Local_Search/traveling_salesman_problem_example_2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Solving the Traveling Salesman  Problem using Local Search\n",
    "\n",
    "Points: 10"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## The Traveling Salesman Problem\n",
    "\n",
    "Problem definition (see [TSP on Wikipedia](https://en.wikipedia.org/wiki/Travelling_salesman_problem)):\n",
    "\n",
    "* __Goal:__ Find the shortest tour visiting each of $n$ cities exactly once and returning back to the starting city. Given are pairwise distances between cities, where $d_{i,j}$ is the distance from city $i$ to city $j$. \n",
    "\n",
    "* __State space:__ Each state represents a tour. The cities are numbered and a tour can be expressed as vector  $\\pi$ with the order in which the cities are visited (a [permutation](https://en.wikipedia.org/wiki/Permutation)). That is, $\\pi(1)$ is the index of the first city to visit, $\\pi(2)$ the index of the second, and so on.\n",
    "\n",
    "* __Objective function:__ Minimize the tour length. The optimization problem is to find the optimal tour $\\pi^*$ through the $n$ cities and returning to the starting city:\n",
    "\n",
    "  > minimize: $\\mathrm{tourlength}(\\pi) = d_{\\pi(n),\\pi(1)} + \\sum_{i = 1}^{n-1} d_{\\pi(i),\\pi(i+1)}$\n",
    "  > \n",
    "  > subject to: $\\pi \\ \\text{is a valid permutation vector}$\n",
    "\n",
    "* __Local moves:__ Exchange two cities in the order."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Helper functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import math\n",
    "import random\n",
    "\n",
    "np.set_printoptions(precision = 2)\n",
    "pd.set_option('precision', 2)\n",
    "\n",
    "# make the results repeatable\n",
    "np.random.seed(1234)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[4, 6, 2, 5, 0, 7, 9, 3, 8, 1]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def random_tour(n):\n",
    "    \"\"\"Create a random tour\"\"\"\n",
    "    \n",
    "    tour = list(range(n))\n",
    "    random.shuffle(tour)\n",
    "    return(tour)\n",
    "\n",
    "random_tour(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Positions:\n",
      "      x     y\n",
      "0  0.19  0.36\n",
      "1  0.62  0.50\n",
      "2  0.44  0.68\n",
      "3  0.79  0.71\n",
      "4  0.78  0.37\n",
      "5  0.27  0.56\n",
      "6  0.28  0.50\n",
      "7  0.80  0.01\n",
      "8  0.96  0.77\n",
      "9  0.88  0.88\n",
      "Distance matrix:\n",
      "      0     1     2     3     4     5     6     7     8     9\n",
      "0  0.00  0.45  0.41  0.69  0.59  0.22  0.17  0.70  0.87  0.86\n",
      "1  0.45  0.00  0.26  0.27  0.20  0.35  0.35  0.52  0.43  0.46\n",
      "2  0.41  0.26  0.00  0.35  0.46  0.21  0.24  0.76  0.53  0.48\n",
      "3  0.69  0.27  0.35  0.00  0.34  0.53  0.55  0.70  0.18  0.19\n",
      "4  0.59  0.20  0.46  0.34  0.00  0.54  0.52  0.36  0.44  0.52\n",
      "5  0.22  0.35  0.21  0.53  0.54  0.00  0.06  0.76  0.72  0.68\n",
      "6  0.17  0.35  0.24  0.55  0.52  0.06  0.00  0.72  0.73  0.71\n",
      "7  0.70  0.52  0.76  0.70  0.36  0.76  0.72  0.00  0.77  0.87\n",
      "8  0.87  0.43  0.53  0.18  0.44  0.72  0.73  0.77  0.00  0.14\n",
      "9  0.86  0.46  0.48  0.19  0.52  0.68  0.71  0.87  0.14  0.00)\n"
     ]
    }
   ],
   "source": [
    "from scipy.spatial.distance import pdist\n",
    "from scipy.spatial.distance import squareform\n",
    "\n",
    "def random_tsp(n):\n",
    "    \"\"\"\n",
    "    Create a random (Euclidean) traveling salesman problem. Choose n points randomly in a 1 x 1 unit square and calulates a \n",
    "    pairwise Euclidean distance matrix.\n",
    "    \"\"\"\n",
    "    \n",
    "    pos = pd.DataFrame({\n",
    "        \"x\" : np.random.uniform(size = n),\n",
    "        \"y\" : np.random.uniform(size = n)\n",
    "    })\n",
    "    \n",
    "    dist = squareform(pdist(pos))\n",
    "    \n",
    "    return({\"pos\": pos, \"dist\": dist})\n",
    "    \n",
    "tsp = random_tsp(10)\n",
    "\n",
    "print(f\"Positions:\\n{tsp['pos']}\")\n",
    "print(f\"Distance matrix:\\n{pd.DataFrame(tsp['dist'])})\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5.629207918602361"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def tour_length(tsp, tour):\n",
    "    \"\"\"Caclulate the length of a tour, i.e., the objective function.\"\"\"\n",
    "    \n",
    "    # make sure tour is a Python list (not an array or a numpy.array)\n",
    "    if not isinstance(tour, list): tour = tour.tolist()\n",
    "    \n",
    "    tl = 0\n",
    "    dist = tsp[\"dist\"]\n",
    "    \n",
    "    for i in range(len(tour)-1):\n",
    "        tl += dist[tour[i], tour[i+1]]\n",
    "    \n",
    "    tl += dist[tour[-1], tour[0]]\n",
    "    \n",
    "    return(tl)\n",
    "        \n",
    "tour = random_tour(10)\n",
    "tour_length(tsp, tour)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tour length: 5.63\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def show_tsp(tsp, tour = None):  \n",
    "    \"\"\"display the traveling salesman problem and a tour.\"\"\"\n",
    "    \n",
    "    pos = tsp[\"pos\"]\n",
    "    \n",
    "    plt.scatter(pos[\"x\"], pos[\"y\"])\n",
    "    \n",
    "    if tour is not None:\n",
    "        # make sure tour is a Python list (not an array or a numpy.array)\n",
    "        if not isinstance(tour, list): tour = tour.tolist()\n",
    "        \n",
    "        print(f\"Tour length: {round(tour_length(tsp, tour), 2)}\")\n",
    "        \n",
    "        pos_ = pos.reindex(tour)\n",
    "        pos_ = pos_.append(pos_.head(1))\n",
    "        plt.plot(pos_[\"x\"], pos_[\"y\"])\n",
    "    \n",
    "    plt.show()\n",
    "    \n",
    "show_tsp(tsp, tour)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Use R to find a solution\n",
    "\n",
    "Load rpy2, make sure the R [TSP package](https://CRAN.R-project.org/package=TSP) is installed and prepare the distance matrix."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext rpy2.ipython\n",
    "\n",
    "%R if(!\"TSP\" %in% rownames(installed.packages())) install.packages(\"TSP\", repos=\"http://cran.us.r-project.org\")\n",
    "%R if(!\"microbenchmark\" %in% rownames(installed.packages())) install.packages(\"microbenchmark\", repos=\"http://cran.us.r-project.org\")\n",
    "\n",
    "d = tsp[\"dist\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Solve the TSP using [`solve_TSP`](https://www.rdocumentation.org/packages/TSP/versions/1.1-10/topics/solve_TSP) with the default heuristic. Note that 2-opt is steepest ascend hill climbing with exchanging two cities. `rep = 10` means 10 random restarts."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "object of class ‘TSP’ \n",
      "10 cities (distance ‘unknown’) \n",
      "object of class ‘TOUR’ \n",
      "result of method ‘arbitrary_insertion+two_opt_rep_10’ for 10 cities\n",
      "tour length: 2.763574 \n"
     ]
    }
   ],
   "source": [
    "%%R -i d -o tour\n",
    "\n",
    "library(\"TSP\")\n",
    "\n",
    "tsp <- TSP(d)\n",
    "print(tsp)\n",
    "\n",
    "tour <- solve_TSP(tsp, rep = 10)\n",
    "print(tour)\n",
    "\n",
    "# R starts index with 1, but Python starts at 0\n",
    "tour <- tour - 1L"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tour length: 2.76\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "show_tsp(tsp, tour)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "How long does it take to solve the problem?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Unit: milliseconds\n",
      "                   expr     min       lq     mean   median       uq      max\n",
      " solve_TSP(d, rep = 10) 7.99531 8.096485 10.63028 9.047341 12.84836 19.56482\n",
      " neval\n",
      "   100\n"
     ]
    }
   ],
   "source": [
    "%%R -i d\n",
    "\n",
    "library(\"microbenchmark\")\n",
    "d <- TSP(d)\n",
    "\n",
    "\n",
    "microbenchmark(solve_TSP(d, rep = 10))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Steepest-ascend Hill Climbing Search [3 Points]\n",
    "\n",
    "Calculate the objective function for all local moves (move each queen within its column) and always choose the best among all local moves."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Code goes here"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Steepest-ascend Hill Climbing Search with Random Restarts [1 Point]\n",
    "\n",
    "Steepest-ascend with random restarts."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Code goes here"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Stochastic Hill Climbing [1 Points]\n",
    "\n",
    "Chooses randomly from among all uphill moves."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Code goes here"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Stochastic Hill Climbing (First-choice) [1 Point]\n",
    "\n",
    "First-choice hill climbing is a type of stochastic hill climbing that generates one random local neighbor at a time and accept it if it has a better objective function value than the current state."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Local Moves\n",
    "I implement the following three different local move strategies: \n",
    "* Swap two random cities. \n",
    "* Swap two neighboring cities.\n",
    "* Reverse the subtour between two cities."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[7, 1, 2, 3, 4, 5, 6, 0, 8, 9]\n",
      "[9, 1, 2, 3, 4, 5, 6, 7, 8, 0]\n",
      "[0, 1, 3, 2, 4, 5, 6, 7, 8, 9]\n"
     ]
    }
   ],
   "source": [
    "def move_swap(tour):\n",
    "    new_tour = tour.copy()\n",
    "    \n",
    "    # np.random.randint is inclusive lower limit and exclusive upper limit\n",
    "    [a, b] = np.random.randint(0, len(tour), 2)\n",
    "    #print(f\"a={a}, b={b}\")\n",
    "    \n",
    "    new_tour[a] = tour[b]\n",
    "    new_tour[b] = tour[a]\n",
    "    return(new_tour)\n",
    "\n",
    "print(move_swap(list(range(10))))\n",
    "print(move_swap(list(range(10))))\n",
    "print(move_swap(list(range(10))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0, 1, 2, 4, 3, 5, 6, 7, 8, 9]\n",
      "[0, 2, 1, 3, 4, 5, 6, 7, 8, 9]\n",
      "[0, 1, 2, 4, 3, 5, 6, 7, 8, 9]\n"
     ]
    }
   ],
   "source": [
    "def move_swap_neighbors(tour):\n",
    "    new_tour = tour.copy()\n",
    "    \n",
    "    a = np.random.randint(0, len(tour)-1, 1)[0]\n",
    "    if a > 0: b = a+1\n",
    "    else: b = len(tour)-1  \n",
    "    #print(f\"a={a}, b={b}\")\n",
    "    \n",
    "    new_tour[a] = tour[b]\n",
    "    new_tour[b] = tour[a]\n",
    "    return(new_tour)\n",
    "\n",
    "print(move_swap_neighbors(list(range(10))))\n",
    "print(move_swap_neighbors(list(range(10))))\n",
    "print(move_swap_neighbors(list(range(10))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0, 9, 8, 7, 6, 5, 4, 3, 2, 1]\n",
      "[0, 1, 2, 6, 5, 4, 3, 7, 8, 9]\n",
      "[0, 9, 8, 7, 6, 5, 4, 3, 2, 1]\n"
     ]
    }
   ],
   "source": [
    "def move_reverse(tour):\n",
    "    new_tour = tour.copy()\n",
    "    \n",
    "    ab = np.random.randint(0, len(tour)+1, 2)\n",
    "    ab.sort()\n",
    "    #print(f\"a={ab[0]}, b={ab[1]}\")\n",
    "        \n",
    "    new_tour[ab[0]:ab[1]] = new_tour[ab[0]:ab[1]][::-1]\n",
    "    return(new_tour)\n",
    "\n",
    "print(move_reverse(list(range(10))))\n",
    "print(move_reverse(list(range(10))))\n",
    "print(move_reverse(list(range(10))))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Search Algorithm\n",
    "\n",
    "I don't know what the optimal tour is so I run the algorithm for at most `max_steps` steps, and stop if there is no improvement for at least `stop_after` steps."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "def FCSHC(tsp, tour = None, local_move_method = move_swap, \n",
    "         max_steps = 1000000, stop_after = 1000, verbose = True, keep_history = False):\n",
    "    \n",
    "    n = len(tsp['pos'])\n",
    "    \n",
    "    if keep_history: history = list() \n",
    "    \n",
    "    # 1. initialize current tour\n",
    "    # current tour = initial tour (random if no tour is given)\n",
    "    if not tour is None:\n",
    "        current_tour = tour\n",
    "    else:\n",
    "        current_tour = random_tour(n)\n",
    "    \n",
    "    \n",
    "    # initial tour length\n",
    "    current_length = tour_length(tsp, current_tour)    \n",
    "    if verbose: print(f\"initial tour length: {current_length}\")\n",
    "    \n",
    "    # step when we last improved the length\n",
    "    last_improvement_step = 0\n",
    "    \n",
    "    # 2. repeat local moves\n",
    "    for step in range(max_steps):\n",
    "        \n",
    "        new_tour = local_move_method(current_tour)\n",
    "        new_length = tour_length(tsp, new_tour)  \n",
    "        \n",
    "        # check if the new tour is better   \n",
    "        if new_length < current_length: \n",
    "            current_tour = new_tour\n",
    "            current_length = new_length\n",
    "            last_improvement_step = step\n",
    "            if verbose: print(f\"step: {step} - new tour length: {current_length}\")\n",
    "    \n",
    "        if keep_history: history.append(current_length)\n",
    "    \n",
    "        # stop if we did not improve for stop_after steps\n",
    "        if step - last_improvement_step > stop_after: \n",
    "            if verbose: print(f\"step: {step} - no improvement for {stop_after} steps.\")\n",
    "            break\n",
    "    \n",
    "    \n",
    "    if keep_history: return(current_tour, history)\n",
    "    return(current_tour)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Comparison of Local Move Strategies Using a Single Problem"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tour length: 5.7\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "initial_tour = random_tour(len(tsp['pos']))\n",
    "show_tsp(tsp, initial_tour)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "initial tour length: 5.700632953232227\n",
      "step: 0 - new tour length: 5.203695501219807\n",
      "step: 2 - new tour length: 4.236200825575449\n",
      "step: 6 - new tour length: 3.8329097885735353\n",
      "step: 11 - new tour length: 3.6126414736105072\n",
      "step: 12 - new tour length: 3.4473341532649924\n",
      "step: 13 - new tour length: 3.3599137254290525\n",
      "step: 20 - new tour length: 3.153223959037089\n",
      "step: 26 - new tour length: 3.047219924886294\n",
      "step: 39 - new tour length: 2.9525230410605117\n",
      "step: 112 - new tour length: 2.9062078251965873\n",
      "step: 1113 - no improvement for 1000 steps.\n",
      "CPU times: user 61.5 ms, sys: 8.93 ms, total: 70.4 ms\n",
      "Wall time: 57.6 ms\n",
      "Tour length: 2.91\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%time tour, history_swap = FCSHC(tsp, initial_tour, local_move_method = move_swap, keep_history = True)\n",
    "show_tsp(tsp, tour)\n",
    "\n",
    "plt.plot(range(len(history_swap)), history_swap)\n",
    "plt.xlabel(\"Step\")\n",
    "plt.ylabel(\"Path length\")\n",
    "plt.ylim(bottom = 0)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "initial tour length: 5.700632953232227\n",
      "step: 1 - new tour length: 5.382293548102927\n",
      "step: 10 - new tour length: 4.849330216848048\n",
      "step: 11 - new tour length: 4.289357382893998\n",
      "step: 12 - new tour length: 4.218441776207307\n",
      "step: 1013 - no improvement for 1000 steps.\n",
      "CPU times: user 56.4 ms, sys: 9.39 ms, total: 65.8 ms\n",
      "Wall time: 53.4 ms\n",
      "Tour length: 4.22\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%time tour, history_swap_neighbors = FCSHC(tsp, initial_tour, local_move_method = move_swap_neighbors, keep_history = True)\n",
    "show_tsp(tsp, tour)\n",
    "\n",
    "plt.plot(range(len(history_swap_neighbors)), history_swap_neighbors)\n",
    "plt.xlabel(\"Step\")\n",
    "plt.ylabel(\"Path length\")\n",
    "plt.ylim(bottom = 0)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "initial tour length: 5.700632953232227\n",
      "step: 1 - new tour length: 5.515419269149118\n",
      "step: 3 - new tour length: 5.497565020001184\n",
      "step: 5 - new tour length: 5.110778804148519\n",
      "step: 10 - new tour length: 4.856162321689546\n",
      "step: 12 - new tour length: 4.296189487735497\n",
      "step: 16 - new tour length: 3.7632261564806164\n",
      "step: 24 - new tour length: 3.528456460172212\n",
      "step: 35 - new tour length: 3.5215705748430226\n",
      "step: 40 - new tour length: 3.521570574843022\n",
      "step: 47 - new tour length: 3.080087487389965\n",
      "step: 48 - new tour length: 3.0001192492049693\n",
      "step: 55 - new tour length: 2.822090126858674\n",
      "step: 96 - new tour length: 2.7635735777577013\n",
      "step: 1097 - no improvement for 1000 steps.\n",
      "CPU times: user 52.3 ms, sys: 0 ns, total: 52.3 ms\n",
      "Wall time: 49.5 ms\n",
      "Tour length: 2.76\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%time tour, history_reverse = FCSHC(tsp, initial_tour, local_move_method = move_reverse, keep_history = True)\n",
    "show_tsp(tsp, tour)\n",
    "\n",
    "plt.plot(range(len(history_reverse)), history_reverse)\n",
    "plt.xlabel(\"Step\")\n",
    "plt.ylabel(\"Path length\")\n",
    "plt.ylim(bottom = 0)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "All three learning curves"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(range(len(history_swap)), history_swap, label = \"swap\")\n",
    "plt.plot(range(len(history_swap_neighbors)), history_swap_neighbors, label = \"swap_neighbors\")\n",
    "plt.plot(range(len(history_reverse)), history_reverse, label = \"reverse\")\n",
    "plt.xlabel(\"Step\")\n",
    "plt.ylabel(\"Path length\")\n",
    "plt.ylim(bottom = 0)\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Comparison of Local Move Strategies\n",
    "\n",
    "We compare the local move strategies using 100 random problems."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "N = 100 # replications\n",
    "n = 20 # number of cities\n",
    "\n",
    "results = { \n",
    "    'swap' : np.zeros(shape = [N]),\n",
    "    'swap_neighbors' : np.zeros(shape = [N]),\n",
    "    'reverse' : np.zeros(shape = [N]) \n",
    "}\n",
    "\n",
    "for i in range(N):\n",
    "    tsp_test = random_tsp(n)\n",
    "    initial_tour = random_tour(n)\n",
    "    \n",
    "    results['swap'][i] = tour_length(tsp_test, \n",
    "        FCSHC(tsp_test, initial_tour, local_move_method = move_swap, verbose = False))\n",
    "    \n",
    "    results['swap_neighbors'][i] = tour_length(tsp_test, \n",
    "        FCSHC(tsp_test, initial_tour, local_move_method = move_swap_neighbors, verbose = False))\n",
    "    \n",
    "    results['reverse'][i] = tour_length(tsp_test, \n",
    "        FCSHC(tsp_test, initial_tour, local_move_method = move_reverse, verbose = False))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    swap  swap_neighbors  reverse\n",
      "0   4.10            7.17     3.58\n",
      "1   3.30            6.59     3.65\n",
      "2   4.54            6.66     3.75\n",
      "3   4.43            8.49     4.11\n",
      "4   4.13            7.09     3.66\n",
      "..   ...             ...      ...\n",
      "95  4.86            7.22     4.67\n",
      "96  5.08            9.18     4.43\n",
      "97  4.37            7.72     4.15\n",
      "98  4.93            8.90     3.66\n",
      "99  4.14            7.87     3.95\n",
      "\n",
      "[100 rows x 3 columns]\n",
      "swap              4.56\n",
      "swap_neighbors    8.00\n",
      "reverse           3.92\n",
      "dtype: float64\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# compare averages\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "results = pd.DataFrame(results)\n",
    "print(results)\n",
    "\n",
    "means = np.mean(results, axis = 0)\n",
    "print(means)\n",
    "\n",
    "plt.bar(means.keys(), means)\n",
    "plt.ylabel(\"mean tour length\")\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAEGCAYAAABiq/5QAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8rg+JYAAAACXBIWXMAAAsTAAALEwEAmpwYAAAcMUlEQVR4nO3de3RU9bn/8ffDpUbkImroQWggqFBiAohRQApCFeGoRwShwpIqqE0Btag/KtB2qafwW9YjXo616KJi4/rJQQTxsqgieCtyqlwCwaCoCKIG5KoeQQ7K5fn9MUMMIZcBZs+eZH9ea7GYvWfn+30y6Cc7373nGXN3REQkOuqFXYCIiKSWgl9EJGIU/CIiEaPgFxGJGAW/iEjENAi7gEScdtpp3rZt27DLEBGpVYqKina4e2bF/bUi+Nu2bcuKFSvCLkNEpFYxs08r26+lHhGRiFHwi4hEjIJfRCRiasUav4gk1759+ygtLWXv3r1hlyJJkJGRQevWrWnYsGFCxyv4RSKotLSUJk2a0LZtW8ws7HLkOLg7O3fupLS0lOzs7IS+Rks9IhG0d+9eTj31VIV+HWBmnHrqqUf121tgwW9mT5jZNjNbU8lz483Mzey0oOYXkeop9OuOo/23DPKMvxAYUHGnmf0E6Ad8FuDcIiJShcDW+N19sZm1reSpB4E7gBeCmltEjs6Diz5K6ni39Wuf1PEkuVJ6cdfMrgA2ufvqmn41MbMCoAAgKysrBdVVb1rxtLLHY7uMDbESEZHjk7KLu2bWCPg9cGcix7v7dHfPd/f8zMwjWk2ISC337bffctlll9G5c2dyc3O59957GTx4MAAvvPACJ554It9//z179+6lXbt2APz1r3/lvPPOo3Pnzlx11VXs2bMHgJEjRzJ69Gh69epF+/btmT9/fmjfV22Qyrt6zgCygdVmthFoDaw0s39JYQ0ikiYWLFjA6aefzurVq1mzZg2jR49m1apVALz11lvk5uayfPlyli5dSrdu3QAYPHgwy5cvZ/Xq1XTs2JEZM2aUjbdx40b+8Y9/8Pe//53Ro0frPQrVSFnwu3uJu7dw97bu3hYoBbq6+5ZU1SAi6SMvL49XX32VCRMm8NZbb9GsWTPOPPNM1q5dy7Jly7j99ttZvHgxb731Fr169QJgzZo19OrVi7y8PGbOnMl7771XNt4vfvEL6tWrx1lnnUW7du344IMPwvrW0l6Qt3POAt4GOphZqZndENRcIlL7tG/fnqKiIvLy8pg0aRJ//OMf6dWrFy+//DINGzbk4osvZsmSJSxZsoTevXsDsSWdRx55hJKSEu66667DzuorXjfU7apVCyz43X24u7d094bu3trdZ1R4vq277whqfhFJb5s3b6ZRo0aMGDGC8ePHs3LlSnr37s1DDz1Ejx49yMzMZOfOnXzwwQecffbZAOzatYuWLVuyb98+Zs6cedh4c+bM4eDBg6xfv54NGzbQoUOHML6tWkEtG0QklNsvS0pK+O1vf0u9evVo2LAhjz76KGeffTZbt24tO8Pv1KkTLVq0KDt7nzx5Mt26daNNmzbk5eWxa9eusvE6dOjAhRdeyNatW3nsscfIyMhI+fdUWyj4RSQU/fv3p3///kfs/+6778oeT58+/bDnxowZw5gxYyodr2fPnjz44IPJLbKOUq8eEZGI0Rm/iNR6hYWFYZdQq+iMX0QkYhT8IiIRo+AXEYkYBb+ISMTo4q6IwBv3JHe8vpOSO54klYI/DVXWG139zSMg0fBVqKbcjTfeyO23305OTk6Vx4wcOZLLL7+cIUOGHLb/zTffZOrUqWnVMVTBLyJSg8cffzy0uQ8cOED9+vWTOqbW+EUkFGH14y8sLGTw4MEMGDCAs846izvuuKPsuYULF9KjRw+6du3K0KFD2b17NwB9+vRhxYoVAMyYMYP27dvTp08ffvWrX3HzzTeXff3ixYu54IILaNeuHXPnzi3b/8033zBo0CBycnIYPXo0Bw8eBGDWrFnk5eWRm5vLhAkTyo5v3Lgxd955J926dePtt99m4sSJ5OTk0KlTJ8aPH39crzso+EUkJGH24y8uLmb27NmUlJQwe/ZsPv/8c3bs2MGUKVN49dVXWblyJfn5+TzwwAOHfd3mzZuZPHky77zzDosWLTqi9fMXX3zBkiVLmD9/PhMnTizbv2zZMu6//35KSkpYv3498+bNY/PmzUyYMIHXX3+d4uJili9fzvPPPw/Efijm5uaydOlScnJyeO6553jvvfd49913+cMf/nBcrzso+EUkJGH247/oooto1qwZGRkZ5OTk8Omnn/LOO+/w/vvv07NnT7p06cKTTz7Jp59+etjXLVu2jAsvvJBTTjmFhg0bMnTo0MOev/LKK6lXrx45OTls3bq1bP/5559Pu3btqF+/PsOHD2fJkiUsX76cPn36kJmZSYMGDbjmmmtYvHgxAPXr1+eqq64CoGnTpmRkZHDjjTcyb948GjVqdHwvPFrjF5GQHOrH/9JLLzFp0iQuueSSI/rxjxw5kgMHDjB16lQgtqTz/PPP07lzZwoLC3nzzTfLxjuafvwnnHBC2eP69euzf/9+3J1+/foxa9asKr/O3av9nsqPW/7YymqrbqyMjIyydf0GDRqwbNkyXnvtNZ5++mkeeeQRXn/99WrrqImCX0RCuVNo8+bNnHLKKYwYMYLGjRtTWFjIrbfeyrXXXsu1115b1o9/y5YtVfbjb9WqVdl4c+bM4brrruOTTz45pn783bt356abbuLjjz/mzDPPZM+ePZSWltK+/Q931J1//vncdtttfPXVVzRp0oRnn32WvLy8GsdetmwZn3zyCW3atGH27NkUFBTQrVs3xo0bx44dO2jevDmzZs3illtuOeJrd+/ezZ49e7j00kvp3r07Z5555lF9X5VR8ItIKNKtH39mZiaFhYUMHz68rDX0lClTDgv+Vq1a8bvf/Y5u3bpx+umnk5OTQ7NmzWocu0ePHkycOJGSkhJ69+7NoEGDqFevHvfccw99+/bF3bn00ksZOHDgEV+7a9cuBg4cyN69e3H3pLSetpp+dUkH+fn5fuiKelimFU8rezy2y9hA59J9/BGVwvv4165dS8eOHY97nHRR1T30Qdi9ezeNGzdm//79DBo0iOuvv55BgwYFPm9NKvs3NbMid8+veKwu7oqIHIW7776bLl26kJubS3Z2NldeeWXYJR01LfWISK1XWT/+V1555bB74wGys7N57rnnjmuuQxeaa7PAgt/MngAuB7a5e258333AvwHfA+uBUe7+dVA1iEh0VfXRjhLsUk8hMKDCvkVArrt3Aj4C1HRERCTFAgt+d18MfFlh30J33x/ffAdoHdT8IiJSuTDX+K8HZlf1pJkVAAUAWVlZqapJ5AeJ3GWjTplSC4US/Gb2e2A/MLOqY9x9OjAdYrdzpqg0kUgqf7tyMgR9y7Mcn5QHv5ldR+yi70VeG95EICKBc3fcnXr1krf6HMSYdUVKXxEzGwBMAK5w9z2pnFtE0svGjRvp2LEjY8eOpWvXrkyePJnzzjuPTp06cddddwEwYcIEpk374beRu+++m/vvvx+A++6774jjK475+eefM3LkSHJzc8nLyyt71+v69esZMGAA5557Lr169aq2oVtdFFjwm9ks4G2gg5mVmtkNwCNAE2CRmRWb2WNBzS8i6e/DDz/k2muv5d5772XTpk0sW7aM4uJiioqKWLx4McOGDWP27B8uBT7zzDMMHTqUhQsXsm7duiOOLz/mqlWr2LFjB5s2bWLNmjWUlJQwatQoAAoKCvjzn/9MUVERU6dOZezYaC1NBbbU4+7DK9k9o5J9IhJRbdq0oXv37owfP56FCxdyzjnnALG2COvWreOGG25g27ZtbN68me3bt9O8eXOysrJ4+OGHKz0+KyurbEyAdu3asWHDBm655RYuu+wyLrnkEnbv3s0///nPw1oqH+rNExV6566IhOakk04CYuvxkyZN4te//vURxwwZMoS5c+eyZcsWhg0bVu3xGzduLBsToHnz5qxevZpXXnmFv/zlLzzzzDM89NBDnHzyyRQXFwf3jaU5XfUQkdD179+fJ554ouyjDjdt2sS2bdsAGDZsGE8//TRz584ta8JW3fHl7dixg4MHD3LVVVcxefJkVq5cSdOmTcnOzmbOnDlA7IfI6tWrU/Ftpg2d8YtI6LdfXnLJJaxdu5YePXoAsc+cfeqpp2jRogVnn302u3btolWrVrRs2bLa4yt+KPmmTZsYNWpU2Wfc3nNP7L0ZM2fOZMyYMUyZMoV9+/YxbNgwOnfunKpvN3QKfhEJRdu2bVmzZk3Z9rhx4xg3blylx5aUlByxr6rjy4/ZuXNnVq5cecQx2dnZLFiw4FjKrhO01CMiEjEKfhGRiFHwi0SU3jhfdxztv6WCXySCMjIy2Llzp8K/DnB3du7ceVSfMayLuyIR1Lp1a0pLS9m+fXvYpUgSZGRk0Lp14l3uFfwiqZDoB6mnSMOGDcnOzg67DAmJlnpERCJGwS8iEjEKfhGRiFHwi4hEjIJfRCRiFPwiIhGj4BcRiRgFv4hIxCj4RUQiRsEvIhIxCn4RkYgJLPjN7Akz22Zma8rtO8XMFpnZuvjfzYOaX0REKhfkGX8hMKDCvonAa+5+FvBafFtERFIosOB398XAlxV2DwSejD9+ErgyqPlFRKRyqW7L/GN3/wLA3b8wsxZVHWhmBUABQFZWVorKS55pxdPKHo/tMrbaYx9c9FFS5qw4zm392idlXEkzibR47jsp+Dqk1krbi7vuPt3d8909PzMzM+xyRETqjFQH/1YzawkQ/3tbiucXEYm8VAf/i8B18cfXAS+keH4RkcgL8nbOWcDbQAczKzWzG4A/Af3MbB3QL74tIiIpFNjFXXcfXsVTFwU1p4iI1CxtL+6KiEgwFPwiIhGj4BcRiRgFv4hIxCj4RUQiRsEvIhIxCn4RkYhR8IuIRIyCX0QkYlLdllkqsfKb2WWPuza9utJj1HI5TSXSIlkkzeiMX0QkYmoMfjNbYWY36fNxRUTqhkTO+IcBpwPLzexpM+tvZhZwXSIiEpAag9/dP3b33wPtgf8CngA+M7N/N7NTgi5QRESSK6E1fjPrBNwP3Ac8CwwBvgFeD640EREJQo139ZhZEfA1MAOY6O7fxZ9aamY9A6xNREQCkMjtnEPdfUNlT7j74CTXIyIiAUtkqedGMzv50IaZNTezKcGVJCIiQUok+P/V3b8+tOHuXwGXBlaRiIgEKpHgr29mJxzaMLMTgROqOV5ERNJYImv8TwGvmdnfAAeuB548nknN7Dbgxvh4JcAod997PGOKiEhiagx+d/8PMysBLgIMmOzurxzrhGbWCvgNkOPu/2tmzxB7k1jhsY4pIiKJS6hJm7u/DLyc5HlPNLN9QCNgcxLHFhGRaiRyH/9g4F6gBbEzfgPc3Zsey4TuvsnMpgKfAf8LLHT3hZXMWwAUAGRlZR3LVFJBxQ6foC6fIlGUyMXd/wCucPdm7t7U3Zsca+hD7HZQYCCQTawH0ElmNqLice4+3d3z3T0/MzPzWKcTEZEKEgn+re6+NolzXgx84u7b3X0fMA+4IInji4hINRJZ419hZrOB54FD7Rpw93nHOOdnQHcza0RsqeciYMUxjiUiIkcpkeBvCuwBLim3z4mdqR81d19qZnOBlcB+YBUw/VjGEhGRo5fI7Zyjkj2pu98F3JXscUVEpGaJfAJXezN7zczWxLc7mdkfgi9NRESCkMjF3b8Ck4B9AO7+LrE3XImISC2USPA3cvdlFfbtD6IYEREJXiLBv8PMziB2QRczGwJ8EWhVIiISmETu6rmJ2F03PzWzTcAnwBFvuBIRkdohkbt6NgAXm9lJQD133xV8WSIiEpREevXcWWEbAHf/Y0A1iYhIgBJZ6vm23OMM4HIgmS0cREQkhRJZ6rm//Ha8s+aLgVUkIiKBSqgffwWNgHbJLqS2mlY8rezx2C5jEz4+kWOrU1mLZQHeuKfmY/pOCr6OuiSR1xT0utYiiazxlxC/lROoD2QCWt8XEamlEjnjv7zc4/3E2jTrDVwiIrVUIsFf8fbNpofu7AFw9y+TWpGIiAQqkeBfCfwE+IrYxy6eTKynPsSWgLTeLyJSiyTSsmEB8G/ufpq7n0ps6Weeu2e7u0JfRKSWSST4z3P3lw5tuPvLwIXBlSQiIkFKZKlnR7z//lPElnZGADsDrUpERAKTyBn/cGK3cD4X/5MZ3yciIrVQIu/c/RIYZ2aN3X13CmoSEZEAJfLRixeY2fvA+/HtzmY2rYYvExGRNJXIUs+DQH/i6/ruvhroHWRRIiISnESCH3f/vMKuA8czqZmdbGZzzewDM1trZj2OZzwREUlcInf1fG5mFwBuZj8CfsPxt2X+T2CBuw+Jj9noOMcTEZEEJXLGP5rYxy+2AkqBLvHtY2JmTYktFc0AcPfv3f3rYx1PRESOTrVn/GZWH3jI3a9J4pztgO3A38ysM1AEjHP38h/4gpkVAAUAWVlZSZw+9cq3bgZ4e/1OvtuutsqSBhJtuSx1SrVn/O5+AMiML8ckSwOgK/Cou59D7BO+JlYy93R3z3f3/MzMzCROLyISbYms8W8E/tvMXqTcxzC6+wPHOGcpUOruS+Pbc6kk+EVEJBhVnvGb2f+LP7wamB8/tkm5P8fE3bcQu2DcIb7rIuLvERARkeBVd8Z/rpm1IdaC+c9JnvcWYGZ8CWkDMCrJ44uISBWqC/7HiLVkzgZWlNtvHGcffncvBvKP9etFROTYVbnU4+4Pu3tH4G/u3q7cH/XhFxGpxWq8j9/dx6SiEBERSY2EWjaIiEjdoeAXEYkYBb+ISMQo+EVEIkbBLyISMQp+EZGIUfCLiESMuXvYNdQoPz/fV6xYUfOBSfbLZycftt3jjFNr/Jq31+9M2vxdm16dtLGqclu/9oHPkVJqM1w39J0UdgV1gpkVufsRXRJ0xi8iEjEKfhGRiFHwi4hEjIJfRCRiFPwiIhGj4BcRiRgFv4hIxCj4RUQiRsEvIhIxCn4RkYgJLfjNrL6ZrTKz+WHVICISRWGe8Y8D1oY4v4hIJIUS/GbWGrgMeDyM+UVEoiysM/6HgDuAgyHNLyISWQ1SPaGZXQ5sc/ciM+tTzXEFQAFAVlZW0uuYVjyt7PHYLmOTPr6ISLoK44y/J3CFmW0EngZ+bmZPVTzI3ae7e76752dmZqa6RhGROivlwe/uk9y9tbu3BYYBr7v7iFTXISISVbqPX0QkYlK+xl+eu78JvBlmDSIiUaMzfhGRiFHwi4hEjIJfRCRiFPwiIhGj4BcRiRgFv4hIxCj4RUQiRsEvIhIxCn4RkYgJ9Z27qZBIF87yx0Tdg4s+Ou4xbuvXPgmViEhQdMYvIhIxCn4RkYhR8IuIRIyCX0QkYhT8IiIRo+AXEYkYBb+ISMQo+EVEIkbBLyISMQp+EZGIUfCLiERMyoPfzH5iZm+Y2Voze8/MxqW6BhGRKAujSdt+4P+4+0ozawIUmdkid38/hFpERCIn5Wf87v6Fu6+MP94FrAVapboOEZGoCrUts5m1Bc4BllbyXAFQAJCVlZXawo7C2+t3Bjb2ym9mH9XxXZtefdRzJKMNs0jSvXFPzcf0nRR8HXVUaBd3zawx8Cxwq7t/U/F5d5/u7vnunp+ZmZn6AkVE6qhQgt/MGhIL/ZnuPi+MGkREoiqMu3oMmAGsdfcHUj2/iEjUhXHG3xP4JfBzMyuO/7k0hDpERCIp5Rd33X0JYKmeV0REYvTOXRGRiFHwi4hEjIJfRCRiFPwiIhGj4BcRiRgFv4hIxCj4RUQiRsEvIhIxCn4RkYgJtS1zqk0rnlb2uGI75R5nnJrqcuRoJNKmV6Il0f8m0rV9c4j164xfRCRiFPwiIhGj4BcRiRgFv4hIxCj4RUQiRsEvIhIxCn4RkYhR8IuIRIyCX0QkYhT8IiIRo+AXEYmYUILfzAaY2Ydm9rGZTQyjBhGRqEp58JtZfeAvwL8COcBwM8tJdR0iIlEVxhn/+cDH7r7B3b8HngYGhlCHiEgkmbundkKzIcAAd78xvv1LoJu731zhuAKgIL7ZAfgwpYUmz2nAjrCLSHN6jRKj1ykxep1+0MbdMyvuDKMfv1Wy74ifPu4+HZgefDnBMrMV7p4fdh3pTK9RYvQ6JUavU83CWOopBX5Sbrs1sDmEOkREIimM4F8OnGVm2Wb2I2AY8GIIdYiIRFLKl3rcfb+Z3Qy8AtQHnnD391JdRwrV+uWqFNBrlBi9TonR61SDlF/cFRGRcOmduyIiEaPgFxGJGAV/gMysvpmtMrP5YdeSrszsZDOba2YfmNlaM+sRdk3pyMxuM7P3zGyNmc0ys4ywa0oHZvaEmW0zszXl9p1iZovMbF387+Zh1piOFPzBGgesDbuINPefwAJ3/ynQGb1eRzCzVsBvgHx3zyV2U8SwcKtKG4XAgAr7JgKvuftZwGvxbSlHwR8QM2sNXAY8HnYt6crMmgK9gRkA7v69u38dalHpqwFwopk1ABqh974A4O6LgS8r7B4IPBl//CRwZSprqg0U/MF5CLgDOBhyHemsHbAd+Ft8SexxMzsp7KLSjbtvAqYCnwFfAP/j7gvDrSqt/djdvwCI/90i5HrSjoI/AGZ2ObDN3YvCriXNNQC6Ao+6+znAt+jX8iPE16gHAtnA6cBJZjYi3KqkNlPwB6MncIWZbSTWffTnZvZUuCWlpVKg1N2XxrfnEvtBIIe7GPjE3be7+z5gHnBByDWls61m1hIg/ve2kOtJOwr+ALj7JHdv7e5tiV2Ee93ddYZWgbtvAT43sw7xXRcB74dYUrr6DOhuZo3MzIi9TroIXrUXgevij68DXgixlrQURndOkfJuAWbG+zZtAEaFXE/acfelZjYXWAnsB1ahtgQAmNksoA9wmpmVAncBfwKeMbMbiP3QHBpehelJLRtERCJGSz0iIhGj4BcRiRgFv4hIxCj4RUQiRsEvIhIxCn6pc+IdP8cGOP7dZjY+gHFvNbNG5bZ3J3sOEVDwS910MpC04Dez+skaqwa3EmvAJhIoBb/URX8CzjCzYjO7z2Lui/eyLzGzqwHMrE/5z0ows0fMbGT88UYzu9PMllDNG4DM7AwzW2BmRWb2lpn9NL6/0MweNrN/mtkGMxsS31/PzKbFe+vPN7OXzGyImf2GWB+eN8zsjXLj/18zW21m75jZjwN4rSSCFPxSF00E1rt7F3f/LTAY6EKs3//FwH2HernUYK+7/8zdn67mmOnALe5+LjAemFbuuZbAz4DLif0wIl5LWyAPuBHoAeDuDxNrtdzX3fvGjz0JeMfdOwOLgV8lULNIjdSyQaLgZ8Asdz9ArIHXP4DzgG9q+LrZ1T1pZo2JNUubE2uhA8AJ5Q553t0PAu+XO1v/GTAnvn9L+bP7SnwPHPqNpAjoV0O9IglR8EsUWBX793P4b70VP87w2xrGrQd87e5dqnj+u0pqqKqWyuzzH3qqHED/v0qSaKlH6qJdQJNy24uBq+OfgZxJ7FO/lgGfAjlmdoKZNSPW9TJh7v4N8ImZDQWIX0voXMOXLQGuiq/1/5hYg7Gq6hYJhM4gpM5x951m9t/xD+B+mdgnofUAVgMO3BFvCY2ZPQO8C6wj1vXyaF0DPGpmfwAaEvv8hdXVHP8ssR8wa4CPgKXA/8Sfmw68bGZflFvnF0k6decUSTEza+zuu83sVGK/efQ89INIJBV0xi+SevPN7GTgR8Bkhb6kms74RUQiRhd3RUQiRsEvIhIxCn4RkYhR8IuIRIyCX0QkYv4/nTLH+3eAlWsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# histogram\n",
    "\n",
    "plt.hist(results['swap'], bins = 20, alpha = .5)\n",
    "plt.hist(results['swap_neighbors'], bins = 20, alpha = .5)\n",
    "plt.hist(results['reverse'], bins = 20, alpha = .5)\n",
    "plt.xlabel(\"tour length\")\n",
    "plt.ylabel(\"frequency\")\n",
    "plt.legend(labels = [\"swap\", \"swap_neighbors\", \"reverse\"])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXAAAAD5CAYAAAA+0W6bAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8rg+JYAAAACXBIWXMAAAsTAAALEwEAmpwYAAAQa0lEQVR4nO3df3Ac9X3G8efBhgI2EFoHJaEZFCgBEbsQEOQHhkg1EBqYQlpSoklmQqvW01+0pUmLGYcSOnExLdMp005CXJzYHVIxwUDo4NSYMrqAUwO1AYOxEujwqwVayCQxETWNMZ/+cUs4hGSdbk+6+5zerxmNpL3vfvezWu2ze9+923NECACQzz6tLgAA0BgCHACSIsABICkCHACSIsABIKm5M7mwBQsWRHd390wucka9/PLLmjdvXqvLQAPYdrl1+vbbunXr9yPi7WOnz2iAd3d3a8uWLTO5yBlVqVTU19fX6jLQALZdbp2+/Ww/Pd50hlAAICkCHACSIsABICkCHACSIsABICkCHLPa0NCQFi5cqCVLlmjhwoUaGhpqdUlA3Wb0ZYRAOxkaGtLy5cu1evVq7dmzR3PmzNHg4KAkaWBgoMXVAZPjDByz1ooVK7R69Wr19/dr7ty56u/v1+rVq7VixYpWlwbUhQDHrDUyMqLFixe/adrixYs1MjLSooqAqSHAMWv19PRo06ZNb5q2adMm9fT0tKgiYGoIcMxay5cv1+DgoIaHh/Xqq69qeHhYg4ODWr58eatLA+rCRUzMWq9fqLz44os1MjKinp4erVixgguYSIMAx6w2MDCggYGBjr8ZEjoTQygAkBQBDgBJEeAAkBQBDgBJEeAAkBQBDgBJEeAAkBQBDgBJEeAAkBQBDgBJEeAAkBQBDgBJEeAAkBQBDgBJEeAAkBQBDgBJEeAAkBQBDgBJEeAAkBQBDgBJEeAAkBQBDgBJEeAAkBQBDgBJEeAAkBQBDgBJTRrgtr9q+wXb22um/aztO20/Xnw/dHrLBACMVc8Z+BpJZ4+ZtkzSXRFxtKS7it8BADNo0gCPiLsl/WDM5PMkrS1+Xivp/OaWBQCYzNwG5+uKiOclKSKet33YRA1tL5W0VJK6urpUqVQaXGT7Gx0d7ej1y6y/v78p/QwPDzelHzTXbN33Gg3wukXEKkmrJKm3tzf6+vqme5EtU6lU1Mnrl1lE7PXx7mXr9dTKc2aoGjTbbN33Gn0Vyv/YfqckFd9faF5JAIB6NBrg/yzpM8XPn5F0W3PKAQDUq56XEQ5J2izpGNv/ZXtQ0kpJZ9p+XNKZxe8AgBk06Rh4RAxM8NCSJtcCAJgC3okJAEkR4ACQFAEOAEkR4ACQFAEOAEkR4ACQFAEOAEkR4ACQFAEOAEkR4ACQFAEOAEkR4ACQFAEOAEkR4ACQFAEOAEkR4ACQFAEOAEkR4ACQFAEOAEkR4ACQFAEOAEkR4ACQFAEOAEkR4ACQFAEOAEkR4ACQ1NxWFwA0w/FXbtTOXbtL9dG9bH2p+Q85YF9tu+KsUn0AU0GAoyPs3LVbT608p+H5K5WK+vr6StVQ9gAATBVDKACQFAEOAEkR4ACQFAEOAEkR4ACQFAEOAEkR4ACQFAEOAEkR4ACQFAEOAEkR4ACQVKkAt32J7Udtb7c9ZHv/ZhUGANi7hgPc9uGS/lBSb0QslDRH0iebVRgAYO/KDqHMlXSA7bmSDpT0XPmSAAD1aPh2shHxrO1rJD0jaZekjRGxcWw720slLZWkrq4uVSqVRhfZ9kZHRzt6/dpdmb99s7Yd2781Zuu+13CA2z5U0nmS3iPpR5Jusv3piLihtl1ErJK0SpJ6e3uj7D2X21kz7imNBm1YX+pv35RtV7IGNG627ntlPtDhDElPRsSLkmT7FkkflnTDXucCpsFBPcu0aO2ycp2sLVuDJDX+oRLAVJUJ8GckfdD2gaoOoSyRtKUpVQFT9OORlXwiD2adhi9iRsR9ktZJekDSI0Vfq5pUFwBgEqU+EzMirpB0RZNqAQBMAe/EBICkCHAASIoAB4CkCHAASIoAB4CkCHAASIoAB4CkCHAASIoAB4CkCHAASIoAB4CkSt0LBWgnpe8GuKHc/IccsG+55QNTRICjI5S5laxUDf+yfQAzjSEUAEiKAAeApAhwAEiKAAeApAhwAEiKAAeApAhwAEiKAAeApAhwAEiKAAeApAhwAEiKAAeApAhwAEiKAAeApAhwAEiKAAeApAhwAEiKAAeApAhwAEiKAAeApAhwAEiKAAeApAhwAEhqbqsLAGaC7cnbXD15PxHRhGqA5uAMHLNCROz1a3h4eNI2hDfaTakAt/022+tsf9f2iO0PNaswAMDelR1CuVbShoi4wPZ+kg5sQk0AgDo0HOC2D5Z0uqSLJCkifiLpJ80pCwAwmTJDKEdKelHS12w/aPt62/OaVBcAYBJu9MKM7V5J90o6NSLus32tpJci4vIx7ZZKWipJXV1dJ914440lS25fo6Ojmj9/fqvLQAPYdrl1+vbr7+/fGhG9Y6eXCfB3SLo3IrqL30+TtCwizplont7e3tiyZUtDy8ugUqmor6+v1WWgAWy73Dp9+9keN8AbHkKJiP+W9J+2jykmLZG0o9H+AABTU/ZVKBdL+nrxCpQnJP1G+ZIAAPUoFeAR8ZCkt5zWAwCmH+/EBICkCHAASIqbWdWpnpsh1YP7aQBoFs7A61TPjY6OuPR2boYEYMYQ4ACQFAEOAEkR4ACQFAEOAEkR4ACQFAEOAEkR4ACQFAEOAEkR4ACQFAEOAEkR4ACQFAEOAEkR4ACQFAEOAEkR4ACQFAEOAEnxiTyF46/cqJ27dpfup3vZ+obnPeSAfbXtirNK1wBgdiDACzt37dZTK88p1UelUlFfX1/D85cJfwCzD0MoAJAUAQ4ASRHgAJAUY+AA2p7t0n1ERBMqaS+cgQNoexGx168jLr190jadiAAHgKQIcABIigAHgKS4iFk4qGeZFq1dVr6jtWVqkKRybyYCMHsQ4IUfj6zknZgAUmEIBQCSIsABICkCHACSIsABICkuYtZoykXEDeXuBw4A9SLAC2VfgSJVDwDN6AeYTdrhw1SknB+oQoADaKl2+DAVKefLeEuPgdueY/tB27c3oyAAQH2acRHzjySNNKEfAMAUlApw2z+v6nu/r29OOQCAepUdA/9bSX8m6aCJGtheKmmpJHV1dalSqZRcZHvr9PXrVKOjo2y7Fir7t2/W9sv2P9BwgNs+V9ILEbHVdt9E7SJilaRVktTb2xtlLzS0tQ3rS19IQWs04yIYGtSE/aYp2y/h/lvmDPxUSb9i+2OS9pd0sO0bIuLTzSkNwGzQDncCrdYhZbsbaMMBHhGXSbpMkooz8M8R3gCmqh3uBCrN0pcRAgBaoylv5ImIiqRKM/oCANSHM3AASIoAB4CkCHAASIqbWQFouVbfylnKeTtnAhxAS3Er58YxhAIASRHgAJAUAQ4ASRHgAJAUAQ4ASRHgAJAULyOsk+362l2998cjognVAABn4HWLiEm/hoeHJ20DAM1CgANAUgQ4ACTFGDiAtlfPNajZeP2JM3AAbY/rT+MjwAEgKQIcAJIiwAEgKQIcAJIiwAEgKQIcAJIiwAEgKQIcAJLyTL7A3faLkp6esQXOvAWSvt/qItAQtl1unb79joiIt4+dOKMB3ulsb4mI3lbXgalj2+U2W7cfQygAkBQBDgBJEeDNtarVBaBhbLvcZuX2YwwcAJLiDBwAkiLAASApAhwAkiLAgXHYvt72cZO0WWP7gnGm99m+ffqq60yuamomTUef7aRjV6ws2/Nsr7e9zfZ225favqV47Dzbu2zvZ3t/208U03/b9r8X89xs+8Bi+hrb19m+x/Zjts9t5bphchHxWxGxoxXLtj2nFcttBdvdtkdsf0nSA5IuL/ahh21fWbS52vbv1czzBdufLX7+03Haj+3z3cU+uN32I7YvKdodZXuD7a3FvnnsTK9/WQT4xM6W9FxEHB8RCyVdJ+n9xWOnSdou6WRJH5B0XzH9log4OSKOlzQiabCmv25JH5F0jqTrbO8//avQXlp1ULR9ke1bip31cdt/VfPYWbY3237A9k225xfTK7Z7i58Hi2VUbP+D7b+v6f502/9m+4kxZ+MH277V9o6izn2KvgaKENluv/ExvLZHbf+F7fskfcj2ymLeh21fU/qP396OkfSPki6VdLikUySdIOkk26dLulHShTXtf13STbbPknT0OO1/2mdEvF/Vt9kfHhELI2KRpK8VbVZJujgiTpL0OUlfmrY1nCYE+MQekXRGcfQ/LSJ2SvoP2z2q/sP8jaTTVQ3ze4p5FhaB8oikT0l6X01/34iI1yLicUlPSEp3tG+CVh4UT1A1BBZJutD2u20vkPR5SWdExImStkj6k9qZbL9L0uWSPijpTL11u71T0mJJ50paWTP9FEmfLZZ3lKRfLfq6WtIvFfWcbPv8ov08Sdsj4gOSdkj6uKT3RcQvSvriXtarEzwdEfdKOqv4elDVM+djJR0dEQ9KOsz2u2wfL+mHEfHMRO3H9ClV97cjbf+d7bMlvVQcqD+s6oHgIUlfUXVbpjK31QW0q4h4zPZJkj4m6SrbG1UN6l+WtFvSv0paI2mOqkdvFb+fHxHbbF8kqa+2y7GLmK7a29gjkq4pzjxvj4h7bI93UJyjNx8UvyjpbZLmS7qjpr9vRMRrkh4vztiPlfTQBMu+qzgIy/YOSUcUfR4n6Tu2JWk/SZvHzHeKpG9HxA+KeW+S9N6ax79Z1LDDdlfN9Psj4vVnEUOqhvxuSZWIeLGY/vVifb8paY+km4t5X5L0iqTrba+X1Onj6S8X3y3pqoj4yjht1km6QNI7VD0jn7C97e6aPhURPyyC/6OSfl/VM/g/lvSjiDihaWvRApyBT6A4W/rfiLhB0jWSTpR0t6obfnOxE/6cqqHxaDHbQZKet72vqmfgtT5hex/bR0k6UtL3pn8t2ktEPCbpJFWD/Crbf663HhQXF193F7OtkfQHxVPfKyXVnmVP5aD4fzU/71H15MWS7oyIE4qv4yJicMx8nmS1avutbTtebXvr65WI2CNJEfGqqgeOmyWdL2nDJDV0ijsk/WbNMNbhtg8rHrtR0idVDfF1dbT/qeKZ1j4RcbOqz6ZOjIiXJD1p+xNFGxchnwoBPrFFku4vnl4tV/Vp7H2SuvRGuDws6eF44+2slxdt7pT03TH9fU/StyX9i6TfiYhXprX6NtSGB8V7JZ1q+xeK+g60/d4xbe6X9BHbh9qeK+nX6uz7FNvvKca+L5S0SdX/jY/YXuDqhcoBVf8n3qQIpEMi4luq/m1OmOJ6pRQRGyX9k6TNxTDkOlW3vyLi0eLnZyPi+cnaj3G4pEqxL6+RdFkx/VOSBm1vU/X/7bzpWbPpwxDKBCLiDr356frrfqamzdIx83xZ0pcn6PI7EXFJ8ypMaZGkv7b9mqpn3L+r6o4z9qD4wjgHxadVPXOv3UFfPyh2qYGDYkS8WAx1Ddl+fbt+XtJjNW2etf2XRQ3PqTo+vbOO7jerOia+qFi3WyPiNduXSRpW9Wz8WxFx2zjzHiTptmJM35I69v8mIp6StLDm92slXTtB20XjTJuofW2f21Q9WRg775OqXpdJi3uhzADba1Qd8103WVvUZyb/prbnR8RocQZ+q6SvRsSt071cYDKcgc+AiLio1TWglC/YPkPV8feNql50BFqOM3B0DNsfVfVlerWejIiPt6IeYLoR4ACQFK9CAYCkCHAASIoAB4CkCHAASOr/AWyodke6j1Q1AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# boxplot (from pandas.DataFrame)\n",
    "\n",
    "results.boxplot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ttest_relResult(statistic=12.807612401763953, pvalue=9.914781933804352e-23)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# is reverse better than swap?\n",
    "# perform a paired t-test\n",
    "\n",
    "from scipy import stats\n",
    "\n",
    "stats.ttest_rel(results['swap'], results['reverse'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Simulated Annealing [2 Points]\n",
    "\n",
    "Simulated annealing is a form of stochastic hill climbing that also allows downhill moves with a probability proportional to the temperature. This is done to avoid local optima. The temperature is decreased in every iteration following an annealing schedule.\n",
    "\n",
    "Finding a good cooling schedule for the problem is the most challenging part of simulated annealing. Some guidance can be found [here](http://what-when-how.com/artificial-intelligence/a-comparison-of-cooling-schedules-for-simulated-annealing-artificial-intelligence/).\n",
    "\n",
    "The initial temperature $T_0$ should be chosen such that initially any move, no matter how bad, has a high probability of being performed. For $P = exp(-\\Delta E/T_0)$ we get a probability of $exp(-1) = 0.37$ if $T_0$ is equal to the worst $\\Delta E$. This is typically enough.\n",
    "\n",
    "For the schedule, $T_t = T_0 \\alpha^t$ is popular with $\\alpha$ less but close to 1 and $t$ being the time step. Note that this is equivalent to multiplying the current temperature with $\\alpha$ at every step."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "def schedule(t, T0, alpha):\n",
    "    return(T0 * alpha ** t)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ts = range(0,10000)\n",
    "plt.plot(ts, [schedule(t, 100, 0.999) for t in ts])\n",
    "plt.xlabel(\"Step t\")\n",
    "plt.ylabel(\"Temperature\")\n",
    "plt.ylim(bottom = 0)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "def SA(tsp, tour = None, local_move_method = move_reverse, \n",
    "         T0 = None, alpha = 0.999, epsilon = 1e-3, verbose = True, keep_history = False):\n",
    "    \n",
    "    n = len(tsp['pos'])\n",
    "    \n",
    "    if keep_history: history = list() \n",
    "    \n",
    "    # 1. initialize current tour\n",
    "    # current tour = initial tour (random if no tour is given)\n",
    "    if not tour is None:\n",
    "        current_tour = tour\n",
    "    else:\n",
    "        current_tour = random_tour(n)\n",
    "    \n",
    "    # initial tour length\n",
    "    current_length = tour_length(tsp, current_tour)    \n",
    "    if verbose: print(f\"initial tour length: {current_length:3.3}\")\n",
    "    \n",
    "    # use the worst case length. We use the maximal distance for each row of the distance matrix. \n",
    "    # This may not be a valid tour, but it is an upper limit for deltaE.\n",
    "    if T0 is None:\n",
    "        T0 = np.sum(np.amax(tsp[\"dist\"], axis = 0))\n",
    "    \n",
    "    # 2. repeat local moves till temperature is low enough\n",
    "    T = T0\n",
    "    t = 0\n",
    "    while T > epsilon:\n",
    "        \n",
    "        # calculate temperature from schedule\n",
    "        T = schedule(t, T0, alpha)\n",
    "        \n",
    "        # create random move\n",
    "        new_tour = local_move_method(current_tour)\n",
    "        new_length = tour_length(tsp, new_tour)\n",
    "        \n",
    "        deltaE = new_length - current_length\n",
    "        \n",
    "        # check if the new tour is better   \n",
    "        if deltaE < 0 or np.random.rand() < math.exp(-deltaE/T): \n",
    "            current_tour = new_tour\n",
    "            current_length = new_length\n",
    "            if verbose: print(f\"step: {t} \\t temp: {T:5.3f} \\t deltaE: {deltaE:+3.3f} \\t new tour length: {current_length:3.3f}\")\n",
    "      \n",
    "        if keep_history: history.append(current_length)\n",
    "    \n",
    "        t += 1\n",
    "    \n",
    "    if keep_history: return(current_tour, history)\n",
    "    return(current_tour)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 3.77 s, sys: 65.3 ms, total: 3.84 s\n",
      "Wall time: 3.74 s\n",
      "Tour length: 2.76\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%time tour, history = SA(tsp, local_move_method = move_swap, alpha = 0.9999, verbose = False, keep_history = True)\n",
    "show_tsp(tsp, tour)\n",
    "\n",
    "plt.plot(range(len(history)), history)\n",
    "plt.xlabel(\"Step\")\n",
    "plt.ylabel(\"Path length\")\n",
    "plt.ylim(bottom = 0)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Compare Performance [2 Points]\n",
    "\n",
    "Use runtime, scalability (number of cities), and best objective function value to compare the algorithms on boards of different sizes.  \n",
    "\n",
    "For timing you can use the `time` package."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Do something\n",
      "This took: 0.1957416534423828 milliseconds\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "\n",
    "t0 = time.time()\n",
    "print(\"Do something\")\n",
    "t1 = time.time()\n",
    "\n",
    "print(f\"This took: {(t1-t0) * 1e3} milliseconds\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Code and results go here"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Bonus: Genetic Algorithm [+1 Point]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Code goes here"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}