experiments/01_grid/07_very_bad_example.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true,
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "import math\n",
    "\n",
    "from shapely.geometry import Polygon\n",
    "import matplotlib.pyplot as plt\n",
    "from pcpptc import PolygonInstance\n",
    "from pcpptc.instance_converter import RegularSquare\n",
    "from pcpptc.plot import setup_plot, plot_polygon_instance, plot_solution\n",
    "from pcpptc.plot.polgyon_instance import plot_coverage\n",
    "from pcpptc.solver_selection.abstract_solver import PolygonInstanceSolver\n",
    "\n",
    "\n",
    "class OrientedSquare(RegularSquare):\n",
    "    def __init__(self, point_based, full_coverage, angle, translation):\n",
    "        super().__init__(point_based=point_based, full_coverage=full_coverage)\n",
    "        self.angle = angle\n",
    "        self.translation = translation\n",
    "\n",
    "    def __call__(self, pi):\n",
    "        return super().__call__(pi, angle=self.angle, translation=self.translation)\n",
    "\n",
    "\n",
    "class SquareAlgorithm(PolygonInstanceSolver):\n",
    "    def __init__(\n",
    "        self,\n",
    "        full_coverage=False,\n",
    "        point_based=False,\n",
    "        angle: float = 0.0,\n",
    "        translation=(0.0, 0.0),\n",
    "    ):\n",
    "        super().__init__(\n",
    "            OrientedSquare(\n",
    "                point_based=point_based,\n",
    "                full_coverage=full_coverage,\n",
    "                angle=angle,\n",
    "                translation=translation,\n",
    "            )\n",
    "        )\n",
    "\n",
    "    def identifier(self) -> str:\n",
    "        return f\"FOOBAR\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "p = Polygon([(1.0, 1.0), (41.0, 1.0), (41.0, 9.0), (1.0, 9.0)])\n",
    "instance = PolygonInstance(\n",
    "    original_area=p,\n",
    "    feasible_area=p.buffer(-0.9),\n",
    "    valuable_areas=[(p, 1.0)],\n",
    "    tool_radius=1.0,\n",
    "    turn_cost=50.0,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "def plot(solution, ax):\n",
    "    ax.set_aspect(\"equal\", adjustable=\"box\")\n",
    "    plot_polygon_instance(ax, instance)\n",
    "    plot_solution(ax, solution, w=2)\n",
    "    plot_coverage(ax, instance, solution)\n",
    "    ax.plot(\n",
    "        [p.x for p in solution.waypoints],\n",
    "        [p.y for p in solution.waypoints],\n",
    "        \".\",\n",
    "        c=\"black\",\n",
    "    )\n",
    "    plt.tight_layout()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Converting polygon instance to grid instance...\n",
      "Grid Solver:\n",
      "==============================\n",
      "Cycle Cover Solver:\n",
      "Using linear programming and blossom matching to compute a cycle cover with turn-aware, heterogeneous touring costs.\n",
      "IntegralizingFractionalSolverComputing a fractional solution with some BnB steps to improve integrality.Up to 50 BnB steps are performed.\n",
      "NeighborBasedStripStrategy selecting 4 atomic strips first on usage than on minimizing the worst case overhead.\n",
      "------------------------------\n",
      "Local Relaxation CC Optimization:\n",
      " - AreaSelector with area size 50.\n",
      "         Selects the vertex that in sum with its direct neighbors has the highest\n",
      "         coverage costs plus opportunity loss.\n",
      "         \n",
      "- Repeating 25 times.\n",
      "------------------------------\n",
      "Local Relaxation Tour Optimization:\n",
      "- Like CC version but trying to reconnect unconnected solutions up to 10 times.\n",
      " - AreaSelector with area size 50.\n",
      "         Selects the vertex that in sum with its direct neighbors has the highest\n",
      "         coverage costs plus opportunity loss.\n",
      "         \n",
      "- Repeating 25 times.\n",
      "==============================\n",
      "\n",
      "Instance\n",
      "---------------------------------\n",
      "Vertices: 67\n",
      "Edges: 104\n",
      "Max Degree: 4\n",
      "Max value: inf\n",
      "Cycle Cover: Computing fractional solution...\n",
      "Initial fractional solution: 2564.1944901923457\n",
      "Integral fractional solution: 2564.1944901923457\n",
      "Fractional solution: 2564.1944901923457\n",
      "Cycle Cover: Creating matching problem...\n",
      "Using adaptive strip selection strategy.\n",
      "Cycle Cover: Solving matching...\n",
      "Solve matching on 4941 edges.\n",
      "Optimize CC around PointVertex[140646942653968]@(3.042640687119286, 6.414213562373096).\n",
      "Optimize CC around PointVertex[140646933551504]@(5.871067811865476, 6.414213562373096).\n",
      "Optimize CC around PointVertex[140646933551760]@(3.042640687119287, 3.585786437626904).\n",
      "Optimize CC around PointVertex[140646929431824]@(5.871067811865476, 3.585786437626904).\n",
      "Optimize CC around PointVertex[140646929429840]@(8.699494936611666, 6.414213562373096).\n",
      "Optimize CC around PointVertex[140646929431376]@(8.699494936611666, 3.585786437626904).\n",
      "Optimize CC around PointVertex[140646929431952]@(11.527922061357856, 6.414213562373094).\n",
      "Optimize CC around PointVertex[140646929428816]@(11.527922061357856, 3.585786437626906).\n",
      "Optimize CC around PointVertex[140646929430032]@(14.356349186104046, 6.414213562373096).\n",
      "Optimize CC around PointVertex[140646929432272]@(14.356349186104046, 3.585786437626904).\n",
      "Optimize CC around PointVertex[140646929432400]@(17.184776310850236, 6.414213562373096).\n",
      "Optimize CC around PointVertex[140646872224144]@(17.184776310850236, 3.585786437626906).\n",
      "Optimize CC around PointVertex[140646872225808]@(20.013203435596424, 6.414213562373096).\n",
      "Optimize CC around PointVertex[140646872226320]@(20.013203435596427, 3.585786437626904).\n",
      "Optimize CC around PointVertex[140646942616144]@(22.841630560342615, 6.414213562373096).\n",
      "Optimize CC around PointVertex[140646942614416]@(22.841630560342615, 3.585786437626906).\n",
      "Optimize CC around PointVertex[140646942615632]@(25.670057685088807, 6.414213562373096).\n",
      "Optimize CC around PointVertex[140646942615312]@(25.670057685088807, 3.585786437626906).\n",
      "Optimize CC around PointVertex[140646942614096]@(28.498484809835, 6.414213562373096).\n",
      "Optimize CC around PointVertex[140646942616592]@(28.498484809834995, 3.585786437626904).\n",
      "Optimize CC around PointVertex[140646942615440]@(31.326911934581187, 6.414213562373096).\n",
      "Optimize CC around PointVertex[140646942617232]@(31.326911934581187, 3.5857864376269077).\n",
      "Optimize CC around PointVertex[140646942614160]@(34.155339059327375, 6.414213562373098).\n",
      "Optimize CC around PointVertex[140646942614544]@(34.15533905932738, 3.5857864376269077).\n",
      "Optimize CC around PointVertex[140646942616336]@(36.983766184073566, 6.414213562373096).\n",
      "Solution is already connected! :)\n",
      "Optimize tour around PointVertex[140646942653968]@(3.042640687119286, 6.414213562373096).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646929431824]@(5.871067811865476, 3.585786437626904).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646929431376]@(8.699494936611666, 3.585786437626904).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646929431952]@(11.527922061357856, 6.414213562373094).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646929430032]@(14.356349186104046, 6.414213562373096).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646872224144]@(17.184776310850236, 3.585786437626906).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646872226320]@(20.013203435596427, 3.585786437626904).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646942616144]@(22.841630560342615, 6.414213562373096).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646942615632]@(25.670057685088807, 6.414213562373096).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646942616592]@(28.498484809834995, 3.585786437626904).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646942617232]@(31.326911934581187, 3.5857864376269077).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646942614160]@(34.155339059327375, 6.414213562373098).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646942616336]@(36.983766184073566, 6.414213562373096).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646893909072]@(39.81219330881976, 3.5857864376269077).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646929266128]@(36.983766184073566, 3.5857864376269095).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646942615440]@(31.326911934581187, 6.414213562373096).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646942615312]@(25.670057685088807, 3.585786437626906).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646872225808]@(20.013203435596424, 6.414213562373096).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646929432272]@(14.356349186104046, 3.585786437626904).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646929429840]@(8.699494936611666, 6.414213562373096).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646933551760]@(3.042640687119287, 3.585786437626904).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646933551504]@(5.871067811865476, 6.414213562373096).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646929428816]@(11.527922061357856, 3.585786437626906).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646929432400]@(17.184776310850236, 6.414213562373096).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646942614416]@(22.841630560342615, 3.585786437626906).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Touring costs in grid: 2564.194490192347\n",
      "Opportunity loss in grid: nan\n",
      "Done. Converting to tour.\n",
      "Computing coverage polygon\n",
      "Plotting coverage polygon\n",
      "Converting polygon instance to grid instance...\n",
      "Grid Solver:\n",
      "==============================\n",
      "Cycle Cover Solver:\n",
      "Using linear programming and blossom matching to compute a cycle cover with turn-aware, heterogeneous touring costs.\n",
      "IntegralizingFractionalSolverComputing a fractional solution with some BnB steps to improve integrality.Up to 50 BnB steps are performed.\n",
      "NeighborBasedStripStrategy selecting 4 atomic strips first on usage than on minimizing the worst case overhead.\n",
      "------------------------------\n",
      "Local Relaxation CC Optimization:\n",
      " - AreaSelector with area size 50.\n",
      "         Selects the vertex that in sum with its direct neighbors has the highest\n",
      "         coverage costs plus opportunity loss.\n",
      "         \n",
      "- Repeating 25 times.\n",
      "------------------------------\n",
      "Local Relaxation Tour Optimization:\n",
      "- Like CC version but trying to reconnect unconnected solutions up to 10 times.\n",
      " - AreaSelector with area size 50.\n",
      "         Selects the vertex that in sum with its direct neighbors has the highest\n",
      "         coverage costs plus opportunity loss.\n",
      "         \n",
      "- Repeating 25 times.\n",
      "==============================\n",
      "\n",
      "Instance\n",
      "---------------------------------\n",
      "Vertices: 80\n",
      "Edges: 136\n",
      "Max Degree: 4\n",
      "Max value: inf\n",
      "Cycle Cover: Computing fractional solution...\n",
      "Initial fractional solution: 788.3185307179585\n",
      "Integral fractional solution: 788.3185307179585\n",
      "Fractional solution: 788.3185307179585\n",
      "Cycle Cover: Creating matching problem...\n",
      "Using adaptive strip selection strategy.\n",
      "Cycle Cover: Solving matching...\n",
      "Solve matching on 6928 edges.\n",
      "Optimize CC around PointVertex[140646942362448]@(1.95, 1.95).\n",
      "Optimize CC around PointVertex[140646942360848]@(3.95, 3.95).\n",
      "Optimize CC around PointVertex[140646942361872]@(1.95, 5.95).\n",
      "Optimize CC around PointVertex[140646942360592]@(3.95, 7.95).\n",
      "Optimize CC around PointVertex[140646942363600]@(5.95, 1.95).\n",
      "Optimize CC around PointVertex[140646942361296]@(5.95, 5.95).\n",
      "Optimize CC around PointVertex[140646942359760]@(7.95, 3.95).\n",
      "Optimize CC around PointVertex[140646942359696]@(7.95, 7.95).\n",
      "Optimize CC around PointVertex[140646942363024]@(9.95, 1.95).\n",
      "Optimize CC around PointVertex[140646933282256]@(9.95, 5.95).\n",
      "Optimize CC around PointVertex[140646933281616]@(11.95, 3.95).\n",
      "Optimize CC around PointVertex[140646933279248]@(11.95, 7.95).\n",
      "Optimize CC around PointVertex[140647091016144]@(13.95, 1.95).\n",
      "Optimize CC around PointVertex[140646933087696]@(13.95, 5.95).\n",
      "Optimize CC around PointVertex[140646933086416]@(15.95, 3.95).\n",
      "Optimize CC around PointVertex[140646942362384]@(15.95, 7.95).\n",
      "Optimize CC around PointVertex[140646942363280]@(17.95, 1.95).\n",
      "Optimize CC around PointVertex[140646929054800]@(17.95, 5.95).\n",
      "Optimize CC around PointVertex[140646877811600]@(19.95, 3.95).\n",
      "Optimize CC around PointVertex[140646877811792]@(19.95, 7.95).\n",
      "Optimize CC around PointVertex[140646877811280]@(21.95, 1.95).\n",
      "Optimize CC around PointVertex[140646876815504]@(21.95, 5.95).\n",
      "Optimize CC around PointVertex[140646876818256]@(23.95, 3.95).\n",
      "Optimize CC around PointVertex[140646931661712]@(23.95, 7.95).\n",
      "Optimize CC around PointVertex[140646931663760]@(25.95, 1.95).\n",
      "Solution is already connected! :)\n",
      "Optimize tour around PointVertex[140646942362448]@(1.95, 1.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646942363600]@(5.95, 1.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646942363024]@(9.95, 1.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140647091016144]@(13.95, 1.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646942363280]@(17.95, 1.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646877811280]@(21.95, 1.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646931663760]@(25.95, 1.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646929131920]@(29.95, 1.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140647834987728]@(33.95, 1.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646928931856]@(37.95, 1.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646929155920]@(39.95, 3.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646928932816]@(35.95, 3.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646929132560]@(31.95, 3.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646929131984]@(27.95, 3.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646876818256]@(23.95, 3.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646877811600]@(19.95, 3.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646933086416]@(15.95, 3.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646933281616]@(11.95, 3.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646942359760]@(7.95, 3.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646942360848]@(3.95, 3.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646942361296]@(5.95, 5.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646933282256]@(9.95, 5.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646933087696]@(13.95, 5.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646929054800]@(17.95, 5.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Optimize tour around PointVertex[140646876815504]@(21.95, 5.95).\n",
      "Checking for cycles in solution.\n",
      "Found 1 cycles.\n",
      "Touring costs in grid: 788.3185307179585\n",
      "Opportunity loss in grid: nan\n",
      "Done. Converting to tour.\n",
      "Computing coverage polygon\n",
      "Plotting coverage polygon\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 288x288 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(3, 1, sharex=True, figsize=(4, 4))\n",
    "axes[0].set_aspect(\"equal\", adjustable=\"box\")\n",
    "plot_polygon_instance(axes[0], instance)\n",
    "solution = SquareAlgorithm(point_based=False, full_coverage=True, angle=0.25 * math.pi)(\n",
    "    instance\n",
    ")\n",
    "plot(solution, axes[2])\n",
    "solution = SquareAlgorithm(\n",
    "    point_based=False, full_coverage=True, angle=0.0, translation=(-1.95, 0.0)\n",
    ")(instance)\n",
    "plot(solution, axes[1])\n",
    "plt.savefig(\"./output/07_very_bad_example.pdf\")\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}