graphing2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "code_folding": [
     0
    ]
   },
   "outputs": [],
   "source": [
    "#Imports\n",
    "%matplotlib inline\n",
    "import numpy as np\n",
    "import re\n",
    "from collections import defaultdict\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "code_folding": []
   },
   "outputs": [],
   "source": [
    "def parse_line(line):\n",
    "    m = re.match(\"^pbkdf2-sha256/pbkdf2-sha256-(.+)\\.txt:(ALICE|BOB):(OTS|FIPP|FDPP|SG|INP|ON|OUT):(\\d+|\\d\\.\\d+e\\+\\d+)\\n\",\n",
    "            line)\n",
    "    if m == None:\n",
    "        print(line)\n",
    "    size = int(float(m.group(1)))\n",
    "    party = m.group(2)\n",
    "    fpart = m.group(3)\n",
    "    if \"e\" in m.group(4):\n",
    "        time = int(float(m.group(4)))\n",
    "    else:\n",
    "        time = int(m.group(4))\n",
    "    return (size, party, fpart, time)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "code_folding": []
   },
   "outputs": [],
   "source": [
    "#Write dictionary dt, means, stdevs\n",
    "dt = defaultdict(dict)\n",
    "filenames = [\"results/{:d}_{:#02d}_{:s}.txt\".format(sz, i, party) for i in range(1, 11) for party in\n",
    "        [\"ALICE\", \"BOB\"] for sz in [1, 10, 100]]\n",
    "for fn in filenames:\n",
    "    with open(fn, \"r\") as f:\n",
    "        for line in f:\n",
    "            (size, party, fpart, time) = parse_line(line)\n",
    "            if (size, party, fpart) in dt:\n",
    "                dt[(size, party, fpart)].append(time)\n",
    "            else:\n",
    "                dt[(size, party, fpart)] = [time]\n",
    "\n",
    "nonpw_alice_means = defaultdict(dict)\n",
    "nonpw_alice_stdevs = defaultdict(dict)\n",
    "pw_alice_means = defaultdict(dict)\n",
    "pw_alice_stdevs = defaultdict(dict)\n",
    "nonpw_bob_means = defaultdict(dict)\n",
    "nonpw_bob_stdevs = defaultdict(dict)\n",
    "pw_bob_means = defaultdict(dict)\n",
    "pw_bob_stdevs = defaultdict(dict)\n",
    "for size in [1, 10, 100]:\n",
    "    NONPW = [\"OTS\", \"FIPP\", \"FDPP\", \"ON\"]\n",
    "    PW = [\"SG\", \"INP\", \"OUT\"]\n",
    "    nonpw_alice_means[size] = np.average([sum(l) for l in zip(*[dt[(size, \"ALICE\", fp)] for fp in NONPW])])\n",
    "    nonpw_alice_stdevs[size] = np.std([sum(l) for l in zip(*[dt[(size, \"ALICE\", fp)] for fp in NONPW])])\n",
    "    pw_alice_means[size] = np.average([sum(l) for l in zip(*[dt[(size, \"ALICE\", fp)] for fp in PW])])\n",
    "    pw_alice_stdevs[size] = np.std([sum(l) for l in zip(*[dt[(size, \"ALICE\", fp)] for fp in PW])])\n",
    "    nonpw_bob_means[size] = np.average([sum(l) for l in zip(*[dt[(size, \"BOB\", fp)] for fp in NONPW])])\n",
    "    nonpw_bob_stdevs[size] = np.std([sum(l) for l in zip(*[dt[(size, \"BOB\", fp)] for fp in NONPW])])\n",
    "    pw_bob_means[size] = np.average([sum(l) for l in zip(*[dt[(size, \"BOB\", fp)] for fp in PW])])\n",
    "    pw_bob_stdevs[size] = np.std([sum(l) for l in zip(*[dt[(size, \"BOB\", fp)] for fp in PW])])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "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": [
    "fig = plt.figure()\n",
    "ax = fig.add_subplot(1,1,1)\n",
    "x = [1, 10, 100]\n",
    "plt.errorbar(x, [nonpw_alice_means[sz] for sz in x], yerr=[nonpw_alice_stdevs[sz] for sz in x], \n",
    "             label='non-modification cost (Alice)')\n",
    "plt.errorbar(x, [pw_alice_means[sz] for sz in x], yerr=[pw_alice_stdevs[sz] for sz in x], \n",
    "             label='modification cost (Alice)')\n",
    "plt.errorbar(x, [nonpw_bob_means[sz] for sz in x], yerr=[nonpw_bob_stdevs[sz] for sz in x], \n",
    "             label='non-modification cost (Bob)')\n",
    "plt.errorbar(x, [pw_bob_means[sz] for sz in x], yerr=[pw_bob_stdevs[sz] for sz in x], \n",
    "             label='modification cost (Bob)')\n",
    "ax.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "defaultdict(<class 'dict'>, {1: 389.8602, 10: 3084.3147000000004, 100: 33040.7402})\n",
      "defaultdict(<class 'dict'>, {1: 1161.4926, 10: 1673.246, 100: 2278.5095})\n",
      "defaultdict(<class 'dict'>, {1: 402.51590000000004, 10: 3194.8113, 100: 34106.1559})\n",
      "defaultdict(<class 'dict'>, {1: 1003.8154000000001, 10: 1417.17, 100: 1072.5995})\n",
      "defaultdict(<class 'dict'>, {1: 7.760078566612583, 10: 34.49999377985451, 100: 1089.104427792285})\n",
      "defaultdict(<class 'dict'>, {1: 14.01009194973395, 10: 15.738124945494619, 100: 175.85061847900906})\n",
      "defaultdict(<class 'dict'>, {1: 7.269978052924231, 10: 33.6305227317388, 100: 1126.5165023040229})\n",
      "defaultdict(<class 'dict'>, {1: 15.365810315111926, 10: 17.978620080528984, 100: 134.1363605054573})\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "1.1614926"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "for sz in [1, 10, 100]:\n",
    "    nonpw_alice_means[sz] = nonpw_alice_means[sz] / 1000\n",
    "    pw_alice_means[sz] = pw_alice_means[sz] / 1000\n",
    "    nonpw_bob_means[sz] = nonpw_bob_means[sz] / 1000\n",
    "    pw_bob_means[sz] = pw_bob_means[sz] / 1000\n",
    "    nonpw_alice_stdevs[sz] = nonpw_alice_stdevs[sz] / 1000\n",
    "    pw_alice_stdevs[sz] = pw_alice_stdevs[sz] / 1000\n",
    "    nonpw_bob_stdevs[sz] = nonpw_bob_stdevs[sz] / 1000\n",
    "    pw_bob_stdevs[sz] = pw_bob_stdevs[sz] / 1000\n",
    "\n",
    "print(nonpw_alice_means)\n",
    "print(pw_alice_means)\n",
    "print(nonpw_bob_means)\n",
    "print(pw_bob_means)\n",
    "\n",
    "print(nonpw_alice_stdevs)\n",
    "print(pw_alice_stdevs)\n",
    "print(nonpw_bob_stdevs)\n",
    "print(pw_bob_stdevs)\n",
    "\n",
    "1161492.6 / 1000000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "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": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}