Cumulative Sum)\")\n",
    "  plt.grid(True)\n",
    "  plt.show()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average Rewards: -1.79598\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<module 'matplotlib.pyplot' from 'c:\\\\Users\\\\ASUS\\\\AppData\\\\Local\\\\Programs\\\\Python\\\\Python310\\\\lib\\\\site-packages\\\\matplotlib\\\\pyplot.py'>"
      ]
     },
     "execution_count": 153,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(\"Average Rewards:\", np.mean(rewards))\n",
    "rewards_per_episode_plot_2(rewards_per_episode, 'Single Intersection Simple', epsilon_values, window_size=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_results_csv_per_step(folder_path=\"nets/results_csv\", plot_steps=True, plot_episodes=False)\n",
    "plot_results_csv_per_episode(folder_path=\"nets/results_csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [],
   "source": [
    "torch.save(agent.q_network.state_dict(), 'sumo_dqn.pth')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Greedy Evaluation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [],
   "source": [
    "def greedy_agent(agent, model, env, n_episodes=100):\n",
    "    epsilon = 0 # For greedy, it chooses only exploitation\n",
    "    reward_list = []\n",
    "    rewards_per_episode = {}\n",
    "    epsilon_values = []\n",
    "\n",
    "    p_bar = tqdm(range(n_episodes), colour='red', desc='Testing Progress', unit='Episode')\n",
    "\n",
    "    model.eval()\n",
    "    total_rewards = []\n",
    "\n",
    "    for episode in p_bar:\n",
    "        state = env.reset()\n",
    "        done = False\n",
    "        episode_reward = 0\n",
    "\n",
    "        while not done:\n",
    "            state_tensor = torch.FloatTensor(state).unsqueeze(0)\n",
    "            with torch.no_grad():\n",
    "                q_values = model(state_tensor)\n",
    "                action = torch.argmax(q_values).item()\n",
    "\n",
    "            next_state, reward, done = env.step(action)\n",
    "            episode_reward += reward\n",
    "            state = next_state\n",
    "\n",
    "        reward_list.append(episode_reward)\n",
    "        rewards_per_episode[episode] = episode_reward\n",
    "\n",
    "    print(rewards_per_episode)\n",
    "    rewards_per_ep_array = np.array(list(rewards_per_episode.values())).flatten()\n",
    "    avg_rewards_over_eps = np.mean(rewards_per_ep_array)\n",
    "\n",
    "    return avg_rewards_over_eps, rewards_per_episode, epsilon_values\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ASUS\\AppData\\Local\\Temp\\ipykernel_38112\\690269224.py:2: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.\n",
      "  model.load_state_dict(torch.load('sumo_dqn.pth'))\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Observation size Network 11\n",
      "Action size Network 2\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Testing Progress:   0%|\u001b[31m          \u001b[0m| 0/100 [00:00<?, ?Episode/s]"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Testing Progress: 100%|\u001b[31m██████████\u001b[0m| 100/100 [09:30<00:00,  5.71s/Episode]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{0: -0.2, 1: -0.79, 2: -0.10000000000000012, 3: -0.36, 4: -0.2200000000000001, 5: -0.05000000000000003, 6: -0.14999999999999988, 7: -0.86, 8: -0.66, 9: -0.4899999999999999, 10: -0.45000000000000023, 11: -1.39, 12: -0.41, 13: -0.48, 14: -0.21000000000000002, 15: -0.34999999999999987, 16: -0.14000000000000012, 17: -0.26, 18: -0.45, 19: -0.66, 20: -0.15, 21: -0.6, 22: -0.34, 23: -0.42, 24: -0.73, 25: -0.18, 26: -0.1200000000000001, 27: -0.51, 28: -0.2699999999999999, 29: -0.06, 30: -0.3800000000000002, 31: -0.5699999999999998, 32: -0.33, 33: -0.7900000000000001, 34: -0.05, 35: -0.3900000000000001, 36: -0.22, 37: -1.1100000000000003, 38: -0.040000000000000216, 39: -0.16000000000000023, 40: -0.030000000000000006, 41: -0.52, 42: -0.9, 43: -0.68, 44: -0.27, 45: -0.78, 46: -0.55, 47: -0.5799999999999998, 48: -0.26, 49: -0.13, 50: -0.020000000000000004, 51: -0.19, 52: -0.55, 53: -0.22, 54: -0.87, 55: -0.6199999999999998, 56: -0.38, 57: -0.29, 58: -0.62, 59: -0.4, 60: -0.8000000000000002, 61: -0.4, 62: -0.1899999999999999, 63: -0.009999999999999981, 64: -0.40000000000000013, 65: -0.66, 66: -0.21999999999999997, 67: -0.63, 68: -0.3599999999999999, 69: -0.1100000000000001, 70: -0.32, 71: -0.41, 72: -0.8900000000000002, 73: -0.8, 74: -0.42, 75: -0.75, 76: -0.34, 77: -0.5799999999999998, 78: -0.73, 79: -0.47, 80: -0.36, 81: -0.04000000000000001, 82: -0.1700000000000001, 83: -18.94, 84: -0.010000000000000002, 85: -0.31, 86: -0.73, 87: -0.48, 88: -0.54, 89: -0.13, 90: -0.07999999999999999, 91: -0.43999999999999984, 92: -0.52, 93: -0.52, 94: -0.52, 95: -0.48, 96: -0.53, 97: -0.58, 98: -0.5799999999999998, 99: -89.0}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "model = DQN(sumo_env.observation_space, sumo_env.action_space.n)\n",
    "model.load_state_dict(torch.load('sumo_dqn.pth'))\n",
    "\n",
    "greedy_avg_rewards_over_eps, greedy_rewards_per_eps, epsilon_values = greedy_agent(agent,  model, sumo_env ,n_episodes=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Avg Rewards per episode for Greedy Agent:  -1.4941\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<module 'matplotlib.pyplot' from 'c:\\\\Users\\\\ASUS\\\\AppData\\\\Local\\\\Programs\\\\Python\\\\Python310\\\\lib\\\\site-packages\\\\matplotlib\\\\pyplot.py'>"
      ]
     },
     "execution_count": 152,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(\"Avg Rewards per episode for Greedy Agent: \", greedy_avg_rewards_over_eps)\n",
    "rewards_per_episode_plot_2(greedy_rewards_per_eps, 'Sumo-v0', window_size=5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training Double DQN"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Double DQN Agent"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Setting up the Agent Class for doing the DQN \n",
    "class DDQNAgent:\n",
    "    def __init__(self, env, epsilon, gamma, learning_rate, epsilon_decay, mem_size=5000, use_double_dqn=False):\n",
    "        self.env = env\n",
    "        self.state_size = env.observation_space\n",
    "        self.action_size = env.action_space.n\n",
    "        self.epsilon = epsilon\n",
    "        self.gamma = gamma\n",
    "        self.learning_rate = learning_rate\n",
    "        self.min_epsilon = 0.01\n",
    "        self.max_epsilon = 1\n",
    "        self.epsilon_decay = epsilon_decay\n",
    "        self.use_double_dqn = use_double_dqn\n",
    "\n",
    "        print(self.state_size, self.action_size)\n",
    "\n",
    "        self.memory_size_max = mem_size\n",
    "        self.memory_buffer = ReplayMemory(self.memory_size_max)\n",
    "        self.q_network = DQN(self.state_size, self.action_size)\n",
    "        self.q_target_network = DQN(self.state_size, self.action_size)\n",
    "        self.q_target_network.load_state_dict(self.q_network.state_dict())\n",
    "        self.loss_function = nn.MSELoss()\n",
    "        self.optimizer = torch.optim.Adam(self.q_network.parameters(), lr=self.learning_rate)\n",
    "    \n",
    "    def choose_action(self, state):\n",
    "        if np.random.random() < self.epsilon:\n",
    "            return np.random.choice(self.action_size)\n",
    "        else:\n",
    "            state = torch.FloatTensor(state).unsqueeze(0)    \n",
    "            with torch.no_grad():\n",
    "                q_values = self.q_network(state)\n",
    "                return torch.argmax(q_values).item()\n",
    "    \n",
    "    def update_experience(self, state, action, reward, next_state, done):\n",
    "        state = torch.FloatTensor(state).unsqueeze(0)\n",
    "        next_state = torch.FloatTensor(next_state).unsqueeze(0)\n",
    "        action = torch.LongTensor([action]).view(-1, 1)\n",
    "        reward = torch.FloatTensor([reward]).to(torch.float32)\n",
    "        done = torch.FloatTensor([done]).to(torch.float32)\n",
    "        self.memory_buffer.push(state, action, next_state, reward, done)\n",
    "\n",
    "        if self.memory_buffer.__len__() > self.memory_size_max:\n",
    "            self.memory_buffer.pop(0)\n",
    "\n",
    "    def update_exploration_probability(self, episode):\n",
    "        self.epsilon = self.min_epsilon + (self.max_epsilon - self.min_epsilon) * np.exp(-self.epsilon_decay * episode)\n",
    "    \n",
    "    def optimize(self, batch_size, sample_threshold=0.1):\n",
    "        if len(self.memory_buffer) < batch_size:\n",
    "            return\n",
    "        \n",
    "        if len(self.memory_buffer) < sample_threshold * self.memory_size_max:\n",
    "            return\n",
    "\n",
    "        transitions = self.memory_buffer.sample(batch_size)\n",
    "        batch = Transition(*zip(*transitions))\n",
    "        \n",
    "        # Convert the parameters to torch tensors\n",
    "        # state_batch = torch.cat(batch.state)\n",
    "        # action_batch = torch.tensor(batch.action).view(-1, 1)\n",
    "        # reward_batch = torch.tensor(batch.reward).float()\n",
    "        # next_state_batch = torch.cat(batch.next_state)\n",
    "\n",
    "        state_batch = torch.cat(batch.state)\n",
    "        action_batch = torch.cat(batch.action)\n",
    "        reward_batch = torch.cat(batch.reward)\n",
    "        next_state_batch = torch.cat(batch.next_state)\n",
    "        done_batch = torch.cat(batch.done)\n",
    "\n",
    "        curr_q_vals = self.q_network(state_batch)\n",
    "        curr_q_vals = curr_q_vals.gather(1, action_batch)\n",
    "\n",
    "        # update target q value and current q values, backpropagate the loss\n",
    "        with torch.no_grad():\n",
    "            if self.use_double_dqn:\n",
    "                next_q_vals_online = self.q_network(next_state_batch)\n",
    "                max_next_q_vals_online = torch.argmax(next_q_vals_online, dim=1, keepdim=True)\n",
    "                next_q_vals_target = self.q_target_network(next_state_batch).gather(1, max_next_q_vals_online)\n",
    "                target_q_vals = reward_batch + self.gamma * next_q_vals_target.squeeze() * (1 - done_batch)\n",
    "\n",
    "            else:\n",
    "                next_q_vals = self.q_target_network(next_state_batch)\n",
    "                max_next_q_vals = torch.max(next_q_vals, 1)[0]\n",
    "                target_q_vals = reward_batch + self.gamma * max_next_q_vals * (1 - done_batch)\n",
    "\n",
    "        loss = self.loss_function(curr_q_vals.squeeze(), target_q_vals)\n",
    "        self.optimizer.zero_grad()\n",
    "        loss.backward()\n",
    "        # Handling the exploding gradient problem\n",
    "        nn.utils.clip_grad_norm_(self.q_network.parameters(), 10)\n",
    "        self.optimizer.step()\n",
    "\n",
    "        return loss.item()\n",
    "    \n",
    "    # def update_target_network(self):\n",
    "    #     self.q_target_network.load_state_dict(self.q_network.state_dict())\n",
    "\n",
    "    def update_target_network(self, tau=0.005):\n",
    "        target_net_state_dict = self.q_target_network.state_dict()\n",
    "        policy_net_state_dict = self.q_network.state_dict()\n",
    "        counter = 0\n",
    "\n",
    "        target_net_state_dict = self.q_target_network.state_dict()\n",
    "        policy_net_state_dict = self.q_network.state_dict()\n",
    "\n",
    "        for key in policy_net_state_dict:\n",
    "            target_net_state_dict[key] = policy_net_state_dict[key] * tau + target_net_state_dict[key] * (1 - tau)\n",
    "\n",
    "        self.q_target_network.load_state_dict(target_net_state_dict)\n",
    "\n",
    "        # for q_weight, target_weight in zip(policy_net_state_dict.values(), target_net_state_dict.values()):\n",
    "        #     target_weight = target_weight * (1 - tau) + q_weight * tau\n",
    "        #     target_net_state_dict[counter] = target_weight\n",
    "        #     counter += 1\n",
    "\n",
    "        # self.q_target_network.load_state_dict(target_net_state_dict)\n",
    "\n",
    "        # for key in policy_net_state_dict:\n",
    "        #     target_net_state_dict[key] = policy_net_state_dict[key] * tau + target_net_state_dict[key] * (1 - tau)\n",
    "        # self.q_target_network.load_state_dict(target_net_state_dict)\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training DDQN Agent"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\gymnasium\\core.py:311: UserWarning: \u001b[33mWARN: env.traffic_signals to get variables from other wrappers is deprecated and will be removed in v1.0, to get this variable you can do `env.unwrapped.traffic_signals` for environment variables or `env.get_wrapper_attr('traffic_signals')` that will search the reminding wrappers.\u001b[0m\n",
      "  logger.warn(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Observation Space: 11\n",
      "Action Space: Discrete(2)\n",
      "11 2\n",
      "Observation size Network 11\n",
      "Action size Network 2\n",
      "Observation size Network 11\n",
      "Action size Network 2\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Training progress:   0%|\u001b[32m          \u001b[0m| 1/1000 [00:13<3:39:10, 13.16s/Episode]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Keyboard interrupt detected. Closing the environment.\n",
      "Error while closing the environment: Connection closed by SUMO.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "# Trainer method\n",
    "n_episodes = 1000\n",
    "batch_size = 64\n",
    "epsilon = 1\n",
    "gamma = 0.99\n",
    "learning_rate = 1e-3\n",
    "epsilon_decay = 0.995\n",
    "C = 10\n",
    "num_seconds = 500\n",
    "tau=0.005\n",
    "\n",
    "mem_size = 5000\n",
    "use_double_dqn = True\n",
    "\n",
    "nets_dir = 'nets'\n",
    "\n",
    "file_name = 'single_intersection_simple'\n",
    "out_csv_folder = 'ddqn_results_csv'\n",
    "\n",
    "nets_file = os.path.join(nets_dir, f'{file_name}.net.xml')\n",
    "routes_file = os.path.join(nets_dir, f'{file_name}.rou.xml')\n",
    "out_csv_name = os.path.join(nets_dir+\"/\"+out_csv_folder, f'{file_name}.passenger.csv')\n",
    "\n",
    "file_exists = lambda file_path: os.path.exists(file_path)\n",
    "\n",
    "results_dir = os.path.join(nets_dir, 'ddqn_results_csv')\n",
    "if os.path.exists(results_dir):\n",
    "    shutil.rmtree(results_dir)\n",
    "os.makedirs(results_dir, exist_ok=True)\n",
    "\n",
    "if not file_exists(nets_file):\n",
    "    raise FileNotFoundError(f\"Net file not found: {net_file}\")\n",
    "if not file_exists(routes_file):\n",
    "    raise FileNotFoundError(f\"Route file not found: {route_file}\")\n",
    "\n",
    "sumo_env_ddqn = Environment('sumo-rl-v0', net_file=nets_file, route_file=routes_file, out_csv_name=out_csv_name, render_mode=None, num_seconds=num_seconds)\n",
    "\n",
    "print(\"Observation Space:\", sumo_env.observation_space)\n",
    "print(\"Action Space:\", sumo_env.action_space)\n",
    "# print(\"Initial State:\", sumo_env.state)\n",
    "\n",
    "ddqn_agent = DDQNAgent(sumo_env_ddqn, epsilon, gamma, learning_rate, epsilon_decay, use_double_dqn=use_double_dqn, mem_size=mem_size)\n",
    "\n",
    "ddqn_rewards, ddqn_losses, ddqn_rewards_per_episode, ddqn_epsilon_values = trainer_dqn(sumo_env_ddqn, ddqn_agent, n_episodes, batch_size, epsilon, gamma, learning_rate, epsilon_decay, C, tau=tau)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training A3C"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Retrying in 1 seconds\n",
      "Observation Space: 11\n",
      "Action Space: Discrete(2)\n",
      "Initial State: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Training started...\n"
     ]
    },
    {
     "ename": "TypeError",
     "evalue": "cannot pickle '_thread.lock' object",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[1;32mIn[134], line 21\u001b[0m\n\u001b[0;32m     18\u001b[0m actions \u001b[38;5;241m=\u001b[39m sumo_env\u001b[38;5;241m.\u001b[39maction_space\u001b[38;5;241m.\u001b[39mn\n\u001b[0;32m     19\u001b[0m \u001b[38;5;66;03m# Train the model\u001b[39;00m\n\u001b[0;32m     20\u001b[0m \u001b[38;5;66;03m# train_a3c(sumo_env, max_steps=1000, num_episodes=10, gamma=0.99, lr=0.001, beta=0.01, num_processes=4)\u001b[39;00m\n\u001b[1;32m---> 21\u001b[0m \u001b[43mtrain_a3c\u001b[49m\u001b[43m(\u001b[49m\u001b[43msumo_env\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43minput_dims\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43m[\u001b[49m\u001b[43minput_dims\u001b[49m\u001b[43m]\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mn_actions\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mactions\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mn_episodes\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;241;43m10\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43muse_wandb\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mFalse\u001b[39;49;00m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mgrad_clip\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;241;43m1\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mC\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;241;43m10\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43menv_id\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m'\u001b[39;49m\u001b[38;5;124;43mSUMO-RL\u001b[39;49m\u001b[38;5;124;43m'\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mlr\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;241;43m1e-4\u001b[39;49m\u001b[43m)\u001b[49m\n",
      "File \u001b[1;32md:\\RL_Project\\ReinforcementLearning_FinalCourseProject\\a3c_cartpole.py:261\u001b[0m, in \u001b[0;36mtrain_a3c\u001b[1;34m(env, env_id, input_dims, n_actions, n_episodes, gamma, use_wandb, grad_clip, C, lr)\u001b[0m\n\u001b[0;32m    252\u001b[0m rewards_list \u001b[38;5;241m=\u001b[39m thread_manager\u001b[38;5;241m.\u001b[39mlist()\n\u001b[0;32m    254\u001b[0m workers \u001b[38;5;241m=\u001b[39m [Agent(global_actor_critic, optim, input_dims, n_actions, gamma,\n\u001b[0;32m    255\u001b[0m                  lr, i,\n\u001b[0;32m    256\u001b[0m                  global_ep_idx\u001b[38;5;241m=\u001b[39mglobal_ep,\n\u001b[0;32m    257\u001b[0m                  env_id\u001b[38;5;241m=\u001b[39menv_id,\n\u001b[0;32m    258\u001b[0m                  env\u001b[38;5;241m=\u001b[39menv,\n\u001b[0;32m    259\u001b[0m                  n_episodes\u001b[38;5;241m=\u001b[39mn_episodes, rewards_list\u001b[38;5;241m=\u001b[39mrewards_list, grad_clip\u001b[38;5;241m=\u001b[39mgrad_clip, C\u001b[38;5;241m=\u001b[39mC) \u001b[38;5;28;01mfor\u001b[39;00m i \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mrange\u001b[39m(mp\u001b[38;5;241m.\u001b[39mcpu_count())]\n\u001b[1;32m--> 261\u001b[0m [w\u001b[38;5;241m.\u001b[39mstart() \u001b[38;5;28;01mfor\u001b[39;00m w \u001b[38;5;129;01min\u001b[39;00m workers]\n\u001b[0;32m    262\u001b[0m [w\u001b[38;5;241m.\u001b[39mjoin() \u001b[38;5;28;01mfor\u001b[39;00m w \u001b[38;5;129;01min\u001b[39;00m workers]\n\u001b[0;32m    265\u001b[0m folder_path \u001b[38;5;241m=\u001b[39m \u001b[38;5;124m'\u001b[39m\u001b[38;5;124ma3c_models\u001b[39m\u001b[38;5;124m'\u001b[39m\n",
      "File \u001b[1;32md:\\RL_Project\\ReinforcementLearning_FinalCourseProject\\a3c_cartpole.py:261\u001b[0m, in \u001b[0;36m<listcomp>\u001b[1;34m(.0)\u001b[0m\n\u001b[0;32m    252\u001b[0m rewards_list \u001b[38;5;241m=\u001b[39m thread_manager\u001b[38;5;241m.\u001b[39mlist()\n\u001b[0;32m    254\u001b[0m workers \u001b[38;5;241m=\u001b[39m [Agent(global_actor_critic, optim, input_dims, n_actions, gamma,\n\u001b[0;32m    255\u001b[0m                  lr, i,\n\u001b[0;32m    256\u001b[0m                  global_ep_idx\u001b[38;5;241m=\u001b[39mglobal_ep,\n\u001b[0;32m    257\u001b[0m                  env_id\u001b[38;5;241m=\u001b[39menv_id,\n\u001b[0;32m    258\u001b[0m                  env\u001b[38;5;241m=\u001b[39menv,\n\u001b[0;32m    259\u001b[0m                  n_episodes\u001b[38;5;241m=\u001b[39mn_episodes, rewards_list\u001b[38;5;241m=\u001b[39mrewards_list, grad_clip\u001b[38;5;241m=\u001b[39mgrad_clip, C\u001b[38;5;241m=\u001b[39mC) \u001b[38;5;28;01mfor\u001b[39;00m i \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mrange\u001b[39m(mp\u001b[38;5;241m.\u001b[39mcpu_count())]\n\u001b[1;32m--> 261\u001b[0m [\u001b[43mw\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstart\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m \u001b[38;5;28;01mfor\u001b[39;00m w \u001b[38;5;129;01min\u001b[39;00m workers]\n\u001b[0;32m    262\u001b[0m [w\u001b[38;5;241m.\u001b[39mjoin() \u001b[38;5;28;01mfor\u001b[39;00m w \u001b[38;5;129;01min\u001b[39;00m workers]\n\u001b[0;32m    265\u001b[0m folder_path \u001b[38;5;241m=\u001b[39m \u001b[38;5;124m'\u001b[39m\u001b[38;5;124ma3c_models\u001b[39m\u001b[38;5;124m'\u001b[39m\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\multiprocessing\\process.py:121\u001b[0m, in \u001b[0;36mBaseProcess.start\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m    118\u001b[0m \u001b[38;5;28;01massert\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m _current_process\u001b[38;5;241m.\u001b[39m_config\u001b[38;5;241m.\u001b[39mget(\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mdaemon\u001b[39m\u001b[38;5;124m'\u001b[39m), \\\n\u001b[0;32m    119\u001b[0m        \u001b[38;5;124m'\u001b[39m\u001b[38;5;124mdaemonic processes are not allowed to have children\u001b[39m\u001b[38;5;124m'\u001b[39m\n\u001b[0;32m    120\u001b[0m _cleanup()\n\u001b[1;32m--> 121\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_popen \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_Popen\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[43m)\u001b[49m\n\u001b[0;32m    122\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_sentinel \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_popen\u001b[38;5;241m.\u001b[39msentinel\n\u001b[0;32m    123\u001b[0m \u001b[38;5;66;03m# Avoid a refcycle if the target function holds an indirect\u001b[39;00m\n\u001b[0;32m    124\u001b[0m \u001b[38;5;66;03m# reference to the process object (see bpo-30775)\u001b[39;00m\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\multiprocessing\\context.py:224\u001b[0m, in \u001b[0;36mProcess._Popen\u001b[1;34m(process_obj)\u001b[0m\n\u001b[0;32m    222\u001b[0m \u001b[38;5;129m@staticmethod\u001b[39m\n\u001b[0;32m    223\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21m_Popen\u001b[39m(process_obj):\n\u001b[1;32m--> 224\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43m_default_context\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget_context\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mProcess\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_Popen\u001b[49m\u001b[43m(\u001b[49m\u001b[43mprocess_obj\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\multiprocessing\\context.py:327\u001b[0m, in \u001b[0;36mSpawnProcess._Popen\u001b[1;34m(process_obj)\u001b[0m\n\u001b[0;32m    324\u001b[0m \u001b[38;5;129m@staticmethod\u001b[39m\n\u001b[0;32m    325\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21m_Popen\u001b[39m(process_obj):\n\u001b[0;32m    326\u001b[0m     \u001b[38;5;28;01mfrom\u001b[39;00m \u001b[38;5;21;01m.\u001b[39;00m\u001b[38;5;21;01mpopen_spawn_win32\u001b[39;00m \u001b[38;5;28;01mimport\u001b[39;00m Popen\n\u001b[1;32m--> 327\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mPopen\u001b[49m\u001b[43m(\u001b[49m\u001b[43mprocess_obj\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\multiprocessing\\popen_spawn_win32.py:93\u001b[0m, in \u001b[0;36mPopen.__init__\u001b[1;34m(self, process_obj)\u001b[0m\n\u001b[0;32m     91\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[0;32m     92\u001b[0m     reduction\u001b[38;5;241m.\u001b[39mdump(prep_data, to_child)\n\u001b[1;32m---> 93\u001b[0m     \u001b[43mreduction\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mdump\u001b[49m\u001b[43m(\u001b[49m\u001b[43mprocess_obj\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mto_child\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m     94\u001b[0m \u001b[38;5;28;01mfinally\u001b[39;00m:\n\u001b[0;32m     95\u001b[0m     set_spawning_popen(\u001b[38;5;28;01mNone\u001b[39;00m)\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\multiprocessing\\reduction.py:60\u001b[0m, in \u001b[0;36mdump\u001b[1;34m(obj, file, protocol)\u001b[0m\n\u001b[0;32m     58\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mdump\u001b[39m(obj, file, protocol\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m):\n\u001b[0;32m     59\u001b[0m \u001b[38;5;250m    \u001b[39m\u001b[38;5;124;03m'''Replacement for pickle.dump() using ForkingPickler.'''\u001b[39;00m\n\u001b[1;32m---> 60\u001b[0m     \u001b[43mForkingPickler\u001b[49m\u001b[43m(\u001b[49m\u001b[43mfile\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mprotocol\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mdump\u001b[49m\u001b[43m(\u001b[49m\u001b[43mobj\u001b[49m\u001b[43m)\u001b[49m\n",
      "\u001b[1;31mTypeError\u001b[0m: cannot pickle '_thread.lock' object"
     ]
    }
   ],
   "source": [
    "from a3c_sumo import train_a3c\n",
    "\n",
    "nets_dir = 'nets'\n",
    "\n",
    "file_name = 'single_intersection_simple'\n",
    "\n",
    "nets_file = os.path.join(nets_dir, f'{file_name}.net.xml')\n",
    "routes_file = os.path.join(nets_dir, f'{file_name}.rou.xml')\n",
    "\n",
    "num_seconds = 1000\n",
    "n_episodes = 10\n",
    "sumo_env = Environment('sumo-rl-v0', net_file=nets_file, route_file=routes_file, out_csv_name=out_csv_name, render_mode=None, num_seconds=num_seconds)\n",
    "\n",
    "print(\"Observation Space:\", sumo_env.observation_space)\n",
    "print(\"Action Space:\", sumo_env.action_space)\n",
    "print(\"Initial State:\", sumo_env.state)\n",
    "\n",
    "# max_steps = 1000\n",
    "\n",
    "input_dims = sumo_env.observation_space\n",
    "actions = sumo_env.action_space.n\n",
    "# Train the models\n",
    "# train_a3c(sumo_env, max_steps=1000, num_episodes=10, gamma=0.99, lr=0.001, beta=0.01, num_processes=4)\n",
    "train_a3c(sumo_env, input_dims=[input_dims], n_actions=actions, n_episodes=n_episodes, use_wandb=False, grad_clip=1, C=10, env_id='SUMO-RL', lr=1e-4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SUMO_HOME is set to: C:\\Program Files (x86)\\Eclipse\\Sumo\n",
      " Retrying in 1 seconds\n",
      "Discrete(2)\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0.        1.        1.        0.0400128 0.        0.        0.\n",
      " 0.        0.        0.        0.       ]\n",
      "Reward: 0.0\n",
      "Observation: [0.        1.        1.        0.0400128 0.        0.        0.\n",
      " 0.        0.        0.        0.       ]\n",
      "Reward: 0.0\n",
      "Observation: [0.        1.        1.        0.0400128 0.        0.        0.\n",
      " 0.        0.        0.        0.       ]\n",
      "Reward: 0.0\n",
      "Observation: [1.        0.        0.        0.0400128 0.        0.        0.\n",
      " 0.        0.        0.        0.       ]\n",
      "Reward: -0.04\n",
      "Observation: [1.        0.        1.        0.0400128 0.        0.        0.\n",
      " 0.0400128 0.        0.        0.       ]\n",
      "Reward: 0.04\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1.         0.         1.         0.         0.         0.06479482\n",
      " 0.         0.         0.         0.06479482 0.        ]\n",
      "Reward: 0.0\n",
      "Observation: [0.         1.         0.         0.         0.         0.06479482\n",
      " 0.         0.         0.         0.         0.        ]\n",
      "Reward: 0.0\n",
      "Observation: [0.         1.         1.         0.         0.         0.06479482\n",
      " 0.         0.         0.         0.         0.        ]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0.         1.         1.         0.         0.         0.06479482\n",
      " 0.         0.         0.         0.         0.        ]\n",
      "Reward: 0.0\n",
      "Observation: [0.         1.         1.         0.         0.         0.06479482\n",
      " 0.         0.         0.         0.         0.        ]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0.        1.        1.        0.0400128 0.        0.        0.\n",
      " 0.        0.        0.        0.       ]\n",
      "Reward: 0.0\n",
      "Observation: [0.        1.        1.        0.0400128 0.        0.        0.\n",
      " 0.        0.        0.        0.       ]\n",
      "Reward: 0.0\n",
      "Observation: [0.        1.        1.        0.0400128 0.        0.        0.\n",
      " 0.        0.        0.        0.       ]\n",
      "Reward: 0.0\n",
      "Observation: [1.        0.        0.        0.0400128 0.        0.        0.\n",
      " 0.        0.        0.        0.       ]\n",
      "Reward: -0.04\n",
      "Observation: [1.        0.        1.        0.0400128 0.        0.        0.\n",
      " 0.0400128 0.        0.        0.       ]\n",
      "Reward: -0.049999999999999996\n",
      "Observation: [1.        0.        1.        0.0400128 0.        0.        0.\n",
      " 0.0400128 0.        0.        0.       ]\n",
      "Reward: 0.09\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0.        1.        1.        0.0400128 0.        0.        0.\n",
      " 0.        0.        0.        0.       ]\n",
      "Reward: 0.0\n",
      "Observation: [1.        0.        0.        0.0400128 0.        0.        0.\n",
      " 0.        0.        0.        0.       ]\n",
      "Reward: 0.0\n",
      "Observation: [1.        0.        1.        0.0400128 0.        0.        0.\n",
      " 0.        0.        0.        0.       ]\n",
      "Reward: -0.03\n",
      "Observation: [1.        0.        1.        0.0400128 0.        0.        0.\n",
      " 0.0400128 0.        0.        0.       ]\n",
      "Reward: 0.03\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0.         1.         1.         0.         0.         0.\n",
      " 0.05614193 0.         0.         0.         0.        ]\n",
      "Reward: 0.0\n",
      "Observation: [0.         1.         1.         0.         0.         0.\n",
      " 0.05614193 0.         0.         0.         0.        ]\n",
      "Reward: 0.0\n",
      "Observation: [0.         1.         1.         0.         0.         0.\n",
      " 0.05614193 0.         0.         0.         0.        ]\n",
      "Reward: -0.05\n",
      "Observation: [0.         1.         1.         0.         0.         0.\n",
      " 0.05614193 0.         0.         0.         0.05614193]\n",
      "Reward: 0.05\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [1. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      "Reward: 0.0\n",
      "Observation: [0. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "Cell \u001b[1;32mIn[62], line 25\u001b[0m\n\u001b[0;32m     22\u001b[0m \u001b[38;5;28mprint\u001b[39m(env\u001b[38;5;241m.\u001b[39maction_space)\n\u001b[0;32m     24\u001b[0m \u001b[38;5;28;01mwhile\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m done:\n\u001b[1;32m---> 25\u001b[0m     next_obs, reward, terminated, truncated, info \u001b[38;5;241m=\u001b[39m \u001b[43menv\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstep\u001b[49m\u001b[43m(\u001b[49m\u001b[43menv\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43maction_space\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43msample\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m     26\u001b[0m     \u001b[38;5;28mprint\u001b[39m(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mReward:\u001b[39m\u001b[38;5;124m\"\u001b[39m, reward)\n\u001b[0;32m     27\u001b[0m     \u001b[38;5;28mprint\u001b[39m(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mObservation:\u001b[39m\u001b[38;5;124m\"\u001b[39m, next_obs)\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\gymnasium\\wrappers\\order_enforcing.py:56\u001b[0m, in \u001b[0;36mOrderEnforcing.step\u001b[1;34m(self, action)\u001b[0m\n\u001b[0;32m     54\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_has_reset:\n\u001b[0;32m     55\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m ResetNeeded(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mCannot call env.step() before calling env.reset()\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[1;32m---> 56\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43menv\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstep\u001b[49m\u001b[43m(\u001b[49m\u001b[43maction\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\gymnasium\\wrappers\\env_checker.py:51\u001b[0m, in \u001b[0;36mPassiveEnvChecker.step\u001b[1;34m(self, action)\u001b[0m\n\u001b[0;32m     49\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m env_step_passive_checker(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39menv, action)\n\u001b[0;32m     50\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m---> 51\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43menv\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mstep\u001b[49m\u001b[43m(\u001b[49m\u001b[43maction\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sumo_rl\\environment\\env.py:320\u001b[0m, in \u001b[0;36mSumoEnvironment.step\u001b[1;34m(self, action)\u001b[0m\n\u001b[0;32m    318\u001b[0m terminated \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mFalse\u001b[39;00m  \u001b[38;5;66;03m# there are no 'terminal' states in this environment\u001b[39;00m\n\u001b[0;32m    319\u001b[0m truncated \u001b[38;5;241m=\u001b[39m dones[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124m__all__\u001b[39m\u001b[38;5;124m\"\u001b[39m]  \u001b[38;5;66;03m# episode ends when sim_step >= max_steps\u001b[39;00m\n\u001b[1;32m--> 320\u001b[0m info \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_compute_info\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    322\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msingle_agent:\n\u001b[0;32m    323\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m observations[\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mts_ids[\u001b[38;5;241m0\u001b[39m]], rewards[\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mts_ids[\u001b[38;5;241m0\u001b[39m]], terminated, truncated, info\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sumo_rl\\environment\\env.py:361\u001b[0m, in \u001b[0;36mSumoEnvironment._compute_info\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m    359\u001b[0m     info\u001b[38;5;241m.\u001b[39mupdate(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_get_system_info())\n\u001b[0;32m    360\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39madd_per_agent_info:\n\u001b[1;32m--> 361\u001b[0m     info\u001b[38;5;241m.\u001b[39mupdate(\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_get_per_agent_info\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m)\n\u001b[0;32m    362\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mmetrics\u001b[38;5;241m.\u001b[39mappend(info\u001b[38;5;241m.\u001b[39mcopy())\n\u001b[0;32m    363\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m info\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sumo_rl\\environment\\env.py:430\u001b[0m, in \u001b[0;36mSumoEnvironment._get_per_agent_info\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m    428\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21m_get_per_agent_info\u001b[39m(\u001b[38;5;28mself\u001b[39m):\n\u001b[0;32m    429\u001b[0m     stopped \u001b[38;5;241m=\u001b[39m [\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mtraffic_signals[ts]\u001b[38;5;241m.\u001b[39mget_total_queued() \u001b[38;5;28;01mfor\u001b[39;00m ts \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mts_ids]\n\u001b[1;32m--> 430\u001b[0m     accumulated_waiting_time \u001b[38;5;241m=\u001b[39m [\n\u001b[0;32m    431\u001b[0m         \u001b[38;5;28msum\u001b[39m(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mtraffic_signals[ts]\u001b[38;5;241m.\u001b[39mget_accumulated_waiting_time_per_lane()) \u001b[38;5;28;01mfor\u001b[39;00m ts \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mts_ids\n\u001b[0;32m    432\u001b[0m     ]\n\u001b[0;32m    433\u001b[0m     average_speed \u001b[38;5;241m=\u001b[39m [\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mtraffic_signals[ts]\u001b[38;5;241m.\u001b[39mget_average_speed() \u001b[38;5;28;01mfor\u001b[39;00m ts \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mts_ids]\n\u001b[0;32m    434\u001b[0m     info \u001b[38;5;241m=\u001b[39m {}\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sumo_rl\\environment\\env.py:431\u001b[0m, in \u001b[0;36m<listcomp>\u001b[1;34m(.0)\u001b[0m\n\u001b[0;32m    428\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21m_get_per_agent_info\u001b[39m(\u001b[38;5;28mself\u001b[39m):\n\u001b[0;32m    429\u001b[0m     stopped \u001b[38;5;241m=\u001b[39m [\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mtraffic_signals[ts]\u001b[38;5;241m.\u001b[39mget_total_queued() \u001b[38;5;28;01mfor\u001b[39;00m ts \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mts_ids]\n\u001b[0;32m    430\u001b[0m     accumulated_waiting_time \u001b[38;5;241m=\u001b[39m [\n\u001b[1;32m--> 431\u001b[0m         \u001b[38;5;28msum\u001b[39m(\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mtraffic_signals\u001b[49m\u001b[43m[\u001b[49m\u001b[43mts\u001b[49m\u001b[43m]\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget_accumulated_waiting_time_per_lane\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m) \u001b[38;5;28;01mfor\u001b[39;00m ts \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mts_ids\n\u001b[0;32m    432\u001b[0m     ]\n\u001b[0;32m    433\u001b[0m     average_speed \u001b[38;5;241m=\u001b[39m [\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mtraffic_signals[ts]\u001b[38;5;241m.\u001b[39mget_average_speed() \u001b[38;5;28;01mfor\u001b[39;00m ts \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mts_ids]\n\u001b[0;32m    434\u001b[0m     info \u001b[38;5;241m=\u001b[39m {}\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\sumo_rl\\environment\\traffic_signal.py:223\u001b[0m, in \u001b[0;36mTrafficSignal.get_accumulated_waiting_time_per_lane\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m    221\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m veh \u001b[38;5;129;01min\u001b[39;00m veh_list:\n\u001b[0;32m    222\u001b[0m     veh_lane \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msumo\u001b[38;5;241m.\u001b[39mvehicle\u001b[38;5;241m.\u001b[39mgetLaneID(veh)\n\u001b[1;32m--> 223\u001b[0m     acc \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43msumo\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mvehicle\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mgetAccumulatedWaitingTime\u001b[49m\u001b[43m(\u001b[49m\u001b[43mveh\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    224\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m veh \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39menv\u001b[38;5;241m.\u001b[39mvehicles:\n\u001b[0;32m    225\u001b[0m         \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39menv\u001b[38;5;241m.\u001b[39mvehicles[veh] \u001b[38;5;241m=\u001b[39m {veh_lane: acc}\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\_vehicle.py:575\u001b[0m, in \u001b[0;36mVehicleDomain.getAccumulatedWaitingTime\u001b[1;34m(self, vehID)\u001b[0m\n\u001b[0;32m    570\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mgetAccumulatedWaitingTime\u001b[39m(\u001b[38;5;28mself\u001b[39m, vehID):\n\u001b[0;32m    571\u001b[0m \u001b[38;5;250m    \u001b[39m\u001b[38;5;124;03m\"\"\"getAccumulatedWaitingTime(string) -> double\u001b[39;00m\n\u001b[0;32m    572\u001b[0m \u001b[38;5;124;03m    The accumulated waiting time of a vehicle collects the vehicle's waiting time\u001b[39;00m\n\u001b[0;32m    573\u001b[0m \u001b[38;5;124;03m    over a certain time interval (interval length is set per option '--waiting-time-memory')\u001b[39;00m\n\u001b[0;32m    574\u001b[0m \u001b[38;5;124;03m    \"\"\"\u001b[39;00m\n\u001b[1;32m--> 575\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_getUniversal\u001b[49m\u001b[43m(\u001b[49m\u001b[43mtc\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mVAR_ACCUMULATED_WAITING_TIME\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mvehID\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\domain.py:149\u001b[0m, in \u001b[0;36mDomain._getUniversal\u001b[1;34m(self, varID, objectID, format, *values)\u001b[0m\n\u001b[0;32m    147\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_deprecatedFor:\n\u001b[0;32m    148\u001b[0m     warnings\u001b[38;5;241m.\u001b[39mwarn(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mThe domain \u001b[39m\u001b[38;5;132;01m%s\u001b[39;00m\u001b[38;5;124m is deprecated, use \u001b[39m\u001b[38;5;132;01m%s\u001b[39;00m\u001b[38;5;124m instead.\u001b[39m\u001b[38;5;124m\"\u001b[39m \u001b[38;5;241m%\u001b[39m (\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_name, \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_deprecatedFor))\n\u001b[1;32m--> 149\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m _parse(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_retValFunc, varID, \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_getCmd\u001b[49m\u001b[43m(\u001b[49m\u001b[43mvarID\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mobjectID\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mformat\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mvalues\u001b[49m\u001b[43m)\u001b[49m)\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\domain.py:154\u001b[0m, in \u001b[0;36mDomain._getCmd\u001b[1;34m(self, varID, objID, format, *values)\u001b[0m\n\u001b[0;32m    152\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_connection \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[0;32m    153\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m FatalTraCIError(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mNot connected.\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[1;32m--> 154\u001b[0m r \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_connection\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_sendCmd\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_cmdGetID\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mvarID\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mobjID\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mformat\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mvalues\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    155\u001b[0m r\u001b[38;5;241m.\u001b[39mreadLength()\n\u001b[0;32m    156\u001b[0m response, retVarID \u001b[38;5;241m=\u001b[39m r\u001b[38;5;241m.\u001b[39mread(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124m!BB\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\connection.py:232\u001b[0m, in \u001b[0;36mConnection._sendCmd\u001b[1;34m(self, cmdID, varID, objID, format, *values)\u001b[0m\n\u001b[0;32m    230\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_string \u001b[38;5;241m+\u001b[39m\u001b[38;5;241m=\u001b[39m struct\u001b[38;5;241m.\u001b[39mpack(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124m!i\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;28mlen\u001b[39m(objID)) \u001b[38;5;241m+\u001b[39m objID\n\u001b[0;32m    231\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_string \u001b[38;5;241m+\u001b[39m\u001b[38;5;241m=\u001b[39m packed\n\u001b[1;32m--> 232\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_sendExact\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\connection.py:130\u001b[0m, in \u001b[0;36mConnection._sendExact\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m    128\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m _DEBUG:\n\u001b[0;32m    129\u001b[0m     \u001b[38;5;28mprint\u001b[39m(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124msending\u001b[39m\u001b[38;5;124m\"\u001b[39m, Storage(length \u001b[38;5;241m+\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_string)\u001b[38;5;241m.\u001b[39mgetDebugString())\n\u001b[1;32m--> 130\u001b[0m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_socket\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43msend\u001b[49m\u001b[43m(\u001b[49m\u001b[43mlength\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m+\u001b[39;49m\u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_string\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    131\u001b[0m result \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_recvExact()\n\u001b[0;32m    132\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m _DEBUG:\n",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "import gymnasium as gym\n",
    "from sumo_rl import SumoEnvironment\n",
    "import os\n",
    "\n",
    "# Print SUMO_HOME for confirmation\n",
    "print(\"SUMO_HOME is set to:\", os.environ['SUMO_HOME'])\n",
    "\n",
    "# Initialize SUMO Gym environment\n",
    "env = gym.make(\n",
    "    'sumo-rl-v0',\n",
    "    net_file=net_file,\n",
    "    route_file=route_file,\n",
    "    out_csv_name=out_csv_name,\n",
    "    use_gui=True,\n",
    "    num_seconds=100000\n",
    ")\n",
    "\n",
    "# Reset the environment and start simulation\n",
    "obs, info = env.reset()\n",
    "done = False\n",
    "\n",
    "print(env.action_space)\n",
    "\n",
    "while not done:\n",
    "    next_obs, reward, terminated, truncated, info = env.step(env.action_space.sample())\n",
    "    print(\"Reward:\", reward)\n",
    "    print(\"Observation:\", next_obs)\n",
    "    done = terminated or truncated\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Edge IDs: ['n_t', 't_e', 't_s', 'w_t']\n",
      "Traffic Light IDs: ['t']\n"
     ]
    }
   ],
   "source": [
    "import sumolib\n",
    "\n",
    "# Load the network file\n",
    "net = sumolib.net.readNet(nets_file)  # Replace with your network file\n",
    "\n",
    "# Get all edge IDs\n",
    "edge_ids = [edge.getID() for edge in net.getEdges()]\n",
    "print(\"Edge IDs:\", edge_ids)\n",
    "\n",
    "# List all traffic light IDs\n",
    "tls_ids = [tls.getID() for tls in net.getTrafficLights()]\n",
    "print(\"Traffic Light IDs:\", tls_ids)\n",
    "\n",
    "# # Get all vehicle IDs in the current step\n",
    "# vehicle_ids = traci.vehicle.getIDList()\n",
    "# print(\"Vehicle IDs:\", vehicle_ids)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Retrying in 1 seconds\n",
      "TraCI connected successfully.\n",
      "Step 0, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 1, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 2, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 3, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 4, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 5, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 6, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 7, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 8, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 9, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 10, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 11, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 12, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 13, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 14, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 15, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 16, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 17, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 18, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 19, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 20, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 21, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 22, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 23, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 24, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 25, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 26, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 27, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 28, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 29, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 30, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 31, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 32, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 33, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 34, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 35, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 36, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 37, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 38, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 39, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 40, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 41, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 42, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 1\n",
      "Step 43, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 1\n",
      "Step 44, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 2\n",
      "Step 45, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 2\n",
      "Step 46, Vehicle count on t_e: 2\n",
      "Traffic light t phase: 2\n",
      "Step 47, Vehicle count on t_e: 2\n",
      "Traffic light t phase: 2\n",
      "Step 48, Vehicle count on t_e: 2\n",
      "Traffic light t phase: 2\n",
      "Step 49, Vehicle count on t_e: 4\n",
      "Traffic light t phase: 2\n",
      "Step 50, Vehicle count on t_e: 4\n",
      "Traffic light t phase: 2\n",
      "Step 51, Vehicle count on t_e: 6\n",
      "Traffic light t phase: 2\n",
      "Step 52, Vehicle count on t_e: 6\n",
      "Traffic light t phase: 2\n",
      "Step 53, Vehicle count on t_e: 8\n",
      "Traffic light t phase: 2\n",
      "Step 54, Vehicle count on t_e: 8\n",
      "Traffic light t phase: 2\n",
      "Step 55, Vehicle count on t_e: 10\n",
      "Traffic light t phase: 2\n",
      "Step 56, Vehicle count on t_e: 10\n",
      "Traffic light t phase: 2\n",
      "Step 57, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 58, Vehicle count on t_e: 11\n",
      "Traffic light t phase: 2\n",
      "Step 59, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 60, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 61, Vehicle count on t_e: 13\n",
      "Traffic light t phase: 2\n",
      "Step 62, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 63, Vehicle count on t_e: 13\n",
      "Traffic light t phase: 2\n",
      "Step 64, Vehicle count on t_e: 13\n",
      "Traffic light t phase: 2\n",
      "Step 65, Vehicle count on t_e: 14\n",
      "Traffic light t phase: 2\n",
      "Step 66, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 67, Vehicle count on t_e: 14\n",
      "Traffic light t phase: 2\n",
      "Step 68, Vehicle count on t_e: 13\n",
      "Traffic light t phase: 2\n",
      "Step 69, Vehicle count on t_e: 14\n",
      "Traffic light t phase: 2\n",
      "Step 70, Vehicle count on t_e: 13\n",
      "Traffic light t phase: 2\n",
      "Step 71, Vehicle count on t_e: 13\n",
      "Traffic light t phase: 2\n",
      "Step 72, Vehicle count on t_e: 14\n",
      "Traffic light t phase: 2\n",
      "Step 73, Vehicle count on t_e: 13\n",
      "Traffic light t phase: 2\n",
      "Step 74, Vehicle count on t_e: 14\n",
      "Traffic light t phase: 2\n",
      "Step 75, Vehicle count on t_e: 13\n",
      "Traffic light t phase: 2\n",
      "Step 76, Vehicle count on t_e: 14\n",
      "Traffic light t phase: 2\n",
      "Step 77, Vehicle count on t_e: 13\n",
      "Traffic light t phase: 2\n",
      "Step 78, Vehicle count on t_e: 13\n",
      "Traffic light t phase: 2\n",
      "Step 79, Vehicle count on t_e: 13\n",
      "Traffic light t phase: 2\n",
      "Step 80, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 81, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 82, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 83, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 84, Vehicle count on t_e: 11\n",
      "Traffic light t phase: 2\n",
      "Step 85, Vehicle count on t_e: 11\n",
      "Traffic light t phase: 2\n",
      "Step 86, Vehicle count on t_e: 10\n",
      "Traffic light t phase: 3\n",
      "Step 87, Vehicle count on t_e: 10\n",
      "Traffic light t phase: 3\n",
      "Step 88, Vehicle count on t_e: 9\n",
      "Traffic light t phase: 0\n",
      "Step 89, Vehicle count on t_e: 8\n",
      "Traffic light t phase: 0\n",
      "Step 90, Vehicle count on t_e: 8\n",
      "Traffic light t phase: 0\n",
      "Step 91, Vehicle count on t_e: 7\n",
      "Traffic light t phase: 0\n",
      "Step 92, Vehicle count on t_e: 6\n",
      "Traffic light t phase: 0\n",
      "Step 93, Vehicle count on t_e: 6\n",
      "Traffic light t phase: 0\n",
      "Step 94, Vehicle count on t_e: 4\n",
      "Traffic light t phase: 0\n",
      "Step 95, Vehicle count on t_e: 2\n",
      "Traffic light t phase: 0\n",
      "Step 96, Vehicle count on t_e: 2\n",
      "Traffic light t phase: 0\n",
      "Step 97, Vehicle count on t_e: 1\n",
      "Traffic light t phase: 0\n",
      "Step 98, Vehicle count on t_e: 1\n",
      "Traffic light t phase: 0\n",
      "Step 99, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 100, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 101, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 102, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 103, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 104, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 105, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 106, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 107, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 108, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 109, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 110, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 111, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 112, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 113, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 114, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 115, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 116, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 117, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 118, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 119, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 120, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 121, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 122, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 123, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 124, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 125, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 126, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 127, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 128, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 129, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 130, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 1\n",
      "Step 131, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 1\n",
      "Step 132, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 2\n",
      "Step 133, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 2\n",
      "Step 134, Vehicle count on t_e: 2\n",
      "Traffic light t phase: 2\n",
      "Step 135, Vehicle count on t_e: 2\n",
      "Traffic light t phase: 2\n",
      "Step 136, Vehicle count on t_e: 2\n",
      "Traffic light t phase: 2\n",
      "Step 137, Vehicle count on t_e: 4\n",
      "Traffic light t phase: 2\n",
      "Step 138, Vehicle count on t_e: 4\n",
      "Traffic light t phase: 2\n",
      "Step 139, Vehicle count on t_e: 6\n",
      "Traffic light t phase: 2\n",
      "Step 140, Vehicle count on t_e: 6\n",
      "Traffic light t phase: 2\n",
      "Step 141, Vehicle count on t_e: 8\n",
      "Traffic light t phase: 2\n",
      "Step 142, Vehicle count on t_e: 8\n",
      "Traffic light t phase: 2\n",
      "Step 143, Vehicle count on t_e: 10\n",
      "Traffic light t phase: 2\n",
      "Step 144, Vehicle count on t_e: 10\n",
      "Traffic light t phase: 2\n",
      "Step 145, Vehicle count on t_e: 11\n",
      "Traffic light t phase: 2\n",
      "Step 146, Vehicle count on t_e: 11\n",
      "Traffic light t phase: 2\n",
      "Step 147, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 148, Vehicle count on t_e: 11\n",
      "Traffic light t phase: 2\n",
      "Step 149, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 150, Vehicle count on t_e: 14\n",
      "Traffic light t phase: 2\n",
      "Step 151, Vehicle count on t_e: 13\n",
      "Traffic light t phase: 2\n",
      "Step 152, Vehicle count on t_e: 14\n",
      "Traffic light t phase: 2\n",
      "Step 153, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 154, Vehicle count on t_e: 14\n",
      "Traffic light t phase: 2\n",
      "Step 155, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 156, Vehicle count on t_e: 14\n",
      "Traffic light t phase: 2\n",
      "Step 157, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 158, Vehicle count on t_e: 14\n",
      "Traffic light t phase: 2\n",
      "Step 159, Vehicle count on t_e: 13\n",
      "Traffic light t phase: 2\n",
      "Step 160, Vehicle count on t_e: 13\n",
      "Traffic light t phase: 2\n",
      "Step 161, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 162, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 163, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 164, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 165, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 166, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 167, Vehicle count on t_e: 11\n",
      "Traffic light t phase: 2\n",
      "Step 168, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 169, Vehicle count on t_e: 11\n",
      "Traffic light t phase: 2\n",
      "Step 170, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 171, Vehicle count on t_e: 13\n",
      "Traffic light t phase: 2\n",
      "Step 172, Vehicle count on t_e: 13\n",
      "Traffic light t phase: 2\n",
      "Step 173, Vehicle count on t_e: 12\n",
      "Traffic light t phase: 2\n",
      "Step 174, Vehicle count on t_e: 14\n",
      "Traffic light t phase: 3\n",
      "Step 175, Vehicle count on t_e: 14\n",
      "Traffic light t phase: 3\n",
      "Step 176, Vehicle count on t_e: 13\n",
      "Traffic light t phase: 0\n",
      "Step 177, Vehicle count on t_e: 11\n",
      "Traffic light t phase: 0\n",
      "Step 178, Vehicle count on t_e: 11\n",
      "Traffic light t phase: 0\n",
      "Step 179, Vehicle count on t_e: 9\n",
      "Traffic light t phase: 0\n",
      "Step 180, Vehicle count on t_e: 8\n",
      "Traffic light t phase: 0\n",
      "Step 181, Vehicle count on t_e: 7\n",
      "Traffic light t phase: 0\n",
      "Step 182, Vehicle count on t_e: 6\n",
      "Traffic light t phase: 0\n",
      "Step 183, Vehicle count on t_e: 5\n",
      "Traffic light t phase: 0\n",
      "Step 184, Vehicle count on t_e: 4\n",
      "Traffic light t phase: 0\n",
      "Step 185, Vehicle count on t_e: 3\n",
      "Traffic light t phase: 0\n",
      "Step 186, Vehicle count on t_e: 1\n",
      "Traffic light t phase: 0\n",
      "Step 187, Vehicle count on t_e: 1\n",
      "Traffic light t phase: 0\n",
      "Step 188, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 189, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 190, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 191, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 192, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 193, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 194, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 195, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 196, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 197, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 198, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 199, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 200, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 201, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 202, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 203, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 204, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 205, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 206, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 207, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 208, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 209, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 210, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 211, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 212, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 213, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 214, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "Step 215, Vehicle count on t_e: 0\n",
      "Traffic light t phase: 0\n",
      "\n",
      "Simulation interrupted by user.\n"
     ]
    },
    {
     "ename": "FatalTraCIError",
     "evalue": "Connection closed by SUMO.",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mFatalTraCIError\u001b[0m                           Traceback (most recent call last)",
      "Cell \u001b[1;32mIn[6], line 63\u001b[0m\n\u001b[0;32m     61\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[0;32m     62\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m traci\u001b[38;5;241m.\u001b[39misLoaded():\n\u001b[1;32m---> 63\u001b[0m         \u001b[43mtraci\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mclose\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m     64\u001b[0m         \u001b[38;5;28mprint\u001b[39m(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mTraCI connection closed safely.\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[0;32m     65\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m traci\u001b[38;5;241m.\u001b[39mexceptions\u001b[38;5;241m.\u001b[39mTraCIException \u001b[38;5;28;01mas\u001b[39;00m e:\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\main.py:262\u001b[0m, in \u001b[0;36mclose\u001b[1;34m(wait)\u001b[0m\n\u001b[0;32m    258\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mclose\u001b[39m(wait\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mTrue\u001b[39;00m):\n\u001b[0;32m    259\u001b[0m \u001b[38;5;250m    \u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[0;32m    260\u001b[0m \u001b[38;5;124;03m    Tells TraCI to close the connection.\u001b[39;00m\n\u001b[0;32m    261\u001b[0m \u001b[38;5;124;03m    \"\"\"\u001b[39;00m\n\u001b[1;32m--> 262\u001b[0m     \u001b[43mconnection\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcheck\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mclose\u001b[49m\u001b[43m(\u001b[49m\u001b[43mwait\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\connection.py:397\u001b[0m, in \u001b[0;36mConnection.close\u001b[1;34m(self, wait)\u001b[0m\n\u001b[0;32m    395\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mremoveStepListener(listenerID)\n\u001b[0;32m    396\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_socket \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[1;32m--> 397\u001b[0m     \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_sendCmd\u001b[49m\u001b[43m(\u001b[49m\u001b[43mtc\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mCMD_CLOSE\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43;01mNone\u001b[39;49;00m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43;01mNone\u001b[39;49;00m\u001b[43m)\u001b[49m\n\u001b[0;32m    398\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_socket\u001b[38;5;241m.\u001b[39mclose()\n\u001b[0;32m    399\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_socket \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\connection.py:232\u001b[0m, in \u001b[0;36mConnection._sendCmd\u001b[1;34m(self, cmdID, varID, objID, format, *values)\u001b[0m\n\u001b[0;32m    230\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_string \u001b[38;5;241m+\u001b[39m\u001b[38;5;241m=\u001b[39m struct\u001b[38;5;241m.\u001b[39mpack(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124m!i\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;28mlen\u001b[39m(objID)) \u001b[38;5;241m+\u001b[39m objID\n\u001b[0;32m    231\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_string \u001b[38;5;241m+\u001b[39m\u001b[38;5;241m=\u001b[39m packed\n\u001b[1;32m--> 232\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_sendExact\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[1;32mc:\\Users\\ASUS\\AppData\\Local\\Programs\\Python\\Python310\\lib\\site-packages\\traci\\connection.py:137\u001b[0m, in \u001b[0;36mConnection._sendExact\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m    135\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_socket\u001b[38;5;241m.\u001b[39mclose()\n\u001b[0;32m    136\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_socket \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[1;32m--> 137\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m FatalTraCIError(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mConnection closed by SUMO.\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[0;32m    138\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m command \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_queue:\n\u001b[0;32m    139\u001b[0m     prefix \u001b[38;5;241m=\u001b[39m result\u001b[38;5;241m.\u001b[39mread(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124m!BBB\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n",
      "\u001b[1;31mFatalTraCIError\u001b[0m: Connection closed by SUMO."
     ]
    }
   ],
   "source": [
    "import traci\n",
    "import sumolib\n",
    "import sys\n",
    "import time\n",
    "\n",
    "try:\n",
    "    # Define paths to SUMO binary, network file, and route file\n",
    "    sumo_binary = \"sumo-gui\"  # Use \"sumo\" for non-GUI mode\n",
    "    # Command to start SUMO with TraCI connection\n",
    "    sumo_cmd = [sumo_binary, \"-n\", nets_file, \"-r\", routes_file, \"--start\"]\n",
    "    \n",
    "    # Attempt to start SUMO with the specified command\n",
    "    try:\n",
    "        traci.start(sumo_cmd, label=\"sim1\")\n",
    "    except traci.exceptions.TraCIException as e:\n",
    "        print(f\"Failed to start TraCI: {e}\")\n",
    "        sys.exit(1)\n",
    "\n",
    "    print(\"TraCI connected successfully.\")\n",
    "    \n",
    "    # Main simulation loop with safe TraCI operations\n",
    "    for step in range(1000):\n",
    "        try:\n",
    "            traci.simulationStep()  # Advance simulation by one step\n",
    "            \n",
    "            # Check vehicle count on a specific edge\n",
    "            edge_id = \"t_e\"  # Replace with actual edge ID\n",
    "            try:\n",
    "                vehicle_count = traci.edge.getLastStepVehicleNumber(edge_id)\n",
    "                print(f\"Step {step}, Vehicle count on {edge_id}: {vehicle_count}\")\n",
    "            except traci.exceptions.TraCIException as e:\n",
    "                print(f\"Error retrieving vehicle count for edge {edge_id}: {e}\")\n",
    "            \n",
    "            # Get traffic light phase and queue length\n",
    "            tls_id = \"t\"  # Replace with actual traffic light ID\n",
    "            try:\n",
    "                phase = traci.trafficlight.getPhase(tls_id)\n",
    "                print(f\"Traffic light {tls_id} phase: {phase}\")\n",
    "            except traci.exceptions.TraCIException as e:\n",
    "                print(f\"Error retrieving traffic light phase for {tls_id}: {e}\")\n",
    "            \n",
    "            # Pause to inspect values for each step\n",
    "            # input(\"Press Enter to continue to the next step...\")\n",
    "            time.sleep(0.5)\n",
    "            \n",
    "            if step >= max_steps -1:\n",
    "                break            \n",
    "\n",
    "        except KeyboardInterrupt:\n",
    "            print(\"\\nSimulation interrupted by user.\")\n",
    "            break  # Safely exit the loop if user interrupts\n",
    "\n",
    "except (FileNotFoundError, traci.exceptions.TraCIException) as e:\n",
    "    print(f\"Error in setup or simulation: {e}\")\n",
    "\n",
    "except Exception as e:\n",
    "    print(f\"Unexpected error occurred: {e}\")\n",
    "\n",
    "finally:\n",
    "    # Ensure that TraCI closes safely even if an error occurs\n",
    "    try:\n",
    "        if traci.isLoaded():\n",
    "            traci.close()\n",
    "            print(\"TraCI connection closed safely.\")\n",
    "    except traci.exceptions.TraCIException as e:\n",
    "        print(f\"Error while closing TraCI: {e}\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}