PerformanceTest_02.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-04-12T11:53:30.250034",
     "start_time": "2016-04-12T11:53:30.235961"
    },
    "collapsed": true,
    "run_control": {
     "marked": false
    }
   },
   "source": [
    "# Implicit Trajectory combined with RRT #\n",
    "- Work space: 3D with obstacles\n",
    "- Obstacle type: static and known\n",
    "- Agent Model: Dubins airplane (Owen2015)\n",
    "- Integration Algorithm: 4th Order Runge Kutta\n",
    "- Implicit Curve: Time-invariant Circle\n",
    "- Distance Metric: admissionable heuristic, based on time-optimal dubins-airplane paths (altitude difference is integrated in a linear fashion)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-05-10T16:40:14.354210",
     "start_time": "2016-05-10T16:40:11.459345"
    },
    "collapsed": true,
    "run_control": {
     "marked": false
    }
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "# load the code\n",
    "%run 'MAIN_CODE.ipynb'\n",
    "\n",
    "import random\n",
    "random.seed( 10 ) #arbitrary"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-04-12T11:00:30.033027",
     "start_time": "2016-04-12T11:00:29.977472"
    },
    "collapsed": false,
    "run_control": {
     "marked": false
    }
   },
   "source": [
    "# Performance Experiments"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true,
    "run_control": {
     "marked": false
    }
   },
   "source": [
    "## 2nd stage: bidirectional VF-RRT+impl.Curve vs. bidirectional RRT with goal-bias\n",
    "compare the subproblem solver with a naive algorithm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-05-10T16:40:14.368006",
     "start_time": "2016-05-10T16:40:14.356035"
    },
    "collapsed": true,
    "run_control": {
     "marked": true
    }
   },
   "outputs": [],
   "source": [
    "import time\n",
    "\n",
    "#Parameters in General\n",
    "NumOfExperiments = 50\n",
    "qstart = np.array([10.,10.,3.,np.pi])# specific initial point\n",
    "MaxNodes = 300\n",
    "MAXiter = MaxNodes*3\n",
    "\n",
    "# Experiment parameters for VF-RRT\n",
    "lambdaNsteps=10\n",
    "Es=0.8\n",
    "lamdb=0.01\n",
    "\n",
    "# Experiment parameters for naive RRT\n",
    "goalBias=0.15"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Prepare Test\n",
    "### Naive Solution, without obstacles"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-05-10T16:40:15.391067",
     "start_time": "2016-05-10T16:40:14.370050"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "finished after 186 iterations\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAV0AAADtCAYAAAAcNaZ2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nNTCVoERNuioqLiAIwl3AOl+oF1SGJv8P2jcGyBoCPcAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fda468ddfd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXEAAAEKCAYAAADkYmWmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fda468d5048>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#%matplotlib inline\n",
    "import numpy as np\n",
    "\n",
    "MAXnodes = 200\n",
    "\n",
    "# start on curve\n",
    "qstart = np.array([xc+r,yc,zc,0.])\n",
    "qstart[-1] = VFangle(qstart)\n",
    "\n",
    "qList = [qstart]\n",
    "distList = []\n",
    "q = qstart\n",
    "for i in range(0,MAXnodes):\n",
    "    #import pdb; pdb.set_trace()\n",
    "    # steer in the direction of the VF\n",
    "    offset = VF(q)*inputBounds[\"Vmax\"]*TimeStep\n",
    "    newangle = VFangle(q[0:3]+offset)\n",
    "    qVF = np.append(q[0:3]+offset,newangle)\n",
    "    qnew = steer(q,qVF)\n",
    "    q = qnew\n",
    "    qList.append(qnew)\n",
    "    \n",
    "    distList.append(dubinsDist(qnew,qstart)) \n",
    "    if distList[-1]<=DistStep*2:\n",
    "        print('finished after %d iterations' % i)\n",
    "        break\n",
    "\n",
    "def plotQlist(qList,text):\n",
    "    # split the List of arrays into plottable data\n",
    "    x,y,z,t = zip(*qList)\n",
    "\n",
    "    fig = plt.figure()\n",
    "    ax = fig.gca(projection='3d')\n",
    "    #ax.plot(curve[2000:,0], curve[2000:,1], curve[2000:,2], label='integral curve', color='green')# the VF traj\n",
    "    ax.plot(x, y, z, label=text,color='red')\n",
    "    ax.set_xlabel('x')\n",
    "    ax.set_ylabel('y')\n",
    "    ax.set_zlabel('z')\n",
    "    ax.legend()\n",
    "\n",
    "\n",
    "plotQlist(qList,'naive solution w/o obstacles')\n",
    "\n",
    "# Plot dubins airplane distance from the newest node to the start node\n",
    "fig = plt.figure()\n",
    "ax = fig.gca()\n",
    "ax.plot(distList)\n",
    "plt.title('dubins airplane distance from node_i to the start node')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Find blocked Nodes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-05-10T16:40:15.543211",
     "start_time": "2016-05-10T16:40:15.416846"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[(45, 50)]\n"
     ]
    }
   ],
   "source": [
    "# Spherical Obstacle Definition\n",
    "ob = dict()\n",
    "ob[\"pos\"] = np.array([xc,yc+r,zc])\n",
    "ob[\"r\"] = 0.2\n",
    "\n",
    "Obstacles = list()\n",
    "Obstacles.append(ob)\n",
    "\n",
    "# Split the qList in subsets whenever an obstacle blocks a node\n",
    "oldResult = isBlocked(qList[0],Obstacles)\n",
    "toConnect = []\n",
    "for i in range(1,len(qList)):\n",
    "    newResult = isBlocked(qList[i],Obstacles)\n",
    "    if newResult and not oldResult:\n",
    "        ilast = i-1\n",
    "    if not newResult and oldResult:\n",
    "        ifirst = i\n",
    "        toConnect.append((ilast,ifirst))\n",
    "    oldResult = newResult\n",
    "\n",
    "print(toConnect)\n",
    "\n",
    "# starting and final configuration\n",
    "l,f=toConnect[0]\n",
    "offset = 2\n",
    "qs = qList[l-offset] # that number is arbitrary\n",
    "qf = qList[f+offset] # that number is arbitrary\n",
    "\n",
    "# sample bounds\n",
    "bounds = dict()\n",
    "edge = 1.0\n",
    "bounds[\"xmin\"] = np.min([qs[0],qf[0]])-edge\n",
    "bounds[\"xmax\"] = np.max([qs[0],qf[0]])+edge\n",
    "bounds[\"ymin\"] = np.min([qs[1],qf[1]])-edge\n",
    "bounds[\"ymax\"] = np.max([qs[1],qf[1]])+edge\n",
    "bounds[\"zmin\"] = np.min([qs[2],qf[2]])-edge\n",
    "bounds[\"zmax\"] = np.max([qs[2],qf[2]])+edge"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-05-10T16:40:15.708200",
     "start_time": "2016-05-10T16:40:15.558198"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "\n",
    "#T1.plot(obs=Obstacles,DrawCollisionNodes=True,DrawTrajectory=False)\n",
    "#T2.plot(obs=Obstacles,DrawCollisionNodes=True,DrawTrajectory=False)\n",
    "#print(\"Min Distance from T1 to T2 %.3f\" % np.min(T1.dists))\n",
    "#print(\"Min Distance from T1 to T2 %.3f\" % np.min(T2.dists))\n",
    "#DrawStats(T1)\n",
    "#DrawStats(T2)\n",
    "\n",
    "#T.plot(obs=Obstacles,DrawCollisionNodesDrawStats(T)=True,DrawTrajectory=False)\n",
    "#DrawStats(T)\n",
    "\n",
    "#plt.figure()\n",
    "#plt.plot(dists)\n",
    "#plt.title('dubins distances min=%.2f' % np.min(dists))\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Run the tests"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-05-10T16:41:15.399376",
     "start_time": "2016-05-10T16:40:15.710452"
    },
    "collapsed": false,
    "run_control": {
     "marked": true
    },
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "#\tt\tsuccess\twork\tupstr\tnodes\titer\teuDist\tangDist\n",
      "VFRRT\n",
      "0\t0.060\tTrue\t0.812\t3.464\t39\t53\t1.192\t7.845\n",
      "1\t0.109\tTrue\t0.858\t1.228\t61\t80\t1.009\t6.086\n",
      "2\t0.434\tTrue\t0.898\t1.688\t138\t153\t1.076\t7.007\n",
      "3\t0.081\tTrue\t0.882\t4.511\t51\t57\t1.354\t6.738\n",
      "4\t1.885\tTrue\t1.011\t20.391\t334\t352\t3.217\t16.530\n",
      "5\t0.015\tTrue\t0.843\t2.350\t15\t17\t1.087\t7.040\n",
      "6\t0.289\tTrue\t0.798\t3.497\t129\t152\t1.192\t7.871\n",
      "7\t0.781\tTrue\t0.790\t4.420\t220\t234\t1.308\t9.105\n",
      "8\t0.418\tTrue\t1.061\t14.017\t146\t169\t2.552\t12.569\n",
      "9\t0.092\tTrue\t0.804\t4.150\t57\t71\t1.273\t7.972\n",
      "10\t0.291\tTrue\t0.831\t2.247\t117\t126\t1.088\t6.113\n",
      "11\t0.311\tTrue\t0.817\t7.779\t128\t136\t1.652\t10.403\n",
      "12\t1.908\tTrue\t0.785\t10.949\t381\t401\t2.007\t11.904\n",
      "13\t0.084\tTrue\t0.903\t1.720\t49\t73\t1.085\t6.649\n",
      "14\t4.219\tTrue\t0.606\t16.476\t573\t593\t2.350\t13.729\n",
      "15\t2.856\tTrue\t1.151\t18.934\t466\t506\t3.248\t16.389\n",
      "16\t0.026\tTrue\t0.912\t2.819\t25\t29\t1.229\t8.435\n",
      "17\t0.052\tTrue\t0.743\t5.361\t39\t55\t1.313\t9.105\n",
      "18\t0.201\tTrue\t0.907\t2.087\t108\t117\t1.129\t6.666\n",
      "19\t0.178\tTrue\t0.793\t4.838\t96\t105\t1.319\t9.025\n",
      "20\t0.038\tTrue\t0.882\t1.784\t28\t42\t1.082\t5.981\n",
      "21\t0.047\tTrue\t0.801\t5.037\t42\t43\t1.342\t7.895\n",
      "22\t4.508\tFalse\t\t\t594\t621\t\t\n",
      "23\t0.090\tTrue\t0.905\t1.807\t57\t66\t1.126\t6.237\n",
      "24\t1.181\tTrue\t1.079\t17.364\t285\t295\t2.944\t15.054\n",
      "25\t0.024\tTrue\t0.829\t2.775\t21\t26\t1.130\t6.839\n",
      "26\t0.109\tTrue\t0.917\t1.738\t67\t82\t1.139\t5.084\n",
      "27\t0.136\tTrue\t0.824\t7.550\t81\t89\t1.630\t10.771\n",
      "28\t0.799\tTrue\t0.882\t1.948\t232\t269\t1.088\t6.941\n",
      "29\t0.090\tTrue\t0.847\t6.762\t56\t70\t1.569\t9.619\n",
      "30\t0.864\tTrue\t0.844\t5.604\t240\t274\t1.416\t9.105\n",
      "31\t0.400\tTrue\t0.879\t3.214\t155\t163\t1.227\t7.373\n",
      "32\t0.068\tTrue\t0.825\t2.156\t46\t57\t1.046\t7.722\n",
      "33\t0.041\tTrue\t0.833\t2.083\t30\t36\t1.053\t6.426\n",
      "34\t0.030\tTrue\t0.811\t4.604\t26\t40\t1.329\t8.289\n",
      "35\t0.770\tTrue\t0.844\t2.088\t226\t268\t1.064\t6.750\n",
      "36\t3.773\tTrue\t0.851\t6.307\t539\t562\t1.582\t9.254\n",
      "37\t0.289\tTrue\t0.807\t4.761\t131\t149\t1.299\t7.972\n",
      "38\t0.645\tTrue\t0.785\t9.623\t208\t229\t1.816\t10.529\n",
      "39\t0.107\tTrue\t0.759\t5.413\t65\t78\t1.358\t9.018\n",
      "40\t0.463\tTrue\t1.006\t12.020\t172\t199\t2.300\t13.120\n",
      "41\t0.045\tTrue\t0.896\t1.900\t33\t42\t1.111\t6.557\n",
      "42\t4.947\tFalse\t\t\t598\t619\t\t\n",
      "43\t0.037\tTrue\t0.825\t2.462\t28\t36\t1.117\t7.842\n",
      "44\t3.277\tTrue\t0.869\t3.621\t497\t523\t1.255\t8.506\n",
      "45\t0.075\tTrue\t0.835\t1.681\t47\t65\t1.007\t6.839\n",
      "46\t0.714\tTrue\t0.895\t2.565\t222\t236\t1.163\t8.319\n",
      "47\t0.979\tTrue\t0.858\t6.043\t261\t304\t1.496\t9.444\n",
      "48\t0.057\tTrue\t0.894\t2.032\t40\t47\t1.113\t7.027\n",
      "49\t0.046\tTrue\t0.918\t1.889\t33\t38\t1.140\t6.885\n",
      "#\tt\tsuccess\twork\tupstr\tnodes\titer\teuDist\tangDist\n",
      "\n",
      "Stats for successful runs only\n",
      "96%\t0.015\tTrue\t0.606\t1.228\t15\t17\t1.007\t5.084\tmin\n",
      "96%\t0.614\tTrue\t0.863\t5.412\t146\t162\t1.450\t8.720\tmean\n",
      "96%\t4.219\tTrue\t1.151\t20.391\t573\t593\t3.248\t16.530\tmax\n",
      "\n",
      "96%\t0.991\tTrue\t0.089\t4.815\t143\t149\t0.557\t2.655\tstd\n",
      "\n",
      "RRT\n",
      "0\t0.771\tFalse\t\t\t300\t306\t\t\n",
      "1\t0.056\tTrue\t0.722\t2.145\t60\t66\t0.950\t3.822\n",
      "2\t0.771\tFalse\t\t\t300\t304\t\t\n",
      "3\t0.030\tTrue\t0.733\t2.228\t40\t47\t0.973\t3.716\n",
      "4\t0.779\tFalse\t\t\t300\t303\t\t\n",
      "5\t0.777\tFalse\t\t\t300\t315\t\t\n",
      "6\t0.045\tTrue\t0.771\t1.976\t51\t53\t0.985\t2.866\n",
      "7\t0.792\tFalse\t\t\t300\t304\t\t\n",
      "8\t0.014\tTrue\t0.679\t2.670\t26\t27\t0.967\t3.029\n",
      "9\t0.021\tTrue\t0.690\t1.789\t26\t36\t0.886\t2.426\n",
      "10\t0.044\tTrue\t0.748\t2.002\t53\t55\t0.957\t3.507\n",
      "11\t0.036\tTrue\t0.766\t1.397\t44\t51\t0.913\t2.879\n",
      "12\t0.932\tFalse\t\t\t300\t314\t\t\n",
      "13\t0.826\tFalse\t\t\t300\t314\t\t\n",
      "14\t0.021\tTrue\t0.723\t3.212\t29\t44\t1.068\t2.705\n",
      "15\t0.014\tTrue\t0.723\t3.951\t26\t25\t1.125\t4.274\n",
      "16\t0.766\tFalse\t\t\t300\t315\t\t\n",
      "17\t0.029\tTrue\t0.729\t1.258\t37\t46\t0.866\t2.407\n",
      "18\t0.776\tFalse\t\t\t300\t302\t\t\n",
      "19\t0.073\tTrue\t0.793\t1.454\t70\t81\t0.954\t2.825\n",
      "20\t0.808\tFalse\t\t\t300\t305\t\t\n",
      "21\t0.043\tTrue\t0.725\t1.564\t46\t64\t0.901\t2.411\n",
      "22\t0.776\tFalse\t\t\t300\t309\t\t\n",
      "23\t0.013\tTrue\t0.670\t1.781\t23\t28\t0.864\t3.753\n",
      "24\t0.027\tTrue\t0.773\t2.321\t42\t43\t1.010\t4.378\n",
      "25\t0.015\tTrue\t0.722\t1.451\t31\t35\t0.882\t2.650\n",
      "26\t0.017\tTrue\t0.711\t2.085\t33\t33\t0.936\t4.155\n",
      "27\t0.032\tTrue\t0.754\t2.656\t43\t49\t1.036\t3.035\n",
      "28\t0.013\tTrue\t0.673\t2.303\t23\t25\t0.907\t4.443\n",
      "29\t0.042\tTrue\t0.613\t4.981\t47\t62\t1.143\t5.744\n",
      "30\t0.784\tFalse\t\t\t300\t310\t\t\n",
      "31\t0.766\tFalse\t\t\t300\t311\t\t\n",
      "32\t0.005\tTrue\t0.676\t1.517\t13\t14\t0.837\t2.566\n",
      "33\t0.801\tFalse\t\t\t300\t311\t\t\n",
      "34\t0.768\tFalse\t\t\t300\t306\t\t\n",
      "35\t0.793\tFalse\t\t\t300\t304\t\t\n",
      "36\t0.005\tTrue\t0.687\t1.738\t13\t15\t0.876\t2.509\n",
      "37\t0.788\tFalse\t\t\t300\t321\t\t\n",
      "38\t0.024\tTrue\t0.694\t2.165\t35\t36\t0.923\t2.887\n",
      "39\t0.762\tFalse\t\t\t300\t311\t\t\n",
      "40\t0.777\tFalse\t\t\t300\t303\t\t\n",
      "41\t0.784\tFalse\t\t\t300\t305\t\t\n",
      "42\t0.018\tTrue\t0.695\t1.772\t30\t34\t0.891\t2.453\n",
      "43\t0.038\tTrue\t0.725\t1.399\t48\t57\t0.880\t2.204\n",
      "44\t0.016\tTrue\t0.707\t1.370\t32\t35\t0.859\t2.048\n",
      "45\t0.086\tTrue\t0.770\t1.874\t85\t90\t0.977\t2.496\n",
      "46\t0.771\tFalse\t\t\t300\t311\t\t\n",
      "47\t0.007\tTrue\t0.715\t2.074\t15\t23\t0.932\t3.575\n",
      "48\t0.775\tFalse\t\t\t300\t302\t\t\n",
      "49\t0.837\tFalse\t\t\t300\t300\t\t\n",
      "#\tt\tsuccess\twork\tupstr\tnodes\titer\teuDist\tangDist\n",
      "\n",
      "Stats for successful runs only\n",
      "54%\t0.005\tTrue\t0.613\t1.258\t13\t14\t0.837\t2.048\tmin\n",
      "54%\t0.029\tTrue\t0.718\t2.116\t37\t43\t0.944\t3.176\tmean\n",
      "54%\t0.086\tTrue\t0.793\t4.981\t85\t90\t1.143\t5.744\tmax\n",
      "\n",
      "54%\t0.019\tTrue\t0.039\t0.815\t16\t18\t0.077\t0.857\tstd\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# initilize\n",
    "VFRRTnodes = list()\n",
    "VFRRTupstream = list()\n",
    "VFRRTnodes = list()\n",
    "VFRRTexetime = list()\n",
    "VFRRTwork = list()\n",
    "VFRRTsuccess = list()\n",
    "VFRRTiter = list()\n",
    "VFRRTeuclLength = list()\n",
    "VFRRTanglLength = list()\n",
    "    \n",
    "RRTnodes = list()\n",
    "RRTupstream = list()\n",
    "RRTnodes = list()\n",
    "RRTexetime = list()\n",
    "RRTwork = list()\n",
    "RRTsuccess = list()\n",
    "RRTiter = list()\n",
    "RRTeuclLength = list()\n",
    "RRTanglLength = list()\n",
    "\n",
    "print('#\\tt\\tsuccess\\twork\\tupstr\\tnodes\\titer\\teuDist\\tangDist')\n",
    "print('VFRRT')\n",
    "for IofExp in range(0,NumOfExperiments):\n",
    "    # Execute and return Tree\n",
    "    stampBegin = time.process_time()\n",
    "    T1,T2,path = VFRRT(qs,VF,lambdaNsteps,Es,lamdb,MaxNodes,MAXiter,Obstacles,bounds=bounds,connect=True,qfinal=qf)\n",
    "    CPUtimePassed = time.process_time()-stampBegin\n",
    "\n",
    "    \n",
    "    if T1.successful:\n",
    "        VFRRTexetime.append(CPUtimePassed)\n",
    "        VFRRTiter.append(T1.iterations+T2.iterations)\n",
    "        VFRRTsuccess.append(IofExp)\n",
    "        IDpath,Datapath,Edgepath = path #T.get_solution()\n",
    "        work,upstream = T1.get_WorkAndUpstreamOfPath(Datapath)\n",
    "        euclDist,anglDist = T1.get_length(Datapath)\n",
    "        VFRRTnodes.append(T1.NumNodes+T2.NumNodes)\n",
    "        VFRRTupstream.append(upstream)\n",
    "        VFRRTwork.append(work)\n",
    "        VFRRTeuclLength.append(euclDist)\n",
    "        VFRRTanglLength.append(anglDist)\n",
    "        \n",
    "        print('%d\\t%.3f\\tTrue\\t%.3f\\t%.3f\\t%d\\t%d\\t%.3f\\t%.3f' % (IofExp,\n",
    "                                                                  CPUtimePassed,\n",
    "                                                                  work,\n",
    "                                                                  upstream,\n",
    "                                                                  T1.NumNodes+T2.NumNodes,\n",
    "                                                                  T1.iterations+T2.iterations,\n",
    "                                                                  euclDist,\n",
    "                                                                  anglDist))\n",
    "    else:\n",
    "        print('%d\\t%.3f\\tFalse\\t\\t\\t%d\\t%d\\t\\t' % (IofExp,CPUtimePassed,T1.NumNodes+T2.NumNodes,T1.iterations+T2.iterations))\n",
    "    \n",
    "    #T.plot(DrawTrajectory=True,DrawCollisionNodes=True)\n",
    "    #T.plot(obs=Obstacles,OnlyClosestObs=True,DrawCollisionNodes=True)\n",
    "    #DrawStats(T)\n",
    "if len(VFRRTsuccess):\n",
    "    print('#\\tt\\tsuccess\\twork\\tupstr\\tnodes\\titer\\teuDist\\tangDist')\n",
    "    print('\\nStats for successful runs only\\n%.1d%%\\t%.3f\\tTrue\\t%.3f\\t%.3f\\t%d\\t%d\\t%.3f\\t%.3f\\tmin' % (\n",
    "                                                                    len(VFRRTsuccess)*100./NumOfExperiments,\n",
    "                                                                    np.min(VFRRTexetime),\n",
    "                                                                    np.min(VFRRTwork),\n",
    "                                                                    np.min(VFRRTupstream),\n",
    "                                                                    np.min(VFRRTnodes),\n",
    "                                                                    np.min(VFRRTiter),\n",
    "                                                                    np.min(VFRRTeuclLength),\n",
    "                                                                    np.min(VFRRTanglLength)))\n",
    "    print('%.1d%%\\t%.3f\\tTrue\\t%.3f\\t%.3f\\t%d\\t%d\\t%.3f\\t%.3f\\tmean' % (\n",
    "                                                                    len(VFRRTsuccess)*100./NumOfExperiments,\n",
    "                                                                    np.mean(VFRRTexetime),\n",
    "                                                                    np.mean(VFRRTwork),\n",
    "                                                                    np.mean(VFRRTupstream),\n",
    "                                                                    np.mean(VFRRTnodes),\n",
    "                                                                    np.mean(VFRRTiter),\n",
    "                                                                    np.mean(VFRRTeuclLength),\n",
    "                                                                    np.mean(VFRRTanglLength)))\n",
    "    print('%.1d%%\\t%.3f\\tTrue\\t%.3f\\t%.3f\\t%d\\t%d\\t%.3f\\t%.3f\\tmax' % (\n",
    "                                                                    len(VFRRTsuccess)*100./NumOfExperiments,\n",
    "                                                                    np.max(VFRRTexetime),\n",
    "                                                                    np.max(VFRRTwork),\n",
    "                                                                    np.max(VFRRTupstream),\n",
    "                                                                    np.max(VFRRTnodes),\n",
    "                                                                    np.max(VFRRTiter),\n",
    "                                                                    np.max(VFRRTeuclLength),\n",
    "                                                                    np.max(VFRRTanglLength)))\n",
    "    print('\\n%.1d%%\\t%.3f\\tTrue\\t%.3f\\t%.3f\\t%d\\t%d\\t%.3f\\t%.3f\\tstd\\n' % (\n",
    "                                                                    len(VFRRTsuccess)*100./NumOfExperiments,\n",
    "                                                                    np.std(VFRRTexetime),\n",
    "                                                                    np.std(VFRRTwork),\n",
    "                                                                    np.std(VFRRTupstream),\n",
    "                                                                    np.std(VFRRTnodes),\n",
    "                                                                    np.std(VFRRTiter),\n",
    "                                                                    np.std(VFRRTeuclLength),\n",
    "                                                                    np.std(VFRRTanglLength)))\n",
    "print('RRT')\n",
    "for IofExp in range(0,NumOfExperiments):\n",
    "    stampBegin = time.process_time()\n",
    "    #pdb.set_trace()\n",
    "    T = RRT(qs,qf,MaxNodes,MAXiter,Obstacles,bounds,goalBias)\n",
    "    CPUtimePassed = time.process_time()-stampBegin\n",
    "\n",
    "    \n",
    "    if T.successful:\n",
    "        RRTexetime.append(CPUtimePassed)\n",
    "        RRTiter.append(T.iterations)\n",
    "        T.VF = VF\n",
    "        RRTsuccess.append(IofExp)\n",
    "        IDpath,Datapath,Edgepath = T.get_solution()\n",
    "        work,upstream = T.get_WorkAndUpstreamOfPath(Datapath)\n",
    "        euclDist,anglDist = T.get_length(Datapath)\n",
    "        RRTnodes.append(T.NumNodes)\n",
    "        RRTupstream.append(upstream)\n",
    "        RRTwork.append(work)\n",
    "        RRTeuclLength.append(euclDist)\n",
    "        RRTanglLength.append(anglDist)\n",
    "        \n",
    "        print('%d\\t%.3f\\tTrue\\t%.3f\\t%.3f\\t%d\\t%d\\t%.3f\\t%.3f' % (IofExp,CPUtimePassed,work,upstream,T.NumNodes,T.iterations,euclDist,anglDist))\n",
    "    else:\n",
    "        print('%d\\t%.3f\\tFalse\\t\\t\\t%d\\t%d\\t\\t' % (IofExp,CPUtimePassed,T.NumNodes,T.iterations))\n",
    "    #T.plot(obs=Obstacles,OnlyClosestObs=True,DrawCollisionNodes=True)\n",
    "#plt.show()\n",
    "\n",
    "if len(RRTsuccess):\n",
    "    print('#\\tt\\tsuccess\\twork\\tupstr\\tnodes\\titer\\teuDist\\tangDist')\n",
    "    print('\\nStats for successful runs only\\n%.1d%%\\t%.3f\\tTrue\\t%.3f\\t%.3f\\t%d\\t%d\\t%.3f\\t%.3f\\tmin' % (\n",
    "                                                                    len(RRTsuccess)*100./NumOfExperiments,\n",
    "                                                                    np.min(RRTexetime),\n",
    "                                                                    np.min(RRTwork),\n",
    "                                                                    np.min(RRTupstream),\n",
    "                                                                    np.min(RRTnodes),\n",
    "                                                                    np.min(RRTiter),\n",
    "                                                                    np.min(RRTeuclLength),\n",
    "                                                                    np.min(RRTanglLength)))\n",
    "\n",
    "    print('%.1d%%\\t%.3f\\tTrue\\t%.3f\\t%.3f\\t%d\\t%d\\t%.3f\\t%.3f\\tmean' % (\n",
    "                                                                    len(RRTsuccess)*100./NumOfExperiments,\n",
    "                                                                    np.mean(RRTexetime),\n",
    "                                                                    np.mean(RRTwork),\n",
    "                                                                    np.mean(RRTupstream),\n",
    "                                                                    np.mean(RRTnodes),\n",
    "                                                                    np.mean(RRTiter),\n",
    "                                                                    np.mean(RRTeuclLength),\n",
    "                                                                    np.mean(RRTanglLength)))\n",
    "    print('%.1d%%\\t%.3f\\tTrue\\t%.3f\\t%.3f\\t%d\\t%d\\t%.3f\\t%.3f\\tmax' % (\n",
    "                                                                    len(RRTsuccess)*100./NumOfExperiments,\n",
    "                                                                    np.max(RRTexetime),\n",
    "                                                                    np.max(RRTwork),\n",
    "                                                                    np.max(RRTupstream),\n",
    "                                                                    np.max(RRTnodes),\n",
    "                                                                    np.max(RRTiter),\n",
    "                                                                    np.max(RRTeuclLength),\n",
    "                                                                    np.max(RRTanglLength)))\n",
    "    print('\\n%.1d%%\\t%.3f\\tTrue\\t%.3f\\t%.3f\\t%d\\t%d\\t%.3f\\t%.3f\\tstd\\n' % (\n",
    "                                                                    len(RRTsuccess)*100./NumOfExperiments,\n",
    "                                                                    np.std(RRTexetime),\n",
    "                                                                    np.std(RRTwork),\n",
    "                                                                    np.std(RRTupstream),\n",
    "                                                                    np.std(RRTnodes),\n",
    "                                                                    np.std(RRTiter),\n",
    "                                                                    np.std(RRTeuclLength),\n",
    "                                                                    np.std(RRTanglLength)))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.1"
  },
  "notify_time": "5",
  "toc": {
   "toc_cell": false,
   "toc_number_sections": false,
   "toc_section_display": "block",
   "toc_threshold": 6,
   "toc_window_display": true
  },
  "toc_position": {
   "height": "461px",
   "left": "1148px",
   "right": "13px",
   "top": "141px",
   "width": "279px"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}