Evaluation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import glob\n",
    "import numpy as np\n",
    "import csv\n",
    "import os, sys, getopt\n",
    "import csv\n",
    "\n",
    "def readfile(tfile_path, c, t):\n",
    "    f = open(tfile_path)\n",
    "    csv_f = csv.reader(f)\n",
    "    for time, quality in csv_f:\n",
    "        if int(quality) <= c and float(time) <= t:\n",
    "            return True\n",
    "    return False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "opt_cutoff = {'karate':14, 'football':94, 'jazz':158, 'email':594, 'delaunay_n10':703,'netscience':899, \\\n",
    "              'power':2203,'as-22july06':3303,'hep-th':3926,'star2':4542,'star':6902}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'star2': [327.78982758522034, 541.9585990905762, 429.50568675994873, 591.917680978775, 472.9427752494812, 378.23771929740906, 557.0073528289795, 553.8911187648773, 431.783704996109, 482.1038465499878]}\n",
      "{'star2': [4551.0, 4547.0, 4548.0, 4549.0, 4545.0, 4547.0, 4550.0, 4547.0, 4547.0, 4547.0]}\n"
     ]
    }
   ],
   "source": [
    "algo = 'SA'\n",
    "dataset = 'star2'\n",
    "record_time, record_nV = {}, {}\n",
    "record_time_detail = {}\n",
    "for fn in glob.glob('./{}_output/{}*.trace'.format(algo, dataset)):\n",
    "    sub_detail = {}\n",
    "    fin = open(fn, 'r').readlines()\n",
    "    name = fn.split('/')[-1].split('_')[0]\n",
    "    try:\n",
    "        if algo == 'approx':\n",
    "            time, nV = fin[-1].split()\n",
    "        else:\n",
    "            time, nV = fin[-1].split()[0].split(',')\n",
    "    except:\n",
    "        print (fn)\n",
    "        continue\n",
    "\n",
    "    time, nV = float(time), float(nV)\n",
    "    if name not in record_time:\n",
    "        record_time[name] = [time]\n",
    "        record_nV[name] = [nV]\n",
    "    else:\n",
    "        record_time[name].append(time)\n",
    "        record_nV[name].append(nV)\n",
    "        \n",
    "#     for i in range(1, len(fin)):\n",
    "#         time, nV = fin[i].split()[0].split(',')\n",
    "#         print (time, nV)\n",
    "print (record_time)\n",
    "print (record_nV)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "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": [
    "percents=[.1,.2,.5,1,2]\n",
    "maxtime = np.max(record_time[dataset])\n",
    "interval = maxtime/100.0\n",
    "optimal = opt_cutoff[dataset]\n",
    "fig = plt.figure(1)\n",
    "ax = fig.add_subplot(111)\n",
    "x = np.arange(0,maxtime+interval,interval)\n",
    "for p in percents:\n",
    "    c = optimal + (optimal * p/100.0)\n",
    "    y = []\n",
    "    for xval in x:\n",
    "        num_runs = 0\n",
    "        num_solutions = 0\n",
    "#         for tfile in os.listdir(tracefile):\n",
    "        for tfile in glob.glob('./{}_output/{}*.trace'.format(algo, dataset)):\n",
    "            result = readfile(tfile, c, xval)\n",
    "            num_runs += 1\n",
    "            if result:\n",
    "                num_solutions += 1\n",
    "        y.append(num_solutions/float(num_runs))\n",
    "    # plot\n",
    "    ax.semilogx(x,y,label=str(p)+\"%\")\n",
    "\n",
    "# making plot look pretty\n",
    "ax.set_xlabel('Time (sec)')\n",
    "ax.set_ylabel('P(solve)')\n",
    "ax.set_title('Qualified RTD')\n",
    "handles, labels = ax.get_legend_handles_labels()\n",
    "box = ax.get_position()\n",
    "ax.set_position([box.x0, box.y0, box.width*0.8, box.height])\n",
    "lgd = ax.legend(handles, labels, loc='center left', bbox_to_anchor=(1,0.5))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "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": [
    "times=[0.5, 1, 5, 10, 20]\n",
    "maxquality = .5\n",
    "interval = maxquality/100.0\n",
    "fig = plt.figure(2)\n",
    "ax = fig.add_subplot(111)\n",
    "x = np.arange(0,maxquality+interval,interval)\n",
    "for t in times:\n",
    "    y = []\n",
    "    for xval in x:\n",
    "        c = (xval + 1) * optimal\n",
    "        num_runs = 0\n",
    "        num_solutions = 0\n",
    "        for tfile in glob.glob('./{}_output/{}*.trace'.format(algo, dataset)):\n",
    "            result = readfile(tfile, c, t)\n",
    "            num_runs += 1\n",
    "            if result:\n",
    "                num_solutions += 1\n",
    "        y.append(num_solutions/float(num_runs))\n",
    "    # plot\n",
    "    ax.plot(x,y,label=str(t)+\"s\")\n",
    "\n",
    "# making plot look pretty\n",
    "ax.set_xlabel('Relative Solution Quality (%)')\n",
    "ax.set_ylabel('P(solve)')\n",
    "ax.set_title('Solution Quality Distributions')\n",
    "handles, labels = ax.get_legend_handles_labels()\n",
    "box = ax.get_position()\n",
    "ax.set_position([box.x0, box.y0, box.width*0.8, box.height])\n",
    "lgd = ax.legend(handles, labels, loc='center left', bbox_to_anchor=(1,0.5))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "hep-th\t89.58912029266358\t143.51016524685127\t3926.0\t0.0\t0.0\t3926\n",
      "dummy1\t0.00025081634521484375\t4.4425257487686596e-06\t2.0\t0.0\t1999999999999998.8\t1e-15\n",
      "as-22july06\t38.619063067436215\t13.192087864031762\t3303.0\t0.0\t0.0\t3303\n",
      "delaunay\t75.33566763401032\t55.16584893930376\t703.4\t0.48989794855663565\t0.0005689900426742208\t703\n",
      "star2\t476.7138312101364\t81.30110196739818\t4547.8\t1.661324772583615\t0.0012769704975781994\t4542\n",
      "netscience\t0.049217510223388675\t0.00026792314320424934\t899.0\t0.0\t0.0\t899\n",
      "dummy2\t0.0002535581588745117\t3.5543564595879492e-06\t3.0\t0.0\t2999999999999999.0\t1e-15\n",
      "karate\t0.0005883455276489257\t7.532706627836184e-06\t14.0\t0.0\t0.0\t14\n",
      "jazz\t0.060322809219360354\t0.027541626240624936\t158.0\t0.0\t0.0\t158\n",
      "email\t4.105719828605652\t2.545893085279006\t594.0\t0.0\t0.0\t594\n",
      "star\t583.3891662836074\t12.420653034591693\t6963.4\t5.2\t0.008895972181976186\t6902\n",
      "power\t85.02687439918517\t40.980767834065546\t2203.0\t0.0\t0.0\t2203\n",
      "football\t0.023671698570251466\t0.02016909276421845\t94.0\t0.0\t0.0\t94\n"
     ]
    }
   ],
   "source": [
    "algo = 'SA'\n",
    "record_time, record_nV = {}, {}\n",
    "for fn in glob.glob('./{}_output/*.trace'.format(algo)):\n",
    "\n",
    "    fin = open(fn, 'r').readlines()\n",
    "    name = fn.split('/')[-1].split('_')[0]\n",
    "    try:\n",
    "        if algo == 'approx':\n",
    "            time, nV = fin[-1].split()\n",
    "        else:\n",
    "            time, nV = fin[-1].split()[0].split(',')\n",
    "    except:\n",
    "        print (fn)\n",
    "        continue\n",
    "\n",
    "    time, nV = float(time), float(nV)\n",
    "    if name not in record_time:\n",
    "        record_time[name] = [time]\n",
    "        record_nV[name] = [nV]\n",
    "    else:\n",
    "        record_time[name].append(time)\n",
    "        record_nV[name].append(nV)\n",
    "\n",
    "opt_cutoff = {'karate':14, 'football':94, 'jazz':158, 'email':594, 'delaunay':703,'netscience':899, \\\n",
    "      'power':2203,'as-22july06':3303,'hep-th':3926,'star2':4542,'star':6902}\n",
    "for key in record_time.keys():\n",
    "    opt_nv = opt_cutoff.get('{}'.format(key), 1e-15)\n",
    "    err = (np.mean(record_nV[key]) - opt_nv)/opt_nv\n",
    "    print ('{}\\t{}\\t{}\\t{}\\t{}\\t{}\\t{}'.format(key, \n",
    "                                   np.mean(record_time[key]), np.std(record_time[key]), \n",
    "                                   np.mean(record_nV[key]), np.std(record_nV[key]), err, opt_nv))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'whiskers': [<matplotlib.lines.Line2D at 0x7f0c7d249310>,\n",
       "  <matplotlib.lines.Line2D at 0x7f0c7d20e150>,\n",
       "  <matplotlib.lines.Line2D at 0x7f0c7d249f90>,\n",
       "  <matplotlib.lines.Line2D at 0x7f0c7d248f10>],\n",
       " 'caps': [<matplotlib.lines.Line2D at 0x7f0c7d20e5d0>,\n",
       "  <matplotlib.lines.Line2D at 0x7f0c7d20ead0>,\n",
       "  <matplotlib.lines.Line2D at 0x7f0c7d235950>,\n",
       "  <matplotlib.lines.Line2D at 0x7f0c7d248490>],\n",
       " 'boxes': [<matplotlib.lines.Line2D at 0x7f0c7d2492d0>,\n",
       "  <matplotlib.lines.Line2D at 0x7f0c7d248a50>],\n",
       " 'medians': [<matplotlib.lines.Line2D at 0x7f0c7d248050>,\n",
       "  <matplotlib.lines.Line2D at 0x7f0c7d04a390>],\n",
       " 'fliers': [<matplotlib.lines.Line2D at 0x7f0c7d248550>,\n",
       "  <matplotlib.lines.Line2D at 0x7f0c7d04a890>],\n",
       " 'means': []}"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "algo = 'SA'\n",
    "record_time, record_nV = {}, {}\n",
    "record_time_detail = {}\n",
    "for fn in glob.glob('./{}_output/{}*.trace'.format(algo, dataset))[:]:\n",
    "    sub_detail = {}\n",
    "    fin = open(fn, 'r').readlines()\n",
    "    name = fn.split('/')[-1].split('_')[0]\n",
    "    try:\n",
    "        if algo == 'approx':\n",
    "            time, nV = fin[-1].split()\n",
    "        else:\n",
    "            time, nV = fin[-1].split()[0].split(',')\n",
    "    except:\n",
    "        print (fn)\n",
    "        continue\n",
    "\n",
    "    time, nV = float(time), float(nV)\n",
    "    if name not in record_time:\n",
    "        record_time[name] = [time]\n",
    "        record_nV[name] = [nV]\n",
    "    else:\n",
    "        record_time[name].append(time)\n",
    "        record_nV[name].append(nV)\n",
    "ls1 = np.array(record_time[dataset])\n",
    "\n",
    "algo = 'GA'\n",
    "record_time, record_nV = {}, {}\n",
    "record_time_detail = {}\n",
    "for fn in glob.glob('./{}_output/{}*.trace'.format(algo, dataset))[:]:\n",
    "    sub_detail = {}\n",
    "    fin = open(fn, 'r').readlines()\n",
    "    name = fn.split('/')[-1].split('_')[0]\n",
    "    try:\n",
    "        if algo == 'approx':\n",
    "            time, nV = fin[-1].split()\n",
    "        else:\n",
    "            time, nV = fin[-1].split()[0].split(',')\n",
    "    except:\n",
    "        print (fn)\n",
    "        continue\n",
    "\n",
    "    time, nV = float(time), float(nV)\n",
    "    if name not in record_time:\n",
    "        record_time[name] = [time]\n",
    "        record_nV[name] = [nV]\n",
    "    else:\n",
    "        record_time[name].append(time)\n",
    "        record_nV[name].append(nV)\n",
    "ls2 = np.array(record_time[dataset])\n",
    "        \n",
    "fig1, ax1 = plt.subplots()\n",
    "ax1.set_title('Plot')\n",
    "ax1.boxplot([ls1, ls2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}