nb-cache_deltas.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data preparation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import sqlite3\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import Cdf\n",
    "import Pmf\n",
    "\n",
    "\n",
    "# define global plot parameters\n",
    "params = {'axes.labelsize' : 12, 'axes.titlesize' : 12,\n",
    "          'font.size' : 12, 'legend.fontsize' : 12,\n",
    "          'xtick.labelsize' : 12, 'ytick.labelsize' : 12}\n",
    "plt.rcParams.update(params)\n",
    "plt.rcParams.update({'figure.max_open_warning': 0})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# helper function to get CDFs for an array of dataframes for a specific column name\n",
    "def get_cdfs(dfs, col):\n",
    "    cdfs = []\n",
    "    for df in dfs:\n",
    "        cdfs.append(Cdf.MakeCdfFromList(df[col]))\n",
    "    return cdfs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Cache Identification based on heuristic (see paper)\n",
    "Matching source and destination AS numbers indicate that request stayed within ISP network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# connect to database and get data\n",
    "conn = sqlite3.connect('data/youtube-traceroute.db')\n",
    "\n",
    "# cache for IPv4 only\n",
    "v4_caches = pd.read_sql_query('select * \\\n",
    "                               from pair_medians_meta \\\n",
    "                               where src_asn_v4 = dst_asn_v4 and src_asn_v6 != dst_asn_v6', conn)\n",
    "\n",
    "# cache for IPv6 only\n",
    "v6_caches = pd.read_sql_query('select * \\\n",
    "                               from pair_medians_meta \\\n",
    "                               where src_asn_v4 != dst_asn_v4 and src_asn_v6 = dst_asn_v6', conn)\n",
    "\n",
    "# caches for both\n",
    "both_caches = pd.read_sql_query('select * \\\n",
    "                                 from pair_medians_meta \\\n",
    "                                 where src_asn_v4 = dst_asn_v4 and src_asn_v6 = dst_asn_v6', conn)\n",
    "\n",
    "# no cache\n",
    "no_caches = pd.read_sql_query('select * \\\n",
    "                               from pair_medians_meta \\\n",
    "                               where src_asn_v4 != dst_asn_v4 and src_asn_v6 != dst_asn_v6', conn)  \n",
    "\n",
    "conn.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Putting lists of the four cases together for plotting\n",
    "\n",
    "- IPv4 cache, no IPv6 cache\n",
    "- IPv6 cache, no IPv4 cache\n",
    "- caches for both\n",
    "- no caches for either"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# all lists will represent the four cases above and therefore have length 4\n",
    "data = [v4_caches, \n",
    "        v6_caches, \n",
    "        both_caches,\n",
    "        no_caches]\n",
    "titles = ['(v4 but no v6 cache)', \n",
    "          '(v6 but no v4 cache)',\n",
    "          '(v4 and v6 caches)',\n",
    "          '(no ISP caches)']\n",
    "filenames = ['v4_cache',\n",
    "             'v6_cache',\n",
    "             'both_caches',\n",
    "             'no_cache']\n",
    "\n",
    "# use helper function to get CDFs for each cases, and each metric\n",
    "cdfs_ttl = get_cdfs(data, 'm_ttl_delta')\n",
    "cdfs_rtt = get_cdfs(data, 'm_rtt_delta')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Plotting"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Put all four cases into shared plots\n",
    "- one plot for TTL\n",
    "- one plot for RTT"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x144 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x144 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "markers = ['^',\n",
    "           's',\n",
    "           'o',\n",
    "           'd']\n",
    "colors = ['blue',\n",
    "          'red',\n",
    "          'darkorchid',\n",
    "          'forestgreen']\n",
    "labels = ['IPv4 only',\n",
    "          'IPv6 only',\n",
    "          'Both versions',\n",
    "          'Neither version']\n",
    "\n",
    "\n",
    "## TTL\n",
    "ttl_fig, ttl_ax = plt.subplots(figsize = (5, 2))\n",
    "for i in range(4):\n",
    "    ttl_ax.plot(cdfs_ttl[i].xs, cdfs_ttl[i].ps, \n",
    "                label = '%s' % (labels[i]), \n",
    "                marker = markers[i], \n",
    "                linewidth = 0.5, \n",
    "                markersize = 3.5, \n",
    "                fillstyle = 'none', \n",
    "                color = colors[i])\n",
    "\n",
    "# PLOT FORMATTING\n",
    "ttl_ax.set_xlabel('TTL delta')\n",
    "ttl_ax.set_ylabel('CDF')\n",
    "ttl_ax.set_xlim([-15.3, 15.3])\n",
    "ttl_ax.set_ylim([-0.05, 1.05])\n",
    "ttl_ax.legend(loc = 'lower right', fontsize = 'small')\n",
    "\n",
    "yticks = np.arange(0.0, 1.01, 0.2)\n",
    "ttl_ax.set_yticks(yticks)\n",
    "ttl_ax.grid(False)\n",
    "ttl_ax.spines['right'].set_color('none')\n",
    "ttl_ax.spines['top'].set_color('none')\n",
    "ttl_ax.yaxis.set_ticks_position('left')\n",
    "ttl_ax.xaxis.set_ticks_position('bottom')\n",
    "ttl_ax.spines['bottom'].set_position(('axes', -0.02))\n",
    "ttl_ax.spines['left'].set_position(('axes', -0.02))\n",
    "\n",
    "\n",
    "## RTT\n",
    "rtt_fig, rtt_ax = plt.subplots(figsize = (5, 2))\n",
    "for i in range(4):\n",
    "    rtt_ax.plot(cdfs_rtt[i].xs, cdfs_rtt[i].ps,\n",
    "                label = '%s' % (labels[i]),\n",
    "                marker = markers[i],\n",
    "                linewidth = 0.5,\n",
    "                markersize = 2.5,\n",
    "                fillstyle = 'none',\n",
    "                color = colors[i])\n",
    "\n",
    "# PLOT FORMATTING\n",
    "# customize axes and grid appearance\n",
    "rtt_ax.set_xlabel('RTT delta [ms]')\n",
    "rtt_ax.set_xscale('symlog')\n",
    "rtt_ax.set_ylabel('CDF')\n",
    "rtt_ax.set_ylim([-0.05, 1.05])\n",
    "rtt_ax.set_xlim([-1500, 1500])\n",
    "rtt_ax.tick_params(axis = 'x', which = 'major', labelsize = 11)\n",
    "rtt_ax.legend(loc = 'lower right', fontsize = 'small')\n",
    "\n",
    "rtt_ax.set_yticks(yticks)\n",
    "rtt_ax.grid(False)\n",
    "rtt_ax.spines['right'].set_color('none')\n",
    "rtt_ax.spines['top'].set_color('none')\n",
    "rtt_ax.yaxis.set_ticks_position('left')\n",
    "rtt_ax.xaxis.set_ticks_position('bottom')\n",
    "rtt_ax.spines['bottom'].set_position(('axes', -0.02))\n",
    "rtt_ax.spines['left'].set_position(('axes', -0.02))\n",
    "\n",
    "from matplotlib.ticker import ScalarFormatter\n",
    "rtt_ax.xaxis.set_major_formatter(ScalarFormatter())\n",
    "\n",
    "\n",
    "# add arrows and text which indicate in which region IPv6 was better/worse than IPv4\n",
    "ttl_ax.annotate('', xy = (0.4, 1.1), \n",
    "                xycoords = 'axes fraction',\n",
    "                xytext = (0.1, 1.1),\n",
    "                arrowprops = dict(arrowstyle = \"<-\"))\n",
    "ttl_ax.annotate('', xy = (0.9, 1.1),\n",
    "                xycoords = 'axes fraction',\n",
    "                xytext = (0.6, 1.1),\n",
    "                arrowprops = dict(arrowstyle = '->'))\n",
    "ttl_ax.text(-10.5, 1.2, 'IPv6 slower')\n",
    "ttl_ax.text(4.5, 1.2, 'IPv6 faster')\n",
    "\n",
    "rtt_ax.annotate('', xy = (0.4, 1.1),\n",
    "                xycoords = 'axes fraction',\n",
    "                xytext = (0.1, 1.1),\n",
    "                arrowprops = dict(arrowstyle = '<-'))\n",
    "rtt_ax.annotate('', xy = (0.9, 1.1),\n",
    "                xycoords = 'axes fraction',\n",
    "                xytext = (0.6, 1.1),\n",
    "                arrowprops = dict(arrowstyle = '->'))\n",
    "rtt_ax.text(-70, 1.2, 'IPv6 slower')\n",
    "rtt_ax.text(2.5, 1.2, 'IPv6 faster')\n",
    "\n",
    "# dotted vertical line to separate positive and negative delta region\n",
    "ttl_ax.axvline(0, linestyle = 'dotted', color = 'black', linewidth = 0.5)\n",
    "rtt_ax.axvline(0, linestyle = 'dotted', color = 'black', linewidth = 0.5)\n",
    "\n",
    "# saving and showing plot\n",
    "ttl_fig.savefig('plots/caches_ttl_comparison.pdf', bbox_inches = 'tight')\n",
    "rtt_fig.savefig('plots/caches_rtt_comparison.pdf', bbox_inches = 'tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Distributions of median deltas for all cases"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# separate TTL CDFs out of list\n",
    "v4, v6, both, no = cdfs_ttl"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---- TTL v4 ----\n",
      "m_ttl_delta; cdf\n",
      "----------------\n",
      "-12.00; 0.00695\n",
      "-11.50; 0.00759\n",
      "-11.00; 0.15234\n",
      "-10.50; 0.15360\n",
      "-10.00; 0.18078\n",
      "-9.00; 0.22124\n",
      "-8.00; 0.26991\n",
      "-7.00; 0.29077\n",
      "-6.00; 0.46144\n",
      "-5.50; 0.46207\n",
      "-5.00; 0.54298\n",
      "-4.00; 0.62389\n",
      "-3.00; 0.87484\n",
      "-2.00; 0.94564\n",
      "-1.00; 0.99747\n",
      "2.00; 1.00000\n",
      "\n",
      "--------------------------------------------------------\n",
      "\n",
      "---- TTL v6 ----\n",
      "m_ttl_delta; cdf\n",
      "----------------\n",
      "0.00; 0.00445\n",
      "1.00; 0.09503\n",
      "2.00; 0.15590\n",
      "3.00; 0.58797\n",
      "3.50; 0.58872\n",
      "4.00; 0.79287\n",
      "5.00; 0.79955\n",
      "5.50; 0.80030\n",
      "6.00; 0.83964\n",
      "7.00; 0.89384\n",
      "8.00; 0.96288\n",
      "8.50; 0.96362\n",
      "9.00; 0.97773\n",
      "10.00; 0.98070\n",
      "11.00; 0.98961\n",
      "12.50; 0.99035\n",
      "14.00; 0.99629\n",
      "15.00; 1.00000\n",
      "\n",
      "--------------------------------------------------------\n",
      "\n",
      "--- TTL both ---\n",
      "m_ttl_delta; cdf\n",
      "----------------\n",
      "-3.00; 0.01136\n",
      "-2.50; 0.01190\n",
      "-2.00; 0.04545\n",
      "-1.00; 0.15260\n",
      "-0.50; 0.15963\n",
      "0.00; 0.73052\n",
      "0.50; 0.73268\n",
      "1.00; 0.90801\n",
      "1.50; 0.91017\n",
      "2.00; 0.98485\n",
      "3.00; 0.99838\n",
      "4.00; 1.00000\n",
      "\n",
      "--------------------------------------------------------\n",
      "\n",
      "--- TTL none ---\n",
      "m_ttl_delta; cdf\n",
      "----------------\n",
      "-13.00; 0.00003\n",
      "-12.00; 0.00049\n",
      "-11.00; 0.00065\n",
      "-10.50; 0.00067\n",
      "-10.00; 0.00192\n",
      "-9.00; 0.00449\n",
      "-8.00; 0.00923\n",
      "-7.50; 0.00929\n",
      "-7.00; 0.01497\n",
      "-6.50; 0.01512\n",
      "-6.00; 0.02770\n",
      "-5.50; 0.02781\n",
      "-5.00; 0.04768\n",
      "-4.50; 0.04778\n",
      "-4.00; 0.07992\n",
      "-3.50; 0.08031\n",
      "-3.00; 0.12357\n",
      "-2.50; 0.12368\n",
      "-2.00; 0.15929\n",
      "-1.50; 0.15999\n",
      "-1.00; 0.25444\n",
      "-0.50; 0.25532\n",
      "0.00; 0.60306\n",
      "0.50; 0.60464\n",
      "1.00; 0.75153\n",
      "1.50; 0.75213\n",
      "2.00; 0.82774\n",
      "2.50; 0.82826\n",
      "3.00; 0.88239\n",
      "3.50; 0.88294\n",
      "4.00; 0.92159\n",
      "4.50; 0.92190\n",
      "5.00; 0.95152\n",
      "5.50; 0.95162\n",
      "6.00; 0.97593\n",
      "7.00; 0.98521\n",
      "8.00; 0.99432\n",
      "8.50; 0.99440\n",
      "9.00; 0.99811\n",
      "9.50; 0.99821\n",
      "10.00; 0.99935\n",
      "11.00; 0.99974\n",
      "12.00; 0.99992\n",
      "13.00; 0.99997\n",
      "14.00; 1.00000\n"
     ]
    }
   ],
   "source": [
    "print('---- TTL v4 ----')\n",
    "print('m_ttl_delta; cdf')\n",
    "print('----------------')\n",
    "for x, y in zip(v4.xs, v4.ps):\n",
    "    print('%.2f; %.5f' % (x, y))\n",
    "\n",
    "print('\\n--------------------------------------------------------\\n')\n",
    "\n",
    "print('---- TTL v6 ----')\n",
    "print('m_ttl_delta; cdf')\n",
    "print('----------------')\n",
    "for x, y in zip(v6.xs, v6.ps):\n",
    "    print('%.2f; %.5f' % (x, y))\n",
    "    \n",
    "print('\\n--------------------------------------------------------\\n')\n",
    "\n",
    "print('--- TTL both ---')\n",
    "print('m_ttl_delta; cdf')\n",
    "print('----------------')\n",
    "for x, y in zip(both.xs, both.ps):\n",
    "    print('%.2f; %.5f' % (x, y))\n",
    "    \n",
    "print('\\n--------------------------------------------------------\\n')\n",
    "\n",
    "print('--- TTL none ---')\n",
    "print('m_ttl_delta; cdf')\n",
    "print('----------------')\n",
    "for x, y in zip(no.xs, no.ps):\n",
    "    print('%.2f; %.5f' % (x, y))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# separate RTT CDFs out of list\n",
    "v4, v6, both, no = cdfs_rtt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---- RTT v4 ----\n",
      "m_rtt_delta; cdf\n",
      "----------------\n",
      "-462.17; 0.00063\n",
      "-121.74; 0.00695\n",
      "-111.39; 0.01327\n",
      "-107.39; 0.01960\n",
      "-82.70; 0.02592\n",
      "-51.09; 0.03224\n",
      "-48.85; 0.03856\n",
      "-48.77; 0.04488\n",
      "-48.68; 0.05120\n",
      "-48.57; 0.05815\n",
      "-48.52; 0.06511\n",
      "-48.47; 0.07206\n",
      "-48.43; 0.07965\n",
      "-48.39; 0.08597\n",
      "-48.30; 0.09292\n",
      "-48.26; 0.09924\n",
      "-48.22; 0.10556\n",
      "-48.19; 0.11315\n",
      "-48.15; 0.11947\n",
      "-48.09; 0.12705\n",
      "-48.02; 0.13338\n",
      "-47.97; 0.13970\n",
      "-47.92; 0.14602\n",
      "-47.89; 0.15297\n",
      "-47.83; 0.15929\n",
      "-47.72; 0.16561\n",
      "-47.65; 0.17193\n",
      "-47.54; 0.17889\n",
      "-43.15; 0.18521\n",
      "-34.62; 0.19153\n",
      "-31.88; 0.19848\n",
      "-31.54; 0.20480\n",
      "-30.99; 0.21176\n",
      "-25.50; 0.21808\n",
      "-23.59; 0.22440\n",
      "-23.27; 0.23072\n",
      "-23.08; 0.23704\n",
      "-22.80; 0.24336\n",
      "-22.33; 0.24968\n",
      "-21.55; 0.25601\n",
      "-21.14; 0.26233\n",
      "-18.73; 0.26865\n",
      "-16.92; 0.27497\n",
      "-15.02; 0.28129\n",
      "-14.22; 0.28761\n",
      "-12.80; 0.29393\n",
      "-12.32; 0.30025\n",
      "-11.79; 0.30657\n",
      "-11.22; 0.31290\n",
      "-10.90; 0.31922\n",
      "-10.71; 0.32554\n",
      "-10.52; 0.33186\n",
      "-10.31; 0.33881\n",
      "-10.01; 0.34513\n",
      "-9.84; 0.35145\n",
      "-9.55; 0.35777\n",
      "-9.36; 0.36410\n",
      "-9.19; 0.37105\n",
      "-9.05; 0.37737\n",
      "-8.95; 0.38496\n",
      "-8.86; 0.39254\n",
      "-8.81; 0.39886\n",
      "-8.69; 0.40518\n",
      "-8.57; 0.41150\n",
      "-8.41; 0.41783\n",
      "-8.26; 0.42415\n",
      "-8.02; 0.43110\n",
      "-7.80; 0.43742\n",
      "-7.51; 0.44437\n",
      "-7.37; 0.45133\n",
      "-7.29; 0.45828\n",
      "-7.22; 0.46460\n",
      "-7.21; 0.47472\n",
      "-7.14; 0.48230\n",
      "-7.07; 0.48862\n",
      "-7.05; 0.49684\n",
      "-6.99; 0.50569\n",
      "-6.91; 0.51264\n",
      "-6.90; 0.51960\n",
      "-6.82; 0.52592\n",
      "-6.75; 0.53224\n",
      "-6.68; 0.54046\n",
      "-6.57; 0.54678\n",
      "-6.43; 0.55310\n",
      "-6.27; 0.55942\n",
      "-6.18; 0.56574\n",
      "-6.09; 0.57396\n",
      "-6.03; 0.58091\n",
      "-5.95; 0.58850\n",
      "-5.91; 0.59482\n",
      "-5.86; 0.60114\n",
      "-5.80; 0.60746\n",
      "-5.69; 0.61378\n",
      "-5.62; 0.62073\n",
      "-5.17; 0.62705\n",
      "-4.07; 0.63401\n",
      "-1.98; 0.64033\n",
      "-1.55; 0.64665\n",
      "-1.34; 0.65297\n",
      "-1.10; 0.65929\n",
      "-1.03; 0.66561\n",
      "-0.96; 0.67193\n",
      "-0.90; 0.67826\n",
      "-0.84; 0.68584\n",
      "-0.78; 0.69279\n",
      "-0.72; 0.70101\n",
      "-0.68; 0.70796\n",
      "-0.63; 0.71429\n",
      "-0.60; 0.72250\n",
      "-0.56; 0.72882\n",
      "-0.52; 0.73767\n",
      "-0.48; 0.74463\n",
      "-0.40; 0.75221\n",
      "-0.34; 0.75917\n",
      "-0.30; 0.76549\n",
      "-0.20; 0.77181\n",
      "-0.16; 0.77813\n",
      "-0.10; 0.78445\n",
      "-0.07; 0.79140\n",
      "-0.01; 0.79772\n",
      "0.04; 0.80405\n",
      "0.11; 0.81226\n",
      "0.17; 0.81858\n",
      "0.23; 0.82491\n",
      "0.29; 0.83123\n",
      "0.39; 0.83818\n",
      "0.49; 0.84513\n",
      "0.64; 0.85209\n",
      "0.78; 0.85841\n",
      "0.92; 0.86473\n",
      "1.23; 0.87105\n",
      "1.58; 0.87737\n",
      "1.91; 0.88369\n",
      "2.06; 0.89001\n",
      "2.21; 0.89697\n",
      "2.40; 0.90329\n",
      "2.70; 0.90961\n",
      "2.85; 0.91593\n",
      "2.95; 0.92225\n",
      "3.09; 0.92857\n",
      "3.31; 0.93552\n",
      "3.40; 0.94185\n",
      "3.50; 0.94817\n",
      "3.60; 0.95449\n",
      "3.71; 0.96081\n",
      "3.85; 0.96713\n",
      "3.97; 0.97408\n",
      "4.20; 0.98040\n",
      "5.14; 0.98673\n",
      "8.02; 0.99305\n",
      "1325.02; 0.99937\n",
      "\n",
      "--------------------------------------------------------\n",
      "\n",
      "---- RTT v6 ----\n",
      "m_rtt_delta; cdf\n",
      "----------------\n",
      "-419.30; 0.00074\n",
      "-54.52; 0.00817\n",
      "-21.25; 0.01559\n",
      "-15.28; 0.02301\n",
      "-14.73; 0.03044\n",
      "-9.45; 0.03786\n",
      "-4.82; 0.04529\n",
      "-4.24; 0.05271\n",
      "-3.95; 0.06013\n",
      "-3.78; 0.06756\n",
      "-3.67; 0.07498\n",
      "-3.54; 0.08315\n",
      "-3.45; 0.09057\n",
      "-3.33; 0.09800\n",
      "-3.17; 0.10542\n",
      "-3.00; 0.11359\n",
      "-2.88; 0.12101\n",
      "-2.69; 0.12843\n",
      "-2.38; 0.13586\n",
      "-2.21; 0.14328\n",
      "-2.03; 0.15071\n",
      "-1.77; 0.15813\n",
      "-1.48; 0.16555\n",
      "-1.19; 0.17298\n",
      "-1.02; 0.18040\n",
      "-0.85; 0.18782\n",
      "-0.74; 0.19525\n",
      "-0.64; 0.20267\n",
      "-0.54; 0.21084\n",
      "-0.46; 0.21901\n",
      "-0.40; 0.22717\n",
      "-0.34; 0.23460\n",
      "-0.23; 0.24276\n",
      "-0.18; 0.25019\n",
      "-0.13; 0.25761\n",
      "-0.05; 0.26578\n",
      "-0.01; 0.27320\n",
      "0.04; 0.28285\n",
      "0.07; 0.29027\n",
      "0.11; 0.29770\n",
      "0.17; 0.30512\n",
      "0.26; 0.31329\n",
      "0.30; 0.32146\n",
      "0.39; 0.32888\n",
      "0.44; 0.33705\n",
      "0.51; 0.34521\n",
      "0.54; 0.35338\n",
      "0.59; 0.36080\n",
      "0.65; 0.36897\n",
      "0.73; 0.37639\n",
      "0.76; 0.38382\n",
      "0.78; 0.39198\n",
      "0.80; 0.39941\n",
      "0.86; 0.40683\n",
      "0.91; 0.41500\n",
      "0.95; 0.42242\n",
      "0.97; 0.43133\n",
      "1.01; 0.44024\n",
      "1.03; 0.44766\n",
      "1.08; 0.45509\n",
      "1.16; 0.46251\n",
      "1.21; 0.46993\n",
      "1.26; 0.47736\n",
      "1.36; 0.48478\n",
      "1.46; 0.49220\n",
      "1.55; 0.50037\n",
      "1.59; 0.50854\n",
      "1.66; 0.51745\n",
      "1.70; 0.52635\n",
      "1.77; 0.53452\n",
      "1.80; 0.54269\n",
      "1.84; 0.55011\n",
      "1.87; 0.55754\n",
      "1.90; 0.56570\n",
      "1.94; 0.57313\n",
      "2.02; 0.58129\n",
      "2.07; 0.59169\n",
      "2.13; 0.60059\n",
      "2.23; 0.60876\n",
      "2.33; 0.61618\n",
      "2.48; 0.62361\n",
      "2.61; 0.63103\n",
      "2.79; 0.63846\n",
      "2.92; 0.64588\n",
      "3.25; 0.65405\n",
      "3.79; 0.66147\n",
      "4.86; 0.66889\n",
      "5.13; 0.67632\n",
      "5.82; 0.68448\n",
      "5.96; 0.69191\n",
      "6.06; 0.69933\n",
      "6.20; 0.70676\n",
      "6.46; 0.71492\n",
      "6.82; 0.72235\n",
      "7.15; 0.73051\n",
      "7.56; 0.73794\n",
      "7.87; 0.74536\n",
      "8.57; 0.75278\n",
      "9.40; 0.76095\n",
      "10.16; 0.76837\n",
      "10.76; 0.77580\n",
      "12.71; 0.78322\n",
      "13.92; 0.79065\n",
      "14.04; 0.79807\n",
      "14.08; 0.80549\n",
      "14.14; 0.81292\n",
      "14.18; 0.82183\n",
      "14.25; 0.83073\n",
      "14.31; 0.83816\n",
      "14.39; 0.84558\n",
      "14.73; 0.85301\n",
      "14.89; 0.86043\n",
      "15.03; 0.86785\n",
      "15.07; 0.87528\n",
      "15.21; 0.88270\n",
      "16.63; 0.89013\n",
      "19.33; 0.89755\n",
      "21.03; 0.90497\n",
      "21.64; 0.91314\n",
      "22.83; 0.92056\n",
      "24.70; 0.92799\n",
      "28.80; 0.93541\n",
      "29.66; 0.94284\n",
      "30.44; 0.95026\n",
      "32.29; 0.95768\n",
      "35.12; 0.96511\n",
      "44.68; 0.97253\n",
      "69.75; 0.97996\n",
      "89.81; 0.98738\n",
      "101.42; 0.99480\n",
      "\n",
      "--------------------------------------------------------\n",
      "\n",
      "--- RTT both ---\n",
      "m_rtt_delta; cdf\n",
      "----------------\n",
      "-66.13; 0.00054\n",
      "-19.81; 0.00595\n",
      "-11.71; 0.01136\n",
      "-5.06; 0.01677\n",
      "-3.77; 0.02219\n",
      "-3.18; 0.02760\n",
      "-2.92; 0.03301\n",
      "-2.73; 0.03842\n",
      "-2.61; 0.04383\n",
      "-2.45; 0.04924\n",
      "-2.30; 0.05465\n",
      "-2.16; 0.06061\n",
      "-1.89; 0.06602\n",
      "-1.74; 0.07143\n",
      "-1.58; 0.07684\n",
      "-1.43; 0.08225\n",
      "-1.35; 0.08766\n",
      "-1.26; 0.09307\n",
      "-1.23; 0.09848\n",
      "-1.14; 0.10390\n",
      "-1.05; 0.10931\n",
      "-1.00; 0.11472\n",
      "-0.94; 0.12067\n",
      "-0.89; 0.12716\n",
      "-0.84; 0.13258\n",
      "-0.81; 0.14069\n",
      "-0.78; 0.14665\n",
      "-0.75; 0.15422\n",
      "-0.72; 0.16071\n",
      "-0.70; 0.16775\n",
      "-0.68; 0.17424\n",
      "-0.66; 0.18074\n",
      "-0.64; 0.18669\n",
      "-0.62; 0.19264\n",
      "-0.59; 0.19859\n",
      "-0.56; 0.20455\n",
      "-0.54; 0.21050\n",
      "-0.52; 0.21861\n",
      "-0.50; 0.22457\n",
      "-0.48; 0.22998\n",
      "-0.45; 0.23647\n",
      "-0.44; 0.24459\n",
      "-0.43; 0.25162\n",
      "-0.42; 0.26028\n",
      "-0.40; 0.26677\n",
      "-0.39; 0.27327\n",
      "-0.38; 0.28084\n",
      "-0.36; 0.28734\n",
      "-0.33; 0.29329\n",
      "-0.32; 0.29978\n",
      "-0.31; 0.30628\n",
      "-0.29; 0.31277\n",
      "-0.27; 0.31926\n",
      "-0.25; 0.32522\n",
      "-0.23; 0.33117\n",
      "-0.21; 0.33874\n",
      "-0.20; 0.34794\n",
      "-0.19; 0.35498\n",
      "-0.17; 0.36526\n",
      "-0.16; 0.37175\n",
      "-0.15; 0.37933\n",
      "-0.14; 0.38907\n",
      "-0.13; 0.39827\n",
      "-0.12; 0.40855\n",
      "-0.11; 0.41667\n",
      "-0.10; 0.42695\n",
      "-0.09; 0.44264\n",
      "-0.08; 0.45130\n",
      "-0.08; 0.46212\n",
      "-0.07; 0.47294\n",
      "-0.06; 0.48160\n",
      "-0.05; 0.49188\n",
      "-0.04; 0.50433\n",
      "-0.04; 0.51948\n",
      "-0.03; 0.53301\n",
      "-0.03; 0.54600\n",
      "-0.02; 0.55465\n",
      "-0.01; 0.56926\n",
      "-0.01; 0.57955\n",
      "-0.01; 0.59091\n",
      "-0.00; 0.60606\n",
      "0.00; 0.61797\n",
      "0.01; 0.63366\n",
      "0.01; 0.64232\n",
      "0.02; 0.65584\n",
      "0.03; 0.66558\n",
      "0.03; 0.67424\n",
      "0.04; 0.68561\n",
      "0.05; 0.69481\n",
      "0.06; 0.70455\n",
      "0.07; 0.71374\n",
      "0.08; 0.72078\n",
      "0.09; 0.72781\n",
      "0.10; 0.73701\n",
      "0.11; 0.74459\n",
      "0.12; 0.75108\n",
      "0.13; 0.75812\n",
      "0.15; 0.76515\n",
      "0.17; 0.77273\n",
      "0.18; 0.77922\n",
      "0.20; 0.78626\n",
      "0.22; 0.79275\n",
      "0.24; 0.79924\n",
      "0.27; 0.80574\n",
      "0.29; 0.81277\n",
      "0.30; 0.81872\n",
      "0.32; 0.82468\n",
      "0.35; 0.83117\n",
      "0.38; 0.83766\n",
      "0.40; 0.84307\n",
      "0.43; 0.84848\n",
      "0.47; 0.85444\n",
      "0.51; 0.85985\n",
      "0.56; 0.86580\n",
      "0.59; 0.87121\n",
      "0.63; 0.87771\n",
      "0.67; 0.88366\n",
      "0.74; 0.88961\n",
      "0.85; 0.89502\n",
      "0.94; 0.90097\n",
      "1.07; 0.90639\n",
      "1.17; 0.91180\n",
      "1.21; 0.91775\n",
      "1.26; 0.92316\n",
      "1.40; 0.92857\n",
      "1.61; 0.93452\n",
      "1.85; 0.94048\n",
      "2.10; 0.94643\n",
      "2.31; 0.95184\n",
      "2.52; 0.95725\n",
      "3.04; 0.96266\n",
      "4.10; 0.96807\n",
      "4.52; 0.97348\n",
      "6.35; 0.97890\n",
      "12.78; 0.98431\n",
      "13.32; 0.98972\n",
      "16.48; 0.99513\n",
      "\n",
      "--------------------------------------------------------\n",
      "\n",
      "--- RTT none ---\n",
      "m_rtt_delta; cdf\n",
      "----------------\n",
      "-9961.96; 0.00003\n",
      "-275.57; 0.00265\n",
      "-120.97; 0.00524\n",
      "-71.07; 0.00783\n",
      "-55.36; 0.01043\n",
      "-50.74; 0.01305\n",
      "-46.57; 0.01567\n",
      "-45.85; 0.01842\n",
      "-44.78; 0.02101\n",
      "-40.76; 0.02363\n",
      "-36.85; 0.02630\n",
      "-35.26; 0.02890\n",
      "-33.72; 0.03154\n",
      "-30.56; 0.03416\n",
      "-28.73; 0.03681\n",
      "-27.20; 0.03953\n",
      "-25.98; 0.04228\n",
      "-24.77; 0.04493\n",
      "-23.03; 0.04757\n",
      "-21.50; 0.05024\n",
      "-20.42; 0.05286\n",
      "-19.36; 0.05564\n",
      "-18.77; 0.05841\n",
      "-18.37; 0.06140\n",
      "-18.07; 0.06425\n",
      "-17.83; 0.06739\n",
      "-17.63; 0.07061\n",
      "-17.40; 0.07367\n",
      "-17.17; 0.07673\n",
      "-16.94; 0.07992\n",
      "-16.63; 0.08277\n",
      "-16.20; 0.08560\n",
      "-15.72; 0.08853\n",
      "-15.23; 0.09125\n",
      "-14.97; 0.09468\n",
      "-14.77; 0.09888\n",
      "-14.42; 0.10178\n",
      "-13.82; 0.10466\n",
      "-13.19; 0.10747\n",
      "-12.54; 0.11016\n",
      "-11.92; 0.11296\n",
      "-11.38; 0.11577\n",
      "-10.84; 0.11854\n",
      "-10.41; 0.12152\n",
      "-9.98; 0.12440\n",
      "-9.60; 0.12726\n",
      "-9.23; 0.13021\n",
      "-8.96; 0.13320\n",
      "-8.68; 0.13613\n",
      "-8.46; 0.13908\n",
      "-8.16; 0.14207\n",
      "-7.86; 0.14497\n",
      "-7.58; 0.14783\n",
      "-7.24; 0.15091\n",
      "-6.94; 0.15390\n",
      "-6.64; 0.15693\n",
      "-6.38; 0.15989\n",
      "-6.15; 0.16310\n",
      "-5.92; 0.16632\n",
      "-5.75; 0.16946\n",
      "-5.56; 0.17294\n",
      "-5.37; 0.17620\n",
      "-5.14; 0.17945\n",
      "-4.95; 0.18251\n",
      "-4.75; 0.18557\n",
      "-4.56; 0.18860\n",
      "-4.37; 0.19195\n",
      "-4.20; 0.19542\n",
      "-4.02; 0.19898\n",
      "-3.87; 0.20251\n",
      "-3.71; 0.20585\n",
      "-3.56; 0.20909\n",
      "-3.39; 0.21247\n",
      "-3.22; 0.21553\n",
      "-3.06; 0.21918\n",
      "-2.91; 0.22258\n",
      "-2.75; 0.22572\n",
      "-2.61; 0.22917\n",
      "-2.49; 0.23257\n",
      "-2.37; 0.23646\n",
      "-2.20; 0.23988\n",
      "-2.07; 0.24414\n",
      "-1.96; 0.24782\n",
      "-1.85; 0.25184\n",
      "-1.74; 0.25610\n",
      "-1.66; 0.26097\n",
      "-1.56; 0.26645\n",
      "-1.48; 0.27135\n",
      "-1.40; 0.27698\n",
      "-1.33; 0.28240\n",
      "-1.25; 0.28805\n",
      "-1.18; 0.29417\n",
      "-1.11; 0.29967\n",
      "-1.03; 0.30600\n",
      "-0.96; 0.31210\n",
      "-0.90; 0.31804\n",
      "-0.83; 0.32489\n",
      "-0.77; 0.33103\n",
      "-0.72; 0.33889\n",
      "-0.66; 0.34673\n",
      "-0.60; 0.35495\n",
      "-0.55; 0.36247\n",
      "-0.48; 0.37391\n",
      "-0.44; 0.38569\n",
      "-0.38; 0.39697\n",
      "-0.34; 0.40786\n",
      "-0.29; 0.42107\n",
      "-0.25; 0.43419\n",
      "-0.21; 0.44797\n",
      "-0.17; 0.46394\n",
      "-0.13; 0.47686\n",
      "-0.10; 0.49191\n",
      "-0.06; 0.51126\n",
      "-0.04; 0.53388\n",
      "-0.01; 0.55458\n",
      "0.01; 0.56697\n",
      "0.04; 0.58015\n",
      "0.08; 0.59206\n",
      "0.12; 0.60355\n",
      "0.16; 0.61592\n",
      "0.20; 0.62663\n",
      "0.24; 0.63589\n",
      "0.29; 0.64928\n",
      "0.33; 0.65916\n",
      "0.38; 0.66897\n",
      "0.44; 0.67747\n",
      "0.49; 0.68653\n",
      "0.55; 0.69532\n",
      "0.62; 0.70347\n",
      "0.68; 0.71117\n",
      "0.75; 0.71913\n",
      "0.81; 0.72546\n",
      "0.88; 0.73086\n",
      "0.96; 0.73633\n",
      "1.03; 0.74167\n",
      "1.11; 0.74709\n",
      "1.20; 0.75176\n",
      "1.29; 0.75706\n",
      "1.39; 0.76134\n",
      "1.49; 0.76551\n",
      "1.60; 0.76990\n",
      "1.71; 0.77472\n",
      "1.80; 0.77892\n",
      "1.87; 0.78354\n",
      "1.96; 0.78818\n",
      "2.05; 0.79231\n",
      "2.14; 0.79646\n",
      "2.26; 0.80043\n",
      "2.40; 0.80395\n",
      "2.55; 0.80740\n",
      "2.71; 0.81093\n",
      "2.85; 0.81430\n",
      "3.07; 0.81747\n",
      "3.27; 0.82050\n",
      "3.50; 0.82351\n",
      "3.72; 0.82665\n",
      "3.94; 0.82971\n",
      "4.14; 0.83306\n",
      "4.37; 0.83620\n",
      "4.55; 0.83957\n",
      "4.78; 0.84281\n",
      "4.99; 0.84600\n",
      "5.19; 0.84929\n",
      "5.37; 0.85274\n",
      "5.54; 0.85604\n",
      "5.70; 0.85931\n",
      "5.89; 0.86245\n",
      "6.09; 0.86566\n",
      "6.34; 0.86872\n",
      "6.56; 0.87184\n",
      "6.80; 0.87474\n",
      "7.08; 0.87770\n",
      "7.37; 0.88053\n",
      "7.64; 0.88356\n",
      "7.94; 0.88657\n",
      "8.25; 0.88942\n",
      "8.45; 0.89285\n",
      "8.64; 0.89596\n",
      "8.85; 0.89907\n",
      "9.06; 0.90221\n",
      "9.25; 0.90563\n",
      "9.44; 0.90877\n",
      "9.61; 0.91191\n",
      "9.80; 0.91523\n",
      "10.01; 0.91834\n",
      "10.21; 0.92138\n",
      "10.43; 0.92460\n",
      "10.67; 0.92750\n",
      "10.95; 0.93041\n",
      "11.34; 0.93326\n",
      "11.70; 0.93622\n",
      "12.10; 0.93910\n",
      "12.45; 0.94192\n",
      "12.74; 0.94498\n",
      "13.14; 0.94784\n",
      "13.69; 0.95064\n",
      "14.27; 0.95341\n",
      "14.75; 0.95637\n",
      "15.37; 0.95912\n",
      "16.62; 0.96192\n",
      "18.04; 0.96467\n",
      "19.22; 0.96732\n",
      "21.43; 0.96994\n",
      "24.51; 0.97261\n",
      "25.78; 0.97528\n",
      "28.11; 0.97803\n",
      "31.15; 0.98062\n",
      "35.22; 0.98324\n",
      "39.50; 0.98586\n",
      "45.19; 0.98846\n",
      "52.25; 0.99108\n",
      "67.25; 0.99367\n",
      "79.38; 0.99629\n",
      "187.41; 0.99888\n"
     ]
    }
   ],
   "source": [
    "print('---- RTT v4 ----')\n",
    "print('m_rtt_delta; cdf')\n",
    "print('----------------')\n",
    "for x, y in list(zip(v4.xs, v4.ps))[0::10]:\n",
    "    print('%.2f; %.5f' % (x, y))\n",
    "\n",
    "print('\\n--------------------------------------------------------\\n')\n",
    "\n",
    "print('---- RTT v6 ----')\n",
    "print('m_rtt_delta; cdf')\n",
    "print('----------------')\n",
    "for x, y in list(zip(v6.xs, v6.ps))[0::10]:\n",
    "    print('%.2f; %.5f' % (x, y))\n",
    "    \n",
    "print('\\n--------------------------------------------------------\\n')\n",
    "\n",
    "print('--- RTT both ---')\n",
    "print('m_rtt_delta; cdf')\n",
    "print('----------------')\n",
    "for x, y in list(zip(both.xs, both.ps))[0::10]:\n",
    "    print('%.2f; %.5f' % (x, y))\n",
    "    \n",
    "print('\\n--------------------------------------------------------\\n')\n",
    "\n",
    "print('--- RTT none ---')\n",
    "print('m_rtt_delta; cdf')\n",
    "print('----------------')\n",
    "for x, y in list(zip(no.xs, no.ps))[0::100]:\n",
    "    print('%.2f; %.5f' % (x, y))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}