notebooks/aco_tsp.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "F4lbTHND81gf"
   },
   "source": [
    "<a href=\"https://colab.research.google.com/github/ZosoV/aco_random_walk/blob/main/notebooks/aco_tsp.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>\n",
    "# Python Implementation of ACO for Travelling Salesman Problem\n",
    "This is a python implementation of the Ant Colony Optimization algorithm for the Travelling Salesman Problem"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Swb-gCMIRJar",
    "outputId": "5a10296c-9ab7-485d-8109-0a912586d918"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mounted at /content/gdrive\n"
     ]
    }
   ],
   "source": [
    "from google.colab import drive\n",
    "drive.mount('/content/gdrive')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "68SGeWRERJas",
    "outputId": "23a27468-c9ac-4ef0-c402-f379b24704c5"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/content/gdrive/MyDrive/Proyectos/ACO\n"
     ]
    }
   ],
   "source": [
    "%cd /content/gdrive/MyDrive/Proyectos/aco_random_walk/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "20OuD4E79EF_"
   },
   "outputs": [],
   "source": [
    "# import necessary libraries\n",
    "import networkx as nx\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib.pyplot import figure\n",
    "\n",
    "import enum\n",
    "import numpy as np\n",
    "import random\n",
    "import os\n",
    "import pandas as pd\n",
    "from collections import Counter"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "j8dhvuDVYV2e"
   },
   "source": [
    "For each problem, we have to define a weighted undirected graph $G(N,E)$\n",
    "\n",
    "# Travelling Salesman (TSP) Problem\n",
    "\n",
    "In this part, we create the graph according to well-known datasets from\n",
    "1. [TSP Lib](http://comopt.ifi.uni-heidelberg.de/software/TSPLIB95/)\n",
    "2. [TSP Updated Dataset](https://people.sc.fsu.edu/~jburkardt/datasets/tsp/tsp.html)\n",
    "\n",
    "The datasets have the extension .tsp and could be found in the attached information.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "0aSsk0awYYI8"
   },
   "outputs": [],
   "source": [
    "# loading the information in a graph with the library networkx\n",
    "class TSPGraph(nx.Graph):\n",
    "  def __init__(self, tau_0, dimension, filename = None):\n",
    "    \"\"\"Define a graph environment given a filename, \n",
    "    and tau_0 (initial pheromone)\n",
    "    \"\"\"\n",
    "    self.tau_0 = tau_0\n",
    "    self.n_nodes = dimension\n",
    "    self.dataset = filename\n",
    "    nx.Graph.__init__(self) \n",
    "    self.__adding_nodes(self.dataset)\n",
    "    self.__adding_edges(self.tau_0)\n",
    "  \n",
    "  def __adding_nodes(self,filename):\n",
    "\n",
    "    with open(os.path.join(DATA_DIR,filename), \"r\") as f:\n",
    "      get_nodes = False\n",
    "\n",
    "      counter_nodes = 0\n",
    "      for line in f:\n",
    "        line = line.strip()\n",
    "\n",
    "        # get the best solution value\n",
    "        if \"BEST_SOLUTION\" in line:\n",
    "          str_best = line.split(\":\")[-1].strip()\n",
    "          self.best_solution = float(str_best)\n",
    "          print(self.best_solution, type(self.best_solution))\n",
    "\n",
    "        if get_nodes and counter_nodes < self.n_nodes:\n",
    "          idx, posx, posy = line.split(\" \")\n",
    "\n",
    "          self.add_node(counter_nodes, pos=(float(posx),float(posy))) \n",
    "\n",
    "          counter_nodes += 1\n",
    "\n",
    "        if \"NODE_COORD_SECTION\" in line: \n",
    "          get_nodes = True\n",
    "\n",
    "  def __adding_edges(self, tau_0):\n",
    "    for i in range(self.n_nodes):\n",
    "      for j in range(self.n_nodes):\n",
    "        if i < j:\n",
    "          x1, y1 = self.nodes[i]['pos']\n",
    "          x2, y2 = self.nodes[j]['pos']\n",
    "\n",
    "          distance = np.sqrt((x2-x1)**2 + (y2-y1)**2)\n",
    "\n",
    "          self.add_edge(i, j, \n",
    "                pheromone = tau_0, \n",
    "                distance = distance)\n",
    "\n",
    "  #Function to plot the structure of the graph\n",
    "  def draw_graph(self, node_color='blue', node_size=25, with_labels=False, draw_edges = False):\n",
    "    \"\"\"\n",
    "    Draw the graph.\n",
    "        \n",
    "    node_size: the size of all the nodes\n",
    "    node_color: the color of all the nodes\n",
    "    with_labels: True or False to show the ids of the nodes\n",
    "    \"\"\"\n",
    "    fig, ax = plt.subplots()\n",
    "    ax.set_aspect('equal') #set the x and y axes to the same scale\n",
    "    pos = nx.get_node_attributes(self, 'pos')\n",
    "    if draw_edges:\n",
    "      nx.draw(self, pos, node_color=node_color, node_size=node_size, with_labels=with_labels)\n",
    "    else:\n",
    "      nx.draw(self, pos, node_color=node_color, node_size=node_size, with_labels=with_labels, edgelist = [])\n",
    "\n",
    "  \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 265
    },
    "id": "m0yOdJRgZhvq",
    "outputId": "16f0d711-7be2-4eaa-e950-c1b5354d7ac0"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "33523.0 <class 'float'>\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "tsp_graph = TSPGraph(0.1, 48, \"att48.tsp\")\n",
    "tsp_graph.draw_graph(draw_edges = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "JS-F_HOWYIJf",
    "outputId": "0e1ec407-86bf-407c-d6d1-681e021440f3"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.2"
      ]
     },
     "execution_count": 186,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tsp_graph.edges[(0,1)]['pheromone'] = 0.2\n",
    "tsp_graph.edges[(1,0)]['pheromone']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 265
    },
    "id": "jF_u4WXgmdT3",
    "outputId": "cc768d28-9015-4c1d-da6e-644de673de43"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "33523.0 <class 'float'>\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "tsp_graph = TSPGraph(0.1, 48, \"att48.tsp\")\n",
    "tsp_graph.draw_graph(draw_edges = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Qo3i4yLe0Ugo"
   },
   "source": [
    "### ACO for Travelling Salesman Problem"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "h5RcbiDS0Mvo"
   },
   "outputs": [],
   "source": [
    "class ACOTSP(AntColonyOptimizer):\n",
    "  def __init__(self, graph, ants, alpha, beta, p, local_p = None, intensity = None, q_0 = None):\n",
    "    \"\"\"\n",
    "    Ant colony optimizer for Path Planning.  \n",
    "    Traverses a graph and finds the min weight distance \n",
    "    between a start and target node\n",
    "    \n",
    "    :param graph: graph environment\n",
    "    :param ants:  number of ants to traverse the graph\n",
    "    :param alpha: weighting of pheromone\n",
    "    :param beta:  weighting of heuristic\n",
    "    :param p:     evaporation rate at which pheromone evaporates\n",
    "    :param local_p: local evaporation rate at which pheromone evaporates (optional)\n",
    "    :param intensity: the amount of pheromones to add per edge (optional)\n",
    "    :param q_0: probability to choose the best construction step (optional)\n",
    "    \"\"\"\n",
    "    self.start_node = 0\n",
    "    super().__init__(graph, ants, alpha, beta, p, local_p, intensity, q_0)\n",
    "\n",
    "  def update_state(self):\n",
    "    neighbors = self.graph[self.current_position]\n",
    "\n",
    "    # Process to get the weight of each neighbor\n",
    "    weights = []\n",
    "    aux_weights = []\n",
    "    neighbors_idx = []\n",
    "    for neighbor_node, edge in neighbors.items():\n",
    "\n",
    "      heuristic = 1 / edge['distance']\n",
    "\n",
    "      if not self.is_visited(neighbor_node):\n",
    "        # get the weight per neighbor and append it\n",
    "        weights.append( (edge['pheromone'] ** self.alpha) * \n",
    "                      ( heuristic ** self.beta))\n",
    "        aux_weights.append( edge['pheromone'] * \n",
    "                      (heuristic ** self.beta))\n",
    "        neighbors_idx.append(neighbor_node)\n",
    "\n",
    "    if len(weights) != 0:\n",
    "      acu_weight = sum(weights)\n",
    "      probabilities = np.array(weights)/acu_weight\n",
    "      #print(\"[INFO] probabilities: {}\".format(probabilities))\n",
    "\n",
    "      # with probability q_0 select the best trial\n",
    "      if self.q_0 is not None:\n",
    "        if np.random.rand() < self.q_0:\n",
    "          new_position = neighbors_idx[np.argmax(aux_weights)]\n",
    "        else:\n",
    "          # choose an option following the wheel selection algorithm  \n",
    "          new_position = random.choices(neighbors_idx, weights = probabilities)[0]\n",
    "      else:\n",
    "        # choose an option following the wheel selection algorithm  \n",
    "        new_position = random.choices(neighbors_idx, weights = probabilities)[0]\n",
    "\n",
    "      # perform local evaporation\n",
    "      if self.local_p is not None:\n",
    "        self.local_evaporation(new_position)\n",
    "\n",
    "      self.current_path.append((self.current_position, new_position))\n",
    "      self.visited_nodes.append(new_position)\n",
    "      self.current_position = new_position\n",
    "    else:      \n",
    "      self.current_path.append((self.current_position, self.start_node))\n",
    "      self.visited_nodes.append(self.start_node)\n",
    "      self.current_position =  self.start_node\n",
    "\n",
    "  def end_route(self):\n",
    "    return self.current_position == self.start_node"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "sQYgkvhz3DO5"
   },
   "outputs": [],
   "source": [
    "# Define the parameters of our model\n",
    "ants = 10\n",
    "tau_0 = 1\n",
    "alpha = 2. \n",
    "beta = 1.\n",
    "p = 0.3\n",
    "intensity = None\n",
    "\n",
    "total_iter = 500\n",
    "step_plot = 10\n",
    "\n",
    "local_p = 0.4\n",
    "q_0 = 0.3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 353
    },
    "id": "AUbRBaMF3plx",
    "outputId": "0e8148e5-677d-43f3-a6fd-697e55a1bced"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "33523.0 <class 'float'>\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Create the graph\n",
    "graph = TSPGraph(0.1, 48, \"att48.tsp\")\n",
    "\n",
    "# Create the envioronment\n",
    "optimizer = ACOTSP(graph, ants, alpha, beta, p, local_p, intensity, q_0)\n",
    "#draw_grid(env)\n",
    "display_pheromones(optimizer,10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "8pSM3qCv2QdW",
    "outputId": "ae743184-dd9d-4d04-9441-f3aa3d328dee"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[INFO] iter: 0 best: 72946.8709588917 d_mean: 88956.67004376012 d_stdv: 10966.828835766877\n",
      "[INFO] iter: 10 best: 44768.824581247194 d_mean: 56679.71624998244 d_stdv: 6583.888108539997\n",
      "[INFO] iter: 20 best: 41981.484837360505 d_mean: 53587.145472516655 d_stdv: 5951.260545374647\n",
      "[INFO] iter: 30 best: 38476.87864480143 d_mean: 54413.54196180873 d_stdv: 7567.71074718774\n",
      "[INFO] iter: 40 best: 38476.87864480143 d_mean: 56275.79551874193 d_stdv: 4929.408677903402\n",
      "[INFO] iter: 50 best: 38476.87864480143 d_mean: 57950.888802805006 d_stdv: 5363.800274644261\n",
      "[INFO] iter: 60 best: 38476.87864480143 d_mean: 56331.77684652495 d_stdv: 8436.233363641111\n",
      "[INFO] iter: 70 best: 38476.87864480143 d_mean: 56922.30567429676 d_stdv: 6194.860357531522\n",
      "[INFO] iter: 80 best: 38476.87864480143 d_mean: 52730.955792260844 d_stdv: 4818.932140872716\n",
      "[INFO] iter: 90 best: 38476.87864480143 d_mean: 58963.326630168056 d_stdv: 6252.38631505985\n",
      "[INFO] iter: 100 best: 38476.87864480143 d_mean: 54101.610792354986 d_stdv: 3819.1121198479223\n",
      "[INFO] iter: 110 best: 38476.87864480143 d_mean: 54451.22132842798 d_stdv: 6410.439420520084\n",
      "[INFO] iter: 120 best: 38476.87864480143 d_mean: 58449.07096972902 d_stdv: 7181.3946442561655\n",
      "[INFO] iter: 130 best: 38476.87864480143 d_mean: 55953.189775959974 d_stdv: 5586.945310865678\n",
      "[INFO] iter: 140 best: 38476.87864480143 d_mean: 59884.11910631737 d_stdv: 7132.569210039977\n",
      "[INFO] iter: 150 best: 38476.87864480143 d_mean: 59194.16834748288 d_stdv: 6605.791738174012\n",
      "[INFO] iter: 160 best: 38476.87864480143 d_mean: 56977.44945794543 d_stdv: 6939.105224408726\n",
      "[INFO] iter: 170 best: 38476.87864480143 d_mean: 56515.79665265376 d_stdv: 4868.380762125952\n",
      "[INFO] iter: 180 best: 38476.87864480143 d_mean: 55735.323939168105 d_stdv: 6306.864299180089\n",
      "[INFO] iter: 190 best: 38476.87864480143 d_mean: 58579.53303097292 d_stdv: 6985.586713536326\n",
      "[INFO] iter: 200 best: 38476.87864480143 d_mean: 55166.692384670976 d_stdv: 6998.966378167174\n",
      "[INFO] iter: 210 best: 38476.87864480143 d_mean: 55207.728063709044 d_stdv: 3083.9881989368982\n",
      "[INFO] iter: 220 best: 38476.87864480143 d_mean: 55563.37710374054 d_stdv: 3151.220943190064\n",
      "[INFO] iter: 230 best: 38476.87864480143 d_mean: 55271.02662385367 d_stdv: 3043.2408060408866\n",
      "[INFO] iter: 240 best: 38476.87864480143 d_mean: 53591.775296928055 d_stdv: 3892.1564127065767\n",
      "[INFO] iter: 250 best: 38476.87864480143 d_mean: 56772.457939155676 d_stdv: 4075.2303709225753\n",
      "[INFO] iter: 260 best: 38476.87864480143 d_mean: 62806.74785239816 d_stdv: 7626.580320715873\n",
      "[INFO] iter: 270 best: 38476.87864480143 d_mean: 57800.692621799375 d_stdv: 7103.4443752627885\n",
      "[INFO] iter: 280 best: 38476.87864480143 d_mean: 55433.91998398515 d_stdv: 2374.9476575639223\n",
      "[INFO] iter: 290 best: 38476.87864480143 d_mean: 58791.08497791835 d_stdv: 7490.383210276847\n",
      "[INFO] iter: 300 best: 38476.87864480143 d_mean: 57427.53064293952 d_stdv: 6816.21605387879\n",
      "[INFO] iter: 310 best: 38476.87864480143 d_mean: 55307.04831194079 d_stdv: 6096.518447719102\n",
      "[INFO] iter: 320 best: 38476.87864480143 d_mean: 58581.10608090486 d_stdv: 5045.28216711446\n",
      "[INFO] iter: 330 best: 38476.87864480143 d_mean: 55616.793939917945 d_stdv: 7508.2912154366195\n",
      "[INFO] iter: 340 best: 38476.87864480143 d_mean: 51863.273785629965 d_stdv: 4466.846890353397\n",
      "[INFO] iter: 350 best: 38476.87864480143 d_mean: 53551.19424541942 d_stdv: 5668.982343425629\n",
      "[INFO] iter: 360 best: 38476.87864480143 d_mean: 56513.70134351259 d_stdv: 4433.790244270438\n",
      "[INFO] iter: 370 best: 38476.87864480143 d_mean: 56026.48306948111 d_stdv: 5991.906722481783\n",
      "[INFO] iter: 380 best: 38476.87864480143 d_mean: 54220.84812798896 d_stdv: 5036.05057871566\n",
      "[INFO] iter: 390 best: 38476.87864480143 d_mean: 52671.935306658655 d_stdv: 5129.746932097483\n",
      "[INFO] iter: 400 best: 38476.87864480143 d_mean: 55787.546306564145 d_stdv: 4313.890408748713\n",
      "[INFO] iter: 410 best: 38476.87864480143 d_mean: 54136.42935970835 d_stdv: 5731.79974208912\n",
      "[INFO] iter: 420 best: 38476.87864480143 d_mean: 60264.026700566275 d_stdv: 6054.0147657669895\n",
      "[INFO] iter: 430 best: 38476.87864480143 d_mean: 53640.19910696945 d_stdv: 6864.807705878513\n",
      "[INFO] iter: 440 best: 38476.87864480143 d_mean: 53092.752444608275 d_stdv: 6417.061115800088\n",
      "[INFO] iter: 450 best: 38476.87864480143 d_mean: 56550.97722609302 d_stdv: 6099.7233139561395\n",
      "[INFO] iter: 460 best: 38476.87864480143 d_mean: 54960.58541416493 d_stdv: 5065.523748115824\n",
      "[INFO] iter: 470 best: 38476.87864480143 d_mean: 56286.54116738591 d_stdv: 4624.855924558951\n",
      "[INFO] iter: 480 best: 38476.87864480143 d_mean: 53495.01255529326 d_stdv: 5457.323643574827\n",
      "[INFO] iter: 490 best: 38476.87864480143 d_mean: 59512.916932158594 d_stdv: 8407.219498846347\n"
     ]
    }
   ],
   "source": [
    "list_distances = []\n",
    "list_distances_avg = []\n",
    "\n",
    "for iter in range(total_iter):\n",
    "\n",
    "  distance_per_ants = []\n",
    "\n",
    "  for ant in range(optimizer.ants):\n",
    "    \n",
    "    get_target = False\n",
    "    while not get_target:\n",
    "      optimizer.update_state()\n",
    "      get_target = optimizer.end_route()\n",
    "      \n",
    "    optimizer.update_best_path()\n",
    "    current_distance = optimizer.get_total_distance(optimizer.current_path)\n",
    "    #print(\"[INFO] iter: {} ant: {} current: {}\".format(iter, ant, current_distance))\n",
    "    distance_per_ants.append(current_distance)\n",
    "    optimizer.start_node = np.random.randint(0,48)\n",
    "    optimizer.reset_environment()\n",
    "\n",
    "  #optimizer.draw_grid()\n",
    "  optimizer.offline_pheromone_update()\n",
    "\n",
    "  # Track the best distance so far\n",
    "  best_distance = optimizer.get_total_distance(optimizer.best_path)\n",
    "  if iter % step_plot == 0:\n",
    "    print(\"[INFO] iter: {} best: {} d_mean: {} d_stdv: {}\".format(iter, \n",
    "                                                          best_distance, \n",
    "                                                          np.mean(distance_per_ants),\n",
    "                                                          np.std(distance_per_ants)))\n",
    "    #display_pheromones(env,10)\n",
    "  list_distances.append(best_distance)\n",
    "  list_distances_avg.append(np.mean(distance_per_ants))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "cellView": "form",
    "id": "pglmkmk25Elg"
   },
   "outputs": [],
   "source": [
    "#@title Function to plot the best distance per iteration\n",
    "# plot the best distance per iteration\n",
    "def plot_distance(list_distances, save = False, name2save = \"\"):\n",
    "\n",
    "  list_iterations = np.arange(0, len(list_distances))\n",
    "  #path_to_save = \"stuff/figures/\"\n",
    "  #if not os.path.exists(path_to_save):\n",
    "  #    os.mkdir(path_to_save)\n",
    "\n",
    "  title = 'Average Distance vs Iterations'\n",
    "  title += '\\n ants: {} tau_0: {} alpha: {} beta: {} p: {}'.format(\n",
    "      ants,\n",
    "      tau_0,\n",
    "      alpha,\n",
    "      beta, \n",
    "      p\n",
    "  )\n",
    "\n",
    "  plt.title(title)\n",
    "  plt.plot(list_iterations, list_distances, \"b-\") #, label = label)\n",
    "  plt.legend()\n",
    "  #if save: plt.savefig(os.path.join(path_to_save,name2save))\n",
    "\n",
    "  plt.show()\n",
    "  plt.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 313
    },
    "id": "uK1FwT4v4U9C",
    "outputId": "2d95d1c8-8a22-4082-8d27-53d768d04c04"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No handles with labels found to put in legend.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_distance(list_distances_avg)"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "authorship_tag": "ABX9TyPh75WDCTwzACdbDYwlAUoY",
   "name": "aco_tsp.ipynb",
   "provenance": []
  },
  "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}