DATA MAKER.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Collecting exit relay data\n",
    "\n",
    "Relay properties we're collecting:\n",
    "- __Fingerprint__: relay unique identifier\n",
    "- __Bandwidth__: observed bandwidth\n",
    "- __Age__: # seconds since this relay was first seen\n",
    "- __Family Size__: number of relays reported in effective family\n",
    "- __AS__: the autonomous system this relay belongs to\n",
    "- __Uptime__: time since last restart of this relay\n",
    "- __Consensus Weight__: weight as a fraction of the whole network\n",
    "- __Port/Address uptime__: time since relay changed its port or address\n",
    "- __Country__: the country this relay is located in\n",
    "- __Ports accepted__: the number of ports this relay accepts connections on\n",
    "\n",
    "We also collect which flags are present in each relay description:\n",
    "- __Exit__: Is it an exit relay? (always 1)\n",
    "- __Fast__: Does it have a lot of bandwidth available?\n",
    "- __Running__: Has it been online in the past 45 minutes?\n",
    "- __Stable__: Is it suitable for long-lived circuits?\n",
    "- __Guard__: Is it a guard relay?\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "*** IMPORTANT YOU MUST RUN THIS BLOCK TO GET THE NECESSARY IMPORTS AND HELPER METHODS ***"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import matplotlib\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import requests, datetime, statistics, math, json\n",
    "import pandas as pd\n",
    "\n",
    "import plotly.plotly as py\n",
    "from plotly.graph_objs import *\n",
    "\n",
    "onionoo_url = \"https://onionoo.torproject.org/\"\n",
    "\n",
    "# Need this function to convert port-ranges into arrays\n",
    "def mixrange(arr):\n",
    "    r = []\n",
    "    for i in arr:\n",
    "        if '-' not in i:\n",
    "            r.append(int(i))\n",
    "        else:\n",
    "            l,h = map(int, i.split('-'))\n",
    "            r+= range(l,h+1)\n",
    "    return r\n",
    "\n",
    "# Need this function to compute the age of a given relay\n",
    "def compute_age(relay):\n",
    "    if not 'first_seen' in relay or not 'last_seen' in relay:\n",
    "        return None\n",
    "    first_seen = datetime.datetime.strptime(relay['first_seen'],\"%Y-%m-%d %H:%M:%S\")\n",
    "    last_seen = datetime.datetime.strptime(relay['last_seen'],\"%Y-%m-%d %H:%M:%S\")\n",
    "    age = (last_seen - first_seen).total_seconds()\n",
    "    return age\n",
    "\n",
    "def compute_open_ports(relay):\n",
    "    if not 'exit_policy_summary' in relay:\n",
    "        return None\n",
    "    for policy in relay['exit_policy_summary']:\n",
    "        if policy == 'accept':\n",
    "            ports =  len(mixrange(relay['exit_policy_summary'][policy]))\n",
    "        elif policy == 'reject':\n",
    "            ports = 65535 - len(mixrange(relay['exit_policy_summary'][policy]))\n",
    "    return ports\n",
    "\n",
    "def compute_uptime(relay, pulled):\n",
    "    if not 'last_restarted' in relay:\n",
    "        return None\n",
    "    last_restart = datetime.datetime.strptime(relay['last_restarted'],\"%Y-%m-%d %H:%M:%S\")\n",
    "    return (pulled-last_restart).total_seconds()\n",
    "\n",
    "def compute_addr_port_uptime(relay, pulled):\n",
    "    if not 'last_changed_address_or_port' in relay:\n",
    "        return None\n",
    "    last_change = datetime.datetime.strptime(relay['last_changed_address_or_port'],\"%Y-%m-%d %H:%M:%S\")\n",
    "    return (pulled-last_change).total_seconds()\n",
    "\n",
    "def normalize(x):\n",
    "    return (x - x.mean()) / (x.max() - x.min())\n",
    "\n",
    "def log_normalize(df, columns):\n",
    "    # Drop any 0's\n",
    "    size = len(df)\n",
    "    for col in columns:\n",
    "        df = df[df[col]>0]\n",
    "    print(\"Dropped %d rows.\"%(size-len(df)))\n",
    "    \n",
    "    for col in columns:\n",
    "        df[col] = normalize(df[col].apply(lambda x: math.log(abs(x))))\n",
    "    return df\n",
    "\n",
    "def plot_feature(allDF, exitDF, guardDF, feature_x, feature_y):\n",
    "    fig = plt.figure()\n",
    "    ax1 = fig.add_subplot(111)\n",
    "    ax1.scatter(allDF[feature_x], allDF[feature_y], c='b', label='all',alpha=0.3)\n",
    "    ax1.scatter(exitDF[feature_x], exitDF[feature_y], c='r', label='exit',alpha=0.3)\n",
    "    ax1.scatter(guardDF[feature_x], guardDF[feature_y],c='g',label='guard',alpha=0.3)\n",
    "    plt.legend(loc='upper left')\n",
    "    plt.title(\"%s vs. %s\"%(feature_x,feature_y))\n",
    "    plt.xlabel(feature_x)\n",
    "    plt.ylabel(feature_y)\n",
    "    fig.savefig('./images/%s_v_%s.png'%(feature_x,feature_y),dpi=1000)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Relay Details:\n",
    "\n",
    "There are several ways to construct the relay data make sure to look and choose carefully before running any of the blocks below."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Run this block to collect data using onionoo requests\n",
    "\n",
    "This is good if you don't have any saved .json responses from Onionoo.org or if you just want to pull down a more recent response."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Used to filter the results so that they only contain these fields\n",
    "url = onionoo_url+\"details?limit=500&type=relay&flag=Exit\"\n",
    "response = requests.get(url).json()\n",
    "print(response.status_code)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Run this block to load data from a .json file\n",
    "\n",
    "This is best if you already have a response from Onionoo.org that you'd like to use."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "with open('sample-12-18-16.json') as json_data:\n",
    "    response = json.load(json_data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Run this block to build the relay DataFrame\n",
    "\n",
    "Whether you got the response from requests or by loading a .json, you need to run this to parse the json into a DataFrame object. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "columns=['fingerprint',\n",
    "         'nickname',\n",
    "         'bandwidth',\n",
    "         'age',\n",
    "         'family_size',\n",
    "         'as_number',\n",
    "         'consensus_weight',\n",
    "         'ports_accepted',\n",
    "         'uptime',\n",
    "         'address_port_uptime',\n",
    "         'badexit',\n",
    "         'exit',\n",
    "         'fast',\n",
    "         'guard',\n",
    "         'hsdir',\n",
    "         'stable',\n",
    "         'running',\n",
    "         'valid']\n",
    "\n",
    "relayDF = pd.DataFrame(columns=columns)\n",
    "\n",
    "for relay in response['relays']:\n",
    "    row = [None]*len(columns)\n",
    "    \n",
    "    # IDENTIFIERS\n",
    "    row[0] = relay['fingerprint'] if 'fingerprint' in relay else None\n",
    "    row[1] = relay['nickname'] if 'nickname' in relay else None\n",
    "    \n",
    "    # NETWORK PROPERTIES\n",
    "    row[2] = relay['advertised_bandwidth'] if 'advertised_bandwidth' in relay else None\n",
    "    row[3] = compute_age(relay)\n",
    "    row[4] = len(relay['effective_family']) if 'effective_family' in relay else None\n",
    "    row[5] = relay['as_number'] if 'as_number' in relay else None\n",
    "    row[6] = relay['consensus_weight_fraction'] if 'consensus_weight_fraction' in relay else None\n",
    "    row[7] = compute_open_ports(relay)\n",
    "    row[8] = compute_uptime(relay)\n",
    "    row[9] = compute_addr_port_uptime(relay)\n",
    "    \n",
    "    # FLAGS\n",
    "    if 'flags' in relay:\n",
    "        row[10] = int('BadExit' in relay['flags'])\n",
    "        row[11] = int('Exit' in relay['flags'])\n",
    "        row[12] = int('Fast' in relay['flags'])\n",
    "        row[13] = int('Guard' in relay['flags'])\n",
    "        row[14] = int('HSDir' in relay['flags'])\n",
    "        row[15] = int('Stable' in relay['flags'])\n",
    "        row[16] = int('Running' in relay['flags'])\n",
    "        row[17] = int('Valid' in relay['flags'])\n",
    "    \n",
    "    relayDF = relayDF.append(pd.DataFrame([row],columns=columns))\n",
    "relayDF.reset_index(drop=True,inplace=True)\n",
    "relayDF.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# How many did we get?\n",
    "len(relayDF)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### VERSION 2.0 - MORE FIELDS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>fingerprint</th>\n",
       "      <th>nickname</th>\n",
       "      <th>last_seen</th>\n",
       "      <th>last_restarted</th>\n",
       "      <th>last_changed_addr_port</th>\n",
       "      <th>first_seen</th>\n",
       "      <th>hibernating</th>\n",
       "      <th>age</th>\n",
       "      <th>uptime</th>\n",
       "      <th>address_port_uptime</th>\n",
       "      <th>...</th>\n",
       "      <th>exit_prob</th>\n",
       "      <th>badexit</th>\n",
       "      <th>exit</th>\n",
       "      <th>fast</th>\n",
       "      <th>guard</th>\n",
       "      <th>hsdir</th>\n",
       "      <th>stable</th>\n",
       "      <th>running</th>\n",
       "      <th>valid</th>\n",
       "      <th>v2dir</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2B35CC60F043B3DB6DFFE73F464A67B82B0B0B4E</td>\n",
       "      <td>torjittertrapnet</td>\n",
       "      <td>2016-12-18 14:00:00</td>\n",
       "      <td>2016-12-08 08:30:31</td>\n",
       "      <td>2016-08-10 03:00:00</td>\n",
       "      <td>2016-08-10 03:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>11271600.0</td>\n",
       "      <td>883769.0</td>\n",
       "      <td>11271600.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>055A3CAA2582B87448301367ECD5905D138EF049</td>\n",
       "      <td>verax1</td>\n",
       "      <td>2016-12-18 14:00:00</td>\n",
       "      <td>2016-08-28 01:40:14</td>\n",
       "      <td>2016-08-28 02:00:00</td>\n",
       "      <td>2014-11-29 22:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>64771200.0</td>\n",
       "      <td>9721186.0</td>\n",
       "      <td>9720000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>D5F2C65F4131A1468D5B67A8838A9B7ED8C049E2</td>\n",
       "      <td>che</td>\n",
       "      <td>2016-12-18 14:00:00</td>\n",
       "      <td>2016-12-08 18:02:47</td>\n",
       "      <td>2016-06-20 14:00:00</td>\n",
       "      <td>2007-10-27 12:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>288583200.0</td>\n",
       "      <td>849433.0</td>\n",
       "      <td>15638400.0</td>\n",
       "      <td>...</td>\n",
       "      <td>6.33964e-06</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0F69025EA0A61A96E5444A8BF39A8B1335488610</td>\n",
       "      <td>Uni25</td>\n",
       "      <td>2016-12-18 14:00:00</td>\n",
       "      <td>2016-11-28 23:11:16</td>\n",
       "      <td>2016-10-25 16:00:00</td>\n",
       "      <td>2015-04-24 21:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>52160400.0</td>\n",
       "      <td>1694924.0</td>\n",
       "      <td>4658400.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2A587ED761FB8CF8AA7CB77164455479C002141B</td>\n",
       "      <td>Unnamed</td>\n",
       "      <td>2016-12-18 14:00:00</td>\n",
       "      <td>2016-12-12 23:06:28</td>\n",
       "      <td>2016-12-12 23:00:00</td>\n",
       "      <td>2016-12-12 23:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>486000.0</td>\n",
       "      <td>485612.0</td>\n",
       "      <td>486000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 32 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                fingerprint          nickname  \\\n",
       "0  2B35CC60F043B3DB6DFFE73F464A67B82B0B0B4E  torjittertrapnet   \n",
       "1  055A3CAA2582B87448301367ECD5905D138EF049            verax1   \n",
       "2  D5F2C65F4131A1468D5B67A8838A9B7ED8C049E2               che   \n",
       "3  0F69025EA0A61A96E5444A8BF39A8B1335488610             Uni25   \n",
       "4  2A587ED761FB8CF8AA7CB77164455479C002141B           Unnamed   \n",
       "\n",
       "             last_seen       last_restarted last_changed_addr_port  \\\n",
       "0  2016-12-18 14:00:00  2016-12-08 08:30:31    2016-08-10 03:00:00   \n",
       "1  2016-12-18 14:00:00  2016-08-28 01:40:14    2016-08-28 02:00:00   \n",
       "2  2016-12-18 14:00:00  2016-12-08 18:02:47    2016-06-20 14:00:00   \n",
       "3  2016-12-18 14:00:00  2016-11-28 23:11:16    2016-10-25 16:00:00   \n",
       "4  2016-12-18 14:00:00  2016-12-12 23:06:28    2016-12-12 23:00:00   \n",
       "\n",
       "            first_seen  hibernating          age     uptime  \\\n",
       "0  2016-08-10 03:00:00          0.0   11271600.0   883769.0   \n",
       "1  2014-11-29 22:00:00          0.0   64771200.0  9721186.0   \n",
       "2  2007-10-27 12:00:00          0.0  288583200.0   849433.0   \n",
       "3  2015-04-24 21:00:00          0.0   52160400.0  1694924.0   \n",
       "4  2016-12-12 23:00:00          0.0     486000.0   485612.0   \n",
       "\n",
       "   address_port_uptime  ...     exit_prob  badexit  exit  fast guard  hsdir  \\\n",
       "0           11271600.0  ...             0      0.0   0.0   0.0   0.0    0.0   \n",
       "1            9720000.0  ...             0      0.0   0.0   1.0   0.0    1.0   \n",
       "2           15638400.0  ...   6.33964e-06      0.0   1.0   1.0   0.0    1.0   \n",
       "3            4658400.0  ...             0      0.0   0.0   1.0   0.0    1.0   \n",
       "4             486000.0  ...             0      0.0   0.0   1.0   0.0    1.0   \n",
       "\n",
       "  stable running valid v2dir  \n",
       "0    1.0     1.0   1.0   1.0  \n",
       "1    1.0     1.0   1.0   1.0  \n",
       "2    1.0     1.0   1.0   1.0  \n",
       "3    1.0     1.0   1.0   1.0  \n",
       "4    1.0     1.0   1.0   1.0  \n",
       "\n",
       "[5 rows x 32 columns]"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "columns=['fingerprint',\n",
    "         'nickname',\n",
    "         'last_seen',\n",
    "         'last_restarted',\n",
    "         'last_changed_addr_port',\n",
    "         'first_seen',\n",
    "         'hibernating',\n",
    "         'age', # computed from last_seen and first_seen\n",
    "         'uptime', # computed from last_restarted and date pulled\n",
    "         'address_port_uptime', # computed from last_changed_addr_port and date pulled\n",
    "         'bandwidth_rate',\n",
    "         'bandwidth_burst',\n",
    "         'observed_bandwidth',\n",
    "         'advertised_bandwidth',\n",
    "         'consensus_weight',\n",
    "         'ports_accepted',\n",
    "         'country',\n",
    "         'as_number',\n",
    "         'effective_family',\n",
    "         'indirect_family',\n",
    "         'guard_prob',\n",
    "         'middle_prob',\n",
    "         'exit_prob',\n",
    "         'badexit',\n",
    "         'exit',\n",
    "         'fast',\n",
    "         'guard',\n",
    "         'hsdir',\n",
    "         'stable',\n",
    "         'running',\n",
    "         'valid',\n",
    "         'v2dir']\n",
    "\n",
    "relayDF = pd.DataFrame(columns=columns)\n",
    "\n",
    "date_pulled = datetime.datetime.strptime(response['relays_published'],\"%Y-%m-%d %H:%M:%S\")\n",
    "\n",
    "for relay in response['relays']:\n",
    "    row = [None]*len(columns)\n",
    "    \n",
    "    # IDENTIFIERS\n",
    "    row[0] = relay['fingerprint'] if 'fingerprint' in relay else None\n",
    "    row[1] = relay['nickname'] if 'nickname' in relay else None\n",
    "    \n",
    "    # REPORTED TIMING PROPERTIES\n",
    "    row[2] = relay['last_seen'] if 'last_seen' in relay else None\n",
    "    row[3] = relay['last_restarted'] if 'last_restarted' in relay else None\n",
    "    row[4] = relay['last_changed_address_or_port'] if 'last_changed_address_or_port' in relay else None\n",
    "    row[5] = relay['first_seen'] if 'first_seen' in relay else None\n",
    "    row[6] = 1 if 'hibernating' in relay else 0\n",
    "    \n",
    "    # COMPUTED TIMING PROPERTIES\n",
    "    row[7] = compute_age(relay)\n",
    "    row[8] = compute_uptime(relay, date_pulled)\n",
    "    row[9] = compute_addr_port_uptime(relay, date_pulled)\n",
    "    \n",
    "    # BANDWIDTH PROPERTIES\n",
    "    row[10] = relay['bandwidth_rate'] if 'bandwidth_rate' in relay else None\n",
    "    row[11] = relay['bandwidth_burst'] if 'bandwidth_burst' in relay else None\n",
    "    row[12] = relay['observed_bandwidth'] if 'observed_bandwidth' in relay else None\n",
    "    row[13] = relay['advertised_bandwidth'] if 'advertised_bandwidth' in relay else None\n",
    "    \n",
    "    # CONSENSUS AND PORTS\n",
    "    row[14] = relay['consensus_weight_fraction'] if 'consensus_weight_fraction' in relay else None\n",
    "    row[15] = compute_open_ports(relay)\n",
    "    \n",
    "    # FAMILY PROPERTIES\n",
    "    row[16] = relay['country'] if 'country' in relay else None\n",
    "    row[17] = relay['as_number'] if 'as_number' in relay else None\n",
    "    row[18] = len(relay['effective_family']) if 'effective_family' in relay else None\n",
    "    row[19] = len(relay['indirect_family']) if 'indirect_family' in relay else None\n",
    "    \n",
    "    # POSITIONAL PROBABILITIES\n",
    "    row[20] = relay['guard_probability'] if 'guard_probability' in relay else None\n",
    "    row[21] = relay['middle_probability'] if 'middle_probability' in relay else None\n",
    "    row[22] = relay['exit_probability'] if 'exit_probability' in relay else None\n",
    "    \n",
    "    # FLAGS\n",
    "    if 'flags' in relay:\n",
    "        row[23] = int('BadExit' in relay['flags'])\n",
    "        row[24] = int('Exit' in relay['flags'])\n",
    "        row[25] = int('Fast' in relay['flags'])\n",
    "        row[26] = int('Guard' in relay['flags'])\n",
    "        row[27] = int('HSDir' in relay['flags'])\n",
    "        row[28] = int('Stable' in relay['flags'])\n",
    "        row[29] = int('Running' in relay['flags'])\n",
    "        row[30] = int('Valid' in relay['flags'])\n",
    "        row[31] = int('V2Dir' in relay['flags'])\n",
    "    \n",
    "    relayDF = relayDF.append(pd.DataFrame([row],columns=columns))\n",
    "relayDF.reset_index(drop=True,inplace=True)\n",
    "relayDF.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8382"
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(relayDF)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Run this to save the relay DataFrame as a .csv\n",
    "\n",
    "Make sure you change the name of the file or it will overwrite what's already there!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "relayDF.to_csv('sample-12-18-16-v2dir.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Run this to load a relay dataframe from a .csv file\n",
    "\n",
    "If you already have a parsed .csv representation of the Onionoo response, you can load it with this block."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>fingerprint</th>\n",
       "      <th>nickname</th>\n",
       "      <th>last_seen</th>\n",
       "      <th>last_restarted</th>\n",
       "      <th>last_changed_addr_port</th>\n",
       "      <th>first_seen</th>\n",
       "      <th>hibernating</th>\n",
       "      <th>age</th>\n",
       "      <th>uptime</th>\n",
       "      <th>address_port_uptime</th>\n",
       "      <th>...</th>\n",
       "      <th>middle_prob</th>\n",
       "      <th>exit_prob</th>\n",
       "      <th>badexit</th>\n",
       "      <th>exit</th>\n",
       "      <th>fast</th>\n",
       "      <th>guard</th>\n",
       "      <th>hsdir</th>\n",
       "      <th>stable</th>\n",
       "      <th>running</th>\n",
       "      <th>valid</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2B35CC60F043B3DB6DFFE73F464A67B82B0B0B4E</td>\n",
       "      <td>torjittertrapnet</td>\n",
       "      <td>2016-12-18 14:00:00</td>\n",
       "      <td>2016-12-08 08:30:31</td>\n",
       "      <td>2016-08-10 03:00:00</td>\n",
       "      <td>2016-08-10 03:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>11271600.0</td>\n",
       "      <td>883769.0</td>\n",
       "      <td>11271600.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000001</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>055A3CAA2582B87448301367ECD5905D138EF049</td>\n",
       "      <td>verax1</td>\n",
       "      <td>2016-12-18 14:00:00</td>\n",
       "      <td>2016-08-28 01:40:14</td>\n",
       "      <td>2016-08-28 02:00:00</td>\n",
       "      <td>2014-11-29 22:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>64771200.0</td>\n",
       "      <td>9721186.0</td>\n",
       "      <td>9720000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000028</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>D5F2C65F4131A1468D5B67A8838A9B7ED8C049E2</td>\n",
       "      <td>che</td>\n",
       "      <td>2016-12-18 14:00:00</td>\n",
       "      <td>2016-12-08 18:02:47</td>\n",
       "      <td>2016-06-20 14:00:00</td>\n",
       "      <td>2007-10-27 12:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>288583200.0</td>\n",
       "      <td>849433.0</td>\n",
       "      <td>15638400.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000006</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0F69025EA0A61A96E5444A8BF39A8B1335488610</td>\n",
       "      <td>Uni25</td>\n",
       "      <td>2016-12-18 14:00:00</td>\n",
       "      <td>2016-11-28 23:11:16</td>\n",
       "      <td>2016-10-25 16:00:00</td>\n",
       "      <td>2015-04-24 21:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>52160400.0</td>\n",
       "      <td>1694924.0</td>\n",
       "      <td>4658400.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000010</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2A587ED761FB8CF8AA7CB77164455479C002141B</td>\n",
       "      <td>Unnamed</td>\n",
       "      <td>2016-12-18 14:00:00</td>\n",
       "      <td>2016-12-12 23:06:28</td>\n",
       "      <td>2016-12-12 23:00:00</td>\n",
       "      <td>2016-12-12 23:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>486000.0</td>\n",
       "      <td>485612.0</td>\n",
       "      <td>486000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000015</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 31 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                fingerprint          nickname  \\\n",
       "0  2B35CC60F043B3DB6DFFE73F464A67B82B0B0B4E  torjittertrapnet   \n",
       "1  055A3CAA2582B87448301367ECD5905D138EF049            verax1   \n",
       "2  D5F2C65F4131A1468D5B67A8838A9B7ED8C049E2               che   \n",
       "3  0F69025EA0A61A96E5444A8BF39A8B1335488610             Uni25   \n",
       "4  2A587ED761FB8CF8AA7CB77164455479C002141B           Unnamed   \n",
       "\n",
       "             last_seen       last_restarted last_changed_addr_port  \\\n",
       "0  2016-12-18 14:00:00  2016-12-08 08:30:31    2016-08-10 03:00:00   \n",
       "1  2016-12-18 14:00:00  2016-08-28 01:40:14    2016-08-28 02:00:00   \n",
       "2  2016-12-18 14:00:00  2016-12-08 18:02:47    2016-06-20 14:00:00   \n",
       "3  2016-12-18 14:00:00  2016-11-28 23:11:16    2016-10-25 16:00:00   \n",
       "4  2016-12-18 14:00:00  2016-12-12 23:06:28    2016-12-12 23:00:00   \n",
       "\n",
       "            first_seen  hibernating          age     uptime  \\\n",
       "0  2016-08-10 03:00:00          0.0   11271600.0   883769.0   \n",
       "1  2014-11-29 22:00:00          0.0   64771200.0  9721186.0   \n",
       "2  2007-10-27 12:00:00          0.0  288583200.0   849433.0   \n",
       "3  2015-04-24 21:00:00          0.0   52160400.0  1694924.0   \n",
       "4  2016-12-12 23:00:00          0.0     486000.0   485612.0   \n",
       "\n",
       "   address_port_uptime  ...    middle_prob  exit_prob  badexit  exit  fast  \\\n",
       "0           11271600.0  ...       0.000001   0.000000      0.0   0.0   0.0   \n",
       "1            9720000.0  ...       0.000028   0.000000      0.0   0.0   1.0   \n",
       "2           15638400.0  ...       0.000000   0.000006      0.0   1.0   1.0   \n",
       "3            4658400.0  ...       0.000010   0.000000      0.0   0.0   1.0   \n",
       "4             486000.0  ...       0.000015   0.000000      0.0   0.0   1.0   \n",
       "\n",
       "   guard hsdir stable  running  valid  \n",
       "0    0.0   0.0    1.0      1.0    1.0  \n",
       "1    0.0   1.0    1.0      1.0    1.0  \n",
       "2    0.0   1.0    1.0      1.0    1.0  \n",
       "3    0.0   1.0    1.0      1.0    1.0  \n",
       "4    0.0   1.0    1.0      1.0    1.0  \n",
       "\n",
       "[5 rows x 31 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "relayDF = pd.DataFrame.from_csv('sample-12-18-16.csv')\n",
    "relayDF.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Relay Feature Properties\n",
    "\n",
    "Here we look at the distributions of the relay properties and try different spaces to look for normal distributions."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Advertised Bandwidth Distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 409,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYYAAAEZCAYAAACTsIJzAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x12e5c4390>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import scipy.stats as stats\n",
    "fig = plt.figure()\n",
    "ax1 = fig.add_subplot(111)\n",
    "a = relayDF.advertised_bandwidth[relayDF.advertised_bandwidth>0].apply(lambda x: math.log(x))\n",
    "e = relayDF.advertised_bandwidth[(relayDF.advertised_bandwidth>0)&(relayDF.exit==1)].apply(lambda x: math.log(x))\n",
    "g = relayDF.advertised_bandwidth[(relayDF.advertised_bandwidth>0)&(relayDF.guard==1)].apply(lambda x: math.log(x))\n",
    "ax1.hist((a,e,g),normed=True,label=['all','exit','guard'])\n",
    "fit = stats.norm.pdf(a.sort_values(), a.mean(), a.std())\n",
    "ax1.plot(a.sort_values(),fit,color='b')\n",
    "fit = stats.norm.pdf(e.sort_values(), e.mean(), e.std())\n",
    "ax1.plot(e.sort_values(),fit,color='g')\n",
    "fit = stats.norm.pdf(g.sort_values(), g.mean(), g.std())\n",
    "ax1.plot(g.sort_values(),fit,color='r')\n",
    "plt.title(\"Advertised Bandwidth Distribution\")\n",
    "plt.ylabel(\"Frequency\")\n",
    "plt.xlabel(\"Log-space Bandwidth\")\n",
    "plt.legend(loc='upper left')\n",
    "fig.savefig('./images/12-18-log_bandwidth_distribution',dpi=1000)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 410,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAM4AAACfCAYAAACxzeaWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkhIDISKExISA6HihITEQKg4ISEx8P8CfvtIxvJLwgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1222d54a8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "ax1 = fig.add_subplot(221)\n",
    "x = relayDF[relayDF.advertised_bandwidth>0].advertised_bandwidth.apply(lambda x: math.log(x))\n",
    "res = stats.probplot(x, plot=plt)\n",
    "ax1.set_title(\"Advertised Bandwidth\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 268,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(13.821603721539171, 14.670205264031926, 15.68674609098163)"
      ]
     },
     "execution_count": 268,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a.mean(),e.mean(),g.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 269,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.1875756839413856"
      ]
     },
     "execution_count": 269,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "math.pow(a.std(),2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 440,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "7.568681581995953"
      ]
     },
     "execution_count": 440,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "relayDF[(relayDF.guard==1)].advertised_bandwidth.std()/1000000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 329,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAEZCAYAAACervI0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1412333c8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "ax1 = fig.add_subplot(311)\n",
    "x = relayDF[relayDF.advertised_bandwidth>0].advertised_bandwidth.apply(lambda x: math.log(x))\n",
    "res = stats.probplot(x, plot=plt)\n",
    "ax2 = fig.add_subplot(312)\n",
    "x = relayDF[(relayDF.advertised_bandwidth>0)&(relayDF.exit==1)].advertised_bandwidth.apply(lambda x: math.log(x))\n",
    "res = stats.probplot(x, plot=plt)\n",
    "ax3 = fig.add_subplot(313)\n",
    "x = relayDF[(relayDF.advertised_bandwidth>0)&(relayDF.guard==1)].advertised_bandwidth.apply(lambda x: math.log(x))\n",
    "res = stats.probplot(x, plot=plt)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Consensus Weight Distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 425,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY0AAAEZCAYAAABrUHmEAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n8iYqjUgkEonkTVQakUgkEsmbfwBHFit9PBSxOwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11eb8e128>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "ax1 = fig.add_subplot(111)\n",
    "a = relayDF.consensus_weight[relayDF.consensus_weight>0].apply(lambda x: math.log(x))\n",
    "e = relayDF.consensus_weight[(relayDF.consensus_weight>0)&(relayDF.exit==1)].apply(lambda x: math.log(x))\n",
    "g = relayDF.consensus_weight[(relayDF.consensus_weight>0)&(relayDF.guard==1)].apply(lambda x: math.log(x))\n",
    "ax1.hist((a,e,g),normed=True,label=['all','exit','guard'])\n",
    "fit = stats.norm.pdf(a.sort_values(), a.mean(), a.std())\n",
    "ax1.plot(a.sort_values(),fit,color='b')\n",
    "fit = stats.norm.pdf(e.sort_values(), e.mean(), e.std())\n",
    "ax1.plot(e.sort_values(),fit,color='g')\n",
    "fit = stats.norm.pdf(g.sort_values(), g.mean(), g.std())\n",
    "ax1.plot(g.sort_values(),fit,color='r')\n",
    "plt.title(\"Consensus Weight Distribution\")\n",
    "plt.ylabel(\"Frequency\")\n",
    "plt.xlabel(\"Log-space Fraction of Consensus Weight\")\n",
    "plt.legend(loc='upper left')\n",
    "fig.savefig('./images/12-18-log_consensus_distribution',dpi=1000)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 296,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.0001388696016257246, 0.0002440580011284369, 0.00034393339421295524)"
      ]
     },
     "execution_count": 296,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "relayDF.consensus_weight.mean(),relayDF[relayDF.exit==1].consensus_weight.mean(),relayDF[relayDF.guard==1].consensus_weight.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 328,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEZCAYAAACNebLAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x119d6a278>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "ax1 = fig.add_subplot(311)\n",
    "x = relayDF[relayDF.consensus_weight>0].consensus_weight.apply(lambda x: math.log(x))\n",
    "res = stats.probplot(x, plot=plt)\n",
    "ax2 = fig.add_subplot(312)\n",
    "x = relayDF[(relayDF.consensus_weight>0)&(relayDF.exit==1)].consensus_weight.apply(lambda x: math.log(x))\n",
    "res = stats.probplot(x, plot=plt)\n",
    "ax3 = fig.add_subplot(313)\n",
    "x = relayDF[(relayDF.consensus_weight>0)&(relayDF.guard==1)].consensus_weight.apply(lambda x: math.log(x))\n",
    "res = stats.probplot(x, plot=plt)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 431,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.013900472634234681"
      ]
     },
     "execution_count": 431,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "relayDF[(relayDF.consensus_weight>0)].consensus_weight.mean()*100"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Bandwidth Discrepencies\n",
    "\n",
    "There are several bandwdith values a relay can report. They are typically not the same. The smallest is used to determine consensus weight and therefore the most important with regards to a relay's likelihood of being chosen for a path."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXYAAAEZCAYAAACD/A7qAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x12272bd68>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "adv_mean = [relayDF.advertised_bandwidth.mean(),\n",
    "            relayDF[relayDF.exit==1].advertised_bandwidth.mean(),\n",
    "            relayDF[relayDF.guard==1].advertised_bandwidth.mean()]\n",
    "\n",
    "obs_mean = [relayDF.observed_bandwidth.mean(),\n",
    "            relayDF[relayDF.exit==1].observed_bandwidth.mean(),\n",
    "            relayDF[relayDF.guard==1].observed_bandwidth.mean()]\n",
    "\n",
    "burst_mean = [relayDF.bandwidth_burst.mean(),\n",
    "              relayDF[relayDF.exit==1].bandwidth_burst.mean(),\n",
    "              relayDF[relayDF.guard==1].bandwidth_burst.mean()]\n",
    "\n",
    "rate_mean = [relayDF.bandwidth_rate.mean(),\n",
    "             relayDF[relayDF.exit==1].bandwidth_rate.mean(),\n",
    "             relayDF[relayDF.guard==1].bandwidth_rate.mean()]\n",
    "\n",
    "x = [0,1,2]\n",
    "\n",
    "plt.scatter(x, adv_mean, s=80, marker='o',label='Advertised',color='0.75')\n",
    "plt.scatter(x, obs_mean, marker='+', label='Observed',color='k')\n",
    "plt.scatter(x, burst_mean, marker='_', label='Burst',color='k')\n",
    "plt.scatter(x, rate_mean, marker='x', label='Rate',color='k')\n",
    "plt.xlabel('Type of Relay')\n",
    "plt.ylabel('Reported Bandwidth (B/s)')\n",
    "plt.title('Mean Reported Bandwidth Values')\n",
    "plt.xlim(-0.25,2.25)\n",
    "plt.ylim(ymin=0)\n",
    "plt.legend(loc='upper left')\n",
    "plt.xticks(x, ['All', 'Exit','Guard'])\n",
    "plt.savefig('./images/12-18-reported_bandwidths',dpi=1000)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Family Size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1379"
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Number of relays reporting a family size>0:\n",
    "len(relayDF[relayDF.effective_family>0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "99"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Number of relays reporting an indirect family size>0:\n",
    "len(relayDF[(relayDF.indirect_family>0)&(relayDF.effective_family>0)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "7.5165086222440189"
      ]
     },
     "execution_count": 131,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "relayDF[relayDF.effective_family>0].effective_family.std()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "6.3428629928602946"
      ]
     },
     "execution_count": 132,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "relayDF[(relayDF.indirect_family>0)&(relayDF.effective_family>0)].indirect_family.std()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 252,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAEZCAYAAACEkhK6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x103d64668>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "ax1 = fig.add_subplot(111)\n",
    "a = relayDF.effective_family[relayDF.effective_family>0]\n",
    "e = relayDF.effective_family[(relayDF.effective_family>0)&(relayDF.exit==1)]\n",
    "g = relayDF.effective_family[(relayDF.effective_family>0)&(relayDF.guard==1)]\n",
    "ax1.hist((a,e,g),normed=True,label=['all','exit','guard'])\n",
    "plt.axvline(math.log(relayDF[relayDF.effective_family>0].effective_family.mean()), color='k', linestyle='dashed', linewidth=2)\n",
    "plt.title(\"Family Size Distribution\")\n",
    "plt.legend(loc='upper right')\n",
    "plt.ylabel(\"Frequency\")\n",
    "plt.xlabel(\"Family Size\")\n",
    "fig.savefig('./images/12-18-family_size_distribution',dpi=1000)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## AS Number"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "56\n"
     ]
    }
   ],
   "source": [
    "# How many relays didn't report an AS?\n",
    "count = 0\n",
    "for num in relayDF.as_number:\n",
    "    if num == None:\n",
    "        count += 1\n",
    "print(count)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Flags"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "573"
      ]
     },
     "execution_count": 126,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Number of exit-only relays\n",
    "len(relayDF[(relayDF.exit==1)&(relayDF.guard==0)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2123"
      ]
     },
     "execution_count": 127,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Number of guard-only relays\n",
    "len(relayDF[(relayDF.guard==1)&(relayDF.exit==0)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "458"
      ]
     },
     "execution_count": 128,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Number of exit-guard relays\n",
    "len(relayDF[(relayDF.guard==1)&(relayDF.exit==1)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5228"
      ]
     },
     "execution_count": 129,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Number of non-exit, non-guard relays\n",
    "len(relayDF[(relayDF.guard==0)&(relayDF.exit==0)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11cf0feb8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAccAAAEZCAYAAADiyO7tAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrfcjA0wAAAABSURBVBuiXbQEAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11cd0f3c8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "flags = ['badexit','exit','fast','guard','hsdir','stable','running','valid','v2dir']\n",
    "countsA = [relayDF[flag].sum()/len(relayDF) for flag in flags]\n",
    "countsE = [relayDF[relayDF.exit==1][flag].sum()/len(relayDF[relayDF.exit==1]) for flag in flags]\n",
    "countsG = [relayDF[relayDF.guard==1][flag].sum()/len(relayDF[relayDF.guard==1]) for flag in flags]\n",
    "\n",
    "negA = [1-x for x in countsA]\n",
    "negE = [1-x for x in countsE]\n",
    "negG = [1-x for x in countsG]\n",
    "\n",
    "N=9\n",
    "ind=np.arange(N)\n",
    "width=1\n",
    "\n",
    "fig = plt.figure()\n",
    "fig, axarr = plt.subplots(3,1,sharex=True)\n",
    "axarr[0].bar(ind,countsA,label='all',color='b')\n",
    "axarr[0].bar(ind,negA,bottom=countsA,color='c')\n",
    "axarr[1].bar(ind,countsE,label='all',color='g')\n",
    "axarr[1].bar(ind,negE,bottom=countsE,color='c')\n",
    "axarr[2].bar(ind,countsG,label='all',color='r')\n",
    "axarr[2].bar(ind,negG,bottom=countsG,color='c')\n",
    "\n",
    "plt.xlabel(\"Flag\")\n",
    "axarr[0].set_xticks(ind+0.5)\n",
    "axarr[0].set_xticklabels(flags)\n",
    "axarr[0].set_title('Flags')\n",
    "axarr[1].set_ylabel(\"Frequency\")\n",
    "# Shrink current axis by 20%\n",
    "box = axarr[0].get_position()\n",
    "axarr[0].set_position([box.x0, box.y0, box.width * 0.8, box.height])\n",
    "box = axarr[1].get_position()\n",
    "axarr[1].set_position([box.x0, box.y0, box.width * 0.8, box.height])\n",
    "box = axarr[2].get_position()\n",
    "axarr[2].set_position([box.x0, box.y0, box.width * 0.8, box.height])\n",
    "\n",
    "# Put a legend to the right of the current axis\n",
    "axarr[0].legend(('all - flag','all - no flag'),loc='center left', bbox_to_anchor=(1, 0.5))\n",
    "axarr[1].legend(('exit - flag','exit - no flag'),loc='center left', bbox_to_anchor=(1, 0.5))\n",
    "lgd = axarr[2].legend(('guard - flag','guard - no flag'),loc='center left', bbox_to_anchor=(1, 0.5))\n",
    "\n",
    "fig.savefig('./images/12-18-flags.png',dpi=1000,bbox_extra_artists=(lgd,), bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Ports Accepted\n",
    "\n",
    "Default Exit Policy = blackblist\n",
    "\n",
    "Reduced Exit Policy = whitelist"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 286,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAakAAAEbCAYAAABgLnslAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ndCgjo9Z7jJEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x122736320>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "ax1 = fig.add_subplot(111)\n",
    "# a = relayDF.ports_accepted[(relayDF.exit==0)&(relayDF.guard==0)]\n",
    "e = relayDF.ports_accepted[(relayDF.exit==1)]\n",
    "g = relayDF.ports_accepted[(relayDF.guard==1)]\n",
    "ax1.hist((e,g),normed=True,label=['exit','guard'],color=['g','r'])\n",
    "plt.title(\"Distribution of Ports Accepted\")\n",
    "plt.legend(loc='upper right')\n",
    "plt.ylabel(\"Frequency\")\n",
    "plt.xlabel(\"# Ports Accepted\")\n",
    "plt.tight_layout()\n",
    "plt.savefig('./images/12-18-ports_accepted_distribution',dpi=1000)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Timing Distributions\n",
    "\n",
    "Age, uptime, addr/port uptime distributions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 396,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAakAAAEbCAYAAABgLnslAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x12e5f72e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "ax1 = fig.add_subplot(221)\n",
    "a = relayDF[relayDF.age>0].age.apply(lambda x: math.log(x))\n",
    "e = relayDF[relayDF.age>0].age[(relayDF.exit==1)].apply(lambda x: math.log(x))\n",
    "g = relayDF[relayDF.age>0].age[(relayDF.guard==1)].apply(lambda x: math.log(x))\n",
    "ax1.hist((a,e,g),normed=True,label=['all','exit','guard'],alpha=0.7)\n",
    "fit = stats.norm.pdf(a.sort_values(), a.mean(), a.std())\n",
    "ax1.plot(a.sort_values(),fit,color='b')\n",
    "fit = stats.norm.pdf(e.sort_values(), e.mean(), e.std())\n",
    "ax1.plot(e.sort_values(),fit,color='g')\n",
    "fit = stats.norm.pdf(g.sort_values(), g.mean(), g.std())\n",
    "ax1.plot(g.sort_values(),fit,color='r')\n",
    "ax1.set_title(\"Age Distribution\")\n",
    "ax1.set_ylabel(\"Frequency\")\n",
    "ax1.set_xlabel(\"Age (log(s))\")\n",
    "ax1.set_xlim(xmin=a.min(),xmax=a.max())\n",
    "ax1.set_ylim(ymin=0)\n",
    "\n",
    "ax2 = fig.add_subplot(222)\n",
    "a = relayDF[relayDF.uptime>0].uptime.apply(lambda x: math.log(x))\n",
    "e = relayDF[relayDF.uptime>0].uptime[(relayDF.exit==1)].apply(lambda x: math.log(x))\n",
    "g = relayDF[relayDF.uptime>0].uptime[(relayDF.guard==1)].apply(lambda x: math.log(x))\n",
    "ax2.hist((a,e,g),normed=True,label=['all','exit','guard'],alpha=0.7)\n",
    "fit = stats.norm.pdf(a.sort_values(), a.mean(), a.std())\n",
    "ax2.plot(a.sort_values(),fit,color='b')\n",
    "fit = stats.norm.pdf(e.sort_values(), e.mean(), e.std())\n",
    "ax2.plot(e.sort_values(),fit,color='g')\n",
    "fit = stats.norm.pdf(g.sort_values(), g.mean(), g.std())\n",
    "ax2.plot(g.sort_values(),fit,color='r')\n",
    "ax2.set_title(\"Uptime Distribution\")\n",
    "ax2.set_ylabel(\"Frequency\")\n",
    "ax2.set_xlabel(\"Uptime (log(s))\")\n",
    "ax2.set_xlim(xmin=a.min(),xmax=a.max())\n",
    "ax2.set_ylim(ymin=0)\n",
    "\n",
    "ax3 = fig.add_subplot(223)\n",
    "a = relayDF[relayDF.address_port_uptime>0].uptime.apply(lambda x: math.log(x))\n",
    "e = relayDF[relayDF.address_port_uptime>0].address_port_uptime[(relayDF.exit==1)].apply(lambda x: math.log(x))\n",
    "g = relayDF[relayDF.address_port_uptime>0].address_port_uptime[(relayDF.guard==1)].apply(lambda x: math.log(x))\n",
    "ax3.hist((a,e,g),normed=True,label=['all','exit','guard'],alpha=0.7)\n",
    "fit = stats.norm.pdf(a.sort_values(), a.mean(), a.std())\n",
    "ax3.plot(a.sort_values(),fit,color='b')\n",
    "fit = stats.norm.pdf(e.sort_values(), e.mean(), e.std())\n",
    "ax3.plot(e.sort_values(),fit,color='g')\n",
    "fit = stats.norm.pdf(g.sort_values(), g.mean(), g.std())\n",
    "ax3.plot(g.sort_values(),fit,color='r')\n",
    "ax3.set_title(\"Address or Port Uptime Distribution\")\n",
    "ax3.set_ylabel(\"Frequency\")\n",
    "ax3.set_xlabel(\"Address or Port Uptime (log(s))\")\n",
    "ax3.set_xlim(xmin=a.min(),xmax=a.max())\n",
    "ax3.set_ylim(ymin=0)\n",
    "\n",
    "plt.legend(bbox_to_anchor=(2, 1))\n",
    "plt.tight_layout()\n",
    "# plt.savefig('./images/12-18-log-norm_age_distribution',dpi=1000)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 407,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAaoAAAEbCAYAAACLGcAmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x12c414b00>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "ax1 = fig.add_subplot(221)\n",
    "x = relayDF[relayDF.age>0].age.apply(lambda x: math.log(x))\n",
    "res = stats.probplot(x, plot=plt)\n",
    "ax1.set_title(\"Age\")\n",
    "ax2 = fig.add_subplot(222)\n",
    "x = relayDF[(relayDF.uptime>0)].uptime.apply(lambda x: math.log(x))\n",
    "res = stats.probplot(x, plot=plt)\n",
    "ax2.set_title(\"Uptime\")\n",
    "ax3 = fig.add_subplot(223)\n",
    "x = relayDF[(relayDF.address_port_uptime>0)].address_port_uptime.apply(lambda x: math.log(x))\n",
    "res = stats.probplot(x, plot=plt)\n",
    "ax3.set_title(\"Address or Port Uptime\")\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"./images/12-18-goodness-timing-fit\",dpi=1000)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 341,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAakAAAEbCAYAAABgLnslAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1195e0cc0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "ax1 = fig.add_subplot(221)\n",
    "x = relayDF[relayDF.uptime>0].uptime.apply(lambda x: math.log(x))\n",
    "res = stats.probplot(x, plot=plt)\n",
    "ax2 = fig.add_subplot(222)\n",
    "x = relayDF[(relayDF.uptime>0)&(relayDF.exit==1)].uptime.apply(lambda x: math.log(x))\n",
    "res = stats.probplot(x, plot=plt)\n",
    "ax3 = fig.add_subplot(223)\n",
    "x = relayDF[(relayDF.uptime>0)&(relayDF.guard==1)].uptime.apply(lambda x: math.log(x))\n",
    "res = stats.probplot(x, plot=plt)\n",
    "ax1.set_title('Relay Uptime')\n",
    "ax2.set_title('Exit Uptime')\n",
    "ax3.set_title('Guard Uptime')\n",
    "plt.tight_layout()\n",
    "plt.savefig('./images/12-18-log_uptime_normality_plot',dpi=1000)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Bandwidth change over time\n",
    "\n",
    "Collect bandwidth data for 500 relays. Store exit and non-exit separately. Plot bandwidth and time.\n",
    "\n",
    "\n",
    "TODO: get stdev of bandwidth over time for exit and non-exit relays and compare."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "with open('bw_sample_ALL.json') as json_data:\n",
    "    resp_all = json.load(json_data)\n",
    "    \n",
    "with open('bw_sample_exitflag.json') as json_data:\n",
    "    resp_exit = json.load(json_data)\n",
    "\n",
    "with open('bw_sample_Guard.json') as json_data:\n",
    "    resp_guard = json.load(json_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "samples = {'3_days':{'exit':[],'all':[], 'guard':[]},\n",
    "           '1_week':{'exit':[],'all':[], 'guard':[]},\n",
    "           '1_month':{'exit':[],'all':[], 'guard':[]},\n",
    "           '3_months':{'exit':[],'all':[], 'guard':[]},\n",
    "           '1_year':{'exit':[],'all':[], 'guard':[]},\n",
    "           '5_years':{'exit':[],'all':[], 'guard':[]}}\n",
    "\n",
    "for relay in resp_all['relays']:\n",
    "    if not 'write_history' in relay:\n",
    "        pass\n",
    "    else:\n",
    "        for history in relay['write_history']:\n",
    "            samples[history]['all'].append(relay['write_history'][history]['values'][:10])\n",
    "        \n",
    "for relay in resp_exit['relays']:\n",
    "    if not 'write_history' in relay:\n",
    "        pass\n",
    "    else:\n",
    "        for history in relay['write_history']:\n",
    "            samples[history]['exit'].append(relay['write_history'][history]['values'][:10])\n",
    "            \n",
    "for relay in resp_guard['relays']:\n",
    "    if not 'write_history' in relay:\n",
    "        pass\n",
    "    else:\n",
    "        for history in relay['write_history']:\n",
    "            samples[history]['guard'].append(relay['write_history'][history]['values'][:10])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "allBW = pd.DataFrame()\n",
    "for entry in samples['5_years']['all']:\n",
    "    allBW = allBW.append(pd.DataFrame([entry]))\n",
    "allBW = allBW.transpose()\n",
    "allBW['mean'] = allBW.mean(axis=1)\n",
    "allBW"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "exitBW = pd.DataFrame()\n",
    "for entry in samples['5_years']['exit']:\n",
    "    exitBW = exitBW.append(pd.DataFrame([entry]))\n",
    "exitBW = exitBW.transpose()\n",
    "exitBW['mean'] = exitBW.mean(axis=1)\n",
    "exitBW.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "guardBW = pd.DataFrame()\n",
    "for entry in samples['5_years']['guard']:\n",
    "    guardBW = guardBW.append(pd.DataFrame([entry]))\n",
    "guardBW = guardBW.transpose()\n",
    "guardBW['mean'] = guardBW.mean(axis=1)\n",
    "guardBW"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "fig = plt.figure()\n",
    "ax1 = fig.add_subplot(111)\n",
    "x = [i*10 for i in range(0,10)]\n",
    "ax1.plot(x, exitBW['mean'], c='b', marker=\"s\", label='exit')\n",
    "ax1.plot(x, allBW['mean'], c='r', marker=\"o\", label='all')\n",
    "ax1.plot(x, guardBW['mean'], c='g', marker=\"v\", label='guard')\n",
    "plt.legend(loc='lower right')\n",
    "plt.title('Bandwidth')\n",
    "plt.xlabel('Time (days)')\n",
    "plt.ylabel('Bandwidth')\n",
    "fig.savefig('./images/time_bandwidth.png', dpi=1000)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Correlations\n",
    "\n",
    "Looking for correlation between the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "logDF = log_normalize(relayDF,['bandwidth','consensus_weight'])\n",
    "# logDF = relayDF.copy()\n",
    "# logDF = logDF[(logDF.bandwidth>0)&(logDF.consensus_weight>0)]\n",
    "# logDF.bandwidth = logDF.bandwidth.apply(lambda x: math.log(x))\n",
    "# logDF.consensus_weight = logDF.consensus_weight.apply(lambda x: math.log(x))\n",
    "A = logDF[(logDF.guard==0)&(logDF.exit==0)]\n",
    "E = logDF[logDF.exit==1]\n",
    "G = logDF[logDF.guard==1]\n",
    "plot_feature(A,E,G,'bandwidth','consensus_weight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# logDF = log_normalize(relayDF,['bandwidth','uptime'])\n",
    "logDF = relayDF.copy()\n",
    "logDF = logDF[(logDF.bandwidth>0)&(logDF.uptime>0)]\n",
    "logDF.bandwidth = logDF.bandwidth.apply(lambda x: math.log(x))\n",
    "logDF.uptime = logDF.uptime.apply(lambda x: math.log(x))\n",
    "A = logDF[(logDF.guard==0)&(logDF.exit==0)]\n",
    "E = logDF[logDF.exit==1]\n",
    "G = logDF[logDF.guard==1]\n",
    "plot_feature(A,E,G,'bandwidth','uptime')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "logDF = log_normalize(relayDF,['bandwidth','age'])\n",
    "# logDF = relayDF.copy()\n",
    "# logDF = logDF[(logDF.bandwidth>0)&(logDF.consensus_weight>0)]\n",
    "# logDF.bandwidth = logDF.bandwidth.apply(lambda x: math.log(x))\n",
    "# logDF.consensus_weight = logDF.consensus_weight.apply(lambda x: math.log(x))\n",
    "A = logDF[(logDF.guard==0)&(logDF.exit==0)]\n",
    "E = logDF[logDF.exit==1]\n",
    "G = logDF[logDF.guard==1]\n",
    "plot_feature(A,E,G,'bandwidth','age')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "logDF = log_normalize(relayDF,['bandwidth','ports_accepted'])\n",
    "A = logDF[(logDF.guard==0)&(logDF.exit==0)]\n",
    "E = logDF[logDF.exit==1]\n",
    "G = logDF[logDF.guard==1]\n",
    "plot_feature(A,E,G,'bandwidth','ports_accepted')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "logDF = log_normalize(relayDF,['bandwidth'])\n",
    "A = logDF[(logDF.guard==0)&(logDF.exit==0)]\n",
    "E = logDF[logDF.exit==1]\n",
    "G = logDF[logDF.guard==1]\n",
    "plot_feature(A,E,G,'bandwidth','family_size')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# logDF = log_normalize(relayDF,['family_size','age'])\n",
    "A = relayDF[(relayDF.guard==0)&(relayDF.exit==0)]\n",
    "E = relayDF[relayDF.exit==1]\n",
    "G = relayDF[relayDF.guard==1]\n",
    "plot_feature(A,E,G,'family_size','age')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Machine Learning\n",
    "\n",
    "Here we use regression models to predict lifespans and classification models to classify relays as either exit or non-exit."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(8382, 32)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>fingerprint</th>\n",
       "      <th>nickname</th>\n",
       "      <th>last_seen</th>\n",
       "      <th>last_restarted</th>\n",
       "      <th>last_changed_addr_port</th>\n",
       "      <th>first_seen</th>\n",
       "      <th>hibernating</th>\n",
       "      <th>age</th>\n",
       "      <th>uptime</th>\n",
       "      <th>address_port_uptime</th>\n",
       "      <th>...</th>\n",
       "      <th>exit_prob</th>\n",
       "      <th>badexit</th>\n",
       "      <th>exit</th>\n",
       "      <th>fast</th>\n",
       "      <th>guard</th>\n",
       "      <th>hsdir</th>\n",
       "      <th>stable</th>\n",
       "      <th>running</th>\n",
       "      <th>valid</th>\n",
       "      <th>v2dir</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2B35CC60F043B3DB6DFFE73F464A67B82B0B0B4E</td>\n",
       "      <td>torjittertrapnet</td>\n",
       "      <td>2016-12-18 14:00:00</td>\n",
       "      <td>2016-12-08 08:30:31</td>\n",
       "      <td>2016-08-10 03:00:00</td>\n",
       "      <td>2016-08-10 03:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>11271600.0</td>\n",
       "      <td>883769.0</td>\n",
       "      <td>11271600.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>055A3CAA2582B87448301367ECD5905D138EF049</td>\n",
       "      <td>verax1</td>\n",
       "      <td>2016-12-18 14:00:00</td>\n",
       "      <td>2016-08-28 01:40:14</td>\n",
       "      <td>2016-08-28 02:00:00</td>\n",
       "      <td>2014-11-29 22:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>64771200.0</td>\n",
       "      <td>9721186.0</td>\n",
       "      <td>9720000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>D5F2C65F4131A1468D5B67A8838A9B7ED8C049E2</td>\n",
       "      <td>che</td>\n",
       "      <td>2016-12-18 14:00:00</td>\n",
       "      <td>2016-12-08 18:02:47</td>\n",
       "      <td>2016-06-20 14:00:00</td>\n",
       "      <td>2007-10-27 12:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>288583200.0</td>\n",
       "      <td>849433.0</td>\n",
       "      <td>15638400.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000006</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0F69025EA0A61A96E5444A8BF39A8B1335488610</td>\n",
       "      <td>Uni25</td>\n",
       "      <td>2016-12-18 14:00:00</td>\n",
       "      <td>2016-11-28 23:11:16</td>\n",
       "      <td>2016-10-25 16:00:00</td>\n",
       "      <td>2015-04-24 21:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>52160400.0</td>\n",
       "      <td>1694924.0</td>\n",
       "      <td>4658400.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2A587ED761FB8CF8AA7CB77164455479C002141B</td>\n",
       "      <td>Unnamed</td>\n",
       "      <td>2016-12-18 14:00:00</td>\n",
       "      <td>2016-12-12 23:06:28</td>\n",
       "      <td>2016-12-12 23:00:00</td>\n",
       "      <td>2016-12-12 23:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>486000.0</td>\n",
       "      <td>485612.0</td>\n",
       "      <td>486000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 32 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                fingerprint          nickname  \\\n",
       "0  2B35CC60F043B3DB6DFFE73F464A67B82B0B0B4E  torjittertrapnet   \n",
       "1  055A3CAA2582B87448301367ECD5905D138EF049            verax1   \n",
       "2  D5F2C65F4131A1468D5B67A8838A9B7ED8C049E2               che   \n",
       "3  0F69025EA0A61A96E5444A8BF39A8B1335488610             Uni25   \n",
       "4  2A587ED761FB8CF8AA7CB77164455479C002141B           Unnamed   \n",
       "\n",
       "             last_seen       last_restarted last_changed_addr_port  \\\n",
       "0  2016-12-18 14:00:00  2016-12-08 08:30:31    2016-08-10 03:00:00   \n",
       "1  2016-12-18 14:00:00  2016-08-28 01:40:14    2016-08-28 02:00:00   \n",
       "2  2016-12-18 14:00:00  2016-12-08 18:02:47    2016-06-20 14:00:00   \n",
       "3  2016-12-18 14:00:00  2016-11-28 23:11:16    2016-10-25 16:00:00   \n",
       "4  2016-12-18 14:00:00  2016-12-12 23:06:28    2016-12-12 23:00:00   \n",
       "\n",
       "            first_seen  hibernating          age     uptime  \\\n",
       "0  2016-08-10 03:00:00          0.0   11271600.0   883769.0   \n",
       "1  2014-11-29 22:00:00          0.0   64771200.0  9721186.0   \n",
       "2  2007-10-27 12:00:00          0.0  288583200.0   849433.0   \n",
       "3  2015-04-24 21:00:00          0.0   52160400.0  1694924.0   \n",
       "4  2016-12-12 23:00:00          0.0     486000.0   485612.0   \n",
       "\n",
       "   address_port_uptime  ...    exit_prob  badexit  exit  fast  guard  hsdir  \\\n",
       "0           11271600.0  ...     0.000000      0.0   0.0   0.0    0.0    0.0   \n",
       "1            9720000.0  ...     0.000000      0.0   0.0   1.0    0.0    1.0   \n",
       "2           15638400.0  ...     0.000006      0.0   1.0   1.0    0.0    1.0   \n",
       "3            4658400.0  ...     0.000000      0.0   0.0   1.0    0.0    1.0   \n",
       "4             486000.0  ...     0.000000      0.0   0.0   1.0    0.0    1.0   \n",
       "\n",
       "  stable running  valid  v2dir  \n",
       "0    1.0     1.0    1.0    1.0  \n",
       "1    1.0     1.0    1.0    1.0  \n",
       "2    1.0     1.0    1.0    1.0  \n",
       "3    1.0     1.0    1.0    1.0  \n",
       "4    1.0     1.0    1.0    1.0  \n",
       "\n",
       "[5 rows x 32 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "relayDF = pd.DataFrame.from_csv('sample-12-18-16-v2dir.csv')\n",
    "print(relayDF.shape)\n",
    "relayDF.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(8382, 32)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>fingerprint</th>\n",
       "      <th>nickname</th>\n",
       "      <th>last_seen</th>\n",
       "      <th>last_restarted</th>\n",
       "      <th>last_changed_addr_port</th>\n",
       "      <th>first_seen</th>\n",
       "      <th>hibernating</th>\n",
       "      <th>age</th>\n",
       "      <th>uptime</th>\n",
       "      <th>address_port_uptime</th>\n",
       "      <th>...</th>\n",
       "      <th>exit_prob</th>\n",
       "      <th>badexit</th>\n",
       "      <th>exit</th>\n",
       "      <th>fast</th>\n",
       "      <th>guard</th>\n",
       "      <th>hsdir</th>\n",
       "      <th>stable</th>\n",
       "      <th>running</th>\n",
       "      <th>valid</th>\n",
       "      <th>v2dir</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2B35CC60F043B3DB6DFFE73F464A67B82B0B0B4E</td>\n",
       "      <td>torjittertrapnet</td>\n",
       "      <td>2016-12-18 14:00:00</td>\n",
       "      <td>2016-12-08 08:30:31</td>\n",
       "      <td>2016-08-10 03:00:00</td>\n",
       "      <td>2016-08-10 03:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>11271600.0</td>\n",
       "      <td>883769.0</td>\n",
       "      <td>11271600.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>055A3CAA2582B87448301367ECD5905D138EF049</td>\n",
       "      <td>verax1</td>\n",
       "      <td>2016-12-18 14:00:00</td>\n",
       "      <td>2016-08-28 01:40:14</td>\n",
       "      <td>2016-08-28 02:00:00</td>\n",
       "      <td>2014-11-29 22:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>64771200.0</td>\n",
       "      <td>9721186.0</td>\n",
       "      <td>9720000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>D5F2C65F4131A1468D5B67A8838A9B7ED8C049E2</td>\n",
       "      <td>che</td>\n",
       "      <td>2016-12-18 14:00:00</td>\n",
       "      <td>2016-12-08 18:02:47</td>\n",
       "      <td>2016-06-20 14:00:00</td>\n",
       "      <td>2007-10-27 12:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>288583200.0</td>\n",
       "      <td>849433.0</td>\n",
       "      <td>15638400.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000006</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0F69025EA0A61A96E5444A8BF39A8B1335488610</td>\n",
       "      <td>Uni25</td>\n",
       "      <td>2016-12-18 14:00:00</td>\n",
       "      <td>2016-11-28 23:11:16</td>\n",
       "      <td>2016-10-25 16:00:00</td>\n",
       "      <td>2015-04-24 21:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>52160400.0</td>\n",
       "      <td>1694924.0</td>\n",
       "      <td>4658400.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2A587ED761FB8CF8AA7CB77164455479C002141B</td>\n",
       "      <td>Unnamed</td>\n",
       "      <td>2016-12-18 14:00:00</td>\n",
       "      <td>2016-12-12 23:06:28</td>\n",
       "      <td>2016-12-12 23:00:00</td>\n",
       "      <td>2016-12-12 23:00:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>486000.0</td>\n",
       "      <td>485612.0</td>\n",
       "      <td>486000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 32 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                fingerprint          nickname  \\\n",
       "0  2B35CC60F043B3DB6DFFE73F464A67B82B0B0B4E  torjittertrapnet   \n",
       "1  055A3CAA2582B87448301367ECD5905D138EF049            verax1   \n",
       "2  D5F2C65F4131A1468D5B67A8838A9B7ED8C049E2               che   \n",
       "3  0F69025EA0A61A96E5444A8BF39A8B1335488610             Uni25   \n",
       "4  2A587ED761FB8CF8AA7CB77164455479C002141B           Unnamed   \n",
       "\n",
       "             last_seen       last_restarted last_changed_addr_port  \\\n",
       "0  2016-12-18 14:00:00  2016-12-08 08:30:31    2016-08-10 03:00:00   \n",
       "1  2016-12-18 14:00:00  2016-08-28 01:40:14    2016-08-28 02:00:00   \n",
       "2  2016-12-18 14:00:00  2016-12-08 18:02:47    2016-06-20 14:00:00   \n",
       "3  2016-12-18 14:00:00  2016-11-28 23:11:16    2016-10-25 16:00:00   \n",
       "4  2016-12-18 14:00:00  2016-12-12 23:06:28    2016-12-12 23:00:00   \n",
       "\n",
       "            first_seen  hibernating          age     uptime  \\\n",
       "0  2016-08-10 03:00:00          0.0   11271600.0   883769.0   \n",
       "1  2014-11-29 22:00:00          0.0   64771200.0  9721186.0   \n",
       "2  2007-10-27 12:00:00          0.0  288583200.0   849433.0   \n",
       "3  2015-04-24 21:00:00          0.0   52160400.0  1694924.0   \n",
       "4  2016-12-12 23:00:00          0.0     486000.0   485612.0   \n",
       "\n",
       "   address_port_uptime  ...    exit_prob  badexit  exit  fast  guard  hsdir  \\\n",
       "0           11271600.0  ...     0.000000      0.0   0.0   0.0    0.0    0.0   \n",
       "1            9720000.0  ...     0.000000      0.0   0.0   1.0    0.0    1.0   \n",
       "2           15638400.0  ...     0.000006      0.0   1.0   1.0    0.0    1.0   \n",
       "3            4658400.0  ...     0.000000      0.0   0.0   1.0    0.0    1.0   \n",
       "4             486000.0  ...     0.000000      0.0   0.0   1.0    0.0    1.0   \n",
       "\n",
       "  stable running  valid  v2dir  \n",
       "0    1.0     1.0    1.0    1.0  \n",
       "1    1.0     1.0    1.0    1.0  \n",
       "2    1.0     1.0    1.0    1.0  \n",
       "3    1.0     1.0    1.0    1.0  \n",
       "4    1.0     1.0    1.0    1.0  \n",
       "\n",
       "[5 rows x 32 columns]"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "relayDF = pd.DataFrame.from_csv('sample-12-18-16-v2dir.csv')\n",
    "print(relayDF.shape)\n",
    "relayDF.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "log_cols = ['uptime','address_port_uptime','advertised_bandwidth','consensus_weight']\n",
    "df = relayDF.copy()\n",
    "df = df[(df.uptime>0)&(df.address_port_uptime>0)&(df.advertised_bandwidth>0)&(df.consensus_weight>0)]\n",
    "# for col in log_cols:\n",
    "#     df[col] = df[col].apply(lambda x: math.log(x))\n",
    "# print(df.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(7158, 32)\n"
     ]
    }
   ],
   "source": [
    "label = 'guard'\n",
    "feats = ['advertised_bandwidth',\n",
    "         'fast',\n",
    "         'exit',\n",
    "         'age',\n",
    "         'stable',\n",
    "         'v2dir',\n",
    "#          'family_size',\n",
    "         'consensus_weight',\n",
    "         'ports_accepted',\n",
    "         'uptime',\n",
    "         'address_port_uptime']\n",
    "\n",
    "# Remove NaN feature rows\n",
    "for feat in feats:\n",
    "    df = df[np.isfinite(df[feat])]\n",
    "print(df.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "for feat in feats:\n",
    "    df[feat] = normalize(df[feat])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 465,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(7157, 32)\n"
     ]
    }
   ],
   "source": [
    "# If it's an odd length, we need to drop one entry before splitting\n",
    "df = df[:-1]\n",
    "print(df.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Split our data in half. We train on 50% and test on the other 50%.\n",
    "# There's definitely better ways to split to make sure we get even\n",
    "# distribution of elements but this is good enough for now.\n",
    "\n",
    "# Labels\n",
    "Y_all = np.asarray(df[label])\n",
    "Y_train, Y_test = np.split(Y_all,2)\n",
    "\n",
    "# Features\n",
    "X_all = df[feats].as_matrix()\n",
    "X_train, X_test = np.split(X_all,2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Here's a look at what we end up with:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((3579,), (3579,))"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Y_train.shape,Y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((3579, 10), (3579, 10))"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape,X_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## K-Nearest Neighbours"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5.558411214953271, 32.0, 7.509236669180393)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Pick k based on average family size\n",
    "df.effective_family.mean(), df.effective_family.max(), statistics.stdev(df.effective_family[np.isfinite(df.effective_family)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(26, 0.94300083822296732)\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "\n",
    "scores = []\n",
    "\n",
    "for k in range(1,32):\n",
    "    model = KNeighborsClassifier(n_neighbors=k)\n",
    "    model.fit(X_train,Y_train)\n",
    "    Y_pred = model.predict(X_test)\n",
    "    scores.append((k,accuracy_score(Y_test,Y_pred, normalize=True)))\n",
    "    \n",
    "print(max(scores,key=lambda item:item[1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.93797150041911148"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "model = LogisticRegression()\n",
    "model.fit(X_train,Y_train)\n",
    "Y_pred = model.predict(X_test)\n",
    "accuracy_score(Y_test,Y_pred, normalize=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Naive Bayes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.90108968985750204"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.naive_bayes import BernoulliNB\n",
    "\n",
    "model = BernoulliNB()\n",
    "model.fit(X_train,Y_train)\n",
    "Y_pred = model.predict(X_test)\n",
    "accuracy_score(Y_test,Y_pred, normalize=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.72562168203408772"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.naive_bayes import GaussianNB\n",
    "\n",
    "model = GaussianNB()\n",
    "model.fit(X_train,Y_train)\n",
    "Y_pred = model.predict(X_test)\n",
    "accuracy_score(Y_test,Y_pred, normalize=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.93182453199217663"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "\n",
    "model = DecisionTreeClassifier()\n",
    "model.fit(X_train,Y_train)\n",
    "Y_pred = model.predict(X_test)\n",
    "accuracy_score(Y_test,Y_pred, normalize=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Random Forest\n",
    "\n",
    "This one allows us to get the features that had the biggest impact on decision making."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.95250069851913943"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "model = RandomForestClassifier()\n",
    "model.fit(X_train,Y_train)\n",
    "Y_pred = model.predict(X_test)\n",
    "accuracy_score(Y_test,Y_pred, normalize=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "imp = model.feature_importances_\n",
    "\n",
    "feat_wt = [(feats[i],imp[i]) for i in range(len(imp))]\n",
    "feat_wt = sorted(feat_wt, key=lambda x: x[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.55197606  0.03198492  0.00191007  0.03460119  0.04428406  0.00836103\n",
      "  0.224541    0.00952945  0.05602123  0.03679099]\n",
      "['advertised_bandwidth', 'fast', 'exit', 'age', 'stable', 'v2dir', 'consensus_weight', 'ports_accepted', 'uptime', 'address_port_uptime']\n"
     ]
    }
   ],
   "source": [
    "print(imp)\n",
    "print(feats)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYMAAAEMCAYAAAAmgtofAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nW/9lMrp4qwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x120984400>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "colors = ['#EB7BC0',\n",
    "          '#4AAD52',\n",
    "          '#FF8552',\n",
    "          '#990033',\n",
    "          '#4DCCBD',\n",
    "          '#FFE74C',\n",
    "          '#BC96E6',\n",
    "          '#ED217C',\n",
    "          '#7161EF',\n",
    "          '#D84727']\n",
    "plt.pie([i[1] for i in feat_wt],\n",
    "        labels=[i[0] for i in feat_wt],\n",
    "        shadow=False,\n",
    "        autopct='%1.2f%%',\n",
    "        startangle=200,\n",
    "        pctdistance=0.75,\n",
    "        labeldistance=1.1,\n",
    "        colors=colors[:len(imp)])\n",
    "plt.axis('equal')\n",
    "plt.title(\"Feature Importance\")\n",
    "plt.savefig('./images/feature_importance.png',dpi=1000)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3, 6, 0.95529477507683713)\n"
     ]
    }
   ],
   "source": [
    "scores = []\n",
    "for i in range(2,10):\n",
    "    for j in range(2,10):\n",
    "        estimator = DecisionTreeClassifier(max_depth=i,max_leaf_nodes=j)\n",
    "        estimator.fit(X_train, Y_train)\n",
    "        Y_pred = estimator.predict(X_test)\n",
    "        scores.append((i,j,accuracy_score(Y_test,Y_pred, normalize=True)))\n",
    "print(max(scores,key=lambda item:item[2]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The binary tree structure has 15 nodes and has the following tree structure:\n",
      "node=0 test node: go to node 1 if X[:, 0] <= 2095888.5s else to node 2.\n",
      "\tnode=1 test node: go to node 7 if X[:, 6] <= 6.12003641436e-05s else to node 8.\n",
      "\tnode=2 test node: go to node 3 if X[:, 4] <= 0.5s else to node 4.\n",
      "\t\tnode=3 leaf node.\n",
      "\t\tnode=4 test node: go to node 5 if X[:, 6] <= 4.13314992329e-05s else to node 6.\n",
      "\t\t\tnode=5 leaf node.\n",
      "\t\t\tnode=6 test node: go to node 9 if X[:, 3] <= 703800.0s else to node 10.\n",
      "\t\tnode=7 leaf node.\n",
      "\t\tnode=8 test node: go to node 13 if X[:, 8] <= 800265.0s else to node 14.\n",
      "\t\t\t\tnode=9 leaf node.\n",
      "\t\t\t\tnode=10 test node: go to node 11 if X[:, 8] <= 1700039.5s else to node 12.\n",
      "\t\t\t\t\tnode=11 leaf node.\n",
      "\t\t\t\t\tnode=12 leaf node.\n",
      "\t\t\tnode=13 leaf node.\n",
      "\t\t\tnode=14 leaf node.\n",
      "\n",
      "Rules used to predict sample 0: \n",
      "decision id node 7 : (X[0, -2] (= 28880900.0) > -2.0)\n",
      "\n",
      "The following samples [0, 1] share the node [0 1 7] in the tree\n",
      "It is 20.0 % of all nodes.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/cassiancorey/anaconda/lib/python3.5/site-packages/ipykernel/__main__.py:52: DeprecationWarning:\n",
      "\n",
      "using a non-integer number instead of an integer will result in an error in the future\n",
      "\n",
      "/Users/cassiancorey/anaconda/lib/python3.5/site-packages/ipykernel/__main__.py:48: DeprecationWarning:\n",
      "\n",
      "using a non-integer number instead of an integer will result in an error in the future\n",
      "\n"
     ]
    }
   ],
   "source": [
    "estimator = DecisionTreeClassifier(max_depth=4,max_leaf_nodes=8)\n",
    "estimator.fit(X_train, Y_train)\n",
    "\n",
    "# The decision estimator has an attribute called tree_  which stores the entire\n",
    "# tree structure and allows access to low level attributes. The binary tree\n",
    "# tree_ is represented as a number of parallel arrays. The i-th element of each\n",
    "# array holds information about the node `i`. Node 0 is the tree's root. NOTE:\n",
    "# Some of the arrays only apply to either leaves or split nodes, resp. In this\n",
    "# case the values of nodes of the other type are arbitrary!\n",
    "#\n",
    "# Among those arrays, we have:\n",
    "#   - left_child, id of the left child of the node\n",
    "#   - right_child, id of the right child of the node\n",
    "#   - feature, feature used for splitting the node\n",
    "#   - threshold, threshold value at the node\n",
    "#\n",
    "\n",
    "# Using those arrays, we can parse the tree structure:\n",
    "\n",
    "n_nodes = estimator.tree_.node_count\n",
    "children_left = estimator.tree_.children_left\n",
    "children_right = estimator.tree_.children_right\n",
    "feature = estimator.tree_.feature\n",
    "threshold = estimator.tree_.threshold\n",
    "\n",
    "\n",
    "# The tree structure can be traversed to compute various properties such\n",
    "# as the depth of each node and whether or not it is a leaf.\n",
    "node_depth = np.zeros(shape=n_nodes)\n",
    "is_leaves = np.zeros(shape=n_nodes, dtype=bool)\n",
    "stack = [(0, -1)]  # seed is the root node id and its parent depth\n",
    "while len(stack) > 0:\n",
    "    node_id, parent_depth = stack.pop()\n",
    "    node_depth[node_id] = parent_depth + 1\n",
    "\n",
    "    # If we have a test node\n",
    "    if (children_left[node_id] != children_right[node_id]):\n",
    "        stack.append((children_left[node_id], parent_depth + 1))\n",
    "        stack.append((children_right[node_id], parent_depth + 1))\n",
    "    else:\n",
    "        is_leaves[node_id] = True\n",
    "\n",
    "print(\"The binary tree structure has %s nodes and has \"\n",
    "      \"the following tree structure:\"\n",
    "      % n_nodes)\n",
    "for i in range(n_nodes):\n",
    "    if is_leaves[i]:\n",
    "        print(\"%snode=%s leaf node.\" % (node_depth[i] * \"\\t\", i))\n",
    "    else:\n",
    "        print(\"%snode=%s test node: go to node %s if X[:, %s] <= %ss else to \"\n",
    "              \"node %s.\"\n",
    "              % (node_depth[i] * \"\\t\",\n",
    "                 i,\n",
    "                 children_left[i],\n",
    "                 feature[i],\n",
    "                 threshold[i],\n",
    "                 children_right[i],\n",
    "                 ))\n",
    "print()\n",
    "\n",
    "# First let's retrieve the decision path of each sample. The decision_path\n",
    "# method allows to retrieve the node indicator functions. A non zero element of\n",
    "# indicator matrix at the position (i, j) indicates that the sample i goes\n",
    "# through the node j.\n",
    "\n",
    "node_indicator = estimator.decision_path(X_test)\n",
    "\n",
    "# Similarly, we can also have the leaves ids reached by each sample.\n",
    "\n",
    "leave_id = estimator.apply(X_test)\n",
    "\n",
    "# Now, it's possible to get the tests that were used to predict a sample or\n",
    "# a group of samples. First, let's make it for the sample.\n",
    "\n",
    "sample_id = 0\n",
    "node_index = node_indicator.indices[node_indicator.indptr[sample_id]:\n",
    "                                    node_indicator.indptr[sample_id + 1]]\n",
    "\n",
    "print('Rules used to predict sample %s: ' % sample_id)\n",
    "for node_id in node_index:\n",
    "    if leave_id[sample_id] != node_id:\n",
    "        continue\n",
    "\n",
    "    if (X_test[sample_id, feature[node_id]] <= threshold[node_id]):\n",
    "        threshold_sign = \"<=\"\n",
    "    else:\n",
    "        threshold_sign = \">\"\n",
    "\n",
    "    print(\"decision id node %s : (X[%s, %s] (= %s) %s %s)\"\n",
    "          % (node_id,\n",
    "             sample_id,\n",
    "             feature[node_id],\n",
    "             X_test[i, feature[node_id]],\n",
    "             threshold_sign,\n",
    "             threshold[node_id]))\n",
    "\n",
    "# For a group of samples, we have the following common node.\n",
    "sample_ids = [0, 1]\n",
    "common_nodes = (node_indicator.toarray()[sample_ids].sum(axis=0) ==\n",
    "                len(sample_ids))\n",
    "\n",
    "common_node_id = np.arange(n_nodes)[common_nodes]\n",
    "\n",
    "print(\"\\nThe following samples %s share the node %s in the tree\"\n",
    "      % (sample_ids, common_node_id))\n",
    "print(\"It is %s %% of all nodes.\" % (100 * len(common_node_id) / n_nodes,))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([12])"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "estimator.apply([X_train[1]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Y_train[1]"
   ]
  }
 ],
 "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"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}