Twitter_PageRankAnalysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "When it comes to consuming real-time content, Twitter is the place to be; Be it sending out 'Tweets' real-time, or discovering latest online 'Trends' anywhere, or ability to begin a 'conversation' with anyone, Twitter does it all. In fact, Twitter Management wrote this in their [letter to shareholders][1] last year. \n",
    "[1]: http://files.shareholder.com/downloads/AMDA-2F526X/106725176x0x874459/8A4D1A1D-D184-4AFE-9AC1-F880C5EA06F1/Q415_Shareholder_Letter.pdf\n",
    "\n",
    "> _We’re focused now on what Twitter does best: live. Twitter is live: live commentary, live connections, live conversations. Whether it’s breaking news, entertainment, sports, or everyday topics, hearing about and watching a live event unfold is the fastest way to understand the power of Twitter._ \n",
    "\n",
    "> _Twitter has always been considered a “second screen” for what’s happening in the world and we believe we can become the first screen for everything that’s happening now. And by doing so, we believe we can build the planet’s largest daily connected audience. A connected audience is one that watches together, and can talk with one another in real-time. It’s what Twitter has provided for close to 10 years, and it’s what we will continue to drive in the future_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Embedded in a Twitter User's Social-graph* is a wealth of information on User's likes and interests. Unlike Facebook or LinkedIn, the magic of Twitter is in its 'Follow' structure - where any can follow any without they knowing each other. This directed social-graph, when methodologically summarized, can reveal interesting information on the most influential/central friends in the network and also help personalize/enrich one's Twitter experience by unearthing implicit-clusters in the network.\n",
    "\n",
    "_*Social-graph: [User, 1st degree Connections, and the links between]_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this notebook, we'll look into these: \n",
    "1. Extract the ego-network of 'self' using Twitter API \n",
    "+ Identify the influential nodes using 'Page Rank' formulation \n",
    "+ Identify implicit clusters formed\n",
    "+ Recommend new friends to follow on the basis of cluster of interest "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note: This study is limited to  Ego network: the focal node (\"ego\": here the self-node) and the nodes to whom ego is directly connected to (\"alters\") plus the ties, if any, among the alters. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Extract social-graph strucutre using Twitter API "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "UsageError: Line magic function `%` not found.\n"
     ]
    }
   ],
   "source": [
    "from collections import Counter, defaultdict\n",
    "from datetime import datetime\n",
    "from sklearn.decomposition import PCA\n",
    "\n",
    "import csv\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import os.path\n",
    "import pandas as pd\n",
    "import re\n",
    "import seaborn as sns;  sns.set()\n",
    "import time\n",
    "import twitter\n",
    "\n",
    "% matplotlib inline\n",
    "plt.rcParams['figure.figsize'] = (10,7)\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "#print twitter.__path__  \n",
    "\n",
    "import random\n",
    "random.seed(1000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>body {\n",
       "    margin: 0;\n",
       "    font-family: Helvetica;\n",
       "}\n",
       "table.dataframe {\n",
       "    border-collapse: collapse;\n",
       "    border: none;\n",
       "}\n",
       "table.dataframe tr {\n",
       "    border: none;\n",
       "}\n",
       "table.dataframe td, table.dataframe th {\n",
       "    margin: 0;\n",
       "    border: 1px solid white;\n",
       "    padding-left: 0.25em;\n",
       "    padding-right: 0.25em;\n",
       "}\n",
       "table.dataframe th:not(:empty) {\n",
       "    background-color: #fec;\n",
       "    text-align: left;\n",
       "    font-weight: normal;\n",
       "}\n",
       "table.dataframe tr:nth-child(2) th:empty {\n",
       "    border-left: none;\n",
       "    border-right: 1px dashed #888;\n",
       "}\n",
       "table.dataframe td {\n",
       "    border: 2px solid #ccf;\n",
       "    background-color: #f4f4ff;\n",
       "}\n",
       "h3 {\n",
       "    color: white;\n",
       "    background-color: black;\n",
       "    padding: 0.5em;\n",
       "}\n",
       "</style>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Pandas data-frame print formatting\n",
    "from IPython.core.display import HTML\n",
    "css = open('style-table.css').read() + open('style-notebook.css').read()\n",
    "HTML('<style>{}</style>'.format(css))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "self_screen_name = 'RohanKuntoji'            # Self\n",
    "\n",
    "# Keep appending data\n",
    "fof_filename     = \"edges.csv\"          # 'Alters' and their Source->Sink Edges. \n",
    "cache_filename   = \"cache.csv\"          # local cache of (TwitterId, FullName, UserName)\n",
    "\n",
    "# One-time-use files\n",
    "binaryMap_filename = \"binaryMap.csv\"   # Directed Graph. Adjacencies as 0/1.RowFollowCol\n",
    "cluster_filename   = \"results.csv\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<twitter.api.Api at 0x1081eb8d0>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Twitter auth. https://dev.twitter.com/oauth/overview/application-owner-access-tokens\n",
    "\n",
    "# with open(\"credentials.txt\", \"r\") as f:\n",
    "#     reader = csv.reader(f)\n",
    "#     login_dict = {line[0]: line[1]                       \n",
    "#                     for line in reader}        \n",
    "\n",
    "api = twitter.Api(consumer_key='99mcVQZahMAqkGUbVDhQDmlMb' ,\n",
    "                  consumer_secret='ubyuWD52HKUcEmaeJxdidS1jjb5w0PC26LJaI7SFhu4TLc78tC',\n",
    "                  access_token_key='555800349-2zFuU4HDdr35k5aJPiCuWj3jO3oc50SDEEvOaZbD',\n",
    "                  access_token_secret='DpR0oqFEAyD3bIk8G6LWy9jNQNMQhNj5eSv2rs2EiVabx')\n",
    "api"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['/Users/kuntojirohan/MCG/Twitter/twitter_graph/py36/lib/python3.6/site-packages/twitter']\n"
     ]
    }
   ],
   "source": [
    "print(twitter.__path__)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "555800349\n",
      "['555800349', '376825877', '70831441', '2785337469', '1076526138736951298', '1004265866739867648', '775449094739197953', '3018841323', '216939636', '20167623', '46245868', '62513246', '33838201', '991172011383803905', '80422885', '863720408', '2238652956', '534563976', '288100306', '8515852', '63873759', '254107028', '96951800', '627673190', '177547780', '15639696', '78941611', '29873662', '822215679726100480', '20536157', '57947109', '722214536460173312', '393033324', '303680301', '185142711', '1282946089', '15631238', '3138727082', '775302652427014145', '108252113', '19407053', '183230911', '54829997', '71201743', '8161232', '1636590253', '13298072', '44196397', '944427851280826368', '471745066', '872434564580294657', '1060849038', '937440997', '3007477369', '97865628', '18681139', '51376979', '101695592', '133880286', '101311381', '31678307', '14213195', '92724677', '135421739', '92708272', '18839785', '48410093', '155659213', '113419517', '145125358', '50393960', '88856792']\n"
     ]
    }
   ],
   "source": [
    "# 'Self' and Friends of Self\n",
    "\n",
    "self_node    =  api.GetUser(screen_name = self_screen_name)    \n",
    "self_node_id =  str(self_node.id)                      # Twitter Id of Self\n",
    "print(self_node_id)\n",
    "\n",
    "friends_of_self = api.GetFriendIDs(user_id = self_node_id, \n",
    "                                   screen_name = self_screen_name , \n",
    "                                   stringify_ids = True)\n",
    "index = [self_node_id] + friends_of_self\n",
    "print(index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# GetFriendIDs() API call is rate-limited at 15 req / 15 min \n",
    "# https://dev.twitter.com/rest/public/rate-limiting\n",
    "\n",
    "# For each of the list of nodes, fetch the list of nodes it follows, append to file\n",
    "\n",
    "def update_FoF_File(fileName, to_fetch_list):\n",
    "    \n",
    "    with open(fileName, 'a') as f:\n",
    "        apiReqCount = 0\n",
    "        \n",
    "        for node in to_fetch_list:\n",
    "            friends_of_node = api.GetFriendIDs(user_id = node,  stringify_ids = True) \n",
    "            row = ','.join([str(i) for i in [node] +  friends_of_node ]) + \"\\n\"\n",
    "            f.write(row) \n",
    "            \n",
    "            apiReqCount += 1\n",
    "            \n",
    "            if (apiReqCount == 15): \n",
    "                apiReqCount = 0\n",
    "                \n",
    "                print(\"Off to Sleep :)\")\n",
    "                time.sleep(15*60 + 10)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# parse FoF file and return list of nodes, for whom source->sink Edges are already there.\n",
    "\n",
    "def getFinishList(fileName):\n",
    "        \n",
    "    if not os.path.isfile(fileName):\n",
    "        return [] \n",
    "    with open(fileName, 'r') as f:\n",
    "        return [ row.strip().split(',')[0] for row in f ]  # 1st entry is a user"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Ego-network as adjacency-matrix\n",
    "# Parses FoF file in order of index, create list of adjacencies as 0 | 1\n",
    "# Writes to File. Adjacency-matrix in Row_follows_Column format\n",
    "\n",
    "def updateBinaryMapFile(fof_filename, binaryMap_filename, index):\n",
    "    \n",
    "    with open(fof_filename, \"r\") as f:\n",
    "        stripped_f = (line.replace('\\r', '') for line in f)\n",
    "        reader = csv.reader(stripped_f)\n",
    "        fof_dict = {line[0]: line[1:] for line in reader \n",
    "                    if line[0] in index}   # dict of node:his_followers\n",
    "        if self_node_id not in fof_dict:\n",
    "            fof_dict[self_node_id] = index[1:]             # dict of Self\n",
    "    \n",
    "    bool_list = []\n",
    "    \n",
    "    for user in index:\n",
    "        user_friends = set( fof_dict[user] )  \n",
    "        bool_row = [item in user_friends for item in index]  # for each, fill T/F \n",
    "        bool_list.append(bool_row)\n",
    "    \n",
    "    int_nparray = np.array(bool_list) + 0                    # Bool to int\n",
    "\n",
    "    binaryMap_rfc = pd.DataFrame(data = int_nparray, columns= index, index = index)\n",
    "    binaryMap_rfc.to_csv(binaryMap_filename)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# For list of Ids, fetch Profile details. If not in Offline file, make an API call\n",
    "# Returns ['UserName', 'FullName', 'Followers_count', 'Friends_count', 'Location', 'Created_at']\n",
    "# UsersLookup API 100 Ids/request\n",
    "\n",
    "def lookup_in_cache(friendsIdsList):\n",
    "\n",
    "    cache, delta_cache = pd.DataFrame(), pd.DataFrame()\n",
    "    UserNameList, namesList     = [], []\n",
    "    followers_count, friends_count = [], []\n",
    "    location, created_at           = [], []\n",
    "    \n",
    "    if os.path.isfile(cache_filename):\n",
    "        cache = pd.read_csv(cache_filename, skip_blank_lines=True,\n",
    "                        dtype={'Ids':str, 'Friends_count':int, 'Followers_count':int})\n",
    "        cache.set_index('Ids', inplace=True)\n",
    "        to_fetch_list = list ( set (friendsIdsList) - set(cache.index) )\n",
    "    else :        \n",
    "        to_fetch_list = friendsIdsList   \n",
    "    \n",
    "    i = 0\n",
    "    while (i < len(to_fetch_list) * 1./100):   \n",
    "        print(\"... Cache-Miss for \" + str(len(to_fetch_list)) +  \" nodes. Updating Cache...\")\n",
    "        low, high = i * 100, min( len(to_fetch_list), (i+1)*100 )  # UsersLookup api\n",
    "        twitterObjectsList = api.UsersLookup(user_id = to_fetch_list[low:high])    \n",
    "        \n",
    "        temp = zip(*[( tempObject.screen_name,        #ScreenName\n",
    "                      tempObject.name,                #Name\n",
    "                      tempObject.followers_count,     #Followers\n",
    "                      tempObject.friends_count,       #Friends\n",
    "                      tempObject.location,            #Location\n",
    "                      tempObject.created_at           #CreatedAt\n",
    "                     ) for tempObject in twitterObjectsList])\n",
    "        temp = list(temp)\n",
    "        \n",
    "        UserNameList    += list(temp[0])\n",
    "        namesList       += list(temp[1])\n",
    "        followers_count += list(temp[2])\n",
    "        friends_count   += list(temp[3])\n",
    "        location        += list(temp[4])\n",
    "        created_at      += list(temp[5])\n",
    "        \n",
    "        i = i + 1       \n",
    "    \n",
    "    if len(to_fetch_list) > 0:\n",
    "        delta_cache  = pd.DataFrame({'UserName':UserNameList,\n",
    "                                     'FullName':namesList,\n",
    "                                     'Ids': to_fetch_list,\n",
    "                                     'Followers':followers_count,\n",
    "                                     'Friends': friends_count,\n",
    "                                     'Location':location,\n",
    "                                     'Created':created_at})\n",
    "        \n",
    "        delta_cache['Created'] = delta_cache['Created'].apply(lambda x: \n",
    "                                                datetime.strptime(\n",
    "                                                    re.sub(r\"\\+[0-9]* \", \"\",x),'%c').\n",
    "                                                strftime(\"%b-%Y\"))\n",
    "    \n",
    "        delta_cache.set_index('Ids', inplace=True, drop = True)\n",
    "        cache = cache.append(delta_cache)\n",
    "        cache.to_csv(cache_filename)        \n",
    "\n",
    "    return cache.loc[friendsIdsList]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Display cluster-wise most-influential users, for the given clustering algo\n",
    "\n",
    "def top_nodes_in_cluster(df, cluster_algo, n_clusters):\n",
    "    dummy_df = pd.DataFrame()\n",
    "\n",
    "    for i in range(n_clusters):\n",
    "        nodes_in_cluster = list( df [df[cluster_algo] == i ]['FullName'] )     \n",
    "        if len(nodes_in_cluster) >= 10:            # show only clusters of size > 10        \n",
    "            col_name           = str(i) + \" : \" + str(len(nodes_in_cluster)) + \" Ids\"\n",
    "            dummy_df[col_name] = nodes_in_cluster[:10]      \n",
    "    \n",
    "    return dummy_df   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# identify 20 friends to follow after aggregating friends followed by top 50% in list\n",
    "\n",
    "def discover_Friends_toFollow(ids_of_interest, friend_list, prop = .5, count = 20):    \n",
    "    \n",
    "    ids_of_interest  = ids_of_interest[:int(len(ids_of_interest) * prop)]\n",
    "    \n",
    "    if self_node_id in ids_of_interest:\n",
    "        ids_of_interest.remove(self_node_id)\n",
    "\n",
    "    print(\"'Who-To-Follow' reco after looking at %3d friends' friends:\" %(len(ids_of_interest)))\n",
    "    \n",
    "    with open(fof_filename) as f:\n",
    "        reader = csv.reader(f)\n",
    "        fof_dict = {row[0]:row[0:] for row in reader}  # dict of node:her_followers\n",
    "\n",
    "    friendsToFollow = []\n",
    "    for id in ids_of_interest:\n",
    "        friendsToFollow += list (set(fof_dict[str(id)])  - set(friend_list) ) \n",
    "        \n",
    "\n",
    "    friendsToFollow = Counter(friendsToFollow).most_common(count)    \n",
    "    tuples_list = list(zip(*friendsToFollow) )\n",
    "           \n",
    "    topFriendsToFollowDF = pd.DataFrame()\n",
    "    topFriendsToFollowDF['Ids'] = list(tuples_list[0])\n",
    "    topFriendsToFollowDF['Freq'] = list(tuples_list[1])\n",
    "    topFriendsToFollowDF.set_index('Ids', drop = True, inplace = True)              \n",
    "    index = topFriendsToFollowDF.index\n",
    "    \n",
    "    topFriendsToFollowDF = topFriendsToFollowDF.merge(lookup_in_cache(index), copy = False,\n",
    "                              left_index = True, right_index = True)\n",
    "    \n",
    "    return topFriendsToFollowDF    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 out of 71 Friends details to be fetched\n"
     ]
    }
   ],
   "source": [
    "# For the list of nodes I follow, fetch their friends-list\n",
    "\n",
    "fof_finish_list   = getFinishList(fof_filename )        # Completed nodes\n",
    "fof_to_fetch_list = list ( set(friends_of_self) - set(fof_finish_list) )  # Pending nodes\n",
    "\n",
    "print( str(len(fof_to_fetch_list)) + \" out of \" + str(len(index)  - 1) + \n",
    "      \" Friends details to be fetched\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# For the remaining nodes, populate their details in fof_file\n",
    "update_FoF_File(fof_filename, fof_to_fetch_list) \n",
    "\n",
    "# Build the adjacency matrix in terms of 0 and 1 (if there is an edge)\n",
    "updateBinaryMapFile(fof_filename, binaryMap_filename, index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(72, 72)\n"
     ]
    }
   ],
   "source": [
    "# Read adj-matrix into df. Cell MxN is 1 iff node in Mth row follows node in Nth column \n",
    "binaryMap_rfc = pd.read_csv(binaryMap_filename, skip_blank_lines=True, index_col = 0)\n",
    "print(binaryMap_rfc.shape)\n",
    "\n",
    "\n",
    "outlinks_count = binaryMap_rfc.sum(axis = 1)   # horizontal-sum to count outlinks\n",
    "inlinks_count = binaryMap_rfc.sum(axis = 0)   # vertical-sum to count inlinks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>555800349</th>\n",
       "      <th>376825877</th>\n",
       "      <th>70831441</th>\n",
       "      <th>2785337469</th>\n",
       "      <th>1076526138736951298</th>\n",
       "      <th>1004265866739867648</th>\n",
       "      <th>775449094739197953</th>\n",
       "      <th>3018841323</th>\n",
       "      <th>216939636</th>\n",
       "      <th>20167623</th>\n",
       "      <th>...</th>\n",
       "      <th>92724677</th>\n",
       "      <th>135421739</th>\n",
       "      <th>92708272</th>\n",
       "      <th>18839785</th>\n",
       "      <th>48410093</th>\n",
       "      <th>155659213</th>\n",
       "      <th>113419517</th>\n",
       "      <th>145125358</th>\n",
       "      <th>50393960</th>\n",
       "      <th>88856792</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>555800349</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>376825877</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>70831441</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2785337469</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1076526138736951298</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>155659213</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>113419517</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>145125358</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>50393960</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>88856792</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>72 rows × 72 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                     555800349  376825877  70831441  2785337469  \\\n",
       "555800349                    0          1         1           1   \n",
       "376825877                    0          0         0           0   \n",
       "70831441                     0          0         0           0   \n",
       "2785337469                   0          0         1           0   \n",
       "1076526138736951298          0          0         1           1   \n",
       "...                        ...        ...       ...         ...   \n",
       "155659213                    0          0         0           0   \n",
       "113419517                    0          0         0           0   \n",
       "145125358                    0          0         0           0   \n",
       "50393960                     0          0         0           0   \n",
       "88856792                     0          0         0           0   \n",
       "\n",
       "                     1076526138736951298  1004265866739867648  \\\n",
       "555800349                              1                    1   \n",
       "376825877                              0                    0   \n",
       "70831441                               0                    0   \n",
       "2785337469                             1                    0   \n",
       "1076526138736951298                    0                    0   \n",
       "...                                  ...                  ...   \n",
       "155659213                              0                    0   \n",
       "113419517                              0                    0   \n",
       "145125358                              0                    0   \n",
       "50393960                               0                    0   \n",
       "88856792                               0                    0   \n",
       "\n",
       "                     775449094739197953  3018841323  216939636  20167623  ...  \\\n",
       "555800349                             1           1          1         1  ...   \n",
       "376825877                             0           0          0         0  ...   \n",
       "70831441                              1           0          0         0  ...   \n",
       "2785337469                            1           0          1         1  ...   \n",
       "1076526138736951298                   1           0          1         0  ...   \n",
       "...                                 ...         ...        ...       ...  ...   \n",
       "155659213                             0           0          0         0  ...   \n",
       "113419517                             0           0          0         0  ...   \n",
       "145125358                             0           0          0         0  ...   \n",
       "50393960                              0           0          0         0  ...   \n",
       "88856792                              0           0          0         0  ...   \n",
       "\n",
       "                     92724677  135421739  92708272  18839785  48410093  \\\n",
       "555800349                   1          1         1         1         1   \n",
       "376825877                   0          0         0         0         0   \n",
       "70831441                    0          0         0         0         0   \n",
       "2785337469                  0          0         0         0         0   \n",
       "1076526138736951298         0          0         0         0         0   \n",
       "...                       ...        ...       ...       ...       ...   \n",
       "155659213                   0          0         0         0         0   \n",
       "113419517                   0          0         0         1         0   \n",
       "145125358                   1          1         1         1         0   \n",
       "50393960                    0          0         0         1         0   \n",
       "88856792                    0          1         0         0         0   \n",
       "\n",
       "                     155659213  113419517  145125358  50393960  88856792  \n",
       "555800349                    1          1          1         1         1  \n",
       "376825877                    0          0          0         0         0  \n",
       "70831441                     0          0          0         0         0  \n",
       "2785337469                   0          0          0         1         0  \n",
       "1076526138736951298          0          0          0         0         0  \n",
       "...                        ...        ...        ...       ...       ...  \n",
       "155659213                    0          0          0         0         0  \n",
       "113419517                    0          0          0         0         0  \n",
       "145125358                    1          1          0         1         1  \n",
       "50393960                     0          0          1         0         0  \n",
       "88856792                     0          1          1         0         0  \n",
       "\n",
       "[72 rows x 72 columns]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "binaryMap_rfc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Histogram of number of OutLinks per node, within ego-network\n",
    "sns.distplot(outlinks_count, bins = len(index)//5, kde=False);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Histogram of number of InLinks per node, within ego-network\n",
    "sns.distplot(inlinks_count, bins = len(index)//5, kde=False);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Identify influential friends using 'Page Rank' formulation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From the adjacency matrix generated above, we can construct a column-stochastic matrix  (also called a transition matrix) such that, a column with `m` outlinks will have `1/m` as value in respective `m` cells. Conceptually, a value in the cell `a x b` gives the probability of a random-surfer in node B jumping to node A. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "binaryMap_cfr      = binaryMap_rfc.transpose()            # column-values: Outlinks\n",
    "binaryMap_cfr_norm = binaryMap_cfr / binaryMap_cfr.sum(axis = 0)\n",
    "colStochMatrix = np.matrix( binaryMap_cfr_norm.fillna(0)) # column-stochastic-matrix"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "_Initialize PageRank vector, such that all the nodes have equal PageRank score adding upto 1._"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "pageRankVector = np.matrix([1.0/len(index)] *  len(index)) # iniitialize page-rank-vector \n",
    "pageRankVector = pageRankVector.transpose()                # transpose to column-vector"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "On applying the above Transition-matrix transformation iteratively on the PageRank vector, the vector will eventully converge such that:\n",
    "\n",
    "`Matrix.Vector = Vector`\n",
    "\n",
    "Equivalently, this is Eigen-Vector formulation with PageRank vector being the principal eigen-vector of matrix corresponding to eigen-value `1` [Since M is column-stochastic matrix, principal eigen-value is 1]\n",
    "\n",
    "Here we use Power Iteration method to solve for the PageRank vector. Inorder to handle  the nodes which have zero out-links (dead-ends) and nodes with periodic-loops (spider-traps) in the ego-network, we introduce some randomness through parameter `beta` such that a random-surfer at any node picks a random path approx. every 1 out of 6 times ( 1 - beta = 0.15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sum of Page-Rank Scores: 1.0\n",
      "Converged in 131 iterations\n"
     ]
    }
   ],
   "source": [
    "# PageRank algo: Power Iteration to solve Markov transition matrix \n",
    "# refer this     : http://setosa.io/blog/2014/07/26/markov-chains/index.html\n",
    "\n",
    "beta = 0.85\n",
    "epsilon = 999\n",
    "iteration = 0\n",
    "while epsilon > (1.0/(10**16)):\n",
    "    pageRankVectorUpdating = colStochMatrix * pageRankVector * beta \n",
    "    \n",
    "    # re-insert leaked page-ranks\n",
    "    S = np.array(pageRankVectorUpdating).sum()                      \n",
    "    pageRankVectorUpdated = pageRankVectorUpdating + (\n",
    "        1 - S) * (1.0/len(index)) * np.ones_like(len(index))\n",
    "    \n",
    "    # compute the squared-difference and check for convergence\n",
    "    error = np.array(pageRankVectorUpdated - pageRankVector)\n",
    "    epsilon = np.sqrt((error *  error).sum())   \n",
    "    iteration = iteration + 1\n",
    "    pageRankVector = pageRankVectorUpdated    \n",
    "    \n",
    "print( \"Sum of Page-Rank Scores: \" + str(pageRankVector.sum()) + \n",
    "      \"\\nConverged in \" + str(iteration) + \" iterations\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "... Cache-Miss for 72 nodes. Updating Cache...\n"
     ]
    }
   ],
   "source": [
    "# Collect the results\n",
    "\n",
    "results_df = pd.DataFrame()\n",
    "results_df['Ids'], results_df['PageRank'] = index, pageRankVector\n",
    "results_df['Inlinks'], results_df['Outlinks'] = list(inlinks_count), list(outlinks_count)\n",
    "results_df = results_df.set_index('Ids', drop = True )\n",
    "\n",
    "results_df = results_df.merge(lookup_in_cache(index), copy = False,\n",
    "                              left_index = True, right_index = True)\n",
    "results_df = results_df[['PageRank','UserName', 'FullName', 'Inlinks' , 'Outlinks',\n",
    "                         'Followers','Friends', 'Location', 'Created' ]] "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PageRank</th>\n",
       "      <th>UserName</th>\n",
       "      <th>FullName</th>\n",
       "      <th>Inlinks</th>\n",
       "      <th>Outlinks</th>\n",
       "      <th>Followers</th>\n",
       "      <th>Friends</th>\n",
       "      <th>Location</th>\n",
       "      <th>Created</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ids</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>555800349</td>\n",
       "      <td>0.004957</td>\n",
       "      <td>RohanKuntoji</td>\n",
       "      <td>Rohan Kuntoji</td>\n",
       "      <td>7</td>\n",
       "      <td>71</td>\n",
       "      <td>34</td>\n",
       "      <td>71</td>\n",
       "      <td>Mysore, India</td>\n",
       "      <td>Apr-2012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>376825877</td>\n",
       "      <td>0.002929</td>\n",
       "      <td>TwitterOSS</td>\n",
       "      <td>Twitter Open Source</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>322729</td>\n",
       "      <td>50</td>\n",
       "      <td>Twitter HQ</td>\n",
       "      <td>Sep-2011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>70831441</td>\n",
       "      <td>0.008212</td>\n",
       "      <td>soumithchintala</td>\n",
       "      <td>Soumith Chintala</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>70874</td>\n",
       "      <td>318</td>\n",
       "      <td>New York City</td>\n",
       "      <td>Sep-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2785337469</td>\n",
       "      <td>0.004203</td>\n",
       "      <td>seb_ruder</td>\n",
       "      <td>Sebastian Ruder</td>\n",
       "      <td>3</td>\n",
       "      <td>11</td>\n",
       "      <td>34573</td>\n",
       "      <td>751</td>\n",
       "      <td>London, England</td>\n",
       "      <td>Sep-2014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1076526138736951298</td>\n",
       "      <td>0.004427</td>\n",
       "      <td>paperswithcode</td>\n",
       "      <td>Papers with Code</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>14671</td>\n",
       "      <td>44</td>\n",
       "      <td>London, UK</td>\n",
       "      <td>Dec-2018</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>155659213</td>\n",
       "      <td>0.013575</td>\n",
       "      <td>Cristiano</td>\n",
       "      <td>Cristiano Ronaldo</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>80217624</td>\n",
       "      <td>51</td>\n",
       "      <td>Turim, Piemonte</td>\n",
       "      <td>Jun-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>113419517</td>\n",
       "      <td>0.021098</td>\n",
       "      <td>iHrithik</td>\n",
       "      <td>Hrithik Roshan</td>\n",
       "      <td>19</td>\n",
       "      <td>5</td>\n",
       "      <td>26437644</td>\n",
       "      <td>90</td>\n",
       "      <td></td>\n",
       "      <td>Feb-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>145125358</td>\n",
       "      <td>0.040777</td>\n",
       "      <td>SrBachchan</td>\n",
       "      <td>Amitabh Bachchan</td>\n",
       "      <td>23</td>\n",
       "      <td>32</td>\n",
       "      <td>38663364</td>\n",
       "      <td>1836</td>\n",
       "      <td>Mumbai, India</td>\n",
       "      <td>May-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>50393960</td>\n",
       "      <td>0.026798</td>\n",
       "      <td>BillGates</td>\n",
       "      <td>Bill Gates</td>\n",
       "      <td>20</td>\n",
       "      <td>4</td>\n",
       "      <td>48153147</td>\n",
       "      <td>202</td>\n",
       "      <td>Seattle, WA</td>\n",
       "      <td>Jun-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>88856792</td>\n",
       "      <td>0.016891</td>\n",
       "      <td>aamir_khan</td>\n",
       "      <td>Aamir Khan</td>\n",
       "      <td>16</td>\n",
       "      <td>3</td>\n",
       "      <td>25156209</td>\n",
       "      <td>9</td>\n",
       "      <td>Mumbai</td>\n",
       "      <td>Nov-2009</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>72 rows × 9 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                     PageRank         UserName             FullName  Inlinks  \\\n",
       "Ids                                                                            \n",
       "555800349            0.004957     RohanKuntoji        Rohan Kuntoji        7   \n",
       "376825877            0.002929       TwitterOSS  Twitter Open Source        1   \n",
       "70831441             0.008212  soumithchintala     Soumith Chintala        6   \n",
       "2785337469           0.004203        seb_ruder      Sebastian Ruder        3   \n",
       "1076526138736951298  0.004427   paperswithcode     Papers with Code        3   \n",
       "...                       ...              ...                  ...      ...   \n",
       "155659213            0.013575        Cristiano    Cristiano Ronaldo        8   \n",
       "113419517            0.021098         iHrithik       Hrithik Roshan       19   \n",
       "145125358            0.040777       SrBachchan     Amitabh Bachchan       23   \n",
       "50393960             0.026798        BillGates           Bill Gates       20   \n",
       "88856792             0.016891       aamir_khan           Aamir Khan       16   \n",
       "\n",
       "                     Outlinks  Followers  Friends         Location   Created  \n",
       "Ids                                                                           \n",
       "555800349                  71         34       71    Mysore, India  Apr-2012  \n",
       "376825877                   0     322729       50       Twitter HQ  Sep-2011  \n",
       "70831441                    2      70874      318    New York City  Sep-2009  \n",
       "2785337469                 11      34573      751  London, England  Sep-2014  \n",
       "1076526138736951298         5      14671       44       London, UK  Dec-2018  \n",
       "...                       ...        ...      ...              ...       ...  \n",
       "155659213                   0   80217624       51  Turim, Piemonte  Jun-2010  \n",
       "113419517                   5   26437644       90                   Feb-2010  \n",
       "145125358                  32   38663364     1836    Mumbai, India  May-2010  \n",
       "50393960                    4   48153147      202      Seattle, WA  Jun-2009  \n",
       "88856792                    3   25156209        9           Mumbai  Nov-2009  \n",
       "\n",
       "[72 rows x 9 columns]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 2a. Nodes with high PageRank scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PageRank</th>\n",
       "      <th>UserName</th>\n",
       "      <th>Inlinks</th>\n",
       "      <th>Outlinks</th>\n",
       "      <th>Followers</th>\n",
       "      <th>Friends</th>\n",
       "      <th>Location</th>\n",
       "      <th>Created</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FullName</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>TensorFlow</td>\n",
       "      <td>0.055500</td>\n",
       "      <td>TensorFlow</td>\n",
       "      <td>11</td>\n",
       "      <td>3</td>\n",
       "      <td>172811</td>\n",
       "      <td>96</td>\n",
       "      <td>Mountain View, CA</td>\n",
       "      <td>Feb-2011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Google</td>\n",
       "      <td>0.051477</td>\n",
       "      <td>Google</td>\n",
       "      <td>12</td>\n",
       "      <td>2</td>\n",
       "      <td>21517665</td>\n",
       "      <td>214</td>\n",
       "      <td>Mountain View, CA</td>\n",
       "      <td>Feb-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Google AI</td>\n",
       "      <td>0.047022</td>\n",
       "      <td>GoogleAI</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>1380565</td>\n",
       "      <td>14</td>\n",
       "      <td>Mountain View, CA</td>\n",
       "      <td>Apr-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Amitabh Bachchan</td>\n",
       "      <td>0.040777</td>\n",
       "      <td>SrBachchan</td>\n",
       "      <td>23</td>\n",
       "      <td>32</td>\n",
       "      <td>38663364</td>\n",
       "      <td>1836</td>\n",
       "      <td>Mumbai, India</td>\n",
       "      <td>May-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Tesla</td>\n",
       "      <td>0.037226</td>\n",
       "      <td>Tesla</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>4259108</td>\n",
       "      <td>76</td>\n",
       "      <td></td>\n",
       "      <td>Feb-2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Sachin Tendulkar</td>\n",
       "      <td>0.036718</td>\n",
       "      <td>sachin_rt</td>\n",
       "      <td>26</td>\n",
       "      <td>10</td>\n",
       "      <td>30551511</td>\n",
       "      <td>79</td>\n",
       "      <td>ÜT: 18.986431,72.823769</td>\n",
       "      <td>Apr-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Elon Musk</td>\n",
       "      <td>0.035077</td>\n",
       "      <td>elonmusk</td>\n",
       "      <td>15</td>\n",
       "      <td>1</td>\n",
       "      <td>28473165</td>\n",
       "      <td>81</td>\n",
       "      <td></td>\n",
       "      <td>Jun-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Narendra Modi</td>\n",
       "      <td>0.032643</td>\n",
       "      <td>narendramodi</td>\n",
       "      <td>22</td>\n",
       "      <td>9</td>\n",
       "      <td>50489276</td>\n",
       "      <td>2266</td>\n",
       "      <td>India</td>\n",
       "      <td>Jan-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Virat Kohli</td>\n",
       "      <td>0.027765</td>\n",
       "      <td>imVkohli</td>\n",
       "      <td>21</td>\n",
       "      <td>8</td>\n",
       "      <td>31654577</td>\n",
       "      <td>53</td>\n",
       "      <td></td>\n",
       "      <td>Sep-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Bill Gates</td>\n",
       "      <td>0.026798</td>\n",
       "      <td>BillGates</td>\n",
       "      <td>20</td>\n",
       "      <td>4</td>\n",
       "      <td>48153147</td>\n",
       "      <td>202</td>\n",
       "      <td>Seattle, WA</td>\n",
       "      <td>Jun-2009</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  PageRank      UserName  Inlinks  Outlinks  Followers  \\\n",
       "FullName                                                                 \n",
       "TensorFlow        0.055500    TensorFlow       11         3     172811   \n",
       "Google            0.051477        Google       12         2   21517665   \n",
       "Google AI         0.047022      GoogleAI        8         2    1380565   \n",
       "Amitabh Bachchan  0.040777    SrBachchan       23        32   38663364   \n",
       "Tesla             0.037226         Tesla        5         2    4259108   \n",
       "Sachin Tendulkar  0.036718     sachin_rt       26        10   30551511   \n",
       "Elon Musk         0.035077      elonmusk       15         1   28473165   \n",
       "Narendra Modi     0.032643  narendramodi       22         9   50489276   \n",
       "Virat Kohli       0.027765      imVkohli       21         8   31654577   \n",
       "Bill Gates        0.026798     BillGates       20         4   48153147   \n",
       "\n",
       "                  Friends                 Location   Created  \n",
       "FullName                                                      \n",
       "TensorFlow             96        Mountain View, CA  Feb-2011  \n",
       "Google                214        Mountain View, CA  Feb-2009  \n",
       "Google AI              14        Mountain View, CA  Apr-2009  \n",
       "Amitabh Bachchan     1836            Mumbai, India  May-2010  \n",
       "Tesla                  76                           Feb-2008  \n",
       "Sachin Tendulkar       79  ÜT: 18.986431,72.823769  Apr-2010  \n",
       "Elon Musk              81                           Jun-2009  \n",
       "Narendra Modi        2266                    India  Jan-2009  \n",
       "Virat Kohli            53                           Sep-2009  \n",
       "Bill Gates            202              Seattle, WA  Jun-2009  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df.fillna('').sort_values(by = 'PageRank', ascending =False).set_index('FullName').head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['Rohan Kuntoji', 'Twitter Open Source', 'Soumith Chintala',\n",
       "       'Sebastian Ruder', 'Papers with Code', 'Mysuru Consulting Group',\n",
       "       'Ian Goodfellow', 'Open Data Science', 'Andrew Ng', 'KDnuggets',\n",
       "       'David Robinson', 'J.K. Rowling', 'Google AI', 'Sohan Kuntoji',\n",
       "       'Kaggle', 'Sunil Chhetri', 'Indian Football Team', 'Kirk Borne',\n",
       "       'Chinmay', 'Adam Geitgey', 'Python Software Foundation',\n",
       "       'TensorFlow', 'FC Barcelona', 'UEFA Champions League', 'ICC',\n",
       "       'IndianPremierLeague', 'Harsha Bhogle', 'Marques Brownlee',\n",
       "       'President Trump', 'Google', 'Flipkart', 'IBM Data Science & AI',\n",
       "       'Data Science Central', 'Ninad M G', 'BCCI', 'Amazon.in', 'Virgin',\n",
       "       'OnePlus India', 'Rakshak S | ರಕ್ಷಕ್ ಎಸ್', 'Vir Das',\n",
       "       'Entrepreneur', 'Alia Bhatt', 'Anushka Sharma', 'Virat Kohli',\n",
       "       'Richard Branson', 'Tim Cook', 'Tesla', 'Elon Musk',\n",
       "       'Soumak Chongder', 'Nikhil  P Huilgol', 'Ashika', 'Sandesh Hegde',\n",
       "       'Suhas Raju', 'Bindu smitha', 'Farhan Akhtar', 'PRIYANKA',\n",
       "       'Sonam K Ahuja', 'Deepika Padukone', 'Leonardo DiCaprio',\n",
       "       'Shah Rukh Khan', 'MSN Entertainment', 'MSN India',\n",
       "       'Virender Sehwag', 'Sachin Tendulkar', 'Mahendra Singh Dhoni',\n",
       "       'Narendra Modi', 'Tom Cruise', 'Cristiano Ronaldo',\n",
       "       'Hrithik Roshan', 'Amitabh Bachchan', 'Bill Gates', 'Aamir Khan'],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df.FullName.unique()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 2b. Top 50 influential-nodes in Ego-Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>TensorFlow</td>\n",
       "      <td>Andrew Ng</td>\n",
       "      <td>Aamir Khan</td>\n",
       "      <td>Sonam K Ahuja</td>\n",
       "      <td>Soumith Chintala</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>Google</td>\n",
       "      <td>Shah Rukh Khan</td>\n",
       "      <td>Farhan Akhtar</td>\n",
       "      <td>Tom Cruise</td>\n",
       "      <td>Kirk Borne</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>Google AI</td>\n",
       "      <td>Richard Branson</td>\n",
       "      <td>Tim Cook</td>\n",
       "      <td>UEFA Champions League</td>\n",
       "      <td>Data Science Central</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>Amitabh Bachchan</td>\n",
       "      <td>Mahendra Singh Dhoni</td>\n",
       "      <td>Ian Goodfellow</td>\n",
       "      <td>IndianPremierLeague</td>\n",
       "      <td>Open Data Science</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>Tesla</td>\n",
       "      <td>Marques Brownlee</td>\n",
       "      <td>Anushka Sharma</td>\n",
       "      <td>KDnuggets</td>\n",
       "      <td>Virgin</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>Sachin Tendulkar</td>\n",
       "      <td>Hrithik Roshan</td>\n",
       "      <td>Deepika Padukone</td>\n",
       "      <td>President Trump</td>\n",
       "      <td>Entrepreneur</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>Elon Musk</td>\n",
       "      <td>Leonardo DiCaprio</td>\n",
       "      <td>Cristiano Ronaldo</td>\n",
       "      <td>Alia Bhatt</td>\n",
       "      <td>Vir Das</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>Narendra Modi</td>\n",
       "      <td>Virender Sehwag</td>\n",
       "      <td>J.K. Rowling</td>\n",
       "      <td>Indian Football Team</td>\n",
       "      <td>ICC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>Virat Kohli</td>\n",
       "      <td>PRIYANKA</td>\n",
       "      <td>Kaggle</td>\n",
       "      <td>FC Barcelona</td>\n",
       "      <td>Rohan Kuntoji</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>Bill Gates</td>\n",
       "      <td>BCCI</td>\n",
       "      <td>Harsha Bhogle</td>\n",
       "      <td>Sunil Chhetri</td>\n",
       "      <td>Adam Geitgey</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  0                     1                  2  \\\n",
       "0        TensorFlow             Andrew Ng         Aamir Khan   \n",
       "1            Google        Shah Rukh Khan      Farhan Akhtar   \n",
       "2         Google AI       Richard Branson           Tim Cook   \n",
       "3  Amitabh Bachchan  Mahendra Singh Dhoni     Ian Goodfellow   \n",
       "4             Tesla      Marques Brownlee     Anushka Sharma   \n",
       "5  Sachin Tendulkar        Hrithik Roshan   Deepika Padukone   \n",
       "6         Elon Musk     Leonardo DiCaprio  Cristiano Ronaldo   \n",
       "7     Narendra Modi       Virender Sehwag       J.K. Rowling   \n",
       "8       Virat Kohli              PRIYANKA             Kaggle   \n",
       "9        Bill Gates                  BCCI      Harsha Bhogle   \n",
       "\n",
       "                       3                     4  \n",
       "0          Sonam K Ahuja      Soumith Chintala  \n",
       "1             Tom Cruise            Kirk Borne  \n",
       "2  UEFA Champions League  Data Science Central  \n",
       "3    IndianPremierLeague     Open Data Science  \n",
       "4              KDnuggets                Virgin  \n",
       "5        President Trump          Entrepreneur  \n",
       "6             Alia Bhatt               Vir Das  \n",
       "7   Indian Football Team                   ICC  \n",
       "8           FC Barcelona         Rohan Kuntoji  \n",
       "9          Sunil Chhetri          Adam Geitgey  "
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dummy_df = pd.DataFrame()\n",
    "temp_df  = results_df.sort_values( by = 'PageRank', ascending =False)\n",
    "\n",
    "for i in range(5):\n",
    "    dummy_df[i] = pd.Series(list (temp_df [10*i : 10* i + 10]['FullName']))\n",
    "dummy_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 2c. Histogram of PageRank scores"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "PageRank scores are scaled such that nodes have an average-score of 1. So the scores below give an idea of how influential are the nodes, with respect to an average node. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pageRank_to_plot = len(index) * results_df[\"PageRank\"]\n",
    "sns.distplot(pageRank_to_plot, kde=False, rug=True, bins = len(index)//10);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A joint-plot showing how Inlinks and Outlinks of the nodes are distributed (within the ego-network)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAaUAAAGoCAYAAADmTPpwAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nO3deXgc133m+29vaOwrQSwkuIAAD0lRFKmNtiiJ2hdbsixbskfyEsV2kplknPh5xjdzc+Mkdm6Sa48nVsYZe+yJLCuJx5FkK/Kixav2haQ2ShSXI+4rCIAgQexr9/2jAQqEQKAb6Oqq7no/z8NHQKO6+1ctol7+Tp06FYjH44iIiHhB0O0CREREximURETEMxRKIiLiGQolERHxDIWSiIh4RtjtAlLR3t59ZqpgRUUhp071uVmOq/y+/6DPQPuf3ftfXV0ScLsGL8raTikcDrldgqv8vv+gz0D77+/9z1VZ1SlJenX1DfHm7hPsP95Nd+8Q3X1D5OWFqCrNZ15ZPisWVbC0rpRgUP+gE5HMUCj50PYDJ3n8pQPYw53E41CYH6asKI+iggine4Y40NJNT/8wAMUFES5oquLqdQtprC91uXIRyXUKJR85dqKXh5/ew1t7O6gqzef6SxporCulojhKLB5nfHGPQAAGh0Y52NrDgdYuXt3VzovbjrOktoSb1i/i4hXzCQbUPYlI+imUfOLFbS388y8skXCQWzcs4bwlFcRiiZ+Nxs5eaioeh7xIiOaFZTQvLOPKNfXYw528Ztv5zk+30/DyQT58xVLWNs0joHASkTRSKOW4kdEYDz+1h9+8dgTTUM6tly8lCGcCKRnRSIg1jVWsXlLJ7qOneWFbC//4yDZWLCrnP1zbzKKaEsfqFxF/USjlsJHRGN9+9G227jnB1esW8L5VNe/pilIRDAYwDeU0Lyjj7f0nee6tY3zlgVfYuHYBH93YSFF+JI3Vi4gfKZRy1MhojO/+dDtb95zgoxuX0bywbE6BNFEwGGDNsiqWN5SxeUcrz209ymu2jY9d3cRlq2s1pCcis5a11ynJucVice57bAevvdPO7Vc20rywjFiaAmmi/LwwG9cu4HduMpQXR/ne4zv52g9f5+iJ3rS/l4j4g0IpB/34mb1s2dnGrRuWYBrKHQmkiarLC/kP1zTxgfWLOdzWy5fv38Ijz+5lcHjU0fcVkdyj4bsc89LbLfxiyyGuuKCe1Usq0zZkN5NAIMDqxkoa60t4/q0WHn/5IFt2tvLJGwznN1ZlpAYRyX7qlHLIvmNdPPCkxTSUs2F1bcYCaaLC/Ag3XrqIu65tBuDeh9/k249u42TXQMZrEZHso04pR3T3DfGtR7dRXpzHhy5f4viQ3Uwa5hfzqRsMr73TzkvbjrNt30k+dPkSrr+4gXBI/xYSkanp6JAD4vE4339iF919Q3z82mYCeGP2WzgUZP3KGj7zwRUsri3hR0/v5cv3b2H7gZNulyYiHqVQygFPvX6UrXtOcMtlSyjO917zW1YU5cOXL+WOjY0MDI/y9w9u5Vv/vo32zn63SxMRj/HeEUxScrith4ee2sPqpZWsXFThynmkZDXWl7GopoTX32nnpbeP8+beDm68tIEPvn8x+Xn6qygiCqWsNjIa459+voPC/DA3rV/k6UAaFw4FuXRlDSsXV/Di28d5/OWDvPBWCx+5spEN59fpNhkiPqfhuyz2sxf3c6S9h9uvbMy6VbtLCvO46dJFfOpGQ2lRHt9/chdfeeAVduh8k4ivKZSy1DuHTvHEy4dYv6qGuspCt8uZtbrKxIW3H758KT39w/z3B7dy78NvcqS9x+3SRMQFGr7LQsMjo/zDg69TVpTHxgvqXJ/+PVeBQIDlDeU01pfy5t4TvPT2cf7q/i1sOL+O269opKIk6naJIpIhCqUs9LMXD3C4tYfPfGAleGT6dzqEQ0EuWj6fVYsr2bKrjZe3H2fLjlauv6SBm9cvolCrkIvkPIVSljl4vJsnNx3iirULmF9RkPVd0lQKomE2XlDP2qYqXtreyuMvH+SZN45yy2VLuObCBUTCIbdLFBGH6JxSFhkZjXH/EzspKYxw/aUNORlIE5UVRbn50kXcc/MK6qqKeOipPfw//3szL7zVkvP7LuJXCqUs8uSmgxxu6+G2K5YSDvmnW5hfXsBHrmzkrmubKIiGuP+Jnfzl/Vt4edsx4nGFk0gu0fBdljja3sPPXjzAhcurWVBV5HY5rmiYX8Jd1xaz52gXL2xr4e8eeIWldSV85MplrFpSoZsLiuQAhVIWiMXi3P/ELgqiYa65aEFWXCTrlEAgQPPCMpbVl3LwRC+/2XKIv39oK8sbyrj9ikbMogq3SxSROVAoZYFfvXKY/S1dfOKG5QRzaLbdXASDAdYtn09DVSHbD5zk5beP87UfvsHKxRV8aMMShZNIllIoeVzryT4efX4fFyyroqG6iFjM7Yq8JRwKcsGyeaxaXMnb+zvYtKOVr/3wDVYsKufWy5awYrGG9USyiULJw2KxON97YieRUJBrL2lQIE0jEg6yrrma1Uur2H7gJJt2tPL1B7eyrL6UD75/CWuaqrJuKSYRP1IoedgvXznEniOnueu6ZsI6oCYlEg6ytmkeq5dWsuPgKTbvaOWbj7zFgnlF3LR+EetX1egmgyIeplDyqKPtPTz63D4uaJrH4ppidUkpCoeCrGms4rwllew50snmnW187/GdPPLsXq6+cCFXra2npDDP7TJFZBKFkgeNjMa477GdFETDXH/xQgXSHISCAcyiCpY3lHO4rYfX3mnn0ef28dhLB1i/soZrL1rI4toSt8sUkTEKJQ969Ll9HGzt5tM3GZ0HSZNAIMCimhIW1ZTQ0TXAm3s72LKzlRe2tdBYX8rGtfVcurKGaMQ/FyWLeJFCyWPe3t/Bk5sPseH8WuqrirScjgOqSvO5Zt0CNqyuYefBTrbuOcH3n9jFg7/dw/tW1XD5mjqW1JZo1p6ICxRKHnK6d4j7HttJXVUhV6zJ/ltSeF00EmZt0zwuWFZFy8k+tu3r4IVtLTz9xlEWVBdx2epa3reqVrfOEMkghZJHxGJx7ntsB/2DI3z6JqPzSBkUCASoryqivqqIq9cu4J0jp3l7fwc/enovP35mLysXV/C+VbVcuLyawnz9yog4Sb9hHvHIc3vZvv8kd169jKJoGK0z6o68SIjVSytZvbSSzp5Bdh3qZPv+k9z/xE7+5ZeW8xsruWTlfNY2zSM/T78+Iumm3yoP2LKzlSc3HWLD+XU01Zf5em07LykvjvK+VTWsXzmfts4B7OFT7Dxwijd2nyASDrJ6aSUXmWouaJpHkW5AKJIWCiWXHWrt5v7Hd7JsQSlXrqlVIHlQIBCgpqKAmooCrji/jpaTfew5cpqdhxIBFQombue+tmkea5qqqKkodLtkkaylUHJRW2c/9z78JkUFET58RSOjOo/keRPPP12xpo62U/3sa+ninSOn+bff7ubffrubmooCVjdWsWpJBaahQuehRFKg3xaXnO4d4hsPbmVkNMZnb15JSNOPs04gEKCmspCaykLef14tXX1DHDjezYGWbp578xi/fe0IgQAsrimheWE5zQvLaKwvpaIkqunmIuegUHJBT/8w9z60lc7eQT77wVXkR/S/IReUFuaxprGKNY1VjMZitJ7q53BbD0faenjmjaP8+tXDY9tFWFRbwsJ5xSyoLqK2spDqigJKCiIKK/E9HQ0zrLNnkL9/aCutJ/v49E0rKCnQCfJcFAoGzwzzrV9ZQywW48TpQdo6+2k91UdLRx+7Dp5iZPTdc4j5eSEqS/OpKIlSXpxHaWEeJYV5FBdEKMoPU5gfpjA/QmE0TEE0rOvYJCcplDKovbOf//7gG3T1DvO7N6+ksjSqqd8+EQwGmV9RwPyKAlYvrUw8GIfTfUN09gxyum+Irt4hunuHOd07xNH2Hnr6h88KrckCASiIhimMhhPBVRChtDBCaVEeZUVRKkujVJREmVdWQFlxnpaskqygUMqQXQdP8Z2fvs1oLM5nPriSkoKIAsnvAlBWlEdZ0XtXKx/Pj+GRGINDowwMjzIwNMrwyCiDIzGGhkchEKCrZ5D+wVH6B0fo6RumpaOX7r5hhkfOnjUTDgWpLs+ntrKQuqoi6qoKaZhfTF1VEZGwbuUh3qFQclg8HucXWw7xyDP7mF9RwMeuaSJfi37KDMb/wRIOBQkXBCmaYpi3vLyQzs6+sx57N8xG6ekfobt/mO6+Ybp6hzjZPUBLRx9v7u04M/QXDASom1fIovklLK4tYUltCYtrSojm6e+ouEOh5KC2zn7+9Re72H7gFOuWz+OGixvUHYmj3g2zEOXFIcqL37tuXxzo6h3kxOkBTpweoO1UP9sPnOTl7ceBRLDVVxWxtK6UpXUlLK0vZWF1sW6OKBmhUHLA8Mgov371CD97YT/BYICPbFyGWViq65DEEwJAWVGUsqIoy+rLzjzePzRC26l+Wk/103qyj617TvDCthYg0bE1zC9maV0JS2pLWVJXQl1VIaGggkrSS6GURsMjozy79RiPbzrI6Z4hLmiq4tqLGggHAwok8byCvDCLaxLDd5AYeu4dGD4TVMc6enlx23Geev0okLj1fMP8YhbVlNAwv5iG+cUsmFdEQVSHFZk9/e1Jg8NtPTz/1jE2bW+lp3+Y5oVl3HHVMuaXFWjZIMlagUCA4oI8igvyaDzTUcXp7BmivbOfts5+jp/sY/OO4zzzxuiZ580ry6d+XhH18xITKuoqi6ipLKBY12FJEhRKs9A3MMy+Y11s23eSt/Z10Hqyj1AwwJqmKi5aPp/q8nxGR+MKJMlBAcqLo5QXR2leWD72WJze/hHaTw/Q0ZU4T9Xe2c+OAyfPmtJemB+muryA+eUFzCvPZ15ZAVWl+VSWRqksiVIQDSu0RKF0LsMjMTp7BjnVPcjJrgGOn+zj+Mk+Drf10NKRmPEUDgVZ3lDG+8+rpam+hFAoSDwOo9NcWyKSewIUjV0ntaS25KyfdI1dh9XZO0Rn9yCd3YMcON7N6++0v+cfbdFIiLLiPMqL8igtjlJaGHn34uGCxLVYhdHEhcT50TCl5aPE43EFWY7JplAKB4Nn/+Wb/P1snOoa4KcvHqB3YJjB4VEGBkfpGxxJXAcyQQAoL4mybGE5l6+pp6osn+ryfALgyoy6SDjo+2m7fv8MsmH/q/MKqC4vmPJnA0Oj9PQP0zcwTN/gKL0Dw/QNjtDbP0LvwDAnOvvpGxyZ9vVDwQCRcIhIJEheOEgkHCQSCiam0ocChEJBwsEgwWCAUChAKJAYlgwGAwQDiT+MPTb2ZeJrgAAURsNsOL/OqWu5lgBHgOl30mcC8eyZo7wE2O92ESIiabQUOOB2EV6STaEUBha6XYSISBqpU5okm0JJRERynK58ExERz1AoiYiIZyiURETEMxRKIiLiGQolERHxDIWSiIh4hkJJREQ8I5tCKUxiVYdsWhpJRGQufHfcy6YdXQjs7+joIRaLU1FRyKlTfTM+KVf5ff9Bn4H2P7v3v7q6JJnFO8867uWK6fY9mzqls4TD3l6I0ml+33/QZ6D99/f+56qsDSUREck9CiUREfEMhZKIiHiGQklERDxDoSQiIp6hUBIREc9QKImIiGcolERExDMUSiIi4hkKJRER8QyFkoiIeIZCSUREPCObVgmftXB+ntslpN2p7sGc3K9U+P0z0P57Y/9HBobcLiGnqFMSERHPUCiJiIhnKJRERMQzFEoiIuIZCiUREY8bGBp1u4SMUSiJiHhcZ8+g2yVkjEJJRMTjAoGA2yVkjEJJRMTj4sTdLiFjFEoiIl7nn0xSKImIeF0s7p9UUiiJiHhcLOZ2BZnj6Np3xphPAn829u2T1tovGmPWAv8ElAHPAf/RWjviZB0iItksrk5p7owxhcA3gY3ABcAVxpjrgB8An7fWLgcCwO85VYOISC7Q8F16hMZevwiIjP0ZBgqstZvGtnkAuNPBGkREsp6fhu8cCyVrbTfwF8Au4ChwABgCWiZs1gIsdKoGEZFcMOKjVHLsnJIxZg3wGWAxcJrEsN0NU2ya0qddVVV85uvq6pKknnOq2z9XQ4tIZiV7HJqLaDSSkffxAicnOtwI/NZa2wZgjHkA+CJQO2GbOuBYKi/a0dFDLBanurqE9vbupJ7jhRuBiUhuSvY4NFkqIdPVMzDr9/Gi6fbdyXNKbwLXGWOKjDEB4FbgWWDAGLNhbJtPA086WIOISNYbHfXP8J2T55R+Bfwb8BrwFomJDl8FPgHca4zZSWISxDedqkFEJBeM+CiUHL1OyVr7NeBrkx5+E7jUyfcVEcklQ8OaEi4iIh4xNKz7KYmIiEcMjSiURETEIwaH/XNOSaEkIuJxQ8P+WR5UoSQi4nEDQ+qURETEIwY10UFERLxiYEjDdyIi4hGDQ+qURETEIxRKIiLiGQM6pyQiIl6hiQ4iIuIZIyMxYnF/TAtXKImIZIGBQX90SwolEZEs4JelhhRKIiJZwC/nlRRKIiJZYNgnN/pTKImIZIGREYWSiIh4hF9uia5QEhHJAqMxf9wSXaEkIpIF4nGFkoiIeIRPMkmhJCKSDQIBtyvIDIWSiEgWCAX9kUoKJRGRLBAK+eNw7Y+9FBHJcpGwPw7X/thLEZEsF1GnJCIiXpEXCbldQkYolEREskBUoSQiIl6Rn6dQEhERDwgFA4RDmhIuIiIeEI2ECPjk6lmFkoiIx+VFQlpmSEREvMEv55NAoSQi4nnRvLDbJWSMQklExOOief45VPtnT0VEslQ0ok5JREQ8Ij/in0O1f/ZURCRL5Wmig4iIeIVflhgChZKIiOcplERExDP8skI4KJRERDwv4pN170ChJCLieZGwOiUREfGISFidkoiIeETIJ7dCB4WSiIjnhYPqlERExCPUKYmIiGeEfNQpObrKnzHmVuDLQBHwS2vtnxhjrgO+ARQAD1lrv+RkDSIi2S6Af0LJsU7JGNMIfAe4DTgfuNAYczNw/9hjK4FLxh4TEZFzCPhoTMvJXb2dRCd0xFo7DHwc6AN2W2v3W2tHgB8AdzpYg4hI1gsG/NMpOTl81wQMGWN+CdQCPwe2Ay0TtmkBFqbyolVVxWe+rq4uSeo5p7oHU3kLEZGkJXscmouSkvyMvI8XOBlKYeBK4CqgB/gpiU5pslgqL9rR0UMsFqe6uoT29u7kCsnPS+UtRESSluxxaLJUQqavd3DW7+NF0+27k6F0HPiNtbYdwBjzExJDdaMTtqkDjjlYg4hI9vPP6J2jofQY8M/GmHKgG7gZ+DHwfxtjmoD9wN0kJj6IiMg5BHx0TsmxiQ7W2s3AfwNeAHYAB4H/BdwDPDL22C4SQSUiIuLsdUrW2vt5byf0W+ACJ99XRESyk49mv4uIZCf/DN4plEREPE+hJCIi4gKFkoiIeIZCSUREPEOhJCLidT46qaRQEhERz1AoiYiIZyiURETEMxRKIiLiGQolERHxDIWSiIh4hkJJREQ8Q6EkIiKeoVASERHPUCiJiIhnKJRERMQzFEoiIuIZCiUREfEMhZKIiHiGQklERDxDoSQiIp6hUBIREc9QKImIiGcolERExDMUSiIi4hkKJRER8QyFkoiIeIZCSUREPEOhJCIinqFQEhERz1AoiYiIZyiURES8Lu52AZmjUBIREc9QKImIiGcolERExDMUSiIiHuejU0oKJRER8Q6FkoiIx6lTEhERcYFCSUTE63zUKimUREQ8Lu6jVFIoiYh4XNw/maRQEhHxuriPUkmhJCLicbGY2xVkjkJJRMTjYj7qlMLJbGSMWQFcDnwPeBRYA3zWWvu0g7WJiAgavpvKd4F+4BZgHvAZ4O+cKkpERN41GlMoTZZvrf0/wA3Aw9baZ4BIMk80xnzdGPPA2NdrjTGvGGPeMcbcZ4xJqlMTEfGzUR+dVEo2lKLGmBrgg8Bvxr4umOlJxphrgXsmPPQD4PPW2uVAAPi91MoVEfGf0VG3K8icVIbvDgIvWGt3AK8A/zDdE4wxlcDfMjbMZ4xZDBRYazeNbfIAcOcsahYR8ZURH6VSUsNn1tr/ZYz5rrV2vIdcB4zM8LTvAn8ONIx9Xw+0TPh5C7AwhVoBqKoqPvN1dXVJUs851T2Y6tuIiCQl2ePQXIQi4Yy8jxckO/vu/7PW/tmEh9YB9wOLzrH954DD1trfGmPuGXs4MMWmKQ+UdnT0EIvFqa4uob29O6nnhPPzUn0bEZGkJHscmiyVkOnuGZj1+3jRdPue7ESDy4wxXwT+Efg6cAfwR9Ns/3GgzhizFagEikksKVg7YZs64FiS7y8i4ltDw5p9N9kHgNsBC9QAa6y1j55rY2vt9dba1dbatcBfAj+z1v4uMGCM2TC22aeBJ2dfuoiIPwwNz3S2JHdMG0rGmAuNMRcCBvgLEtPAnwYWjT2eqk8A9xpjdgJFwDdn8RoiIr4yOOyfKeEzDd89Mun7IeC/jn0dBxpnegNr7QMkZtphrX0TuDSlCkVEfG5wyD+d0rShZK1dmqlCRERkav1DmhJ+FmNMLfAHJCYtnJlFZ639Y4fqEhGRMQMKpfd4GDgFvIGvbswrIuK+/kEN301Wba290tFKRERkSn4KpWSnhB80xhQ5WomIiEzJT6GUbKfUAmw1xjxD4hYWgM4piYhkQp9C6T0OjP0REZEMGxmJMTQySl445HYpjkt2QdavOF2IiIicW3ffMFWlPg8lY8wL1trLjTHdTDHrzlpb6lhlIiJyRk//MFWl+W6X4biZOqXx+x2tdroQERE5t67eIbdLyIiZQun9xpjpfn4wjbWIiMg5KJQSPj/Nz+LAv6exFhEROYeuPoUS1tqrAYwxl1prt0z8mTHmOicLExGRhGAwQFfvsNtlZMRMEx3WkVjr7p+NMXfz7rp3EeA+YImj1YmICAXRMN3qlAD4T8D1QD1nD9WNAD9yqigREXlXfl6I7n51Slhrfx/AGPM31tovZaYkERGZKD8vzOmeQbfLyIhkV3T496nuNGutfT3N9YiIyCQFeSGOqVM6y8Q70OYBdcCr6C6yIiKOi+aF6B3wx/p3yS4zdNYdaI0x7wM+60hFIiJylvxIiP7BEeLxOIFAYOYnZLFkb11xFmvtJuCiNNciIiJTiOaFGI3FGRrJ/TvQJns79InnkwLAxUCBIxWJiMhZ8iKJQ3X/4CjRSLJnXbJTKueUxhdkjQPtJKaLi4iIw6KRxKBW/+AI5cVRl6tx1ozDd8aYeuBxoAs4DTwJ3GmtfcbZ0kREBCAvPBZKQ7k/fDdtKBljGoAtwDDwJeBvxn602Riz2OHaREQEiIzd3G9gKPdn4M00fPc3wJ9Za/91wmOPGGNeG/vZpxyrTEREAIiMDd8NDPq8UwIunBRIAFhrv4+uURIRyYgzw3eDud8pzRRK002I98eaFyIiLnt3+E6d0sjYRIezjD2mUBIRyYBIKNEf9PvgnNJMofQd4PvGmNLxB4wx84F/Bb7tZGEiIpIQCQcJBPzRKc20Svh3jDFNwFFjzA4S91FqBr45dl5JREQcFyAvHGTAB+eUZrx41lr7RWPMvcD6sYc2WWuPOVuWiIhMlBcOqVMaZ609ytk3+RMRkQzKywtp9p2IiHhDNBxUKImIiDdE80L0KZRERMQLCvLC9PYrlERExAPyo2F6B3L/lugKJRGRLFAYDTMwNMrIaMztUhylUBIRyQJFBYnJ0l29ud0tKZRERLJAUX4EgNO9ub3Cm0JJRCQLlBYmQqmja8DlSpylUBIRyQKlRXkAtHf2u1yJsxRKIiJZID8vTDQSou1UbodSUssMSfbac7gzqe2aGsodrkRE5qqyNEpLR5/bZThKoZSDkg2icz1HASXiTVWl+Rxq63G7DEdp+C6H7DncOatAcup1RCS9qsry6eodoqd/yO1SHKNQyhFOhIjCScRb6ioLAThwPHe7JYVSlstEcCicRLyhZiyU9h7N3d9HhVIWy3RQKJxE3BWNhKguL2DXodz9PXR0ooMx5q+Aj419+7i19k+NMdcB3wAKgIestV9ysoZc5WY4jL+3JkSIZN6S2hJef6edwaFRonkht8tJO8c6pbHwuQFYB6wFLjLG3AXcD9wGrAQuMcbc7FQNucor3cp45+SVekT8YGldCaOxODZHf++cHL5rAf6LtXbIWjsM7ASWA7uttfuttSPAD4A7Hawh53g1ABRQIpmxsLqY/LwQL7193O1SHOHY8J21dvv418aYZuDjwDdJhNW4FmBhKq9bVVV85uvq6pKknnOqOzcWMMyWA76ueRI/SfY4NBclJfmE8949XK9pqmbrO20UleRTOLZQa65w/OJZY8x5wOPAF4FhwEzaJKWbg3R09BCLxamuLqG9vTup54Tz81J5C0/KlkCa7Fx1K6wkVyR7HJoslTDr7h6gp+/da5Oa60vYsuM4jz+/l6vWLpjV+7tpun13eqLDBuAR4AvW2geNMRuB2gmb1AHHnKwhF2RrIE1ntvukMBOB+nlF1FYW8uSmg1yxpo5QMHcmUjsWSsaYBuAnwMettU+NPbw58SPTBOwH7iYx8UHOIRcDaS5m83koyCTXBAIB3reqhp+8sJ9Xd7WzflWN2yWljZOd0heBfOAbxpwZsfsOcA+J7ikfeAL4sYM1ZDUFUnrM9DkqtCQbNS8so7o8n0ee3cva5nlEI7kxPdzJiQ5/AvzJOX58gVPvmysUSJkz1WetoBKvCwQCXHvhQh58ag+PvXSAj25c5nZJaaFVwj0oU4G050gnTQt18J2KgkqywaKaEs5vrOTJzYe4dGUNDfOLZ36SxymUPCbdgbTnyPSvN9PPxym8FFTiTRvXLmBfSxffenQbf/k7F2f9FHGFkoekK5CSDZq5vqaCStdkifsKo2Fu27CUB5/awz/9fAefv2MNwUDA7bJmTaHkEXMNJCeCaLbv6dewUkCJWxZWF3PNugX85rUj/PiZvdx51TICWRpMCqUs50YYzWSmmvwQWgooybR1zfM42T3ALzYfIhIKcvuVjW6XNCsKJQ+YbZfkxUBKRqp1Z3uIKaAkE8Zn48Vi8POXDhAKBfjQhqVul5UyhZLLZhNI2RpGs5VLIabbfoiTAoEA11+8kNFYnJ88v5++gXD/xewAABUnSURBVBE+dnUTwWD2DOUplFzkdiDtPXp62p8vW1CWtvfKpOk+I68ElsJJnBIIBLjxkgaieSF+9cph2jr7+YNbz8uaey8plFySyUCaKXzm8rxsCy6vTc7Q0J44IRgMcM26BVQU5/Gb147w1R++zuc/cj6VpflulzYjhVKWSDWQZhtEqcqVbmvy5+tGSCmgJN3WNVdTVpTHz146wJe//wqfu2UVa5ZVuV3WtALxeNztGpK1BNifC7euSKVLSiWMMhVE6ZQtoeXmsJ8CyttGBoZm3mgK1dUlyZzoWQLsf2PH8bNuXZGqjq4BHnv5IK0n+7h5/SJuv7KRcMi9lcWn23d1ShmmQDpbtgwRTvx/kemAmvx3RiElqaoqzefua5t59s1jPLn5EO8c6eT3blnF/IpCt0t7D4WSRyUbSHMJo92HU39uc0PmA2K6fXQjsMb/33jhPNRECiuZTiQc5LqLFtJQXcQvXznMX96/hY9d3cRV6xZ4agUIhVIGpXtdu1QCaTYBNNvXyWRwneszyERYudk9TUW36JBkmEUV1M8r4levHuYHv3qH199p53dvXklVmTcmQeicUgYlG0ozdUnJhlG6giid3Oi0IHMdlRfCKRUKqrnLhnNKU4nH42zb18FTbxwlGAhw59VNbFxbn5Guabp9VyhlSCYDyYthlIxMBpbTIZVt4TROIZW6bA2lcZ09g/z61SPsb+lieUMZv3PTCuqqihx5r3EKJZdDKVOBNJswamlpSfk5AHV1dbN6XqoyEVROBlS2hhMooJKV7aEEia5px8FTPPX6EYZHYty6YSk3r1/k2Aw9zb7LAekKpNmGUKqvk67Qmmq/0h1UEz/bdAeU25Mi5kKrTvhHIBDgvCWVLK4p4ZmtR3n0uX1s2dnKPTetyPhkInVKDktHlzRdICUTRukKonRwosNyopty4hcxG4NpIoXT1HKhU5ps37HT/PrVI3T1DnHNhQv5yMZGCqLp62E0fOfxUHIqkLwURtNJZ1ClO6AUTmdTML1XLoYSwODwKC+9fZxXd7VRXhLlkzcsZ11zdVpeW8N3OWi6QEo2jE637k5XOZTVNM/6uZPrnUtITfxc0hFQ4/8oSGc4eW0qeSo0pOcf0UiIq9ctYOWiCn716mH+8ZFtrF9Vw93XNVNS6Nw/9NUpOcipLulcgZRMGKUziJIxl7CC9HRR6eqeNCHiXQqld+VqpzTR6GiMV2wbL247TmF+mE9cv5xLVsyf9d1tNXyXhaGUzkDKdBBNZy4hNdeASkc4ZeKkb7YElIIpwQ+hNO7E6X5++cphjrb3cpGp5lM3Gkpn0TVp+M4Fc+2SpuJ0IHUe35NSPeW1TSltD++tJ5WQmrifswmo8c9vLuHkxHDeZF5YsVxkKvPKCrjrmmZe393Os1uPsefIaT57y0pWL03fyuPqlBySqS7pXIGUTBilGkLJmk1Ywey6qNl2T9nSNU3FSyGlbslfndJE7Z39PL7pIG2n+rnhkgY+unEZkXBy1zVp+M6DoeRWIDkVRNNJNaSyKZzcXsHc7YBSKPk3lACGR2K8sK2FV3a1saimmP98+/nMKy+Y8XnT7bt7N9TIYeleeDUdgdR5fI8rgTSb9z7dujvl82Cznf4+1yWZ3L5lyJ4jnWf+iGRaJBzk6nULuGPjMtpO9fPX//wq9tCpOb2mzim5INUuKVnTBdLMz9076/ctq1mW1HYT60imexrfn2Q7p5aWlowtf+RF2bx6hGS3xvpSPnn9cn7ywn6+/uBWPnn9cq5at2BWr6VOyeOS7ZKmCqSZOpTTrXvP/JmLia+T7Gul0j2l0jXNpmPK9m5pMnVN4obK0nzuvq6ZpXUl/MsvLT9/cf+sXkehlGVSCaRzSUcQTSfVcEruNb0dTH6k80kyWX5emNsvb+T8xkoefX4/T246mPJrKJTSLJ0THJI5UKYSSE6H0WzfL9muyelgmgt1SyIJwWCAGy9ZxKolFfzomb38+pXDKT1f55SySDIH2ukCacbntqZ4nVJNcrPqxt97pnNPncf3zHiu6XTr7jmvEnEuuw+fdu0mhCK5JBgM8IH1ixmNxXnwt7tZXFvC8iQ7a3VKHjXbLmnq7aYPpM7WPSkH0myel2zXNPPrJLff2bIgbTbS0J3MJBgMcPOliygrzuN7j+9kcGg0uec5XJckIZmhn9l2SdMFwWzDaC6vk8nhw0zz0hCekzPwFEiSrLxIiJvWL6K9s59Hnkvud1+hlKWS6RZmCqR0SzacZuzcXLqeSkTSb9H8Es5vrOS5rccYGp65W1IopdFsJzmkQyoHcicCKZOvD95aZHac26s7jFOXJF6zYlEFQyMxdh6c+cJahZIHpWN68rm6kUwEhrhHgSRetGh+MdFIiNffaZ9xW4VSFph8PsmLXcJkM4WfF88tZfvMOwWSeFUoFKQgGqI/ickOCqUc4MdzME5NC58tt4fuFEjidb0DI5QVzbw4tq5TclkmZ2xlcuhupmuYkl0vby5SWQdvLl2Sm4Hk9Dp3CiRJh9O9QwyPxCgrViiJ5KRMLLqqQJJ0eentFkLBAJetrp1xW4WSeM5MqzokM3SXiS7JjQ5JYSTZpqNrgLf3n+SqtQuoLMmfcXuFko+U1zRlZAhvLkN3CqSpZep2FAokSafegWF+8vx+8sIhbrtiaVLPUSjlgPLaJs9MdvDCuaRkeT2QMn1fJAWSpFP/4AgPP7WH071DfOFjF1BamNwdwBVKLlu2oCzlyQ5lNc0zTgsvq1k25bRrp7qlZBZnnSmQMtkleTWQ3LhBn8JI0q27b4hHntvHyZ5B/vija1i1uCLp5yqUskBdXV1aFxdNdzB5IZCcHrJTGIkkZ8/R0zy56SAjsTh/+OHVnN9YldLzFUoe1NxQlvKqDlMN4Z2rW4J3g2S24ZTsbSu81B2BtwLJzduWK5Ak3UZGYzzzxlFe332C+nlF/NHtq6mrKkr5dRRKWSqZIbzEducOJnhvuEwVUskG0OT3nYnXh+ucCCM3gwgURuKMQ63d/Oa1I5w4PcDV6xZw13XNhEOzW5vBlVAyxtwNfAnIA+611n7LjTrSramhfNpFWZsWls96UdZkhvDONeFhpmA66zVmEUCT32vG9/B4d5TuMFIQSa463TPIM1uPYQ93Ul4c5Y/vWMPapnlzes2Mh5IxZgHwt8BFwCDwkjHmaWvtjkzX4hVTTXZIZghvqm5pumACZ9acS3ZG3UxhlHit3OmOFEaSq4ZGRtm8o41XdrUSCAS45bIl3HrZYiLh0Jxf241O6TrgKWvtSQBjzI+BO4C/dqGWrDJVt5RKMCW2fzdAZhtQqUzrTlcQQXaEkdtBBAojcc7IaIw393awaftxegdGuHB5NXdd10xV6cwXxSbLjVCqByYeWVuAS5N9clVV8Zmvq6tLknrOqe7BZF/ececawku2W0olmGD6xVqdvGbIT2GkIPK3ZI9Dc1FSkk84z70pAKOjMV6zbTz96mFO9w6xfFE5n755JRcsn5/293JjLwNTPBZL9skdHT3EYnGqq0tob+9O6jnh/OQu2kqHmc4rpUOywQTJhVM6JBNC49w+ZwS5EUYKIm9I9jg0WSph1t09QE/f0KzeZy5isTg7Dpzkpe3H6ewZYnFtCZ+9ZSWrl1YCAUf23Y1QOgpcMeH7OuCYC3W4Zq7dEpw7mGDq+y1NDo25hFQqATQu3V0RuDOBQUEkfjA6GuPtAyfZvKOVzp4hFlQX8akbV7CuuYqp+4r0cSOUfgN82RhTDfQCHwV+34U6skYqwQTTh9O42QRLKlK535G6omneVyEkGTQyGmPbvg4272ilq2+YhfOLuft6w8VmHk6H0biMh5K19qgx5s+Bp0lMCb/PWrsl03U4KZkhvFS6JUg9mCC5cEqXVG+6l0oQgb+6IgWRZNrwSIy39p5g8842evqHWVxbwj0fWMnaJuc7o8lcOXNmrf0h8EM33jsbTBdMwJSTH+C9t00fN1VgzDaoZnvH11RDCNzpirTcj/jJeBht2tlKb/8IjfWl/MGHVnHe2DkjN2hFB4fMpVuC6Rdqna5rGjfThbZO3048UyEE2RdECiFx2+QwWlZfyu23NXLe0kricXdrUyi5bC7BBO/tmsalElBzNZsAGufmat2ZDCMFkXhBLBbnrX0dvPh2y5Rh5HYggULJUclOD58pmIBZhxOcOzSSDau5hM5Esw2gcdkURAoh8ZJ4PM6eo6d5dusxTnYPsqS2hP902zJPhdE4hZLD0hFMMPN9lyYe8JNdYTxdYTOVuQYQZFcIgYJIvOloew/PbD3G0RO9VJcX8IcfXs3FK6qBgKfCaJxCKQPSGUxw7q5p3GwCarbSET7jsm2pH4WQeFnf4AhPv36U7QdOUlwQ4RPXL+fqCxcQDLgzgSFZCqUMSSWYgKTCCVILKC/JxpW4FUKSDeLxONsPnOTp148yOBzjxksbuP2KRvIic18sNRMUShmUyhJEyd7mItnuyS3ZfoM8BZFkk86eQX655RAHW3tYXFvCZz64goYMrM2XTgqlDEs1mGD6rmnc5IN/pkMqV24XrhCSbLX7SCePbzpIgAB3XdfMdRcvJODStUZzoVByQaqLtqYSTuOSCYmZgsvpoDkXXTckkrxYLM5zbx1jy842FlYX8fk71lBdVuB2WbOmUHLJbFYTn3iwnu0dbCdyK3TGaRkfkbkZGBrh0ef3c7ithw3n1/E7N5lZ34bcKxRKLho/OM7mVhfpDignaWVtkfQbGBrh4af30t7Zzz03r2Dj2npPTvFOlULJA+YSTjD1QT+TQeV26EymEJJcNzg0yo+e2UtbZz9/+OHzuHD5/JwIJFAoecpcw+ms1/JYUDhFASR+Mzoa48fP7qX1VD//8bZEIOUShZIHpTOccokCSARe2NbC0RO9fO6WlVxsciuQQKHkaRMPwn4KKIWPyNQOHu9m8842Lltdy2WrnVsmzE0KpSyRawGl4BFJzchojCc2H6S6vIBP3WjcLscxCqUsNPmA7sWQUuiIpNe2fR109w3zuVtXEc2SJYNmQ6GUA5IJgLkElwJGxF2xWIxNO1pZUlvC+Usr3S7HUQoln1CwiGSvfce66O4b5nc/sBK3blOeKdl96a+IiA/sPdZFUX6YC5bNc7sUxymUREQ87kBrF2ua5uHxWyGlhUJJRMTjhodjrGvO/S4JFEoiIlmhyeUFlDNFoSQi4nGF+RHKi/PcLiMjFEoiIh5XXZ5Prs+6G6dQEhHxuNKiqNslZIxCSUTE4yp8MnQHCiUREc8rKoi4XULGKJRERDyuMOqfxXcUSiIiHlddXuB2CRmjUBIR8bjG+lK3S8gYhZKIiHiGQklERDxDoSQiIp6hUBIREc9QKImIiGf4YvL7yMCQ2yWkXXV1Ce3t3W6X4Sq/fwbaf3/vf65SpyQiIp6hUBIREc9QKImIiGcolERExDMUSiIi4hkKJRER8QyFkoiIeIZCSUREPEOhJCIinqFQEhERz1AoiYiIZ2TT2nchgGAwcOaBiV/7kd/3H/QZaP+zev+XAEeAEZfr8JRAPB53u4ZkXQ4873YRIiJptBQ4MM3PlwD7k9guZ2RTKEWBS4AWYNTlWkRE0mGmTikMLExiu5yRTaEkIiI5ThMdRETEMxRKIiLiGQolERHxDIWSiIh4hkJJREQ8Q6EkIiKeoVASERHPUCiJiIhnZNPad2cxxvw1ELPWfnns+3Lg/wCNQDvwMWvtcfcqdI4x5m7gS0AecK+19lsul5QRxphS4CXgFmvtAWPMdcA3gALgIWvtl1wt0EHGmL8CPjb27ePW2j/12f7/NXAHEAe+Z639hp/230+yrlMyxpQZY74HfHHSj/4GeN5auxL4J+B/ZLy4DDDGLAD+lsRagBcAv2+MWeVuVc4zxqwHXgCWj31fANwP3AasBC4xxtzsXoXOGTv43gCsA9YCFxlj7sI/+78RuAZYA1wMfN4YcwE+2X+/ybpQIvGXcDfw95Me/yCJTgng34CbjTGRTBaWIdcBT1lrT1pre4Efk/gXZK77PeCPgGNj318K7LbW7rfWjgA/AO50qziHtQD/xVo7ZK0dBnaSCGdf7L+19lng6rH9nE9ihKccn+y/32RdKFlr/8Va+1XeuyhrPYlfXsb+knYB1RkuLxPO7OeYFhILNuY0a+3nrLUTV4n3zedgrd1urd0EYIxpBj4OxPDJ/gNYa4eNMV8BdgC/xUf///3Gs+eUjDF3AvdOeniXtfa6czxlqhurxNJblSf4ZT9n4rvPwRhzHvA4iaHrYcBM2iSn999a+1fGmK8BPweap9gkp/ffLzwbStbaHwE/SuEpR4Fa4IgxJgyUAh1O1Oayo8AVE76v490hLT8Z//89Lqc/B2PMBuAR4AvW2gfHzrP4Yv+NMSuAfGvtVmttnzHm30kMWU8cLcnZ/fcbz4bSLDwBfBr4OxLDG8+Pjb/nmt8AXzbGVAO9wEeB33e3JFdsBowxponETdDuJnHiO+cYYxqAnwAft9Y+Nfawb/afxIzarxhjLicx++424LvA132y/76SdeeUpvEXwPuMMduBPyRxUjznWGuPAn8OPA1sBX5ord3iblWZZ60dAO4h0T3sAHaRmPSRi74I5APfMMZsNcZsJbHv9+CD/bfWPkHiH51vAK8BL1lrH8Qn++83usmfiIh4Ri51SiIikuUUSiIi4hkKJRER8QyFkoiIeIZCSUREPEOhJDnHGHPAGHPxND+/xxjz2NjX940teDrd633ZGPM/012niLxXLl08K5Iya+3n3K5BRN6lUJKcZYwZAL4KXE9iAc//Ya39h0nbPAP8T+BVEgt9PgGsByqBP7fWPjRp+y+QuGjzJhIrVX+PxIWtAeA+a+23ndsjkdyn4TvJZVHghLV2A4m10r5qjMmfZvtG4JfW2kuB/wr8t4k/NMb8KYnbI1w1dgPJ/wv4ubX2IuADwJXGGP1OicyBOiXJdT8d++/rJEKqaJpth0l0SuPbV0742UdJLIB6q7W2c+yxR4F/McZcSmJNwj+21mqlapE50L/qJNf1A1hrx9fTmuqWF+OGJoRKfNK2u0l0W982xpSPveZjJG6h8DCJu8JuM8YsS2PtIr6jUBJJzlvW2kdInHf6FoAx5ockVu5+kMQiwF1Ag3slimQ/Dd+JpOYLwNvGmI8B/y9wnzHmD0jc2+dR4Fk3ixPJdlolXEREPEPDdyIi4hkKJRER8QyFkoiIeIZCSUREPEOhJCIinqFQEhERz1AoiYiIZ/z/KFfDfNx62dgAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x432 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.jointplot(x=\"Inlinks\", y=\"Outlinks\", data=results_df, kind = \"kde\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Identify implicit clusters using Clustering algos"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Typically, number of clusters are chosen with a plot of within-cluster sum-of-squares-of-dstances vs number of clusters. \n",
    "Here for simplicity, we use a simple heuristic to fix the number of clusters in advnace."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "6\n"
     ]
    }
   ],
   "source": [
    "n_clusters = min( int( round(np.sqrt(len(index)/2)) ), 10 ) # not more than 6 clusters\n",
    "print(n_clusters)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3a. K-Means clustering"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "K Means is a point-assignment based clustering-algorithm: here we start with k points chosen randomly as centroids,  assign the remaining points to k centroids by using certain distance measure (Euclidean / Cosine / Jaccardi). Then we compute new centroids, re-assign remaining points and repeat, until there is convergence of centroids. \n",
    "\n",
    "Here we are more interested in clustering inherent in who-follows-me (network-driven) graph, rather than who-do-I-follow (self-driven). So the approach is to represent the nodes as Observations and whether other nodes follow them or not (1 or 0) as Features (One observation per row, and one feature per column) Thus the input matrix must be such that any value in cell implies whether node in the column follows node in the row."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>1 : 40 Ids</th>\n",
       "      <th>2 : 10 Ids</th>\n",
       "      <th>4 : 10 Ids</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>Tesla</td>\n",
       "      <td>Shah Rukh Khan</td>\n",
       "      <td>TensorFlow</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>Richard Branson</td>\n",
       "      <td>Hrithik Roshan</td>\n",
       "      <td>Google AI</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>Marques Brownlee</td>\n",
       "      <td>Leonardo DiCaprio</td>\n",
       "      <td>Andrew Ng</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>Cristiano Ronaldo</td>\n",
       "      <td>PRIYANKA</td>\n",
       "      <td>Ian Goodfellow</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>Harsha Bhogle</td>\n",
       "      <td>Aamir Khan</td>\n",
       "      <td>Kaggle</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>Tom Cruise</td>\n",
       "      <td>Farhan Akhtar</td>\n",
       "      <td>KDnuggets</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>UEFA Champions League</td>\n",
       "      <td>Anushka Sharma</td>\n",
       "      <td>Soumith Chintala</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>IndianPremierLeague</td>\n",
       "      <td>Deepika Padukone</td>\n",
       "      <td>Kirk Borne</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>President Trump</td>\n",
       "      <td>Sonam K Ahuja</td>\n",
       "      <td>Data Science Central</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>Indian Football Team</td>\n",
       "      <td>Alia Bhatt</td>\n",
       "      <td>Open Data Science</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              1 : 40 Ids         2 : 10 Ids            4 : 10 Ids\n",
       "0                  Tesla     Shah Rukh Khan            TensorFlow\n",
       "1        Richard Branson     Hrithik Roshan             Google AI\n",
       "2       Marques Brownlee  Leonardo DiCaprio             Andrew Ng\n",
       "3      Cristiano Ronaldo           PRIYANKA        Ian Goodfellow\n",
       "4          Harsha Bhogle         Aamir Khan                Kaggle\n",
       "5             Tom Cruise      Farhan Akhtar             KDnuggets\n",
       "6  UEFA Champions League     Anushka Sharma      Soumith Chintala\n",
       "7    IndianPremierLeague   Deepika Padukone            Kirk Borne\n",
       "8        President Trump      Sonam K Ahuja  Data Science Central\n",
       "9   Indian Football Team         Alia Bhatt     Open Data Science"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.cluster import KMeans\n",
    "est = KMeans(max_iter = 100000, n_clusters = n_clusters, n_init = 200, init='k-means++')  \n",
    "results_df['kmeans'] = est.fit_predict(binaryMap_cfr)\n",
    "\n",
    "top_nodes_in_cluster(results_df.sort_values( by = 'PageRank', ascending =False), \n",
    "                     'kmeans', n_clusters)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3b. Spectral clustering"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "One problem in using K-Means algorithm to cluster the given social-graph is 'Curse of Dimensionality' i.e. at higher-dimensions (here ~400 dimensions), metric like 'Euclidean distance' or 'Centre' would have little meaning in the context of non-convex adjacency matrix.\n",
    "\n",
    "On the other hand, spectral clustering attempts to partition the graph such that number of edges which connect different components are minimized. Below is the output from Spectral Clustering."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/kuntojirohan/MCG/Twitter/twitter_graph/py36/lib/python3.6/site-packages/sklearn/cluster/spectral.py:462: UserWarning: The spectral clustering API has changed. ``fit``now constructs an affinity matrix from data. To use a custom affinity matrix, set ``affinity=precomputed``.\n",
      "  warnings.warn(\"The spectral clustering API has changed. ``fit``\"\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0 : 14 Ids</th>\n",
       "      <th>1 : 23 Ids</th>\n",
       "      <th>2 : 12 Ids</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>Mahendra Singh Dhoni</td>\n",
       "      <td>Google AI</td>\n",
       "      <td>TensorFlow</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>Marques Brownlee</td>\n",
       "      <td>Richard Branson</td>\n",
       "      <td>Andrew Ng</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>Cristiano Ronaldo</td>\n",
       "      <td>Tom Cruise</td>\n",
       "      <td>Ian Goodfellow</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>Harsha Bhogle</td>\n",
       "      <td>FC Barcelona</td>\n",
       "      <td>Kaggle</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>UEFA Champions League</td>\n",
       "      <td>Virgin</td>\n",
       "      <td>KDnuggets</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>IndianPremierLeague</td>\n",
       "      <td>Entrepreneur</td>\n",
       "      <td>Soumith Chintala</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>Indian Football Team</td>\n",
       "      <td>Rohan Kuntoji</td>\n",
       "      <td>Kirk Borne</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>Sunil Chhetri</td>\n",
       "      <td>Papers with Code</td>\n",
       "      <td>Data Science Central</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>Vir Das</td>\n",
       "      <td>IBM Data Science &amp; AI</td>\n",
       "      <td>Open Data Science</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>ICC</td>\n",
       "      <td>Suhas Raju</td>\n",
       "      <td>Adam Geitgey</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              0 : 14 Ids             1 : 23 Ids            2 : 12 Ids\n",
       "0   Mahendra Singh Dhoni              Google AI            TensorFlow\n",
       "1       Marques Brownlee        Richard Branson             Andrew Ng\n",
       "2      Cristiano Ronaldo             Tom Cruise        Ian Goodfellow\n",
       "3          Harsha Bhogle           FC Barcelona                Kaggle\n",
       "4  UEFA Champions League                 Virgin             KDnuggets\n",
       "5    IndianPremierLeague           Entrepreneur      Soumith Chintala\n",
       "6   Indian Football Team          Rohan Kuntoji            Kirk Borne\n",
       "7          Sunil Chhetri       Papers with Code  Data Science Central\n",
       "8                Vir Das  IBM Data Science & AI     Open Data Science\n",
       "9                    ICC             Suhas Raju          Adam Geitgey"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn import cluster\n",
    "\n",
    "spectral = cluster.SpectralClustering(n_clusters=n_clusters,  n_init = 500,\n",
    "                                          eigen_solver='arpack',\n",
    "                                          affinity=\"nearest_neighbors\")\n",
    "spectral.fit(binaryMap_cfr)\n",
    "results_df['spectral'] = spectral.labels_.astype(np.int)\n",
    "\n",
    "top_nodes_in_cluster(results_df.sort_values( by = 'PageRank', ascending =False), \n",
    "                     'spectral', n_clusters)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nodes mainly into Deep Learning community have grouped into Cluster 8, Python Machine Learning community into Cluster 2, Design community into Cluster 9, general Data Science community into Cluster 0 and 5"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "One smaller clusters (1) wasn't shown above."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PageRank</th>\n",
       "      <th>UserName</th>\n",
       "      <th>Inlinks</th>\n",
       "      <th>Outlinks</th>\n",
       "      <th>Followers</th>\n",
       "      <th>Friends</th>\n",
       "      <th>Location</th>\n",
       "      <th>Created</th>\n",
       "      <th>kmeans</th>\n",
       "      <th>spectral</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FullName</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>Google AI</td>\n",
       "      <td>0.047022</td>\n",
       "      <td>GoogleAI</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>1380565</td>\n",
       "      <td>14</td>\n",
       "      <td>Mountain View, CA</td>\n",
       "      <td>Apr-2009</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Richard Branson</td>\n",
       "      <td>0.024579</td>\n",
       "      <td>richardbranson</td>\n",
       "      <td>6</td>\n",
       "      <td>7</td>\n",
       "      <td>12664546</td>\n",
       "      <td>3893</td>\n",
       "      <td></td>\n",
       "      <td>Aug-2007</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Tom Cruise</td>\n",
       "      <td>0.012091</td>\n",
       "      <td>TomCruise</td>\n",
       "      <td>3</td>\n",
       "      <td>9</td>\n",
       "      <td>6740721</td>\n",
       "      <td>53409</td>\n",
       "      <td>Worldwide</td>\n",
       "      <td>Jun-2009</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>FC Barcelona</td>\n",
       "      <td>0.009043</td>\n",
       "      <td>FCBarcelona</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>30930757</td>\n",
       "      <td>79</td>\n",
       "      <td>Barcelona</td>\n",
       "      <td>Dec-2009</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Entrepreneur</td>\n",
       "      <td>0.005913</td>\n",
       "      <td>Entrepreneur</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>3463102</td>\n",
       "      <td>1454</td>\n",
       "      <td>Irvine, CA</td>\n",
       "      <td>Jan-2009</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 PageRank        UserName  Inlinks  Outlinks  Followers  \\\n",
       "FullName                                                                  \n",
       "Google AI        0.047022        GoogleAI        8         2    1380565   \n",
       "Richard Branson  0.024579  richardbranson        6         7   12664546   \n",
       "Tom Cruise       0.012091       TomCruise        3         9    6740721   \n",
       "FC Barcelona     0.009043     FCBarcelona        3         1   30930757   \n",
       "Entrepreneur     0.005913    Entrepreneur        2         4    3463102   \n",
       "\n",
       "                 Friends           Location   Created  kmeans  spectral  \n",
       "FullName                                                                 \n",
       "Google AI             14  Mountain View, CA  Apr-2009       4         1  \n",
       "Richard Branson     3893                     Aug-2007       1         1  \n",
       "Tom Cruise         53409          Worldwide  Jun-2009       1         1  \n",
       "FC Barcelona          79          Barcelona  Dec-2009       1         1  \n",
       "Entrepreneur        1454         Irvine, CA  Jan-2009       1         1  "
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_df [results_df['spectral'].isin([1])].sort_values(by = 'PageRank', ascending =False).set_index('FullName').head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3c. Affinity Propagation clustering"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Unlike K-Means or Spectral clustering, affinity propagation doesn't require the number of clusters to be estimated beforehand. Here the algorithm finds 'exemplars' i.e. members of dataset that are representative of clusters. and tries to find clusters around them.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2 affinity clusters.\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0 : 61 Ids</th>\n",
       "      <th>1 : 11 Ids</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>TensorFlow</td>\n",
       "      <td>Amitabh Bachchan</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>Google</td>\n",
       "      <td>Sachin Tendulkar</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>Google AI</td>\n",
       "      <td>Narendra Modi</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>Tesla</td>\n",
       "      <td>Virat Kohli</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>Elon Musk</td>\n",
       "      <td>Shah Rukh Khan</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>Bill Gates</td>\n",
       "      <td>Hrithik Roshan</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>Andrew Ng</td>\n",
       "      <td>Virender Sehwag</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>Richard Branson</td>\n",
       "      <td>PRIYANKA</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>Mahendra Singh Dhoni</td>\n",
       "      <td>Aamir Khan</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>Marques Brownlee</td>\n",
       "      <td>Farhan Akhtar</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             0 : 61 Ids        1 : 11 Ids\n",
       "0            TensorFlow  Amitabh Bachchan\n",
       "1                Google  Sachin Tendulkar\n",
       "2             Google AI     Narendra Modi\n",
       "3                 Tesla       Virat Kohli\n",
       "4             Elon Musk    Shah Rukh Khan\n",
       "5            Bill Gates    Hrithik Roshan\n",
       "6             Andrew Ng   Virender Sehwag\n",
       "7       Richard Branson          PRIYANKA\n",
       "8  Mahendra Singh Dhoni        Aamir Khan\n",
       "9      Marques Brownlee     Farhan Akhtar"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.cluster import AffinityPropagation\n",
    "\n",
    "af = AffinityPropagation(preference=-50).fit(binaryMap_cfr)\n",
    "results_df['affinity'] = af.labels_\n",
    "n_clusters_affinity = len(af.cluster_centers_indices_)\n",
    "\n",
    "print(str(n_clusters_affinity) + \" affinity clusters.\")\n",
    "\n",
    "top_nodes_in_cluster(results_df.sort_values( by = 'PageRank', ascending =False), \n",
    "                     'affinity', n_clusters_affinity)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3d. Principal Component Analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To handle the 'curse of dimensionality' problem inherent in high-dimension data, PCA is generally used to represent the high-dimensional data in a fewer number of dimensions - this helps in better visualizing of data as well as faster computation while running clustering algorithm."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained-variance and Proportion of Explained-variance in 3 dimensions [dim1 dim2 dim3]\n",
      "[1.61749939 0.48646669 0.41671062] [0.2902709  0.08729965 0.07478146]\n"
     ]
    }
   ],
   "source": [
    "pca = PCA(n_components=3)\n",
    "Xproj = pca.fit_transform(binaryMap_cfr)\n",
    "\n",
    "results_df['dim1'] = Xproj[:,0]\n",
    "results_df['dim2'] = Xproj[:,1]\n",
    "results_df['dim3'] = Xproj[:,2]\n",
    "\n",
    "results_df = results_df.sort_values( by = 'PageRank', ascending =False)\n",
    "results_df.to_csv(cluster_filename)\n",
    "\n",
    "print(\"Explained-variance and Proportion of Explained-variance in 3 dimensions [dim1 dim2 dim3]\")\n",
    "print(pca.explained_variance_, pca.explained_variance_ratio_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A simpler visualization of Spectral-clustering outcome as rendered in 2 dimensions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABIAAAAKECAYAAAB/6UF2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAAMTQAADE0B0s6tTgAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOzdeZSlZX0v+u+7966xJ3qeoGloRsVWkGgEjsHoPSKKKGpIJNyTiBJEOHEZj8blSbK8ZJlzOHowMebktiFGPWo8XhxwNhhUUDQyKYoINFM33fRMD9U17v3eP1o7tk0PdHXXrnrr81mLRdV+n72fX1X9umrVt57neYuyLMsAAAAAUFm1dhcAAAAAwJElAAIAAACoOAEQAAAAQMUJgAAAAAAqTgAEAAAAUHECIAAAAICKEwABAAAAVFyj3QUAAAAAcPDe9a535Qc/+EGmTZuWJDn22GPzN3/zN/t9jgAIAAAAYAK544478vd///c56aSTDvo5RVmW5RGsCQAAAID92LhxY9atW/eU1+bOnZt58+btMfbcc8/Nueeem0cffTRLly7Nn/7pn2bx4sX7ncMKIAAAAIAkrVYrtdrYH5c8Y8aMvPa1r83atWv3unbVVVfl6quv3v3++vXrc8455+Rd73pXFi1alH/4h3/IFVdckc9//vOp1+v7nGPCrQDatq0/zWar3WXQRjNnTsmWLX3tLgMOGz1NFelrqkhfU0X6mqdSr9cyfXpPu8tomyeeWJ+hoeExm6+zsyMLFszLypUrMzAwsNf1X18B9OvKssxzn/vcfOYzn8myZcv2OW7CrQBqNlsZGREATVZFsev/zWYrEyu6hKemp6kifU0V6WuqSF/DUxsaGs7g4NCYz7u/8OZX3XXXXVm3bl3OO++83Y+VZZlGY/8Rj9vAAwAAAEwQQ0NDueaaa7J+/fokycc//vEsW7YsS5Ys2e/zJtwKIAAAAIDJ6vnPf36uvPLK/OEf/mGazWYWLVqUv/7rv07xy2V9+yAAAgAAAJhALrnkklxyySVP6zm2gAEAAABUnAAIAAAAoOIEQAAAAAAVJwACAAAAqDgBEAAAAEDFCYAAAAAAKk4ABAAAAFBxAiAAAACAihMAAQAAAFScAAgAAACg4hqH40U+9alP5ZOf/GSKokhPT0/e/e53Z/ny5XuM+fznP5+/+qu/yoIFC3Y/9vGPfzzTp08/HCUAAAAAsA+jDoDuvPPOrFixIjfccENmzZqVm2++OVdeeWVuueWWFEWxe9ztt9+eq6++Or//+78/2ikBAAAAeBpGHQDNmDEj11xzTWbNmpUkWb58eTZt2pT+/v709vbuHnfHHXdk7dq1+exnP5vu7u689a1vzfOe97xDmvNXciUmmV9+7fUAVaGnqSJ9TRXpa6pIX8PkMuoAaNmyZVm2bFmSpNVq5b3vfW/OPffcPcKfoaGhLFq0KJdddlnOOuus/PCHP8yVV16Zz372sznmmGOe1nwzZ04ZbclUwOzZ09pdAhxWepoq0tdUkb6mivQ1TA5FWZbl4XihHTt25B3veEc2b96cFStWHPBsnyuuuCIvfOEL8/rXv/5pzbNlS1+azdZoSmUCK4pdP6A2bdqew9O50F56mirS11SRvqaK9DX7Uq/XJvXii8ceezyDg0NjNl9XV2eWLFl8xOc5LIdAP/zww3nzm9+c5cuX57rrrktXV9ce19etW5cvf/nLecMb3rD7sbIs09HRcUjz+eZEWeoDqkVPU0X6mirS11SRvobJYdS3gV+zZk0uueSSvO51r8u11167V/iTJL29vfm7v/u73HbbbUmSe+65J3fffXde9KIXjXZ6AAAAAA5g1CuArr/++mzbti033nhjbrzxxt2Pr1ixIpdffnlWrFiR+fPn50Mf+lCuvfbaDA4OptFo5AMf+EDmzJkz2ukBRqUsy2T1qrRu/2HyxJqkVk9OPCm1089IMXNWu8sDAAA4LA7bGUBjZcuWvoyMOANosiqKZM6cadm40T5lRq/csSOtT34s5d13JbVa0tOza/3zzp1JvZ7iZS9P7byXp6iNerHkPulpqkhfU0X6mirS1+xLo+EMIGcAAVREOTCQ1j/8fcqf35csOTZFvb7X9fILn0treDi1V746hfujAgAAE9iR+7M2wDjW+rfvp/zZvcnS4/YKf5Kk6O5Ojjkm5de/ljy+ug0VAgAAHD4CIGDSKZvN5F9vSmbN2u/KnqKzK6kVaf3gtjGsDgAA4PATAAGTz8YNKdevT46aeeCxM2cmd9155GsCAAA4ggRAwOQzPJwUObhzfeqNZGjsDoADAAA4EgRAwOQzdWqSpBwZOfDYgf5dq4AAAAAmMAEQMOkUR81M8axnJxvWH3jwtq3JOS888kUBAAAcQQIgYFKq/daLksGBlAMD+xxTbt6UTJmW2nNOH8PKAAAADj8BEDA5nXJqildelDy+OuXGjSnLcvelcmQk5eOrk4GB1N90RYpp09tYKAAAwOg12l0AQDsURZHaS1+Wcu68lF//SspHH0lZK5KySFKmeNazU3vZy1Mcu7TdpQIAAIyaAAiYtIqiSPHcM1Oefkby6CMpn9ySolZLFixKMX9+u8sDAAA4bARAwKRX1GrJccfnIG4KDwAAMCE5AwgAAACg4gRAAAAAABUnAAIAAACoOAEQAAAAQMUJgAAAAAAqTgAEAAAAUHECIAAAAICKEwABAAAAVJwACAAAAKDiBEAAAAAAFScAAgAAAKg4ARAAAABAxQmAAAAAACqu0e4CAADYv3LL5rTuujO5/75kZCRZsDC1M5+XHLs0RVG0uzwAYAIQAAEAjFNlq5XW17+a8itfTJrNpLc3KWrJfT9L85v/kuK05an933+QYtr0dpcKAIxzAiAAgHGq9ZUvpfzi55PFR6fo7t7jWtlqpbz3J2n9/YdSe8sfp+jtbVOVAMBE4AwgAIBxqFy7JuVXv5Qcfcxe4U+SFLVasuTYlA8+mNb3v9eGCgGAiUQABAAwDrX+7ftJkqKra59jiqJIZs9Obv5mypGRsSoNAJiABEAAAOPR3XclR8088LgZM1Ju3pise+LI1wQATFgCIACA8WhwMGnUDzisKIpdB0NbAQQA7IcACABgPJo1K+nvP+CwXVu/ymTK1CNfEwAwYQmAAADGoeKcFyY7th944MYNKU55Roo5c458UQDAhCUAAgAYh4pnPTvF9KNSbtiwzzHl4EDSvzO1c188doUBABNSo90FAACwt2LKlNQuvyLNv/2blKtXJfMXpOjoSJKUZZls3pQ8+WSKC1+dPPO0NlcLANVR/ujulJs3j918s2YlSxYf8XkEQAAA41Rx/Amp/8k70vrql1PefWfKJEmZtMoURx+T4nden+KM5+46CBoAOCymrHwoXY+vGbP5GosXjc08YzILAACHpFh8dOpv/KOUGzcmax9P2WymmHFUcuzSFDW7+QGAgyMAAgCYAIo5c5I5c2KtDwBwKPzZCAAAAKDiBEAAAAAAFScAAgAAAKg4ARAAAABAxQmAAAAAACpOAAQAAABQcQIgAAAAgIoTAAEAAABUnAAIAAAAoOIEQAAAAAAVJwACAAAAqDgBEAAAAEDFCYAAAAAAKk4ABAAAAFBxAiAAAACAihMAAQAAAFScAAgAAACg4gRAAAAAABUnAAIAAACoOAEQAAAAQMUJgAAAAAAqTgAEAAAAUHECIAAAAICKEwABAAAAVJwACAAAAKDiBEAAAAAAFScAAgAAAKg4ARAAAABAxQmAAAAAACpOAAQAAABQcQIgAAAAgIoTAAEAAABUnAAIAAAAoOIEQAAAAAAVJwACAAAAqDgBEAAAAEDFCYAAAAAAKk4ABAAAAFBxow6APvWpT+WCCy7IK1/5ylx88cX58Y9/vNeYVatW5dJLL83LX/7yXHDBBbnrrrtGOy0AAAAAB6kxmiffeeedWbFiRW644YbMmjUrN998c6688srccsstKYpi97i3ve1tee1rX5uLL744P/nJT3LFFVfkG9/4Rnp7e0f9AQAAAACwf6MKgGbMmJFrrrkms2bNSpIsX748mzZtSn9//+5wZ926dbnvvvty0UUXJUlOO+20LF26NN/61rdy/vnnH9K8v5ItMcn88muvB6gKPU0V6WuqSF9TRfoaJpdRBUDLli3LsmXLkiStVivvfe97c+655+6xsmft2rWZM2dOOjo6dj82f/78rFmz5pDmnDlzymhKpiJmz57W7hLgsNLTVJG+por0NVWkr2FyGFUA9Es7duzIO97xjmzevDkrVqzY41pZlntsB/ulWu3Qjh/asqUvzWbrkJ7LxFcUu35Abdq0PWXZ7mpg9PQ0VaSvqSJ9TRXpa/alXq9ZfFFBow6AHn744bz5zW/O8uXLc91116Wrq2uP64sWLcrGjRszMjKSRmPXdOvXr89LXvKSQ57TNyfKUh9QLXqaKtLXVJG+por0NUwOo7oL2Jo1a3LJJZfkda97Xa699tq9wp9k13avU045JZ/73OeSJPfee28eeOCBnHXWWaOZGgAAAICDNKoVQNdff322bduWG2+8MTfeeOPux1esWJHLL788K1asyPz58/P+978/f/Znf5aPfexjSZL/+T//Z2bMmDG6ygEAAAA4KEVZTqzFflu29GVkxBlAk1VRJHPmTMvGjfYpUw16mirS11SRvqaK9DX70mhM7jOANn7ggxl5/NBuXHUoGosXZc5brz7i84xqCxgAAAAA458ACAAAAKDiBEAAAAAAFScAAgAAAKg4ARAAAABAxQmAAAAAACpOAAQAAABQcQIgAAAAgAnoRz/6UU477bQ88cQTBxwrAAIAAACYYDZt2pS/+Iu/yPDw8EGNbxzhegAAAAA4gJUrV2ZgYGCvx+fOnZt58+bt8djIyEje9ra35b/8l/+SN7zhDQf1+gIgAAAAgF9o7Lgrxdb7x2y++oyTkiRvf/vbc++99+51/aqrrsrVV1+9x2PXXnttnv/85+fss88+6HkEQAAAAAC/0HXR6Wn1zx+z+Wo9i5Ik73vf+/a5AuhXfelLX8pjjz2Wd73rXU9rHgEQAAAAQJstW7bsoMbdcMMNWbduXV71qlftfuyyyy7Le97znpx55pn7fJ4ACAAAAGCC+MhHPrLH+yeffHKuv/76LFiwYL/PcxcwAAAAgIqzAggAAABggvr5z39+UOOsAAIAAACoOAEQAAAAQMUJgAAAAAAqTgAEAAAAUHEOgQbGhXJoKOW9P0l563eSNY8ntXpy0smpnX1OctyyFDV5NQAAwKESAAFtV65fn9aH/1fK1auS3t5k6tSkLJPb/y3N276b4jeen9rrL03R3d3uUgEAACYkARDQVuW2rWn9rw+m3LQxWXpciqL494tTp6VsNlP+4La0ylZqf/gmK4EAAAAOgd+kgLZq3fa9lGvXJIuP3jP8+YWiXk+WHpfy9h8mKx9sQ4UAAAATnwAIaJtyaCj51jeTOXOfMvz5paJeTzo60vruLWNYHQAAQHUIgID22bwp5datybRpBx47c1by858d+ZoAAAAqSAAEtE+rlRTFflf/7FYUKUdGjnxNAAAAFSQAAtpn+oyk0Ug5OHDgsX07UixYeORrAgAAqCABENA2xdSpKZ73m8n69QcevGNHiheee8RrAgAAqCIBENBWtV+EOuW2bfscU65dk2LBwhSnLR+jqgAAAKpFAAS0VXHMktTe8KZk65MpVz2Wcmhw97Wyb0fKhx9KMWVqape/OUVPTxsrBQAAmLga7S4AoHb6c1O8fXZa3/lWyh/+IGWrlZRl0tub4uWvTO3sc1LMnNXuMgEAACYsARAwLhTHLk390j9IeeGrk61bk1otmT0nRXd3u0sDAACY8ARAwLhSTJ+x6+5gAAAAHDbOAAIAAACoOAEQAAAAQMUJgAAAAAAqTgAEAAAAUHECIAAAAICKEwABAAAAVJwACAAAAKDiBEAAAAAAFScAAgAAAKg4ARAAAABAxQmAAAAAACpOAAQAAABQcQIgAAAAgIoTAAEAAABUnAAIAAAAoOIEQAAAAAAVJwACAAAAqDgBEAAAAEDFCYAAAAAAKk4ABAAAAFBxAiAAAACAihMAAQAAAFScAAgAAACg4gRAAAAAABUnAAIAAACoOAEQAAAAQMUJgAAAAAAqTgAEAAAAUHECIAAAAICKEwABAAAAVJwACAAAAKDiBEAAAAAAFScAAgAAAKg4ARAAAABAxQmAAAAAACpOAAQAAABQcY12FwAAAAAwXgw27kqzcf+YzVdvnJSeMZhHAAQAAADwC9sWnZ7B1vwxm6+rtihHjcE8AqAxtH5gQ1b3r06zbGV6Y1qWTT0+jZovAQAAAHBkSR/GwOP9a/KlNV/NPVvvScqkLJKybGV25+z8xwUvztlzzkqtcBwTAAAAcGQIgI6wh3Y8nL978P/NUGs4C7sXpeMXK37KsszW4a3534/+c9YOrMtrj361EAgAAAA4IiQOR9BAcyDXP/yRFClydO/i3eFPkhRFkaM6j8qS3mPyzXU3584td7exUgAAAKDKBEBH0E+23pvNQ1syt3vuPsd01DoyvWN6vrnu5pRlOYbVAQAAAJPFYQuAPvCBD+Td7373U1774Q9/mDPOOCMXXnjh7v8efPDBwzX1uPX9TT/IlPqUA46b3Tkrj+58LGsHnhiDqgAAAIDJZtRnAK1evTp/9Vd/lVtvvTWveMUrnnLM7bffnosvvjjvfOc7RzvdhLJlaEu66t0HHFcraqkVRfpG+sagKgAAAGCyGXUA9OlPfzpnnXVWTjzxxGzYsOEpx9xxxx0ZHBzMa17zmhRFkTe+8Y0577zzDnnOojjkp46prnpXdjZ35kDllmWZsmylo96YMB9bu/zy8+PzRFXoaapIX1NF+poq0tcwuYw6APqTP/mTJMkHP/jBfY6ZNm1aLrjggt1bvy699NLMnz8/p59++tOeb+bMA2+pGi/OPfYF+eeVn01vb+d+x20b2pZ50+fk2UefnM76/seyy+zZ09pdAhxWepoq0tdUkb6mivQ1TA5jchv46667bvfbJ5xwQs4///zcdNNNhxQAbdnSl2azdTjLO2JObJya4aFmNm7bmt5Gz1OOKcsyj+5ck1cvfmW2bRlMMji2RU4wRbHrB9SmTdvjzGyqQE9TRfqaKtLXVJG+Zl/q9dqEWnzBwTniAVBfX18++tGP5vLLL0+jsWu6six3v30oJso3p5kdM3PR4lfl06v+v8zvnpcpjT3/ATXLZlbvXJ3jpxyfc+acM2E+rvGgLCdOH8DB0NNUkb6mivQ1VaSvYXI44gFQb29vbrjhhsybNy+vfe1rs2rVqnzta1/LP/7jPx7pqceFc+e9MI1aPZ9d/YWsH9iQ7np36kUtA82BtFLmmTOekUuP/b1MafS2u1QAAACgoo5YAHThhRfmL//yL/OsZz0rH/rQh3LNNdfkox/9aJrNZv7rf/2vOeWUU47U1ONKURT5D3PPyekzT8+Pn7wnP99+f4Zbw5nfPS9nzDw9R/csTuHUNQAAAOAIKspyYi3227KlLyMjE+MMIA6/okjmzJmWjRvtU6Ya9DRVpK+pIn1NFelr9qXRmNxnAD2244MZbK0Zs/m6aouyZOrVR3ye2hGfAQAAAIC2EgABAAAAVJwACAAAAKDiBEAAAAAAFScAAgAAAKg4ARAAAABAxQmAAAAAACpOAAQAAABQcQIgAAAAgIoTAAEAAABUnAAIAAAAoOIEQAAAAAAVJwACAAAAqDgBEAAAAEDFCYAAAAAAKk4ABAAAAFBxAiAAAACAihMAAQAAAFScAAgAAACg4gRAAAAAABUnAAIAAACoOAEQAAAAQMUJgAAAAAAqTgAEAAAAUHECIAAAAICKEwABAAAAVJwACAAAAKDiBEAAAAAAFScAAgAAAKg4ARAAAABAxQmAAAAAACpOAAQcVuXQUModO1KOjLS7FAAAAH6h0e4CgImvLMvk/p+ndet3Uv7o7qQ5knR1pXj+Wam94KwUS45td4kAAACTmgAIGJWy2Uzrhv+T8uZvJp2dydy5SUdHMjiY8pZvpXnLt1K89uLUfutFKYqi3eUCAABMSgIgYFRaX/9Kypu+kRy7NEVHx79f6OlJlhybcufOlP/8iZTTp6c448z2FQoAADCJOQMIOGTl9m0pv/G1ZPHRe4Y/v6Lo7U2OOirll25M2WyOcYUAAAAkAiBgFFp335UMDqbo6dn/wJmzUq5dk6x8cGwKAwAAYA8CIODQPb5617k/B1AURVKrpdy4YQyKAgAA4NcJgIAxUra7AAAAgElLAAQcusVHJ8PDBxxWlmXSKlPMnjMGRQEAAPDrBEDAIas95/SkoyPlwMD+B27ZkmL+guSEE8emMAAAAPYgADpCtg9vz+2b78y319+S2zZ+P+sG1rW7JDjsimnTU7zkpcnqVSn3sRKo7O9PntyS4hUXpqjXx7hCAAAAkqTR7gKqZudIf7609iv57sbvZag5nKJIkiJFipw24xm56OgLM797frvLhMOmdv4r0trZl/Jb/5qyszOZOy/p6EgGBpKNG3Zt/fqd30vx3DPbXSoAAEAl3HDDDfmnf/qnJMnMmTPznve8J8cdd9x+nyMAOox2jvRnxcoP577tD2RRz8J017t3XxtpjeRn236e6+7/YP7ziW/Jop6FbawUDp+iXk/tdb+bLH9OWrd+J+U9P0pGRpKurhRnn5PaC85JcezSdpcJAABQCQ899FDe//7354tf/GJmz56dj3/84/nzP//zfPzjH9/v8wRAh9HXn/hG7tv+QI6bsnTXba9/RaPWyNG9i7Omf20+9sgn8o5T3pZaYQce1VDUasmpz0j91GekHBxMhoeSru4UHR3tLg0AAGBCWLlyZQae4nzVuXPnZt68ebvfP/744/Ptb387HR0dGRkZyZo1azJz5swDvr4A6DDpG9mZWzZ+Lwu6F+wV/vyqhd0L8kjfo1m546GcOO2EMawQxkbR1ZV0dbW7DAAAgEPyo/KubG7dP2bzzSpOypIkb3/723Pvvffudf2qq67K1VdfvcdjHR0duf322/PHf/zH6e/vz/XXX3/AeQRAh8nKHSsz2BxI7wHO9ymKIvWinh89eY8ACAAAAMaZlQOn5/HhsTu7d3HHomRa8r73vW+fK4Ceyplnnpnvfve7uemmm3L55Zfnm9/8ZqZPn77PeQRAh8nOZn+yn5U/v6qr3pltw9uOcEUAAADARLFs2bKDGrd69eqsWrUqL3jBC5IkL3nJS/Lnf/7neeyxx3Laaaft83kOoTlMumqdSVke1Njh1nCmNHqPcEUAAABA1WzdujVvfetb88QTTyRJvv3tb6dWqx0wQLIC6DA5fupxqReNDDYH01Xf9/knZVlmuDWcZ8545hhWBwAAAFTBM5/5zLzzne/Mm970ptRqtUyfPj0f/vCH09PTs9/nCYAOkxkdM/L82b+R7274XpZOXbrPcRuHNmVu19ycMu2ksSsOAAAAqIyLLrooF1100dN6ji1gh9HLF56XBT0LsmrnqjTL5h7XyrLMhoENGWoN5dKlr0+jJnsDAAAAxoYU4jA6qvOoXHXim/PJR/859267L0mZjlpnmq1mmmlmUffC/N6S38kJ0w7uYCcAAACAw0EAdJjN6pyZt5xwRdYMrM1Pnvxpnhx+Mt31npw87cScOO2E1It6u0sEAAAAJhkB0BFQFEUW9yzK4p5F7S4FAAAAwBlAAAAAAFUnAAIAAACoOAEQAAAAQMUJgAAAAAAqziHQFbJzpD/rBtelVbYyrTEt87rntrskAAAAYBwQAFXAlqEt+fb6W3LLxu9loDmw+/FTp5+cF89/UU6dfkobqwMAAADaTQA0wa0bWJe/feDvs3FwY+Z3z8/87nlJkpHWSB7qezj3PvCzvO7o1+RF83+rzZUCAAAA7SIAmsCGW8P58EMfyfaR7Tlu6nF7XGvUGlnQvSD9zf58ZvUNWdAz30ogAAAAmKQcAj2B/WzbfXl85+NZ2L1wn2N66j3pqnXnX9d9a+wKAwAAAMYVAdAE9r2N3093vSdFUex33NyuOfnZtvuyfmDDGFUGAAAAjCcCoAls3cD6TGlMOeC4Rq2RpMj2ke1HvigAAABg3BEATWCNWiOtsnlQY8uUqRW+3AAAADAZSQQmsGdOPzVbh7cdcFzfSF+6691Z0D1/DKoCAAAAxhsB0AT2vNm/kaTMUGt4v+PWD27IC+eenZ56z9gUBgAAAIwrbgM/wYy0RnL/jgfy060/y/bh7ZnamJp7t96bZ0w/NZ31zj3GlmWZNQNrM7dzTl4495w2VQwAAAC0mwBoAnlox8P5349+Mmv7n0ij1pGOWiMDI4PZOrwt/7Lumzl+6nFZ2L0gtaKWvpGdGWj15+iexXnj8W/IzM6Z7S4fAAAAaBMB0ATxcN8j+eADf5d6Uc/SKUv3uPX7kt5j8nDfw9k89GTSnXTWOnP0tMU5e84Lcsr0k9NR62hj5QAAAEC7CYAmgFbZyqce/XRqRS3zuuftdb1Wq2XZtGXp2vl4FvYsyFUnvHmPgAgAAACY3BwCPQE8tOPhrO5/PPO69g5/ftWC7vm5b9vP88TAE2NUGQAAADARCIAmgPt3PJCkOOCqnkatkbIss3LHw6OaryzL9I3szNbhrRlqDY3qtQAAAID2swVsAtg50p9GUT+osWWSofLQQptm2cxPt96bb6+/JT/ffn9aaaWr1p3fnP28nDXnN3NM79GH9LoAAABAewmAJoAZndMz0ho5qLFFkfTWe5/2HIPNwXzi0U/lB5tvz5TGlCzqWZR6Uc9AazC3bLw1t278bn53ye/k7DkveNqvDQAAALTXYdsC9oEPfCDvfve7n/Lali1bcsUVV+T888/PS1/60tx0002Ha9pJ4bTpz0hRFAcMgQabg2kUjZw6/eSnPcdnV38+/7b59iydcmwWdM9Po9ZIURTpqXdnSe+SzOqclf/96Kfy060/O9QPAwAAAGiTUQdAq1evzlve8pZ85CMf2eeY97znPTnppJPyla98Jddff33+4i/+ImvXrh3t1JPGwp6FOW3GM7Omf03KsnzKMWVZZs3A2jx/9vMyo2PG03r9dQPr891Nt+WY3mNS38dWsymNKZlSn5Ivr/3qPmsAAAAAxtuaM5AAACAASURBVKdRbwH79Kc/nbPOOisnnnhiNmzYsNf1kZGR3HzzzfnqV7+aJDn66KNzzjnn5Itf/GIuv/zyQ5pzMt7h/PeOfV0++MD/yqM7H82i7oXpqnftvtbfHMia/rU5YerxedXRFzztz8+dT96VMmU6ax37HTe3a3Ye6Xskq/pX5dgpSw7lwxi1X35sk7EHqCY9TRXpa6pIX1NF+homl1EHQH/yJ3+SJPngBz/4lNe3bNmSgYGBLFiwYPdjCxYsOOQVQDNnTjmk5010czIt/8/sd+aGh2/Md9d9P5ubzRRlUhZlumpdefUJL8url74iUzqe/udnyxMbMqt3enp7Ow84tqfVlWbPQObMmXYoH8ZhM3t2e+eHw01PU0X6mirS11SRvobJ4YgfAv3L7UK/fgvzWu3Qdp9t2dKXZrM16rompnounPPqnDv9t/Pgjocy2BpMT707J049IVM7pqZ/ayv92f60X7WvbzD9A8PZmQPfPWxwcCRbt+7MxtrTn+dwKIpdP6A2bdoeO9GoAj1NFelrqkhfU0X6mn2p12uTdvFFlR3xAGj27Nnp6urK+vXrM3/+/CTJunXrcsIJJxzya072b07TO2bkjJmn7/HYaD4nS3qW5K4tP86BXqJZNtMqW5nTObftX4Oy1AdUi56mivQ1VaSvqSJ9DZPDYbsL2L7U6/W8+MUvzic/+ckkyeOPP55bbrklL3nJS4701Byk5846I7WiyFBr/yuANg5uzIlTl2Vxz6IxqgwAAAA4HI5YAHThhRfmnnvuSZL82Z/9WVauXJlXvOIVueyyy/Knf/qnWbp06ZGamqdpTtfsnDv3t7Kqb9U+bzW/fXh7+psDOX/heXtt5wMAAADGt8O2Bezqq6/e4/0vfOELu9+eNWtW/vZv//ZwTcURcOHiV2SgOZDvbrotXbWuzO6clUatkYHmQDYObkyj1pHLjvuDnDz9pHaXCgAAADxNR/wMICaGjlpHXn/sxXnurNPz3Q235Z6tP0mzbKa30ZuXLvi/8vzZv5H53fPbXSYAAABwCARA7FYrajl1+ik5dfopGWmNZLgcTletK7XiiB8VBQAAABxBAiCeUqPWSEN7AAAAQCVY2gEAAABQcQIgAAAAgIoTAAEAAABUnAAIAAAAoOKc8jsObR7aks2Dm1MUReZ0zc6MjhntLgkAAACYwARA48gjfY/mX9fdnDuf/FHKskyy69bsz5t1Zn57/rlZ3LOozRUCAAAAE5EAaJy458mf5MMP/WOKFFnQvSCdtY4kyVBrKP+2+fbcueWuXLHsTTl5+kltrhQAAACYaJwBNA6sG1iX6x/+p0zrmJbFvYt3hz9J0lnrzDG9R6er3pV/eOgj2Ty0pY2VAgAAABORAGgcuG3jDzLcGtnvWT+zOmelr7kzd2y+cwwrAwAAAKpAANRmw63h3Lrxe5nTNeeAY2d2zsy3Nnxn9/lAAAAAAAdDANRmO5s7s7O5M731ngOOnVLvzbbhbRlsDY5BZQAAAEBVCIDarJ56kiJlDryqp5VWitTSKJzdDQAAABw8AVCbTWlMybG9S7JpaPMBx24a3JyTpp2QRk0ABAAAABw8AVCbFUWR355/brYNb9vv2T6tspWBZn9+a+5/GMPqAAAAgCoQAI0Dy2eclmdMPyWP9j2aVtna63qzbOaRvkfz7KOenVOnn9KGCgEAAICJzF6icaCr3pXLjvuDfPTRT+SnW3+aRtHI1Ma0JMn2kW1pls2cOeuMvH7J79r+BQAAADxt0oRxYmrH1Lx52ZvycN8juW3j9/PozlWpFUWWz3hmnj/neVnae2yKomh3mQAAAFBpdz/549zf/8CYzXdSz4nJ3CM/jwBoHKkVtSybenyWTT2+3aUAAADApPSc7Wdmbt/RYzbf4pEFYzKPM4AAAAAAKk4ABAAAAFBxAiAAAACAihMAAQAAAFScAAgAAACg4gRAAAAAABUnAAIAAACoOAEQAAAAQMUJgAAAAAAqTgAEAAAAUHECIAAAAICKEwABAAAAVJwACAAAAKDiBEAAAAAAFScAAgAAAKg4ARAAAABAxTXaXQAAo1f29yfDQ0l3T4rOznaXAwAAjDMCIIAJqmy1kp/dm9Yt307503uSkZGkqyvF81+Q2gvOTrH0uHaXCAAAjBMCIIAJqBwZSev/fCrlLd9KurqTBQuTRiMZHEz5vVvTvPU7KV7zO6m96MUpiqLd5QIAAG0mAAKYgFpfvjHlt29OlhyboqPj3y90dyfHLEnZ35/y059MOeOoFM89s32FAgAA44JDoAEmmHLrkym/+S/J4qP3DH9+RdHTk8yclfLLN6ZsNse4QgAAYLwRAAFMMK0f3ZWMjKTo7t7/wJkzU65dk6x8cGwKAwAAxi0BEMBE8/jq5CDu9FUURVLUUm7aOAZFAQAA45kACKDKnP8MAABEAAQw8Ry9JBkaOuCwsiyTVivF3HljUBQAADCeCYAAJpjac05POjpS9vfvf+DmTSkWL06OXzY2hQEAAOOWAAhggimmTU/xH1+WrHk85T5WApV9fcnWrSlecWGKmm/1AAAw2TXaXQDj347+Mvc/NpK+gTIdjSLHLaxn/iy/UEI71c47P62dfSn/9aaUnZ3JnDlJoyMZHEw2bkiKIrVL/lNqzzmj3aUCAADjgACIfeofLPPV2wbznR8NZ2i4TFImKVIkWX5iI6/6D12ZP6ve5iphcirq9dRee3Hy7NPT+u53Ut6969bw6e5Jce5vp/abZ6U4+ph2lwkAAIwTAiCe0sBQmQ9/oT/3PjKShXNq6en69xU/wyNlfrJyJI+saeY/v643C+cIgaAdiqJITjo59ZNOTjk8nAwPJ11dKer+TQIAAHuyj4en9I1/G8xPHxnJcYtq6ena8z7SHY0ix8yvp3+ozD99dSCtVtmmKoFfKjo6UvT2Cn8AAICnJABiLzsHynznruEsmF3btcJgHxbOruWxJ5pZ+XhzDKsDAAAAni4BEHtZ+fhI+gfLTOned/iT7Np+0qgndz0wMkaVAQAAAIdCAMRedg4m2X/2s1tnR5GtO2wBAwAAgPFMAMReujqS8iAznZFmmSndR7YeAAAAYHQEQOzluIX1NOrJ4PD+U6CyLDM0nJx2vJvJAQAAwHgmAGIvM6bW8vxndmTthv0f7rx5W5lZM2o5dakACAAAAMYzARBP6eUv6Mr8WfWsWtdM89du816WZTY+2crOgTKXvrQ7HY2DPDAIAAAAaAtLN3hKR02r5erX9eYTXx/Izx4dSVEkHY1kpJk0m8ncmbVcdkFPTl6ihQAAAGC889s7+zRrei1XvbYnqze0cs+DI9myvZWeriKnLG3kpGPqadSt/AEAAICJQADEfhVFkWPm1XPMvHq7SwEAAAAOkTOAAAAAACpOAAQAAABQcQIgAAAAgIoTAAEAAABUnAAIAAAAoOLcBQyAfSoHB1P+9J6Ud9+V9PUl06endsZzk1OekaKjo93lAQAAB0kABMBTKu/7WVofvT7lk08m3d1JR0cyNJTm97+XYv781P7TZSmOO77dZQIAAAfBFjAA9lLe//M0P/TXKcsyxXHHp1i4KMWcuSkWLU6WHpdyx440P/iBlI892u5SAQCAgyAAAmAPZbOZ1j9/IunuTjFr9l7Xi6JIMW9+0mqmdcP/SVmWbagSAAB4OgRAAOzpgftTrn08mTN3/+PmL0j5wP3J6lVjUxcAAHDIBEAA7KH14P1JrZ6iKPY7rqjXkyTlIw+PRVkAAMAoCIAA2NPgYPKLcOeAyjIZHj6y9QAAAKMmAAJgTzNnJSMjBze2KJKp045sPQAAwKgJgADYQ+2Zz0pqRcoDrOwpBwaSzs4Up546RpUBAACHSgAEwB6K+fNTnP7c5PHV+7zDV1mWyZrHU5z9whTTpo9xhQAAMLl96lOfygUXXJBXvvKVufjii/PjH//4gM9pHI6Jb7nllrzvfe/L4OBgFi5cmP/+3/975s2bt8eYH/7wh/mjP/qjHHPMMbsfe//7358TTjjhcJQAwGFU+53fS2vjxpQPP5RywcIUvb27r5U7tifr1qV45rNSe8Ur21glAABMPnfeeWdWrFiRG264IbNmzcrNN9+cK6+8Mrfccst+b+Qy6gBo8+bNefvb356PfexjOfnkk/Oxj30s73rXu3L99dfvMe7222/PxRdfnHe+852jnRKAI6yYNj21t/xxWv/y9ZS3fjvlhvVJkaRVJtOmpXjVRam96CUpurvbXSoAAEwqM2bMyDXXXJNZs2YlSZYvX55Nmzalv78/vb/yh9tfN+oA6NZbb83JJ5+ck08+OUnyu7/7u7n22muzYcOGzJ07d/e4O+64I4ODg3nNa16Toijyxje+Meedd94hzXmAOxNTYb/82usBqmI893QxbWpqF70m5XkvS/nwQ8nQYNLdk+K44wU/7Nd47ms4VPqaKtLX8NTu3taX+7duG7P5TmruuqnKypUrMzAwsNf1uXPn7rHLatmyZVm2bFmSpNVq5b3vfW/OPffc/YY/yWEIgJ544oksXLhw9/udnZ2ZOXNm1q5du0cANG3atFxwwQW58MIL8+CDD+bSSy/N/Pnzc/rppz+t+WbOnDLakqmA2bPddYhqGd89PS1ZMr/dRTABje++hkOjryemVquV/v7+bNu2LcO/uMlBT09Ppk2blq6urv1umZgM9DXs6TmzXp25XXsHMUfK4im7/rj69re/Pffee+9e16+66qpcffXVez2+Y8eOvOMd78jmzZuzYsWKA84z6gCoLMun/IZZq+15vvR11123++0TTjgh559/fm666aanHQBt2dKXZrN1aMUy4RXFrh9QmzZtzz7OpoUJRU9TRfqaKtLXE9fIyEi2b9+SkZGR1GpFimLX7yk7dw5m48Yt6e7uztSpM3Y/Ppnoa/alXq9ZfNEG73vf+/a5AujXPfzww3nzm9+c5cuX57rrrktXV9cBX3/UAdCiRYvy/e9/f/f7Q0ND2bJlSxYtWrT7sb6+vnz0ox/N5ZdfnkZj15RlWe5+++nyzYmy1AdUi56mivQ1VaSvJ5ZWq5mtWzenLFtpNDr2uFav11KWZQYG+lOWydSpR03alUD6GsaHX27rOpA1a9bkkksuyWWXXZbLLrvsoF9/1DH32WefnXvvvTf3339/kuQzn/lMnv3sZ+8+jChJent7c8MNN+Tzn/98kmTVqlX52te+lpe97GWjnR4AAOApDQzsTKvVTL3+1H94Looi9XpHBgcHMjIyPMbVARya66+/Ptu2bcuNN96YCy+8cPd/69at2+/zirIcfdb7ve99L//jf/yPDA4OZvbs2flv/+2/ZfHixbnwwgvzl3/5l3nWs56V++67L9dcc022bduWZrOZq666Kueff/7TnmvLlr6MjNgCNlkVRTJnzrRs3GiZKtWgp6kifU0V6euJpyxb2bx5Q4qi2Ot4il/XbI6kq6s7U6ceNUbVjQ/6mn1pNCb3FrAP/PjhPN43tmcAvXX5cUd8nlFvAUuSs846K5/73Of2evwLX/jC7rdPOeWUfOITnzgc0wEAAOxXs9lMWbZSr3cccGxR1DI0NDQGVQG0z+Q76QwAAGAvlsAA1SYAAgAAKqdWq6UoihzMiRe7Vgodls0RAOOWAAgAAKicWq2ezs7utFrN/Y4ryzJlWaa7e/KedwJMDgIgAACgknp6elOWSau175vI/PIuYZ2dnWNYGcDYEwABAACV1Gh0ZurUGWm1Wmk2R/bYDtZqtTIyMpxarZbp02emKPxqBFSbja4AAEBldXf3pNFoZGCgL4ODA7tDoFqtlt7eaenu7kmtVm9zlQBHngAIAACotEajI1OnHpXe3ubu7WD1et2qH2BSEQABAACTQq1Wt9oHmLRE3gAAAAAVJwACAAAAqDgBEAAAAEDFCYAAAAAAKk4ABAAAAFBxAiAAAACAihMAAQAAAFScAAgAAACg4gRAAAAAABUnAAIAAACoOAEQAAAAQMUJgAAAAAAqTgAEAAAAUHECIAAAAICKEwABAAAAVJwACAAAAKDiBEAAAAAAFScAAgAAAKg4ARAAAABAxQmAAAAAACpOAMSEsnOgzKNrh/PoE81s62u1uxwAAACYEBrtLgAOxsatrdz6o6Hc8qPhFLWBDA4OpyiSM0/pyLlndGbJ/Hq7SwQAAIBxSwDEuLdqfTN/d8PObO0rM39mLXNmNbJzZyuDw2Vuv284P/zZcN54QU+efWJHu0sFAACAcckWMMa1voEyK77Qn6GRMksX1tPbXey+1tlR5Oh59cyYWuQfv9Sfxzc021gpAAAAjF8CIMa1Hz8wnI1PNjN/1r63eE2fUkuzTG798fAYVgYAAAAThwCIce3bdw9n2pQDt+m8mbXc9pPh7Bwox6AqAAAAmFgEQIxbZVlm/eZWpvYUBxzb3VlkaLjM9p3uDAYAAAC/TgDEuFUURer1pHUQmU5Z7lr5U6sdOCwCAACAyUYAxLh28rH1bNl+4ARoa1+Z2TNqmTlNAAQAAAC/TgDEuHbO8s4MDSfN1r7P9inLMpu3lvntMzrSqAuAANi/ZrOZkZGhjIwMpyxtHQYAJodGuwuA/Tnx6HrOOLmRO+4bybELa2n82havsiyzan0rx8yv5Tee0dmmKgGYCIaHB9Pf35ehocHdj9VqtXR396a7uze12r7vOAkAMNEJgBjX6vUiv//SntSK/tx+30g6G8mCua0M9Leyrb/MwGCydGE9b7yg56AOiwZgcurv70tf37YkRer1Ror/n707j47ruu58/z3n1lyYBwIEKc4iJVuiJFKDNUQSJceSKcVqJ1Hs2Bmk5yl2y8lzt/ulk/XSXnnLeb3ar1c7ttSdxLEz2Ini2IrdtuVJs615oEZKpEZwJgCSmKtQw73nvD9AQoIAEKAwFIbfZy0uQVUHVRvAqap7991nHzP8meGcI58fpFgsUFNTTxDo0EhkMfAUKQUvUrbP4swQ1teSiLYSdxsxKNkrIkuTjnJk3ksnDTdfl+bKLRGP7ixz8JjBYdi4MuCycxKcsTogHlPyR0RExlcqFcnl+rE2wNrRq9+ttRhjcC5iYKCX2trGkeSQiCxMJfsSg4m/x5lu8CkMcaBIIXY/gVtFdeljxPzKSocpIjLnlABawiIfEfmIuInP+4Ndaw3rV8TYsDJGU1M1R48O4CduCyQiIjKiUMhhjBmT/Dlh+L6AMCxTLpdIJJJzHKGIzJSS3UV/8isYX03g1mF48xjXe48zh+lP/g9qiv+JmF9ewUhFROaeEkBLTOQjXhl4lYeOPMLO/pfw3lEdq+by5ss4v2ErjcmGSocoIiIyY6IopFQqTrq0a/hCiKFQyCsBJLJAeUIGE9/C+CyBH3tMazAEvo3I7icf/zdqSrdUIMp3LjLHKAZPEdn9AMTcGhLRlnF/VhGR8SgBtIQUoyL/su9fefzYkySDJI2JBgITYyjK84NDd/Lzznv4+LqbeFfNmZUOVUREZEY4N7zL11QqXa01RFE42yGJyCwp210400ng1p10nHXLKQXPE5kOAt86R9G9c54yudj3KMTvBx9hSANQtI+Tj91BKvxVMuENGJ3aicgktA38EuG954793+fRY0+wKruKtnQbqSBF3MaoidewJruapE3yN69/nb25fZUOV0REpCLm+5JoEZlY2e4GYqOWfY3nRKKkbNvnIKrp8UQMxr/FUPxnBK6VmF9L4FsJ/PDX1rcwFL+TXPxf8LhKhysi85wSQEvE4UIHDx97lNWZVQRm/J0P6hN14A13d9w7x9GJiIjMjhM7fp2oBDoZ5xzxeGIOohKR2eDNEPipVsF4oDyb4cyIsn2RYuxhYm4thrHvT4YkgVtDIfYAZftyBSIUkYVECaAl4snupzAYYvbND8WyCylEBcruzQ+/Zalmnu19jqPFY5UIU+aQ7ziMf+E5/PPP4g8ewKurtogsQtZaUqkMzkUnHXfiPTCZTM9FWCIyC6yvBzP1pI4hO4vRzIxC7AGMT590eZchjvEpCrFfzmFkIrIQaaHoEvH6YDtVsSq893SXetif38/hQgfOO4wxLEsuY3XmNJqSTQAcKR6hKdlY4ahlNvg3XsP95E78rpfgRIm0d5jTN2Lffz3mDPWAkskVSp4X3wh59tWQfMFRW2U5b2OcM1YHxGNaQiPzSyqVpVgsEEXhuM2gvfdEUZlUKjtps2gRmb8S0bnk4z/EE2EYv+IdwJHH+DTx6Iw5jO7UOYYoBS8SuLZJx1rfRCl4Dk/5+Lb3Mh9EpovIHAUg8MsIfFOFI5KlTkc5S4THA57XBl/nlcFXCbBkY1kCExB5x9HiUToLnazOrqYqyB4fL4uNe+4Z3Nf/GmJxWLESExt+C/BRhD+wn+ir/wP7ezdj33NJhSOV+ezF9pBv/WyI3gFPOgXxwPDawYhHXyjR0hhw0/Y0a5ZPfOAtMteCIKCmpp6BgV7CsIwxBmMs4EeWhqVSWbLZGvUAElnAAr+SRHQWpeAlArdq3F5AHoezh0mXt2PnfQVQCYiY2ilbDIiUAJonyvY1hmI/pxQ8B/74PDSQiM4jU76GmF9b2QBlyVICaIlYlTmNx449QWehk+pY9ailYIGxVMeriXzEG4Nv0JRspDGh7SQXG3/kCO4fvgE1dZja2lH3mSCAllZ8bhD3z9/ErFiJOW1VhSKV+WzXnpC//n6eqrRh3YrRSR7vPV09nlvvyPO5D2VYuUxJIJk/YrE4dXWNlEolCoXc8SVhllQqTSqVHukVJCILl8GQLf8ukflLIruHwC3HkBq53zGIs50kos1kwusqGOnUGNIYnwIKwMmXp3oKx5eKJeckNplY0T7NQPJvwMcI3MqR5XvelynZ5ymlnqOm+BkS7uwKRypLkXoALRFb6s7lYP4AaZsalfx5q8AEBAQMRUNkY1VzHKHMNvfkY1Asjkn+vJXJVoH3uEcemsPIZKEII8+37ymQThnqa8Z+fBhjaGmwhJHne78oqK+UzDvGWJLJFLW1jdTXL6O+vpmqqhpisbiSPyKLROAbqC3+B1LhNpw5Rmj3HP/XjjdDZMr/jurSH4xKDM1XhgTJ6BKcPTLpWGePkgp/5aRL32T2heYwg4mvY10dMb9iVO8mQ5yYX4n11QwkvkZkJv+7isw0VQAtEfkoTzaWJR8NkQxS4x7ollwJY6A52czO3p28p+miCkQqs8F7D7/8BTROoa9T8zL8Y4/g/91vYJK6iiRvenlfRFePY83yk187aG2w7N4bceioY0WzDkRFROTUeCJCuxdPHkOCwJ2GnaQC5q0sdVSVP0KmfD2hfQNvShifIe42LIjEz1ulwkspxh7AkZtwyZpjEDAko/fMbXAyRjF4BG9CAj/xBVfr6whtD8XgMTLhr81hdCJKAC0Zx0rdrMmuoafUQ1fxCEmbJB2kMcYQupBclMNgOLfuHCIchwsdlQ5ZZlKphM8NwvLJmwiSSkGxAEN5UAJI3uK1AyGBZdJKiSAweA97OyIlgEREZMo8EYXgYQqxu4nsYYwP8MZjfIZ0eAWp8L1Yqqf8eJYaEu7cWYx49sX8ajKl3yGX+CbeV2N9I+b4Ig6Pw5mjeJOjqnQzMb+iwtEubZ4yhdgvsa550rHWNVGIPUA6vG7k7ykyF5QAWiIMhriNsbVhC52FLtpze+gt9QIQMwGrM6s4LbOSmngN+/MHVAq/2MRiYAOIwuGvT8Y5MHbycbLkFEueYIr5HGOgHM5uPCIisnh4Igbj36QYexDjGwjc2uEmzh4cOfLxOykFL1BT/CyW+kqHO6fS0eXYYg1D8R8R2r2M7OKKI+bWkil9nIQ7q5IhCuDJ400O65dNOtaQITLdw72byMxBdCLDdIa3RDSnmvB4LJa29HLaUq2UXIkIR8LECd7SFyh0ISvSuoKwmJggwJx1Nn7n87Bi5ckHHzuK2XA6ZNUHSkarr7aUwyn29fFQlVYiWUREpqYQu49i7EECt2ZU3xQASxbj1hHavQwmvkl16Q/H3eFrMUu6c0kUNxPadtyJbcXdMgK/Zsn9LuavGMMtdqeyc9vwRgRvn+syfzz7aplXjpXm7Pk2NgawefafRzNuidhYdTpNiSZ6y33UJ+rAGBLB2OU9uTBHJpbh3TXvqkCUMpvsZZcTPf0kPgxHtn9/O+8cDAxgrrxKVWAyxtnrY/zvXxrCyBMLJp4fQ0VPImHYtFofMSIiMjlPiaHYXVi3bMITYoMhcKsoBTuJzD5ifvUcR1l5BkvcrQfWVzoUGYchQzxaT2gPEfiWk4515hiJ6AwMiTmKTk7VuYUGmvNzV86+Ijs3x81acLhExGyMD7RdR0+ph3w4NO6YYlSko9DJ9uXXkIlNvdGeLBCbzsBcfCnsbceXy2Pu9mEIe9oxW7ZizpqD9LMsOK2NAeecHuNgVzThDl/OeQ4ddVx+blwVQCIiMiVl+yrO9GIm6e8z3CvFUAx2zE1gIqfAYEhFV+PNAB434ThPhDd5UuG2OYxOZJguzy4h5zdsZTDM8W8Hvg9FaEw2kbQJyj7kaPEozkdcv/z9XLXsykqHKrPAWIv98O/gkin8gw/gPZA5vptEPjc85uJLsL/1EUw8XrlAZV778HtTdPd72g9FtDZaMqnhJI/3nsEhT1e3Y/OGONddogbiIiIyNd4MgLdTWspkfGpK26KLVEIiOodEdB6l4NnjyxlHN0/0RES2nUR0IXGnFRcy95QAWkKMMWxruYLTqzfw2LEneOzY43S7InET5z2NF3Jx40Wsza7R0p9FzCQSBB/6CP7Kq3E7noD2dvAOVq3Gnn8hLG/T319OqiZrueU30tz9ZImHni9zpNdhAOehJmv44BUptm1NkIxrHomIyFQFGONhCm3mPBHGa9mMzE+GOFWljzGY+CalYAdgMX64ss2bAcCRDC+hqvw76v8jFaFZtwStzKzgNzMf5NdX3kDJlUjYBNZoNeBSYlpaCLb/WqXDkAWqKmP54BUp3ndRkj2HI8qhJ5Uw0ncCtAAAIABJREFUrG0LlPgREZFTFnOr8Rg8ZQyTVCGboionZF6zZKgufYrI7KMQPEZk9wAQuAtJRRcT+JVq3C0VowTQEmaNJRWkKh2GiCxQ2ZTh3Wv1MSIiItMT+GUkovMo2+cI/GkTjnP0Y3yWeKRehTK/GQwxv5qqcOk1K5f5TWUfIiIiIiJSUZnyBzAkiUwHfpy1YI5BnD1CtvzbWLRZiYjIO6EEkIiIiIiIVFTMt1FT/ByWaiLbTmQOEZluItNFaN/Amz6qSh8jGV1U6VBFRBYs1e6LiIiIiMwR7z3lcokwLAOeIIgRjyexVtdlY341dYUvULYvUYg9hjPdGJ8mEZ1HMtqCpabSIS5IHk9k9hHag0CE9XXE3SYMaqYtstQoASQiIiIiMgeKxSHy+QGiKGJ4y6vhRrDGGFKpDJlM9ZLfjdOQIOHOJVE6t9KhLApl+yr52PcpB6+CN2CGl9dZX0u6fC1pt63CEYrIXFICSERERERklhUKeQYH+zDGEgSxUYke7x1DQ4M4F1FVVbfkk0AyM0r2RfqTt2F8jMCtwhCAH64I8gyQS9yOCztp8p+sdKgiMkdUayoiIiIiMouiKCKX68fagCAIxiR4hpNCcYrFIYrFoQpFKYuJY4CBxNewvorALx9O/hxnMFhqCNxqhmL3kvdPVjBSEZlLSgCJiIiIiMyiYnEI7/1J+/wYYzDGUijk8X7sLlgip6IYPIs3Oayvn3CMIY6lhj7303F3XhORxUcJIBERERGRWVQqDU2pybO1AWFYxrloDqKSxawUPIzxkzfNtr6JIq8Tmc45iEpEKk0JIBERERGRWeSc40TD55M5sTRMFUAyXZHtxfjkpOMMFrzFk5+DqESk0pQAEhERERGZRcPVP5MndU4kftQEWqbL+jSYcNJxHg/GaUt4kSVCCSARERERkVmUSKSPVwGdnHMRsVgca4NJx4qcTCK6AEffpOMcvcRoIfDL5yAqEak0JYBERERERGZRMpnGGHPSJJD3Hu8dqVRGFUAybcnoQgwB7iRLuzwOZ49SY64ZtUuYiCxeSgCJiIiIiMyiIAjIZmtwLiKKojE9fpxzRFGZZDJDMpmuUJSymAS+iUz5Qzh7GEf/mPs9ZSK7h0S0mWp7eQUiFJFKiFU6ABERERGRxe5EZU8+P0AUje7NYowhk6kmna5S9Y/MmFR0BaYUJxf/LqE5Cj4BGDAlwJAML6Eq/DDWJIFShaMVkbmgBJCIiIiIyBxIJtMkEinK5RJhWAY8QRAjHk9OaZt4kVNhMKSiS0lEWygHL1A2r4EJsX4ZyWgLgV+G8o0iS4sSQCIiIiIic8QYQyKRJJGYfItukZlgSZOMLiTJhZUORUQqbNqXGh588EFuuOEGrr32Wm6++Wa6urrGjOnp6eEP/uAP2L59O9dccw333HPPdJ9WRERERERERESmaFoJoO7ubj7/+c/zpS99iZ/97Gds27aNP/mTPxkz7s///M/ZuHEjP/nJT/jGN77BF77wBQ4fPjydpxYRERERERERkSma1hKwhx56iE2bNrFp0yYAPvzhD/OlL32JI0eO0NzcDEAYhtx///389Kc/BWDlypVcdtll/OhHP+KTn/zkO3perVVduk787TUHZLHQnJbFSPNaFiPNa1mMNK9FlpZpJYA6OjpYvnz5yP8nEgnq6+s5fPjwSAKop6eHQqFAa2vryLjW1tZ3XAFUX5+dTsiySDQ2Vlc6BJEZpTkti5HmtSxGmteyGGleiywN00oAee/H3aryrbsYeO8Bxox7pzsd9PTkiCL3jr5XFj5jhj+gjh0b4PjUElnQNKdlMdK8lsVI81oWI81rmUgQWBVfLELTSgC1tbXx2GOPjfx/qVSip6eHtra2kdsaGxtJJpN0dXXR0tICQGdnJxs2bHjHz6s3J/Fe80AWF81pWYw0r2Ux0ryWxUjzWmRpmFYT6EsvvZSXXnqJV155BYDvfve7nHPOOTQ0NIyMCYKAq6++mttvvx2AgwcP8uCDD/Le9753Ok8tIiIiIiIiIiJTNK0KoIaGBr785S/zx3/8xxSLRRobG/nSl74EwA033MAXv/hFzj77bP7sz/6M//Jf/gvXX389YRjyn//zf2bNmjUzEb/MI957Dg4doq/cjzWGtvRyauO1lQ5LREREREREZMmbVgII4JJLLuH73//+mNt/8IMfjHzd0NDAbbfdNt2nknnKe88LfTv5ecc9tOf2YI3F4wmwXNRwAe9tvYqWVEulwxQRERERERFZsqadAJKlzXvPPZ338b2D/5uqoIpVmdOwZnhlYTEq8mj34zzX9wK3nP5pVmVOq3C0IiIiIiIiIkvTtHoAiewa2M33D/6A5anlNKeaR5I/AMkgyarMKiIf8bXXv8FQNFTBSEVkuorFIseOHaO3twfntBujiIiIiMhCogogmZb7On9B0qZIBakJx7SkWmgfbOfFvl2c37BlDqMTkZnQ0XGYRx55mIce+iWFQh6A1tY2tm27mq1bLyCTyVQ4QhERERERmYwSQPKOHS0eY/fAbtpSbZOOzcaqeOjoI0oAiSwwL764k7/9278iDEOampppbW3Fe09vbw///M//yMMPP8inPvUZ6usbJn8wERERERGpGC0Bk3dsIBzAe4jZyfOImViao8VjcxCViMyUAwf287Wv/S8ymSyrVq0eqfQxxlBf38DatevZv38ff/d3f0sYhhWOVkRERERETkYJIHnHYiYGeLz3k46NfETcqOBMZCH5xS/uxzlHXV3duPcbY1i1ajWvvfYKu3fvmuPoRERERETkVCgBJO9YS2oZNfEaBsLBScf2lvo4p+7sOYhKRGbCwEA/TzzxGMuWLTvpOGMMqVSahx9+cI4iExERERGRd0IJIHnHEjbBlc2Xc6R45KRVQCVXAjwXNl4wd8GJyLT09PQQhiHJ5MQN3k+orq5m//59cxCVyNyJopBicYhCIU+pVNDOdyIiIrLgaU2OTMulTRezo+cZ9ucPcFpmJcaYUfcXoyIHhg7w/tZrWJ5qrVCUInKq3v5aPhnvwdqpjxeZz8IwJJ/vp1QqjrrdGEM6nSWdzmKMrp+JiIjIwqMEkExLVbyKT2/4BP/Q/i1eHXyNmImTiWVw3jFYHsAay/XLt7O97dpTOqEUkcpqbGwimUyRz+cn3ea9r6+XLVvOn6PIRGZPGJbp7+/Ge08QxEZ9bjnnyOcHCcOQ6uo6faaJiIjIgqMEkExbfaKeP9p4C28MtvN49xMcHuogZmOcsewKtjZsoTnZVOkQReQUZTIZLr30Mu677x5Wr14z4TjnHOVymUsv/ZW5C05kFnjvGRjoHUn+vJ21FmMMxWKBWCxHJlNVgShFRERE3jklgGRGWGPZUL2eDdXrKx2KiMyQyy/fxuOPP0pXVyfLlrWMud85x549ezj33PNYv35DBSIUmTnlcokoCsdN/pxgjCEIAgqFHOl0RkvBRN7GuYhisUAURQDEYjESiRTW6rUiIjIf6N1YRETG1dzczGc+84fE4wn27Gmnp6ebUqlEsVjg8OFD7N3bzubN5/C7v3uTDu5lwSuVhoDJ+18ZY0Yq30RkmPeeXK6fnp4j5HL9FAo5CoUcg4N99PR0MTQ0eNINQ0REZG6oAkhERCa0Zs1a/vRP/4xnnnmaBx64j66uLqw1bNy4iV/5lSs588x3EYvpo0QWvihyU+rrY4zBGPBeu4KJwHDyZ3Cwj2Ixf7x31ugLAs45crkBvPek01XqnyUiUkE6apd5xXnH4UIHQ+EQiSBBW2o5MatpKlJJ1dU1XH75lVx++ZWEYYi1VhU/suicypz2/tR2yhNZzMrlIsXiEEEQH/d1caJ/Vj4/SCKRIhaLVyBKEREBJYBknoh8xJPHnuLergc4OHQIgwE8DYkGrmq5kgvqt9JR6CTv8rTYOqrDRjLByXcmEpGZp2ofWawSiSTF4tCk47wfrhSKxRJzEJXI/Oa9Z2god7wybuKk6InKuUIhT1VV7RxGKCIib6Ujeam40IXcvu9fefjoI9TF61idWYUxBu89PaVevvLKbYQupC3dRiKIkzwQg3LApU0X896Wq6iJ11T6RxARkQUukUhircW5CGuDccd474miiFQqqyo4EYYTouVy6aTN008wJqBYLJDN1qiCTkSkQnT0IhV3T+d9PHz0UVZnVtOYbBw5KHA49ub3cqzYzbFSN7kwx+rMKtZVr6EmVsPdHffy1Vf/F72l3gr/BCIistAZY6mqqsV7RxRFYxrWDid/hncJy2SyFYpSZH458TqZav8sNYIWEaksJYCkogpRgXu77qc11TKm1097bg8Hhw5Rl6ijKdnIwaFD5KM8AJlYmjXZNXQMdfCtvf+iAwoREZm2RCJFdXUDxhiiKCQMy6P+m0gkqalpmLBCSGSpMcZOObHjvcdaVf6IiFSSEkBSUS8PvMJgOUc2NvpqauhC9uT2kI1lscZiTYDH01noGhljjGFFZgW7+nezf+jAXIcuIiKLUCKRpL6+mZqaelKpLMlkikymirq6Jqqr6wkCJX9ETrDWEo8ncS6adKxzEalURsu/REQqSD2ApKL6yv2MdxjQXeqmGJWoS7zZ6DkwAbkwN2pcYAIM8GzPc6zKnDa7wYqIyJJgjCGRSJFIpCodipyE9/545Yk/3mRY1zUrIZ3OUCoV8N5PmNxxzgGGZDI9t8GJiMgoSgBJRQVm/CupJVeCtx1EeDx2nPHJIMmxUvesxCciIiLzi/eecrnI0FCOcrkEDCftksk0qVRG24zPsVgsQSZTRT4/eHzLdzuSCPLe45zDe0dVVe2UmkWLiMjs0buwVNSKdBsYcN5h33LlzpoA89bl5N7jvKMhUTfmMSLvSNnkHEQrIiIileS9Y3Cwj2KxgDHmLQkFT7E4RLGYJ5OpIZ1Wo+65Yowhna7C2oB8fpAoCkfdP9w4vUbVPyIi84ASQFJRqzOrWJddy6FCB62plpHba+M1YIaTO4GxFFyRVJCkOdk85jFKUYlNNRvnMmwRERGpgFxugGKxQBDE3rbcyBAEFuccuVw/QRBoCd8cMsaQSmVIJtOUy6WRnkBBECMWi6vvj4jIPKEEkFSUMYYPrryBr7zyP+kp9VCfqAcgG8vSklxGV7GLVJCmEBU4r/7cMUvGekt91MZreHfNuyoRviwghUKBF198gZ07X6BQKNDQ0MiWLVtYu3Y91qpvhIjIfBdFIYVCfpzkz5ustXjvyOcHiceTSjzMseH+WarKFhGZr5QAkopbX7WOT2/4BH/3xj/SPthOVbyapB2u9tmb30chKrKl7jySNsGe3F4SYYAJAxImSX84wCfW3Uwy0MGGTOzZZ5/m9tv/iYGBfpLJJLFYjGKxyH333c26dev5/d//GMuWLat0mCIichLFYgFg0qSOtQFhWCaKysRiibkITUREpGL+8i//kiNHjvAXf/EXk45VAkjmhTNrzuALZ/3fvND7Ao8ee4KB8gCrsqdx+bLLuLfzfp7re56iKxIQYAMohmVSQZrfPu1GNtedXenwZR579tmn+drX/pq6ujrWrl036j7nHAcO7OfWW7/M5z73eRoaGisUpYiITCYMy1Oq6DkxJooiYjrSFRGRRerAgQP81//6X3nooYe4/vrrp/Q9+liUeaMqluXipvdwcdN7Rm57qvtp7uN+zqjZRC7MU3RFsskk9baRqlgVL/Tv5Nt7v8tvr/6tUU2kRWB42dftt3+Luro66urGNhC31rJixUr27t3DT396Jx/96O9XIEoREZmKU1nNpZVfIiKy2P3rv/4rl1xyCaeffjpHjhyZ0vcoASTzVneph3/aezvNyWaq49UAGCCTSZDPl/BAVayKXx59iDNrN7Gl/ryKxivzz86dzzMwMDCm8uftWluX8/jjj3H99TdQWzs2USQiIpUXBHG8L0w6znuP92jLcREReccOPdfDnvbJP3Nmilmbgo828frrr1MojH3e5ubmMS0r/uN//I8A3HrrrVN+Hn0yyrz1dPfTlFx5JPkznpiNUR2r5v6uX3Je3blq9iij7Nz5PKnU5LvAJJNJoihiz552zjlHiUQRkfkomUyTzw/gvT/p571zEfF4QgkgERF5x97bNkSvz8/Z89W1eQA+//nP89JLL425/5ZbbuGzn/3stJ9Hn4wybz167Anq4rWTjmtKNvLGYDtHS8doTjbNQWSyUBQKRYIgGPc+7z3d3cfYv38fPT099Pf3cccd/0o8nmDjxk3E1DhCRGReCYKAdDrL0NAgQTD+1uLOObz3ZDJVuigkIiILzn//7/99wgqgmaAzHJm3BsNB0kF60nHWWIyBQjR3JXqyMDQ0NFAsFsfcXiqVeO65Z+jq6iQIYiQSw7vE7N+/j1tv/TKbNp3BzTd/XMvBRETmmUymGu89hUIeYwzWWsDgvce5CGMMVVV1xOPaHVRERBae9evXz+rjq2uuzFuZIE3JlScd57zDeU/S6mBPRjvvvK2EYYj3fuS2KIp45pkddHV1UltbR3X18BLDbLaKTZvOZM2atbz66st8/et/M272XUREKscYQzZbQ21tA/F4EuccURQCw1U/dXVNpFKTXzwSERFZilQBJPPWhY0XcOehn1CXOPkysO5SN6syK7X8S8ZYv34Da9as5dChg6xYsRKAI0e6OHLkCHV1dRgzfNU4n89x5pnvHlkutmrVGl599RWee+4ZLrro4kr+CCIi8jbGGOLxJPF48niC3wNGS75ERGRJOpXeQKoAknnr/IatBCZGLpy4+VbkI/rKfVy17Eod+MkY1lpuuunjVFVVs3fvHkqlInv2tJNIxAEoFov09vawfPkK1qxZO/J9xhhqamq59967R1UPiYjI/GKMwRirYwAREZEpUAJI5q3mZBM3nvbrdBY66S31jTkRL0QF9uT2cmHD+doCXibU0tLC5z73n7joovdw+PBhDhzYR7FYor+/D+ccZ5zxLs4997wxzaIbGho4cGA/udxghSIXERERERGZOVoCJvPaZU2XkLJJvn/wh+zJ7SUwlnSUIFcoELdxrml5L9ev2E7MaiqfzEDesedwRKkMmZRh/YqARHzpXC1tamri937v/+B973s//+E/fJaGhibS6RT19Q0T7vZ14mqyc24uQxUREREREZkVOmuWec0YwwWN53NO/WZ297/M/qH9ZLIJkqUsZ9W+m5p4TaVDnNd6Bxw/e7zEoztLhCFgwHuoyRqu2ppg25YE8djSSQS1tLSyfv0GnHOT7vCVy+Woqqomk8nOUXQiIiIiIiKzRwkgWRASNsHmurM5p/5smpqqOXp0gPFas/QMOF54vczRXk8sBmtaA85cE1tSSY4TjvU5brsjT2d3RGtTQCox/Dvw3jM45Pm3+wvsORxx0/b0kqkGMsZwxRVX8b3vfXfSBNCxY0fZvv36CSuEREREREREFhKd2ciiUCh5fvDLAg89XyaMIBYD7yByUF9t+c1tCbZsSsx6HPmCZ9eekN5BRxAYVjRZ1q8IsHZuEyzee771syGO9TvWtI1+mRtjqM4YsinDjt1lVi4L2H5xck7jq6Tzz7+Au+/+GUeOHKG5uXncMd3d3SSTKS68UDuAiYiIiIjI4qAEkCx4pbLn7388xHOvhqxcZkdVs3jv6ct5/vaHBX7//fCes2YnCRRGnrueKHHvUyXyBY8xgB/emHZFs+XXr0zxrjVz93Lb2+F4ZX/EqpaJ+7xba2htDLj/6RJXbU2MVAgtdnV19XzqU5/hr/7qNvbt20tz8zLS6TQAhUKBI0e6iMVi/MEf3DJhgkhEFh/nIorFAqVSAecc1lqSyTSJRAprtWeGiIiILHxKAMmC9/hLJZ55pcy6trGVNsYY6qoMscDzL/cU2LQ6Rn31zB7IR5Hn9rsKPPx8mdYmS0vDm4/vvedYn+d/3pHn4x9Icd7G2a9CAnjutTLWMGnlUTZtONLreHV/xNnrl87bwbp1G/j85/+EX/7yAR5++EG6ujoxxhAEAZdeehmXX76NtrYVlQ5TROZIoTBELje82+TwluIQhhHlcgljBqiuriORWDqVkiIiIrI4LZ0zPlmUoshz71NlGmvsSZMdVWnD0V545uUyV50/swfxO14u8/ALZVYvt8SCsQmopjpDIu745s+KrF0eo26GE1DjOdbvSUyxoscYyA0tvZ2uWlpauPHGD/H+919Hd/cxYHjr96qq6gpHJiJzqVgcYnCwF2sDgmDs+7NzEf39PdTWNhCPz00SX0RERGQ2KAEkC1pHt+NIj2NV6+RJlZqs4fFd4YwmgLz33LejTG3WjEn+jH5uy7G+iGdeKbNt6+xfRU4noFx29A2CcxCPDSfBMOPHGFuCTbJPqKqqoqqqqtJhiEgFeO/J5Qaw1k64zMvaAO9Dcrl+amsbMRO8j0rlOHooxnZQDB7GmX6Mz5CMLiQZXUTgl1U6PBERkXlDCSBZ0IolD4YpHZDHYzBUGGfrsGno6nHs6zx5r50TaqsMj704+wmgfMEzkPM8tTskHgNrDB6oqzKsawtobbQjv69y6DHAmuXBrMYkIjIflctFnIsIgpMfDlkbEIZlwrCsKqB5pmifZjDxd3hTwPp68Cm8KZKP/4h8/MdkSx8mFV2BQYk7ERERJYBkQUslDd6Dc37SfjelMjTWnvoBYKnkuPupMv92f4H2wxEwvL38b2xLsWGlxdrJe+0AJOOGwaFTfvpTMph3/M0Phti1NySZMCRikD7+O8oXPDteLrOuLcaZawKMMXQcc5y7MUZTrRqcisjSE4YlYPKLCCfuj6JQCaB5pGRfYiD511hXT+Db3rzDg/V1OPLkEt/ClBKkoksqF6iIiMg8oQSQLGitDZaVywK6BxxNkyR3BnKe6y45tSl/6GjI574yyGsHouFlVJnh59i1N+QLXx9kZbOlvsbgvZ30BKIUeqozs3cF0nvP7XcXeP1AxOkrA2qzlh27yxjjSSUMmZQhlTC8cTAkm/LE45aqtOHXLk3NWkwiIvOZO8X2Z97PbBWpvHMeRz5+B8ZnsNSOO8aSAddEPn4HyWgrBjXyFhGRpU2X/WVBs9Zw9dY4fYOeMJr4wLx30JFOGc7bGJ/yY/fnHX/05QFePxixotnQ0mDJpgzZ1PDXK5oNB444nn8tpONYNOnj9Q16Ljxz6s9/qg4fczz7ashpLcPJqJYGy5ZNMZw39A54BvKeoZIn8oZnXo1oqrX84Y2ZUbuWiYgsJUFwastftR38/BHa1wntPqxvPuk4Sy3O9FEKds5RZCIiIvOXjmRkwTv/jDiXnh1nX4cj97YeP855jvQ4+gc9N21PUZ2Z+pT/4YNFXj/oaGsy41b3GGNY0WwoluHF9ojITZyAGhzyxAPYsmn2EkBPv1zGGAje0oy6tTFg25Y4522K0dJgqa82rGuzrGsL+OAVSZY3qfePiCxdicRwBeRklT3OOYwxxOOqIJkvInMIMFPs7RMjNHtmOSIREZH5T0vAZMELAsNH3peiqc5y31MlunoceDB4PIbTWgI+fmWSTaumPt2dc9xxf4FM6uS9IYwx1FcZegY8bxwos6o1TvIt26977+kb9PQMOH7v2jQNNbOXc+3sHl7q9XaxwNDWFND2lmTP3o5o1vsRiYjMd0EQkEqlKRSGCILYuO/33nucC8lkqlUBNAe894RhGe+He/sFQXyCz+FTWL/nLZ5wxmIUERFZqJQAkkUhFhi2X5xk25YEu/aE9OU8gYUVzZa1y4MpNWl+q/48HDziWFY/+ffVVRs6uh2Xbk7yzCshpRCCgOPNqaGx1vKJD6TZsml2G4fGY8MVT1PiIdB5jIgImUwNzjmKxSJBYDFmeBntcOLH4b0jmUyTTldVOtRFzTlHoZCnUMjh3tKcaThJlyWVyoxKBFlfN/UHN2VtBy8iIoISQLLIpJNmRpZZlcPhBM4Udpc/PsZw7XuS/PqVKV54PeRonyNmYVVrwBmrY8Rjs7/97OkrAx7dWZp03IleSSublQESEbHWUl1dRzw+xNBQjih6s1IkCGKkUlVjkg8ys5xzDAz0UC4XsTYYqcby3uO9I5frIwxLVFXVjfwd4u5MrK/GMYClesLH9hQxPkYyOneufhwREZF5SwkgkXHUV0FV2pAvTL5zV77gqUobGqoNiYTl0s2V2SJ48+lx0vcXyQ15sumJY+7sdrxrbYzWRvX/EREBMMaSTg9XmURRiPcOY+yEy8JkZuVy/ZTLpTHLvYwxGBNgjKVQKBAEg2Qyw8keQ4p0eC25+LcxLoVh7MUfT0Rk95Mq/yqW+jn7eUREROYrlQCIjCMWs1x7cYK+wcnH9g3C+y5MkEhU9uWUTRk+eHmCjm5HvjD+UrCjfQ5r4dcuUyNTEZG3M8YQi8WJx5PEYhP1npGZFEUhxeLEPZhg+O8SBAFDQ6OXh6XCq0mFVxHZfUSmE8/wjpweR2SOEtl2EtH5ZMNfn5OfRUREZL5TBZDIBG7cluLHD5fo7ncTNm/uGXCkkobfuio1x9GN77JzEpQj+P4vinT1QG3WEItBoQSDeU9tleHf/3qGVS2q/hGRyvLeUSqViKLw+Al+jHg8oaTLElMqFYCTb7gAw0v1wjCiXC6STKaHv4eAbPkjxN2ZFIJ7KAevHx/tibnTSJV/k2R0wbjVQSIiIkuREkAiE1i/Isb/84ks//ff5Dh0xNFQa0Z22SqUPN19nkTc8Ocfz7LhtPnxUjLGcNXWJJs3xNmxu8zTL5cplmFFk+His+KcvT5OJqWTKxGpHO89hUKOfD6H96N3cgqCgEymeuQEXxa/KDqF3bxgVAUQgMGSjLaSiLbgzFE8eQxJrF+GUaG7iIjIKPPjrFVkntq2Jcnf/LHlH39S4MHny3T3DR94BgFcsSXB712bYvOG+XdlsanWcs1FSa65SEu9RGT+8N6Ty/VTKOSwNkYQxEfdN9wMuBfvPalUpoKRylw51YKvCZeJYQh88wxEJCIisngpASQyibPWxfn/bolzpNfRfmivpUolAAAgAElEQVR4d5jVrTFaGnRlUUTkVJTLRQqF/Jhmv/BmnxfnDIOD/cTjCYJAhymLXTyeYGgoh/f+pMvAvB/ubReLaU6IiIi8U/oUFZmi5jpLc11ldvgSEVnovPcMDeWP7+w08Ym+tRbnIorFoZEdn2TxiseTWBvgnCMIJu5P51x0PCk4/6puRUREFgqVMIiIiMis895RLhexdvIm9NZaCoWhOYhKKs0YQzZbjfcO56Jxx0TR8O2ZTLWahIuIiEyDKoBERGRJ6uzsZM+edqIoJJvNsnHjGaTTaj48W5wbXsIzlRN4YwzOuUmXBcnikEym8R5yuT7CsDxSJea9x3uPtZbq6nricVXhioiITIcSQCIisqR0dBzmBz/4Hs8//9zxWzzeQzqd5sorr+aaa95PIqETzZl2IpEzlaSO92CMVfJnCUml0iQSSYrFIUqlwvF5Ykml0seXialoXUREZLqUABIRkSXjwIH93Hrrl8nn86xYsXJUQ9lcLsedd/6AAwf287GPfVJJoBlmrSUeTxCG5UmbOzsXkU5n5ygymS+staTTWf3tRUREZokup4iIyJIQhiF///dfp1wuc9ppq8bsJpTNZlm7dh3PPruD+++/t0JRLl7GGFKp7Miynok45wBDMqnleCIiIiIzSQkgERFZEl555WUOHTpIS0vrhGOstSxb1sL9999LqVSaw+iWhkQiSSqVIYrKeO9G3ee9x7kI5yKy2WpiMe32JCIiIjKTlAASEZEl4ZlnniIej0/aV6aqqpr+/j7eeOP1OYps6Rje8amGTKYa5xxhWB75F0UhAFVVtaRSmQpHKiIiIrL4qAeQSIWFkef1gxF9g57AwopmS2vj5Nski8ip6enpIZlMTTrOGIO1hnw+NwdRLT3GGDKZalKpLKVSEefKgCEI4iQSSTV+FhEREZklSgCJVIhznodfKHPXEyWO9jpGNjjx8K61MX7tsiSrWmY/EeScZ1+no2/QYa1hRbOloUbFgW83MNDP66+/RrFYIp1OsX796WSzalS6kGQyWcIwnNJY57y2nJ5l1g7v8ATq9SMiIiIyF5QAkiWrWPa81B7y6M4Sx/o8ibjhnA0xtp4Rp7ludhMg3nu+94sidz9RpLHOsmb5m9sdh5Hnlf0hX/52yL//jQwbVs7Oy9R7z46Xy9z1RJkDXREnLrobYOsZca69KMHyJlUi9ff38ZOf3Mmjjz5MuVzGGAt4kskUl112Oddeu51MRstVFoLNm8/hyScfn3Tc0NAQiUSStWvXzUFUIiIiIiJzQwkgWZL2d0V8/YdDdPU40ilDOgm5gueHDxa585EiH7gsya9ekJi1pQjPvBJyz5MlVrUGxGOjnyMWGFY0Bxzrc3z9R0P82c1VZFMzG4f3nh8/UuRHD5eorTKsarFYO/wcpbLn6ZfLvNQ+nIBas3zpJoH6+nq59da/5ODBAyxf3kYq9ebyoXw+z89//lP27HmDT33q36saaAE466zNVFfX0NvbQ11d/bhjvPd0dBxm27b3UlVVNccRioiIiIjMHq3zkCWnszvitjvyDA551rYFtDZYarPDy57WtAUsq7f82wMF7t0xOzsAee+5b0eJ6owZk/x5q8ZaS/+gZ+fr5RmPYecbET9+pMTKZZam2jeTPwCJuOG0lgAP/O0P8wwVJ96uebH79rdv5/DhQ6xZs3ZU8gcgk8mwbt06Xn31FX70o+9XKEI5FalUit/93Zvo6+uju7t7zFbkURSxf/8+2tpWcO21769QlCIiIiIis0MJIFlyfv54idyQp6Vh/OmfShjamgJ++GCRvkE37pjp6Ox2tB+KaKydvKonmzE88sLMJoC899z3VJFMypCMTxzDsnpLd79n5xtT65my2HR0HOb5559lxYqVE1aCGWNoa2vjkUcepq+vd44jlHfi7LM38+lPf5YgsOzZ087+/fs4fPgQe/fuYf/+fZxxxru45ZY/oqamttKhioiIiIjMKC0BkyWlZ8Dx5K4yrY0nz32mk4ZyCE+/XGbb1uSMxpAveIxlVNXNRFIJ6B2c2QqcI72eVw5ErGyePP9blTY8/HyJC86Mz2gMC8HOnS/gvScWO/nbZDKZIgzL7Nq1i/e85+I5ik6m46yzzuYLX/giu3e/xO7duymVCtTXN7B587msXHmadqESERERkUVJCSBZUg4ecTg/vMxpMtm0YffeiG1bZzaGeNzg3HAlzmQnmmEENTPcX3hwyGOAIJhCAioJ3f1LcwlYX1/vpMmfE7xHW4YvMIlEgs2bz2Xz5nMrHYqIiIiIyJzQEjBZUqJo6skMa6F8CuOnqrVhuN9Qf37yx+4b8Jy3ceLqm3LoOdAV0X444kivG9PTZDzx2HDCYipjIweJpVf8A0AqlZ7yluHGGBIJbRkuIiIiIiLzlyqAZEmpqbIjyY/Jqm+GCp6W+pnPkcZjhiu3xPnuvQVqMmbCOIaKHhvA+eMsv8oVPI++UOK+HeXhPkVmePv2TasCrtyS5Kx1wYSP21Jvqa2yDOQ9NdmT/w76BhwXnzWzS+AWio0bN3HnnX7SuRJFEYC2DBcRERERkXlNFUCypKxusbQ12UmXNTnnCSM4a23Ex//fXto+cIRl1w3/a7n+CNs/383+Q3nK4TurELr07AQbV8XYe9iNW5WUG/IcOur44OVJmmpHv0z7Bh233ZHnu/cXsdazqtWyujVgRbOlvcPxv76X56ePlSas8EnEDdu2xDna609aBVQsDSc+LlyC/X8A1q/fwIoVK+nq6jzpuI6Ow5xxxpm0ta2Yo8hERERERERO3bQTQDt37uTGG29k+/bt3Hjjjbzxxhvjjjt06BBnn302N9xww8i/Rx55ZLpPL3JKrDW878IEvQOOUnn85If3ngNHHNlEmav/aIgfPlwmjN56Pzy1K2LrJ3J84P86ynfuK7C/Kxr3sSaSTho+dUOa8zbG2N/l2NsR0dntOHw0ov1QSH/O8zvvS3LV1tHLirz3fPNnBfZ1Rqxrs9Rk7Uh1ShAYWhuGE1w/eLDI0y9PvHvYxWcnWLM8YH+nw7mxv4dCaXhp2faLk7Q0BKf0sy0W1lo+8pHfIYoijhzpGpMs897T0XGYZDLJb/7mh9Q4WERERERE5rVpLQErlUrccsstfPGLX+Syyy7j3nvv5ZZbbuHHP/7xmJOhp556im3btvHVr351WgGLTNcFZ8Y5fMzx00dLVGcMjbVmZEeuwSFPV7ejubbE137omay+Z8fLkEoM8OCzST50dYrLzpl6H5iqjOXjH0hz+Jhjx8tljvR4YgGcvjLg7A1xqtJjEwp7Ohy79oSsarETJhySCUNtleGuJ8uctzE+7m5jVWnDpz+Y5h9/UmD33pAgGE5KeT/8O4gF8MErUrzvwqXd12bdug189rP/J9/85j+wZ0878XiceDxOqVQiDENWrFjJTTd9TNU/IiIiIiIy700rAfTCCy8QBAGXXXYZAFdffTVf/OIXef755znnnHNGjd2xYwednZ18+MMfplAocOONN/LRj370HT2vLrQvXSf+9tOZA8YYbviVJKtbA+55ssSewxHGgPPDiZEP/EqS274b4plaVc9Tu+CcDZZ/vqtAXbXl7PVTf1kZY1jRHLCieWpVNk/uKhNYCCbZQr6x2rC3M2J/l2PN8vEfu67a8oe/lab9sOPJXWU6uyPigWHTqoAtm+LUVWuFqPeeeDzOFVdcSXv7Gxw+fJj6+jqam1s499zz2LDhdIJgehVSMzGnReabmZzXpVKJPXvaGRrKk0gkWb16DZnMDG+PKDIFer+WxUjzWmRpmdKZ6i9+8Qs+/elPj7n9M5/5DMuXLx91W0tLC4cOHRqTAEomk1xzzTXcfPPNdHR0cNNNN1FfX8/27dtPKeD6+uwpjZfFqbGxetqP8avN8N6LPYeORAzkHbHAsHJZQCppufkv+qf8OMUQ4oGnuSHB/c96rrywataWAw0WIxrrIJOZPOmQ6Q8xsTRNTcNNnL339PX1USqVyGazZLPDr6XmZrhw86yEu6C1t7fzT//0T7zyyis457DW4r2nv7+b1atXct55757Rk9CZmNMi88105nWpVOLuu+/m5z//OT09PSOvwXQ6zVVXXcV1111HVVXVDEYrMjV6v5bFSPNaZGmYUgLoiiuu4KWXXhpz+5133snjjz8+5nZrx1YO/Omf/unI18uXL+dDH/oQd9111ykngHp6ckSRO6XvkcXDmOEPqGPHBpjCLuZTkrSQPH4OMTgAR48UTvmxn9xd4uKzDLveKPHUTsvaCapupqswVCSXj8gkhvsEdXY7CiVPEDM011qa3rKcrVCI6O/P09GR57nnnuG+++7ljTdexxgIghjnn38Bl112OevXb5iVWBey9vY3+OpXv4z3ntbWtlHvacVikR//+Ge8+mo7n/70LaTT6Wk912zMaZFKm+68LhaL/N3f/S3PPLODZctaaG1dMZJYz+VyfOc7d/Dkk0/zmc98lpqa2hmOXmR8er+WxUjzWiYSBFbFF4vQtJaAtbW10dk5eoecrq6uMVVB3nu+9rWvceONN9LQ0DByWyz2zp5eb04yvJX77D32qSqGYKzBAEd7HWtaZycBtGlVwI7dZd445Dna5zCAtcPL19oPRmRShrPXx6ipGt4XvrG6xDe+8Q/s2PEU1dXVrFq1GmstpVKJp556kscee5Tf+I0bueqqX1UT4+PK5TL/8A/fwFrLsmUtwOg5kUgkWbNmLbt37+Kuu37OBz7w72bkeWdzTotUyjud1z/96Y959tmnWbt2/UgC9sTjZDJZ1q5dz759e/mXf7mdT35ybIWyyGzS+7UsRprXIkvDtJp8bN68mWKxOLKb1wMPPIAxhne/+92jxhljeOihh/jWt74FQHd3N9/5zne4/vrrp/P0IrMilUqd8vesWjb839nOoaxfEfDK/pAjvRE1WaitMlRnDLVZQ20VhJHnyV1lXt4bsnVTnPvu+i47duxgzZq1LFvWMnIilUgkWLnyNFpaWvnOd77Ns88+PbuBLyC7d++is7OD5uZlE44xxtDaupxf/vIBhoaG5jA6kcUvl8vxi188MKb67q2MMaxYsZLnnnuGjo7DcxyhiIiIyMI0rQRQLBbjr/7qr/jKV77Cddddx2233catt9460hT1E5/4BPfeey8A/+2//Td27tzJ9ddfz2//9m/z0Y9+lCuvvHLaP4DIbKg/xWXQ61emcM7jPDTVzl7z5IdfKNNQY4ev0rjR2SZjDJmUIYw8/397dx5cZX3vcfzzPGfNyXqyQRYSSAJBdgQXFgUcUdQCMi6gXSy1thSp2uroOMztdG6nTidtR+da2ylcRq3trVyltxdriy3cKovixmIpCpINQxIInJCF5ORsz/0jmkohS8lyyMP79Q+c8/zO+X0P85DJ+czv9/1V18c0pTCgt97apYKCgm4bFSckJCgtLU1/+tOrisXYWilJ+/btkdvt6XVFVGJiotrb21RRUT5ElQGXhkOHPlQw2N5rj63OVcSWPvhg39AUBgAAMMz1awuYJE2YMEEbN24877X169d3/T03N/esx8DF7PtfT9SDT57p09iS/M4/TzVZGp3jUOHIwQmAmlpjevvvYc0Y79TR+piO1HSeUuZ1d648ikSlUFjyeaTsdFPvvPu+YrHOU6x6kp6eoerqSlVVVaioiH5ALS0tcrvdfR4fDLICCBhIra2tfR7rcrkVCAQGsRoAAAD74Jxn4Dzuut6n62b0no8meqRbr/Wptd1Sa7ulW2b3vnLkQlXWRRWJSl63qXEFTl07zaXiPIecDkOWOreCTRvn1LXTPUpPMbXvQHWfGhQbhiHDMPkS9anExESFw+E+j/d4/vUtgwC611to/XmRSIT/gwAAAH3U7xVAgF29+O9+PfpMk17YEtL5Dp4bNUL6wpwEHa2PyjCke27yauKYwfsvFQqf3WMoyWeqtNBUaeG5Y50OQ5GoIauPEW9nCEQTaEmaPv1y7d69S5Zl9fhv0tbWJo/HqzFjioawOsD+CgtHS5Ki0Wi321elzsMkotGoSkrGDlFlAAAAwxsBENCDsvtTVXa/9MKWNv3v9qCazljyJxsameFUNCadabd03Uy3rp7oUm7m4Jz89ZkEj9Hn0xlCEUsFhcU6/vH+ruei0agMwzinqWo0GlUsFus68epSN378BGVkZOnUqZPKzMw67xjLslRfX6cbbrhJiYkcjwkMpJycXF122QQdOXJEeXl53Y5rbAwoMzNTl102YQirAwAAGL4IgIA++PIin7686B8NSUNhS+Go5HVJDsfQrJwpyXfI5zF0pt1SYkL3c1qWpUhE+sLNV2j9oZd06NBHqq09pmAwKEnKyMhQYeFoZWVlyzRNnTzZoLFjxyk/f9SQfI6Lndvt1j33fE0/+9lTOn68XtnZI85aCRQOh1VbW6OiomLdcMOiOFYK2JNhGFq27HY99dRPdOLEcWVlZZ+zGu/06dNqaWnR6tUPfNoMGgAAYODs37dXNYdPDNl8+eOydYOmDPo8/NYEXAC3y5C7720qBkSCx9C8y136w84Ojcl1yDTPHwLVnYqpYKRDqZ4mBQKNKi8/LL8/Q0lJSbIsS42NjWpoOKHs7BEqKipRR0dQt9yyhC1gnzN27Dg98MB39OKL/6WqqsqulVOxWFSm6dDMmVfpjjuWKykpKd6lAraUnz9Ka9Y8pOee+09VVVXK4/HI4/EoHI6ovb1NycnJWrXqfk2cOCnepQIAABtyTWuWO2voeqS68oampyEBEDCMLLzCo6q6mA5URJSTYcrn/dzKlIil+lMxJSUYuuWKVv3yl09r1KgC+Xw+VVSUKxIJy+dLVEJCgsJhpyoqytXS0qwnnviJxo0rjeOnujgVF4/V44//myoqylVe/rGCwaBSU9M0ceJkZWWdf2sYgIEzevQYrV37fX300UG99947am5uVkKCT9OnX65Jk6b0qck9AAAA/oEACBhGvG5D9y1J0JbdHdq+L6wTjTEZkixJpiFNLXFqyTVevbH1VTU3N8vtdquh4YQikYjOnDmjkydPKikpSSkpqZo+fYYsK6rk5OR4f6yLlmmaKikZS5NZIE5cLpcmT56qyZOnxrsUAACAYY8ACBhmvG5Dt17r1cIrPTp8NKK2oCWn09CYHIey/aZaW1u1bdtrqqioUCjUIY/Hq4yMDKWnpysYbFdHR0g+n08FBYU6ebJBb765Q2PHjhu0epubm7R//z7V1ByVJOXnF2jatOlKTk75l97n9OlGHT9+XJZlye9P14gRNK0GAAAAgL4iAAKGqUSvoenjzm1EVFVVoQMHDsjr9SgtzX/2axKTlJjYGabs2fOeJkyYpMrKikGpLxqN6o9//IO2bn1N4XBYbrdbkrRjx3a9/PJGLVy4SDfddEuPxzxLUm3tMW3d+preffcdWZ8eg2ZZliZMmKSFC29k+xoAAAAA9AEBEGAz+/fvVUdHUJmZmd2OSU1NUyAQUCBwUtnZIwe8hlgsppdf3qht2/6i/PxR8nrPbmoWDAb1yiu/15kzrbrzzru6bUB95MjH+vnP/0PhcFgjRozsCpEikYiOHDmsv//9gO65Z6WuumrWgH8GAAAAALATM94FABg44XBYe/fukdeboEgk0uNYt9ulI0eOqKCgcMDrOHToI73++jYVFo4+J/yRJK/Xq4KCQr3++v/p0KGPzvsezc1NWr/+F3I6XRo1qqAr/JEkp9OpnJxcZWZm6te/fl7V1VUD/hkAAAAAwE4IgAAbaW5uVltbmwoLR6ut7UyPYz0ej5qbmzR79twBr2Pnzu1yuz1yuc7dovYZl8slt9ujXbt2nPf63r171Nzc0uNKps+OYd+5c3v/CgYAAAAAmyMAAmzksx45Y8aMkWEYCgbbux3X3NyslJTUAW8A3dHRob/9bb+ysrJ7HZuVlan9+/cqFAqdc2379r8qLS2t1/fIzh6hd97ZrTNneg68AAAAAOBSRgAE2EhKSoqSk5Nlmg7NmHGFIpGImppOKxQKybIsWZaltrY2nT59Wl5vgubNW9DjKp0LEQ6HFIlE+vS+TqdL0Wj0nADIsiydOnVKPl9Cr+/h8XgUDkcIgAAAAACgBwRAgI243W5dc818nTzZoKysbM2de62Ki8cqHA6rqalJTU2n5fV6NGXKVI0ZM0aLFt084DV4PF55PB4Fg8Fex3Z0dMjj8Zy3T5DT6VQ0Guv1PTpXPVm9niYGAAAAAJcyTgEDbObqq2dr+/a/qqHhhLKyslVaOl4lJZ0hkGFILpdbtbXHNGpUoSZNmjLg87tcLl111Szt2rVDo0YV9Di2oeGErrlmnpzOs38UGYahSZOmaO/e9+Xz+Xp8j8bGgHJzc+X3+3scBwAAAACXMlYAATaTmZmpVavWyDRNVVVVqqWlRaZpyuPxqL29XVVVlcrMzNQ3v3m/EhJ632J1IWbNmtO13aw7n12bNWvOea/Pnj1XoVCox9PMLMtSY2OjFiy4XqbJjzMAAAAA6A7fmAAbKioq1mOPrdWSJcsUiYRVXV2l6uoqud0e3X33l/Td7z6q7OzemzRfqMLC0brttjtVX1+nQOBUV3NqqTO0CQROqb6+TrffvrzbY+hLSsZq1qzZqq6uVDgcPud6LBZTdXWVSkvH6/LLZw7aZwEAAAAAO2ALGGBT6ekZWrToZl1//Q1qbW2RYRhKSkoesl458+dfp5SUVL366mZVV1dJMiRJlhVTTk6u7rzzLk2fPqPb15umqRUrviSn06WdO7fLNE0lJSXLMAydOdOqSCSsyy6bpK9+9WuDtpIJAAAAAOyCAAiwOafTqbS0oe+PYxiGZsyYqWnTpqu8/IhOnmyQJGVlZauoqLhPQZTb7dZdd31J8+dfp7ff3q3y8o8Vi8U0efIUXX31LI0eXcTWLwAAAADoAwIgAIPK4XBo3LhSjRtXekGvNwxDubl5WrbstgGuDAAAAAAuHQRAAHARCgRO6f3339WuXTvV0tKshIQEzZx5pa688mrl5ubFuzwAAAAAwwwBEAD0QSwWU2VluT7++GO1t7crNTVVEydO0ogRIwd8rj173tOvfvWsQqGQ0tL8SklJVSQS1l/+8pq2bv2zlixZpoULb5RhGAM+NwAAAAB7IgACgF5UVlboxRd/o6NHq2UYhhwOh6LRqDZt2qipUy/X8uV3KTU1bUDm+vDDg9qwYZ38/nSlpKR87kqCkpNTFAwGtWnTf8vj8WjevAUDMicAAAAA+yMAAoAeVFSU6+mnn5RhmBo9esxZq24ikYj279+rEyeO64EHvqOUlNR+zRWLxfT732+Sz+f7p/DnH7xer0aMGKnNm/9HV1xxlXw+X7/mBAAAAHBp4PgcAOhGJBLRCy88J4fDoZEjR56z5crpdKqgoFDHjtXo1Vdf6fd8VVUV+uSTamVkZPY4LikpSW1tbTpw4G/9nhMAAADApYEACAC6cfjwIdXX1ykrK7vbMYZhKCcnV7t3v6nm5qZ+zVdfXy/J7NPR9i6XS1VVlf2aDwAAAMClgwAIALpx4MAHcjqdvTZb9nq9CofDKi8/0q/5YrGYJKtPYw3DUCwW7dd8AAAAAC4d9AACgG60trbK5XL3aaxhGOro6OjXfGlpflmWJcuyeg2dwuHwoJxAhviKqUVhx4eydEaSS85YkRxWjgxx4hsAAAD6hwAIALqRnJyicDjUp7GWZcnr9fZrvtLS8fL709Xc3KzU1O4bSodCITkcDk2bNr1f8+HiEVO72pyb1eHcLssISbIky5AMQ+7oRPnCt8tp5cW7TAAAAAxjbAEDgG5MmTJVkUhYltXztqz29na53R6VlIzr13wul0uLFt2skycbFAqdP3iKRqOqqTmquXOvld+f3q/5cHGwFFSL+xcKuv4s08qQMzZGzliRnNYYOWL5CpuH1Oz5iSLG0XiXCgAAgGGMAAgAulFSMlb5+QU6fry+2zGWZamurlZz516jpKSkfs85d+61uvHGm3Xs2Ceqq6tTJBKR1NkfqKGhQVVVlZox40rdeutt/Z4LF4d252sKOw7IESuSobNXkRlyymHly1JELe4NskTfJwAAAFwYAiAA6IbD4dA993xNTqdTx47VKBo9+8t3KNShqqoqFReX6KabvjAgc5qmqWXLbtPq1Q+quLhYx47VqLq6Sp98Uq3MzEzde+83tXLl1+XxeAZkPsRXTO1qd/5VZqznPj+mNVJRs0Zh89AQVgcAAAA7oQcQAPQgP3+UHnzwYb388kYdOnRIlhX79Iohp9OhuXPnaunS25SYmDhgcxqGocmTp2jy5ClqbAx0bTHLyMjotTk0hpeIeUSWzsih7B7HGTIky62QY4/csQlDVB0AAADshAAIAHqRl5evBx74rmprj6m8/IjC4bASExM1fvxlSkvzD+rcfn+6/IM7BeLIMtolw5R6bjMlSTLkUcxoGvyiAAAAYEsEQADQB4ZhKC8vX3l5+fEuBTZiWB71Kf2RZCks0/INbkEAAACwLXoAAQAQJ85YkQzLJUvtPY6zZElGUO7otCGqDAAAAHZDAAQAQJyYSpYnco2iZl1nyNONmBGQaWXKFZs4hNUBAADATgiAAACII1/kZjlioxQ1q2UpctY1S5aiRoMso01Joa/KkDtOVQIAAGC4owcQAABxZCpFqR0PqtX9G4Uc+zqftFySEZEUlSOWq6TQKrlipXGtEwAAAMMbARAAAHFmKk0pofsVMWoVcuxXTI0y5ZMzNk6uWKkMOeJdIgAAAIY5AiAAAC4STitXzkhuvMsAAACADdEDCAAAAAAAwOYIgAAAAAAAAGyOAAgAAAAAAMDmCIAAAAAAAABsjgAIAAAAAADA5giAAAAAAAAAbI4ACAAAAAAAwOYIgAAAAAAAAGyOAAgAAAAAAMDmCIAAAAAAAABsjgAIAAAAAADA5giAAAAAAAAAbI4ACAAAAAAAwOYIgAAAAAAAAGyOAAgAAAAAAMDmCIAAAAAAAABsjgAIAMWPQIEAAAU3SURBVAAAAADA5giAAAAAAAAAbI4ACAAAAAAAYBjZsWOHli5dqkWLFmnlypU6ceJEr68hAAIAAAAAABgmAoGAHnnkEZWVlWnLli1asGCBHn/88V5f5xyC2gaUw0FmBe4D2A/3NOyI+xp2xH0NO+K+xj+71O+J7OzsuMxXXl6uYDB4zvWsrKyzatq5c6dKS0tVWloqSVqxYoXKysrU0NCgrKysbucZdgFQSkpCvEvARcDvT4x3CcCA4p6GHXFfw464r2FH3NfA2e6+e/mQzxkOh3Xvvfeqrq7unGtr1qzRt7/97a7H9fX1ysnJ6Xrsdrvl9/tVV1dnrwAIAAAAAADATpqamvTMM8+c99o/hzqWZckwjHPGmWbPK7cIgAAAAAAAAOIoMzNTmZmZfRqbm5ur3bt3dz0OhUJqbGxUbm5uj6+7tDf2AQAAAAAADCNz5szRwYMHdfjwYUnSSy+9pKlTpyo9Pb3H1xmWZVlDUSAAAAAAAAD6780339SPf/xjdXR0KCMjQz/60Y+Ul5fX42sIgAAAAAAAAGyOLWAAAAAAAAA2RwAEAAAAAABgcwRAAAAAAAAANkcABAAAAAAAYHMEQAAAAAAAADZHAAQAAAAAAGBzBEAYtl566SWtXLky3mUAF2zHjh1aunSpFi1apJUrV+rEiRPxLgkYME899ZTWrl0b7zKAAfHb3/5Wixcv1pIlS7R8+XJ98MEH8S4J6LdNmzZp8eLFWrx4sb7yla+osrIy3iUBGGQEQBh2AoGA1q5dqx/+8IeyLCve5QAXJBAI6JFHHlFZWZm2bNmiBQsW6PHHH493WUC/1dTU6P7779ezzz4b71KAAbFnzx6tW7dOzz//vDZv3qxVq1Zp9erV/A6CYa2iokI//elP9dxzz+mVV17RwoUL9b3vfS/eZQEYZARAGHY2b96s/Px8Pfroo/EuBbhgO3fuVGlpqUpLSyVJK1as0Ntvv62GhoY4Vwb0z8aNGzV79mxWaMI2UlNT9YMf/EDp6emSpClTpujUqVNqb2+Pc2XAhSsqKtIbb7yhjIwMRSIR1dbWyu/3x7ssAIPMGe8CgPN544039K1vfeuc51evXq01a9ZIkn73u98NdVnAgKmvr1dOTk7XY7fbLb/fr7q6OmVlZcWxMqB/Hn74YUnS008/HedKgIFRXFys4uJiSVIsFtMTTzyh+fPny+fzxbkyoH9cLpfee+89Pfjgg2pvb9eGDRviXRKAQUYAhIvSvHnzdPDgwXiXAQway7JkGMY5z5smCzMB4GLU2tqqRx99VIFAQOvWrYt3OcCAmDlzpnbt2qWtW7fqG9/4hrZt26aUlJR4lwVgkPBNAwDiIDc3V8ePH+96HAqF1NjYqNzc3DhWBQA4n8rKSt1+++1KSkrS888/zxdkDHs1NTV66623uh5ff/31crlcOnr0aByrAjDYCIAAIA7mzJmjgwcP6vDhw5I6T7WbOnVqV48JAMDFoba2Vl/84hd1xx13qKysTB6PJ94lAf3W1NSkhx56SPX19ZI62y+Yptm13RGAPbEFDADiID09XU8++aQee+wxdXR0KCMjQ2VlZfEuCwDwTzZs2KDm5mZt3rxZmzdv7np+3bp1GjFiRBwrAy7cxIkT9dhjj+m+++6TaZpKSUnR+vXrlZCQEO/SAAwiw+IMSwAAAAAAAFtjCxgAAAAAAIDNEQABAAAAAADYHAEQAAAAAACAzREAAQAAAAAA2BwBEAAAAAAAgM0RAAEAAAAAANgcARAAAAAAAIDNEQABAAAAAADY3P8DZGSTOniDzGoAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1600x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Spectral clustering | Plot the ego-network in 2 dimensions\n",
    "\n",
    "plt.figure(num=None, figsize=(20, 10), dpi=80, facecolor='w', edgecolor='k')\n",
    "plt.scatter(results_df['dim1'], results_df['dim2'], s = 100  ,c= results_df['spectral'], \n",
    "            alpha=0.5, cmap=plt.cm.get_cmap('nipy_spectral', 10))\n",
    "plt.colorbar();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "More on Clustering Algorithms:  http://scikit-learn.org/stable/modules/clustering.html"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## 4. Recommend new friends to follow"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that we have PageRank in place for nodes in social-graph, we can ask for recommendations on the basis of top-ranked nodes in the graph. E.g. To get 20 recommendations, after looking at friends of top PageRank scoring nodes in my network"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 4a. Reco: After looking at top nodes in full ego-network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'Who-To-Follow' reco after looking at  35 friends' friends:\n",
      "... Cache-Miss for 20 nodes. Updating Cache...\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Freq</th>\n",
       "      <th>UserName</th>\n",
       "      <th>Followers</th>\n",
       "      <th>Friends</th>\n",
       "      <th>Location</th>\n",
       "      <th>Created</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FullName</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>Yann LeCun</td>\n",
       "      <td>11</td>\n",
       "      <td>ylecun</td>\n",
       "      <td>157351</td>\n",
       "      <td>231</td>\n",
       "      <td>New York</td>\n",
       "      <td>Jun-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Andrej Karpathy</td>\n",
       "      <td>10</td>\n",
       "      <td>karpathy</td>\n",
       "      <td>231814</td>\n",
       "      <td>507</td>\n",
       "      <td>Stanford</td>\n",
       "      <td>Apr-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Andreas Mueller</td>\n",
       "      <td>9</td>\n",
       "      <td>amuellerml</td>\n",
       "      <td>35869</td>\n",
       "      <td>894</td>\n",
       "      <td>NYC</td>\n",
       "      <td>Jan-2012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>DeepMind</td>\n",
       "      <td>9</td>\n",
       "      <td>DeepMindAI</td>\n",
       "      <td>258765</td>\n",
       "      <td>115</td>\n",
       "      <td>London, UK</td>\n",
       "      <td>Jan-2016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Stanford NLP Group</td>\n",
       "      <td>9</td>\n",
       "      <td>stanfordnlp</td>\n",
       "      <td>81113</td>\n",
       "      <td>100</td>\n",
       "      <td>Stanford, CA, USA</td>\n",
       "      <td>Feb-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>François Chollet</td>\n",
       "      <td>9</td>\n",
       "      <td>fchollet</td>\n",
       "      <td>162588</td>\n",
       "      <td>595</td>\n",
       "      <td>Mountain View, CA</td>\n",
       "      <td>Aug-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Fei-Fei Li</td>\n",
       "      <td>9</td>\n",
       "      <td>drfeifei</td>\n",
       "      <td>324869</td>\n",
       "      <td>264</td>\n",
       "      <td>Stanford, CA, U.S.A.</td>\n",
       "      <td>Apr-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Hugo Larochelle</td>\n",
       "      <td>8</td>\n",
       "      <td>hugo_larochelle</td>\n",
       "      <td>72871</td>\n",
       "      <td>494</td>\n",
       "      <td></td>\n",
       "      <td>Jun-2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Jeff Dean</td>\n",
       "      <td>8</td>\n",
       "      <td>JeffDean</td>\n",
       "      <td>118243</td>\n",
       "      <td>2156</td>\n",
       "      <td></td>\n",
       "      <td>Sep-2017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Jeremy Howard</td>\n",
       "      <td>8</td>\n",
       "      <td>jeremyphoward</td>\n",
       "      <td>69888</td>\n",
       "      <td>1286</td>\n",
       "      <td>San Francisco</td>\n",
       "      <td>Aug-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Chris Olah</td>\n",
       "      <td>8</td>\n",
       "      <td>ch402</td>\n",
       "      <td>47210</td>\n",
       "      <td>172</td>\n",
       "      <td>San Francisco, CA</td>\n",
       "      <td>Jun-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Sundar Pichai</td>\n",
       "      <td>8</td>\n",
       "      <td>sundarpichai</td>\n",
       "      <td>2483187</td>\n",
       "      <td>336</td>\n",
       "      <td></td>\n",
       "      <td>Mar-2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Moritz Hardt</td>\n",
       "      <td>7</td>\n",
       "      <td>mrtz</td>\n",
       "      <td>8854</td>\n",
       "      <td>327</td>\n",
       "      <td>Berkeley, CA</td>\n",
       "      <td>Aug-2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Pete Warden</td>\n",
       "      <td>7</td>\n",
       "      <td>petewarden</td>\n",
       "      <td>18345</td>\n",
       "      <td>2014</td>\n",
       "      <td>San Francisco, CA</td>\n",
       "      <td>May-2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Oriol Vinyals</td>\n",
       "      <td>7</td>\n",
       "      <td>OriolVinyalsML</td>\n",
       "      <td>77859</td>\n",
       "      <td>48</td>\n",
       "      <td>London, England</td>\n",
       "      <td>Oct-2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Ryan Adams</td>\n",
       "      <td>7</td>\n",
       "      <td>ryan_p_adams</td>\n",
       "      <td>26299</td>\n",
       "      <td>872</td>\n",
       "      <td></td>\n",
       "      <td>Jan-2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Hilary Mason</td>\n",
       "      <td>7</td>\n",
       "      <td>hmason</td>\n",
       "      <td>118024</td>\n",
       "      <td>1654</td>\n",
       "      <td>NYC</td>\n",
       "      <td>Feb-2007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Monica Rogati</td>\n",
       "      <td>7</td>\n",
       "      <td>mrogati</td>\n",
       "      <td>48357</td>\n",
       "      <td>566</td>\n",
       "      <td>Sunnyvale, CA</td>\n",
       "      <td>Nov-2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Gautam Gambhir</td>\n",
       "      <td>7</td>\n",
       "      <td>GautamGambhir</td>\n",
       "      <td>9938638</td>\n",
       "      <td>121</td>\n",
       "      <td></td>\n",
       "      <td>Dec-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Suresh Raina🇮🇳</td>\n",
       "      <td>7</td>\n",
       "      <td>ImRaina</td>\n",
       "      <td>16857878</td>\n",
       "      <td>300</td>\n",
       "      <td>India</td>\n",
       "      <td>Mar-2010</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    Freq         UserName  Followers  Friends  \\\n",
       "FullName                                                        \n",
       "Yann LeCun            11           ylecun     157351      231   \n",
       "Andrej Karpathy       10         karpathy     231814      507   \n",
       "Andreas Mueller        9       amuellerml      35869      894   \n",
       "DeepMind               9       DeepMindAI     258765      115   \n",
       "Stanford NLP Group     9      stanfordnlp      81113      100   \n",
       "François Chollet       9         fchollet     162588      595   \n",
       "Fei-Fei Li             9         drfeifei     324869      264   \n",
       "Hugo Larochelle        8  hugo_larochelle      72871      494   \n",
       "Jeff Dean              8         JeffDean     118243     2156   \n",
       "Jeremy Howard          8    jeremyphoward      69888     1286   \n",
       "Chris Olah             8            ch402      47210      172   \n",
       "Sundar Pichai          8     sundarpichai    2483187      336   \n",
       "Moritz Hardt           7             mrtz       8854      327   \n",
       "Pete Warden            7       petewarden      18345     2014   \n",
       "Oriol Vinyals          7   OriolVinyalsML      77859       48   \n",
       "Ryan Adams             7     ryan_p_adams      26299      872   \n",
       "Hilary Mason           7           hmason     118024     1654   \n",
       "Monica Rogati          7          mrogati      48357      566   \n",
       "Gautam Gambhir         7    GautamGambhir    9938638      121   \n",
       "Suresh Raina🇮🇳         7          ImRaina   16857878      300   \n",
       "\n",
       "                                Location   Created  \n",
       "FullName                                            \n",
       "Yann LeCun                      New York  Jun-2009  \n",
       "Andrej Karpathy                 Stanford  Apr-2009  \n",
       "Andreas Mueller                      NYC  Jan-2012  \n",
       "DeepMind                      London, UK  Jan-2016  \n",
       "Stanford NLP Group     Stanford, CA, USA  Feb-2010  \n",
       "François Chollet       Mountain View, CA  Aug-2009  \n",
       "Fei-Fei Li          Stanford, CA, U.S.A.  Apr-2010  \n",
       "Hugo Larochelle                           Jun-2015  \n",
       "Jeff Dean                                 Sep-2017  \n",
       "Jeremy Howard              San Francisco  Aug-2010  \n",
       "Chris Olah             San Francisco, CA  Jun-2010  \n",
       "Sundar Pichai                             Mar-2008  \n",
       "Moritz Hardt                Berkeley, CA  Aug-2008  \n",
       "Pete Warden            San Francisco, CA  May-2008  \n",
       "Oriol Vinyals            London, England  Oct-2015  \n",
       "Ryan Adams                                Jan-2015  \n",
       "Hilary Mason                         NYC  Feb-2007  \n",
       "Monica Rogati              Sunnyvale, CA  Nov-2008  \n",
       "Gautam Gambhir                            Dec-2009  \n",
       "Suresh Raina🇮🇳                     India  Mar-2010  "
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "discover_Friends_toFollow(ids_of_interest = index,\n",
    "                          friend_list = friends_of_self, \n",
    "                          prop = .5 , count =  20).fillna('').set_index('FullName')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Recommendations on the basis of a specific cluster outcome. \n",
    "E.g. Nodes to follow on the basis of top PageRank nodes in 'Spectral Clustering::Data' clusters. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Reco: 4b. After looking at Research (AI) community clustser (Spectral clustering)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'Who-To-Follow' reco after looking at   6 friends' friends:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Freq</th>\n",
       "      <th>UserName</th>\n",
       "      <th>Followers</th>\n",
       "      <th>Friends</th>\n",
       "      <th>Location</th>\n",
       "      <th>Created</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FullName</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>Yann LeCun</td>\n",
       "      <td>6</td>\n",
       "      <td>ylecun</td>\n",
       "      <td>157351</td>\n",
       "      <td>231</td>\n",
       "      <td>New York</td>\n",
       "      <td>Jun-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Hugo Larochelle</td>\n",
       "      <td>5</td>\n",
       "      <td>hugo_larochelle</td>\n",
       "      <td>72871</td>\n",
       "      <td>494</td>\n",
       "      <td></td>\n",
       "      <td>Jun-2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Jeff Dean</td>\n",
       "      <td>5</td>\n",
       "      <td>JeffDean</td>\n",
       "      <td>118243</td>\n",
       "      <td>2156</td>\n",
       "      <td></td>\n",
       "      <td>Sep-2017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Chris Olah</td>\n",
       "      <td>5</td>\n",
       "      <td>ch402</td>\n",
       "      <td>47210</td>\n",
       "      <td>172</td>\n",
       "      <td>San Francisco, CA</td>\n",
       "      <td>Jun-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>DeepMind</td>\n",
       "      <td>5</td>\n",
       "      <td>DeepMindAI</td>\n",
       "      <td>258765</td>\n",
       "      <td>115</td>\n",
       "      <td>London, UK</td>\n",
       "      <td>Jan-2016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>François Chollet</td>\n",
       "      <td>5</td>\n",
       "      <td>fchollet</td>\n",
       "      <td>162588</td>\n",
       "      <td>595</td>\n",
       "      <td>Mountain View, CA</td>\n",
       "      <td>Aug-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Jeremy Howard</td>\n",
       "      <td>5</td>\n",
       "      <td>jeremyphoward</td>\n",
       "      <td>69888</td>\n",
       "      <td>1286</td>\n",
       "      <td>San Francisco</td>\n",
       "      <td>Aug-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Andrej Karpathy</td>\n",
       "      <td>5</td>\n",
       "      <td>karpathy</td>\n",
       "      <td>231814</td>\n",
       "      <td>507</td>\n",
       "      <td>Stanford</td>\n",
       "      <td>Apr-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Oriol Vinyals</td>\n",
       "      <td>4</td>\n",
       "      <td>OriolVinyalsML</td>\n",
       "      <td>77859</td>\n",
       "      <td>48</td>\n",
       "      <td>London, England</td>\n",
       "      <td>Oct-2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>distillpub</td>\n",
       "      <td>4</td>\n",
       "      <td>distillpub</td>\n",
       "      <td>27915</td>\n",
       "      <td>81</td>\n",
       "      <td></td>\n",
       "      <td>Oct-2016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Douglas Eck</td>\n",
       "      <td>4</td>\n",
       "      <td>douglas_eck</td>\n",
       "      <td>8344</td>\n",
       "      <td>631</td>\n",
       "      <td>Mountain View, CA</td>\n",
       "      <td>Aug-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Ryan Adams</td>\n",
       "      <td>4</td>\n",
       "      <td>ryan_p_adams</td>\n",
       "      <td>26299</td>\n",
       "      <td>872</td>\n",
       "      <td></td>\n",
       "      <td>Jan-2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Stanford NLP Group</td>\n",
       "      <td>4</td>\n",
       "      <td>stanfordnlp</td>\n",
       "      <td>81113</td>\n",
       "      <td>100</td>\n",
       "      <td>Stanford, CA, USA</td>\n",
       "      <td>Feb-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Moritz Hardt</td>\n",
       "      <td>4</td>\n",
       "      <td>mrtz</td>\n",
       "      <td>8854</td>\n",
       "      <td>327</td>\n",
       "      <td>Berkeley, CA</td>\n",
       "      <td>Aug-2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Pete Warden</td>\n",
       "      <td>4</td>\n",
       "      <td>petewarden</td>\n",
       "      <td>18345</td>\n",
       "      <td>2014</td>\n",
       "      <td>San Francisco, CA</td>\n",
       "      <td>May-2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Anthony Goldbloom</td>\n",
       "      <td>4</td>\n",
       "      <td>antgoldbloom</td>\n",
       "      <td>35042</td>\n",
       "      <td>575</td>\n",
       "      <td>San Francisco</td>\n",
       "      <td>Jul-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>mat kelcey</td>\n",
       "      <td>3</td>\n",
       "      <td>mat_kelcey</td>\n",
       "      <td>6646</td>\n",
       "      <td>1071</td>\n",
       "      <td>Victoria, Australia</td>\n",
       "      <td>Mar-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Derek Murray</td>\n",
       "      <td>3</td>\n",
       "      <td>mrry</td>\n",
       "      <td>3643</td>\n",
       "      <td>345</td>\n",
       "      <td>San Francisco, CA</td>\n",
       "      <td>Feb-2007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Taehoon Kim</td>\n",
       "      <td>3</td>\n",
       "      <td>carpedm20</td>\n",
       "      <td>2540</td>\n",
       "      <td>592</td>\n",
       "      <td>San Francisco, CA</td>\n",
       "      <td>Jan-2011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Melody Guan ʕᵔᴥᵔʔ</td>\n",
       "      <td>3</td>\n",
       "      <td>MelodyGuan</td>\n",
       "      <td>3117</td>\n",
       "      <td>699</td>\n",
       "      <td>Stanford, CA</td>\n",
       "      <td>Apr-2011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Sundar Pichai</td>\n",
       "      <td>3</td>\n",
       "      <td>sundarpichai</td>\n",
       "      <td>2483187</td>\n",
       "      <td>336</td>\n",
       "      <td></td>\n",
       "      <td>Mar-2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Luke Metz</td>\n",
       "      <td>3</td>\n",
       "      <td>Luke_Metz</td>\n",
       "      <td>4026</td>\n",
       "      <td>992</td>\n",
       "      <td>Mountain View, CA</td>\n",
       "      <td>Oct-2012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Jasper</td>\n",
       "      <td>3</td>\n",
       "      <td>latentjasper</td>\n",
       "      <td>4198</td>\n",
       "      <td>250</td>\n",
       "      <td></td>\n",
       "      <td>Apr-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Denny Britz</td>\n",
       "      <td>3</td>\n",
       "      <td>dennybritz</td>\n",
       "      <td>39745</td>\n",
       "      <td>380</td>\n",
       "      <td></td>\n",
       "      <td>Jan-2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>👩‍💻 DynamicWebPaige @ #TFDocsSprint ✍️</td>\n",
       "      <td>3</td>\n",
       "      <td>DynamicWebPaige</td>\n",
       "      <td>30519</td>\n",
       "      <td>1205</td>\n",
       "      <td>179°56′39.4″, +0°2′46.2″ 8k pc</td>\n",
       "      <td>Dec-2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Dumitru Erhan</td>\n",
       "      <td>3</td>\n",
       "      <td>doomie</td>\n",
       "      <td>4806</td>\n",
       "      <td>1070</td>\n",
       "      <td>San Francisco, CA</td>\n",
       "      <td>Apr-2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>hardmaru</td>\n",
       "      <td>3</td>\n",
       "      <td>hardmaru</td>\n",
       "      <td>64655</td>\n",
       "      <td>657</td>\n",
       "      <td>東京</td>\n",
       "      <td>Nov-2014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Rajat Monga</td>\n",
       "      <td>3</td>\n",
       "      <td>rajatmonga</td>\n",
       "      <td>13248</td>\n",
       "      <td>214</td>\n",
       "      <td></td>\n",
       "      <td>Nov-2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Eric Jang</td>\n",
       "      <td>3</td>\n",
       "      <td>ericjang11</td>\n",
       "      <td>14398</td>\n",
       "      <td>1434</td>\n",
       "      <td></td>\n",
       "      <td>Jan-2014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>a16z</td>\n",
       "      <td>3</td>\n",
       "      <td>a16z</td>\n",
       "      <td>346731</td>\n",
       "      <td>22</td>\n",
       "      <td>Andreessen Horowitz is a tech VC firm based in...</td>\n",
       "      <td>Aug-2009</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                        Freq         UserName  Followers  \\\n",
       "FullName                                                                   \n",
       "Yann LeCun                                 6           ylecun     157351   \n",
       "Hugo Larochelle                            5  hugo_larochelle      72871   \n",
       "Jeff Dean                                  5         JeffDean     118243   \n",
       "Chris Olah                                 5            ch402      47210   \n",
       "DeepMind                                   5       DeepMindAI     258765   \n",
       "François Chollet                           5         fchollet     162588   \n",
       "Jeremy Howard                              5    jeremyphoward      69888   \n",
       "Andrej Karpathy                            5         karpathy     231814   \n",
       "Oriol Vinyals                              4   OriolVinyalsML      77859   \n",
       "distillpub                                 4       distillpub      27915   \n",
       "Douglas Eck                                4      douglas_eck       8344   \n",
       "Ryan Adams                                 4     ryan_p_adams      26299   \n",
       "Stanford NLP Group                         4      stanfordnlp      81113   \n",
       "Moritz Hardt                               4             mrtz       8854   \n",
       "Pete Warden                                4       petewarden      18345   \n",
       "Anthony Goldbloom                          4     antgoldbloom      35042   \n",
       "mat kelcey                                 3       mat_kelcey       6646   \n",
       "Derek Murray                               3             mrry       3643   \n",
       "Taehoon Kim                                3        carpedm20       2540   \n",
       "Melody Guan ʕᵔᴥᵔʔ                          3       MelodyGuan       3117   \n",
       "Sundar Pichai                              3     sundarpichai    2483187   \n",
       "Luke Metz                                  3        Luke_Metz       4026   \n",
       "Jasper                                     3     latentjasper       4198   \n",
       "Denny Britz                                3       dennybritz      39745   \n",
       "👩‍💻 DynamicWebPaige @ #TFDocsSprint ✍️     3  DynamicWebPaige      30519   \n",
       "Dumitru Erhan                              3           doomie       4806   \n",
       "hardmaru                                   3         hardmaru      64655   \n",
       "Rajat Monga                                3       rajatmonga      13248   \n",
       "Eric Jang                                  3       ericjang11      14398   \n",
       "a16z                                       3             a16z     346731   \n",
       "\n",
       "                                        Friends  \\\n",
       "FullName                                          \n",
       "Yann LeCun                                  231   \n",
       "Hugo Larochelle                             494   \n",
       "Jeff Dean                                  2156   \n",
       "Chris Olah                                  172   \n",
       "DeepMind                                    115   \n",
       "François Chollet                            595   \n",
       "Jeremy Howard                              1286   \n",
       "Andrej Karpathy                             507   \n",
       "Oriol Vinyals                                48   \n",
       "distillpub                                   81   \n",
       "Douglas Eck                                 631   \n",
       "Ryan Adams                                  872   \n",
       "Stanford NLP Group                          100   \n",
       "Moritz Hardt                                327   \n",
       "Pete Warden                                2014   \n",
       "Anthony Goldbloom                           575   \n",
       "mat kelcey                                 1071   \n",
       "Derek Murray                                345   \n",
       "Taehoon Kim                                 592   \n",
       "Melody Guan ʕᵔᴥᵔʔ                           699   \n",
       "Sundar Pichai                               336   \n",
       "Luke Metz                                   992   \n",
       "Jasper                                      250   \n",
       "Denny Britz                                 380   \n",
       "👩‍💻 DynamicWebPaige @ #TFDocsSprint ✍️     1205   \n",
       "Dumitru Erhan                              1070   \n",
       "hardmaru                                    657   \n",
       "Rajat Monga                                 214   \n",
       "Eric Jang                                  1434   \n",
       "a16z                                         22   \n",
       "\n",
       "                                                                                 Location  \\\n",
       "FullName                                                                                    \n",
       "Yann LeCun                                                                       New York   \n",
       "Hugo Larochelle                                                                             \n",
       "Jeff Dean                                                                                   \n",
       "Chris Olah                                                              San Francisco, CA   \n",
       "DeepMind                                                                       London, UK   \n",
       "François Chollet                                                        Mountain View, CA   \n",
       "Jeremy Howard                                                               San Francisco   \n",
       "Andrej Karpathy                                                                  Stanford   \n",
       "Oriol Vinyals                                                             London, England   \n",
       "distillpub                                                                                  \n",
       "Douglas Eck                                                             Mountain View, CA   \n",
       "Ryan Adams                                                                                  \n",
       "Stanford NLP Group                                                      Stanford, CA, USA   \n",
       "Moritz Hardt                                                                 Berkeley, CA   \n",
       "Pete Warden                                                             San Francisco, CA   \n",
       "Anthony Goldbloom                                                           San Francisco   \n",
       "mat kelcey                                                            Victoria, Australia   \n",
       "Derek Murray                                                            San Francisco, CA   \n",
       "Taehoon Kim                                                             San Francisco, CA   \n",
       "Melody Guan ʕᵔᴥᵔʔ                                                            Stanford, CA   \n",
       "Sundar Pichai                                                                               \n",
       "Luke Metz                                                               Mountain View, CA   \n",
       "Jasper                                                                                      \n",
       "Denny Britz                                                                                 \n",
       "👩‍💻 DynamicWebPaige @ #TFDocsSprint ✍️                     179°56′39.4″, +0°2′46.2″ 8k pc   \n",
       "Dumitru Erhan                                                           San Francisco, CA   \n",
       "hardmaru                                                                               東京   \n",
       "Rajat Monga                                                                                 \n",
       "Eric Jang                                                                                   \n",
       "a16z                                    Andreessen Horowitz is a tech VC firm based in...   \n",
       "\n",
       "                                         Created  \n",
       "FullName                                          \n",
       "Yann LeCun                              Jun-2009  \n",
       "Hugo Larochelle                         Jun-2015  \n",
       "Jeff Dean                               Sep-2017  \n",
       "Chris Olah                              Jun-2010  \n",
       "DeepMind                                Jan-2016  \n",
       "François Chollet                        Aug-2009  \n",
       "Jeremy Howard                           Aug-2010  \n",
       "Andrej Karpathy                         Apr-2009  \n",
       "Oriol Vinyals                           Oct-2015  \n",
       "distillpub                              Oct-2016  \n",
       "Douglas Eck                             Aug-2010  \n",
       "Ryan Adams                              Jan-2015  \n",
       "Stanford NLP Group                      Feb-2010  \n",
       "Moritz Hardt                            Aug-2008  \n",
       "Pete Warden                             May-2008  \n",
       "Anthony Goldbloom                       Jul-2010  \n",
       "mat kelcey                              Mar-2009  \n",
       "Derek Murray                            Feb-2007  \n",
       "Taehoon Kim                             Jan-2011  \n",
       "Melody Guan ʕᵔᴥᵔʔ                       Apr-2011  \n",
       "Sundar Pichai                           Mar-2008  \n",
       "Luke Metz                               Oct-2012  \n",
       "Jasper                                  Apr-2009  \n",
       "Denny Britz                             Jan-2008  \n",
       "👩‍💻 DynamicWebPaige @ #TFDocsSprint ✍️  Dec-2008  \n",
       "Dumitru Erhan                           Apr-2008  \n",
       "hardmaru                                Nov-2014  \n",
       "Rajat Monga                             Nov-2008  \n",
       "Eric Jang                               Jan-2014  \n",
       "a16z                                    Aug-2009  "
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "favorite_cluster_df = results_df [results_df['spectral'].isin([2])]\n",
    "favorite_cluster_list = list(favorite_cluster_df.index)\n",
    "\n",
    "discover_Friends_toFollow(ids_of_interest = favorite_cluster_list, \n",
    "                          friend_list = friends_of_self, \n",
    "                          prop = .5, count = 30).fillna('').set_index('FullName')  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Reco: 4c. After looking at Bollywood actors clustser (Spectral clustering)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'Who-To-Follow' reco after looking at   4 friends' friends:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Freq</th>\n",
       "      <th>UserName</th>\n",
       "      <th>Followers</th>\n",
       "      <th>Friends</th>\n",
       "      <th>Location</th>\n",
       "      <th>Created</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FullName</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>Chulbul Pandey</td>\n",
       "      <td>3</td>\n",
       "      <td>BeingSalmanKhan</td>\n",
       "      <td>38222008</td>\n",
       "      <td>24</td>\n",
       "      <td>MUMBAI</td>\n",
       "      <td>Apr-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Abhishek Bachchan</td>\n",
       "      <td>3</td>\n",
       "      <td>juniorbachchan</td>\n",
       "      <td>14707884</td>\n",
       "      <td>410</td>\n",
       "      <td>Mumbai, India</td>\n",
       "      <td>Nov-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Karan Johar</td>\n",
       "      <td>3</td>\n",
       "      <td>karanjohar</td>\n",
       "      <td>16664147</td>\n",
       "      <td>485</td>\n",
       "      <td>Maharashtra, India</td>\n",
       "      <td>Jul-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Barack Obama</td>\n",
       "      <td>3</td>\n",
       "      <td>BarackObama</td>\n",
       "      <td>108917081</td>\n",
       "      <td>611754</td>\n",
       "      <td>Washington, DC</td>\n",
       "      <td>Mar-2007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Preity G Zinta</td>\n",
       "      <td>2</td>\n",
       "      <td>realpreityzinta</td>\n",
       "      <td>4806691</td>\n",
       "      <td>105</td>\n",
       "      <td>Catch me if you can....</td>\n",
       "      <td>Oct-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Dia Mirza</td>\n",
       "      <td>2</td>\n",
       "      <td>deespeak</td>\n",
       "      <td>3056203</td>\n",
       "      <td>1075</td>\n",
       "      <td>India</td>\n",
       "      <td>Feb-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Anil Kapoor</td>\n",
       "      <td>2</td>\n",
       "      <td>AnilKapoor</td>\n",
       "      <td>4994313</td>\n",
       "      <td>60</td>\n",
       "      <td>Mumbai, India</td>\n",
       "      <td>Aug-2013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Ranveer Singh</td>\n",
       "      <td>2</td>\n",
       "      <td>RanveerOfficial</td>\n",
       "      <td>12789980</td>\n",
       "      <td>862</td>\n",
       "      <td>Mumbai</td>\n",
       "      <td>Aug-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Bubbles</td>\n",
       "      <td>2</td>\n",
       "      <td>bubblesbehl</td>\n",
       "      <td>7358</td>\n",
       "      <td>59</td>\n",
       "      <td></td>\n",
       "      <td>Mar-2016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Kunal Kapoor</td>\n",
       "      <td>2</td>\n",
       "      <td>kapoorkkunal</td>\n",
       "      <td>573211</td>\n",
       "      <td>56</td>\n",
       "      <td>mumbai</td>\n",
       "      <td>Jan-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Komal Nahta</td>\n",
       "      <td>2</td>\n",
       "      <td>KomalNahta</td>\n",
       "      <td>986506</td>\n",
       "      <td>33</td>\n",
       "      <td>Bombay</td>\n",
       "      <td>Feb-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Farah Khan</td>\n",
       "      <td>2</td>\n",
       "      <td>FarahKhanAli</td>\n",
       "      <td>686631</td>\n",
       "      <td>977</td>\n",
       "      <td>ÜT: 19.110207,72.825719</td>\n",
       "      <td>Aug-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Shekhar Kapur</td>\n",
       "      <td>2</td>\n",
       "      <td>shekharkapur</td>\n",
       "      <td>2736502</td>\n",
       "      <td>502</td>\n",
       "      <td>Mumbai</td>\n",
       "      <td>Jun-2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Vineet jain</td>\n",
       "      <td>2</td>\n",
       "      <td>vineetjaintimes</td>\n",
       "      <td>130299</td>\n",
       "      <td>194</td>\n",
       "      <td>mumbai</td>\n",
       "      <td>Jan-2013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Shahid Kapoor</td>\n",
       "      <td>2</td>\n",
       "      <td>shahidkapoor</td>\n",
       "      <td>14494052</td>\n",
       "      <td>96</td>\n",
       "      <td></td>\n",
       "      <td>Jul-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Sussanne Khan</td>\n",
       "      <td>2</td>\n",
       "      <td>sussannekroshan</td>\n",
       "      <td>610088</td>\n",
       "      <td>71</td>\n",
       "      <td>ÜT: 19.109427,72.825723</td>\n",
       "      <td>Feb-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Ritesh Sidhwani</td>\n",
       "      <td>2</td>\n",
       "      <td>ritesh_sid</td>\n",
       "      <td>575560</td>\n",
       "      <td>120</td>\n",
       "      <td>Mumbai-India</td>\n",
       "      <td>Oct-2011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Rishi Kapoor</td>\n",
       "      <td>2</td>\n",
       "      <td>chintskap</td>\n",
       "      <td>3354356</td>\n",
       "      <td>27</td>\n",
       "      <td>Home, Mumbai, India</td>\n",
       "      <td>Jan-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Twitterrific</td>\n",
       "      <td>2</td>\n",
       "      <td>Twitterrific</td>\n",
       "      <td>409492</td>\n",
       "      <td>15</td>\n",
       "      <td>a tree in Dubuque, IA</td>\n",
       "      <td>Jan-2007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Anupam Kher</td>\n",
       "      <td>2</td>\n",
       "      <td>AnupamPKher</td>\n",
       "      <td>14268202</td>\n",
       "      <td>175</td>\n",
       "      <td>Mumbai, India</td>\n",
       "      <td>Sep-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Raj Nayak</td>\n",
       "      <td>2</td>\n",
       "      <td>rajcheerfull</td>\n",
       "      <td>566978</td>\n",
       "      <td>741</td>\n",
       "      <td>Earth</td>\n",
       "      <td>Oct-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Uday Chopra</td>\n",
       "      <td>2</td>\n",
       "      <td>udaychopra</td>\n",
       "      <td>1481809</td>\n",
       "      <td>379</td>\n",
       "      <td>Hell, 7th Fire, Oven 666</td>\n",
       "      <td>Jul-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>goldie behl</td>\n",
       "      <td>2</td>\n",
       "      <td>GOLDIEBEHL</td>\n",
       "      <td>134402</td>\n",
       "      <td>666</td>\n",
       "      <td>right here</td>\n",
       "      <td>Jun-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Sonali Bendre Behl</td>\n",
       "      <td>2</td>\n",
       "      <td>iamsonalibendre</td>\n",
       "      <td>321123</td>\n",
       "      <td>129</td>\n",
       "      <td></td>\n",
       "      <td>Oct-2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Jacqueline Fernandez</td>\n",
       "      <td>2</td>\n",
       "      <td>Asli_Jacqueline</td>\n",
       "      <td>14051330</td>\n",
       "      <td>202</td>\n",
       "      <td></td>\n",
       "      <td>Oct-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>King Mika Singh</td>\n",
       "      <td>2</td>\n",
       "      <td>MikaSingh</td>\n",
       "      <td>4207945</td>\n",
       "      <td>173</td>\n",
       "      <td>mumbai</td>\n",
       "      <td>Jun-2012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Dino Morea</td>\n",
       "      <td>2</td>\n",
       "      <td>DinoMorea9</td>\n",
       "      <td>534421</td>\n",
       "      <td>287</td>\n",
       "      <td>Mumbai, Maharashtra</td>\n",
       "      <td>Jul-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>DABBOO RATNANI</td>\n",
       "      <td>2</td>\n",
       "      <td>DabbooRatnani</td>\n",
       "      <td>1736573</td>\n",
       "      <td>583</td>\n",
       "      <td>Mumbai, India</td>\n",
       "      <td>Jun-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Azmi Shabana</td>\n",
       "      <td>2</td>\n",
       "      <td>AzmiShabana</td>\n",
       "      <td>3270467</td>\n",
       "      <td>271</td>\n",
       "      <td>ÜT: 19.111431,72.823609</td>\n",
       "      <td>Mar-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Afsar Zaidi</td>\n",
       "      <td>2</td>\n",
       "      <td>afsarzaidi</td>\n",
       "      <td>22632</td>\n",
       "      <td>137</td>\n",
       "      <td></td>\n",
       "      <td>Jul-2009</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      Freq         UserName  Followers  Friends  \\\n",
       "FullName                                                          \n",
       "Chulbul Pandey           3  BeingSalmanKhan   38222008       24   \n",
       "Abhishek Bachchan        3   juniorbachchan   14707884      410   \n",
       "Karan Johar              3       karanjohar   16664147      485   \n",
       "Barack Obama             3      BarackObama  108917081   611754   \n",
       "Preity G Zinta           2  realpreityzinta    4806691      105   \n",
       "Dia Mirza                2         deespeak    3056203     1075   \n",
       "Anil Kapoor              2       AnilKapoor    4994313       60   \n",
       "Ranveer Singh            2  RanveerOfficial   12789980      862   \n",
       "Bubbles                  2      bubblesbehl       7358       59   \n",
       "Kunal Kapoor             2     kapoorkkunal     573211       56   \n",
       "Komal Nahta              2       KomalNahta     986506       33   \n",
       "Farah Khan               2     FarahKhanAli     686631      977   \n",
       "Shekhar Kapur            2     shekharkapur    2736502      502   \n",
       "Vineet jain              2  vineetjaintimes     130299      194   \n",
       "Shahid Kapoor            2     shahidkapoor   14494052       96   \n",
       "Sussanne Khan            2  sussannekroshan     610088       71   \n",
       "Ritesh Sidhwani          2       ritesh_sid     575560      120   \n",
       "Rishi Kapoor             2        chintskap    3354356       27   \n",
       "Twitterrific             2     Twitterrific     409492       15   \n",
       "Anupam Kher              2      AnupamPKher   14268202      175   \n",
       "Raj Nayak                2     rajcheerfull     566978      741   \n",
       "Uday Chopra              2       udaychopra    1481809      379   \n",
       "goldie behl              2       GOLDIEBEHL     134402      666   \n",
       "Sonali Bendre Behl       2  iamsonalibendre     321123      129   \n",
       "Jacqueline Fernandez     2  Asli_Jacqueline   14051330      202   \n",
       "King Mika Singh          2        MikaSingh    4207945      173   \n",
       "Dino Morea               2       DinoMorea9     534421      287   \n",
       "DABBOO RATNANI           2    DabbooRatnani    1736573      583   \n",
       "Azmi Shabana             2      AzmiShabana    3270467      271   \n",
       "Afsar Zaidi              2       afsarzaidi      22632      137   \n",
       "\n",
       "                                      Location   Created  \n",
       "FullName                                                  \n",
       "Chulbul Pandey                          MUMBAI  Apr-2010  \n",
       "Abhishek Bachchan                Mumbai, India  Nov-2009  \n",
       "Karan Johar                 Maharashtra, India  Jul-2009  \n",
       "Barack Obama                    Washington, DC  Mar-2007  \n",
       "Preity G Zinta         Catch me if you can....  Oct-2009  \n",
       "Dia Mirza                                India  Feb-2010  \n",
       "Anil Kapoor                      Mumbai, India  Aug-2013  \n",
       "Ranveer Singh                           Mumbai  Aug-2009  \n",
       "Bubbles                                         Mar-2016  \n",
       "Kunal Kapoor                            mumbai  Jan-2010  \n",
       "Komal Nahta                             Bombay  Feb-2010  \n",
       "Farah Khan             ÜT: 19.110207,72.825719  Aug-2009  \n",
       "Shekhar Kapur                           Mumbai  Jun-2008  \n",
       "Vineet jain                             mumbai  Jan-2013  \n",
       "Shahid Kapoor                                   Jul-2009  \n",
       "Sussanne Khan          ÜT: 19.109427,72.825723  Feb-2010  \n",
       "Ritesh Sidhwani                   Mumbai-India  Oct-2011  \n",
       "Rishi Kapoor              Home, Mumbai, India   Jan-2010  \n",
       "Twitterrific             a tree in Dubuque, IA  Jan-2007  \n",
       "Anupam Kher                      Mumbai, India  Sep-2009  \n",
       "Raj Nayak                                Earth  Oct-2009  \n",
       "Uday Chopra           Hell, 7th Fire, Oven 666  Jul-2009  \n",
       "goldie behl                         right here  Jun-2009  \n",
       "Sonali Bendre Behl                              Oct-2015  \n",
       "Jacqueline Fernandez                            Oct-2009  \n",
       "King Mika Singh                         mumbai  Jun-2012  \n",
       "Dino Morea                 Mumbai, Maharashtra  Jul-2009  \n",
       "DABBOO RATNANI                   Mumbai, India  Jun-2009  \n",
       "Azmi Shabana           ÜT: 19.111431,72.823609  Mar-2010  \n",
       "Afsar Zaidi                                     Jul-2009  "
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "favorite_cluster_df = results_df [results_df['spectral'].isin([5])]\n",
    "favorite_cluster_list = list(favorite_cluster_df.index)\n",
    "\n",
    "discover_Friends_toFollow(ids_of_interest = favorite_cluster_list, \n",
    "                          friend_list = friends_of_self, \n",
    "                          prop = .5, count = 30).fillna('').set_index('FullName')  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 4d.Reco:  After looking at Sports clustser (Spectral clustering)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'Who-To-Follow' reco after looking at   7 friends' friends:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Freq</th>\n",
       "      <th>UserName</th>\n",
       "      <th>Followers</th>\n",
       "      <th>Friends</th>\n",
       "      <th>Location</th>\n",
       "      <th>Created</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FullName</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>Ashwin Ravichandran</td>\n",
       "      <td>3</td>\n",
       "      <td>ashwinravi99</td>\n",
       "      <td>9698651</td>\n",
       "      <td>236</td>\n",
       "      <td>Chennai</td>\n",
       "      <td>Aug-2011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Gautam Gambhir</td>\n",
       "      <td>3</td>\n",
       "      <td>GautamGambhir</td>\n",
       "      <td>9938638</td>\n",
       "      <td>121</td>\n",
       "      <td></td>\n",
       "      <td>Dec-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Suresh Raina🇮🇳</td>\n",
       "      <td>3</td>\n",
       "      <td>ImRaina</td>\n",
       "      <td>16857878</td>\n",
       "      <td>300</td>\n",
       "      <td>India</td>\n",
       "      <td>Mar-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Twitter Sports</td>\n",
       "      <td>3</td>\n",
       "      <td>TwitterSports</td>\n",
       "      <td>15272596</td>\n",
       "      <td>1992</td>\n",
       "      <td></td>\n",
       "      <td>May-2011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Ishant Sharma</td>\n",
       "      <td>2</td>\n",
       "      <td>ImIshant</td>\n",
       "      <td>2264007</td>\n",
       "      <td>60</td>\n",
       "      <td>New Delhi, India</td>\n",
       "      <td>Dec-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Rohit Sharma</td>\n",
       "      <td>2</td>\n",
       "      <td>ImRo45</td>\n",
       "      <td>14930299</td>\n",
       "      <td>43</td>\n",
       "      <td></td>\n",
       "      <td>Mar-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>R P Singh रुद्र प्रताप सिंह</td>\n",
       "      <td>2</td>\n",
       "      <td>rpsingh</td>\n",
       "      <td>1574542</td>\n",
       "      <td>322</td>\n",
       "      <td>Rae bareli, India</td>\n",
       "      <td>Feb-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Amit Mishra</td>\n",
       "      <td>2</td>\n",
       "      <td>MishiAmit</td>\n",
       "      <td>1354021</td>\n",
       "      <td>54</td>\n",
       "      <td>Delhi</td>\n",
       "      <td>Apr-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Prannoy Roy</td>\n",
       "      <td>2</td>\n",
       "      <td>PrannoyRoyNDTV</td>\n",
       "      <td>758697</td>\n",
       "      <td>487</td>\n",
       "      <td>India</td>\n",
       "      <td>Oct-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Chennaiyin FC 🏆🏆</td>\n",
       "      <td>2</td>\n",
       "      <td>ChennaiyinFC</td>\n",
       "      <td>861166</td>\n",
       "      <td>98</td>\n",
       "      <td>Chennai</td>\n",
       "      <td>Sep-2014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>zaheer khan</td>\n",
       "      <td>2</td>\n",
       "      <td>ImZaheer</td>\n",
       "      <td>5944164</td>\n",
       "      <td>204</td>\n",
       "      <td></td>\n",
       "      <td>Jan-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Anirudh Chaudhry</td>\n",
       "      <td>2</td>\n",
       "      <td>AnirudhChaudhry</td>\n",
       "      <td>287886</td>\n",
       "      <td>332</td>\n",
       "      <td></td>\n",
       "      <td>Mar-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Manchester United</td>\n",
       "      <td>2</td>\n",
       "      <td>ManUtd</td>\n",
       "      <td>20283139</td>\n",
       "      <td>129</td>\n",
       "      <td>Manchester, England</td>\n",
       "      <td>Apr-2012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Harbhajan Turbanator</td>\n",
       "      <td>2</td>\n",
       "      <td>harbhajan_singh</td>\n",
       "      <td>10189709</td>\n",
       "      <td>128</td>\n",
       "      <td>India</td>\n",
       "      <td>Nov-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>yuvraj singh</td>\n",
       "      <td>2</td>\n",
       "      <td>YUVSTRONG12</td>\n",
       "      <td>4720718</td>\n",
       "      <td>378</td>\n",
       "      <td>Mumbai, India</td>\n",
       "      <td>Jul-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Sundar Pichai</td>\n",
       "      <td>2</td>\n",
       "      <td>sundarpichai</td>\n",
       "      <td>2483187</td>\n",
       "      <td>336</td>\n",
       "      <td></td>\n",
       "      <td>Mar-2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Kobe Bryant</td>\n",
       "      <td>2</td>\n",
       "      <td>kobebryant</td>\n",
       "      <td>14499233</td>\n",
       "      <td>509</td>\n",
       "      <td></td>\n",
       "      <td>Jan-2013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Rio Ferdinand</td>\n",
       "      <td>2</td>\n",
       "      <td>rioferdy5</td>\n",
       "      <td>10554871</td>\n",
       "      <td>613</td>\n",
       "      <td>www.neweraglobalsports.com</td>\n",
       "      <td>Jun-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>JuventusFC</td>\n",
       "      <td>2</td>\n",
       "      <td>juventusfc</td>\n",
       "      <td>7341905</td>\n",
       "      <td>57</td>\n",
       "      <td>Allianz Stadium</td>\n",
       "      <td>Feb-2011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Álvaro Arbeloa</td>\n",
       "      <td>2</td>\n",
       "      <td>aarbeloa17</td>\n",
       "      <td>6049324</td>\n",
       "      <td>1236</td>\n",
       "      <td>Madrid, Comunidad de Madrid</td>\n",
       "      <td>Apr-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Xabi Alonso</td>\n",
       "      <td>2</td>\n",
       "      <td>XabiAlonso</td>\n",
       "      <td>11930664</td>\n",
       "      <td>536</td>\n",
       "      <td></td>\n",
       "      <td>Aug-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Kaka</td>\n",
       "      <td>2</td>\n",
       "      <td>KAKA</td>\n",
       "      <td>29667500</td>\n",
       "      <td>305</td>\n",
       "      <td></td>\n",
       "      <td>Jul-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>James Rodríguez</td>\n",
       "      <td>2</td>\n",
       "      <td>jamesdrodriguez</td>\n",
       "      <td>18439360</td>\n",
       "      <td>295</td>\n",
       "      <td>Madrid, Comunidad de Madrid</td>\n",
       "      <td>Apr-2011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Instagram</td>\n",
       "      <td>2</td>\n",
       "      <td>instagram</td>\n",
       "      <td>36380296</td>\n",
       "      <td>62</td>\n",
       "      <td></td>\n",
       "      <td>Aug-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Nani</td>\n",
       "      <td>2</td>\n",
       "      <td>luisnani</td>\n",
       "      <td>3925469</td>\n",
       "      <td>176</td>\n",
       "      <td></td>\n",
       "      <td>Feb-2011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Pepe</td>\n",
       "      <td>2</td>\n",
       "      <td>officialpepe</td>\n",
       "      <td>3564452</td>\n",
       "      <td>42</td>\n",
       "      <td></td>\n",
       "      <td>Jul-2011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Radamel Falcao</td>\n",
       "      <td>2</td>\n",
       "      <td>FALCAO</td>\n",
       "      <td>17433711</td>\n",
       "      <td>619</td>\n",
       "      <td>Estambul, Turquía</td>\n",
       "      <td>Jul-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Dwayne DJ Bravo</td>\n",
       "      <td>2</td>\n",
       "      <td>DJBravo47</td>\n",
       "      <td>1038819</td>\n",
       "      <td>0</td>\n",
       "      <td></td>\n",
       "      <td>Sep-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Michael Clarke</td>\n",
       "      <td>2</td>\n",
       "      <td>MClarke23</td>\n",
       "      <td>2144133</td>\n",
       "      <td>629</td>\n",
       "      <td>Sydney, New South Wales</td>\n",
       "      <td>Nov-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Ajinkya Rahane</td>\n",
       "      <td>2</td>\n",
       "      <td>ajinkyarahane88</td>\n",
       "      <td>5594361</td>\n",
       "      <td>69</td>\n",
       "      <td>Mumbai</td>\n",
       "      <td>Aug-2011</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             Freq         UserName  Followers  Friends  \\\n",
       "FullName                                                                 \n",
       "Ashwin Ravichandran             3     ashwinravi99    9698651      236   \n",
       "Gautam Gambhir                  3    GautamGambhir    9938638      121   \n",
       "Suresh Raina🇮🇳                  3          ImRaina   16857878      300   \n",
       "Twitter Sports                  3    TwitterSports   15272596     1992   \n",
       "Ishant Sharma                   2         ImIshant    2264007       60   \n",
       "Rohit Sharma                    2           ImRo45   14930299       43   \n",
       "R P Singh रुद्र प्रताप सिंह     2          rpsingh    1574542      322   \n",
       "Amit Mishra                     2        MishiAmit    1354021       54   \n",
       "Prannoy Roy                     2   PrannoyRoyNDTV     758697      487   \n",
       "Chennaiyin FC 🏆🏆                2     ChennaiyinFC     861166       98   \n",
       "zaheer khan                     2         ImZaheer    5944164      204   \n",
       "Anirudh Chaudhry                2  AnirudhChaudhry     287886      332   \n",
       "Manchester United               2           ManUtd   20283139      129   \n",
       "Harbhajan Turbanator            2  harbhajan_singh   10189709      128   \n",
       "yuvraj singh                    2      YUVSTRONG12    4720718      378   \n",
       "Sundar Pichai                   2     sundarpichai    2483187      336   \n",
       "Kobe Bryant                     2       kobebryant   14499233      509   \n",
       "Rio Ferdinand                   2        rioferdy5   10554871      613   \n",
       "JuventusFC                      2       juventusfc    7341905       57   \n",
       "Álvaro Arbeloa                  2       aarbeloa17    6049324     1236   \n",
       "Xabi Alonso                     2       XabiAlonso   11930664      536   \n",
       "Kaka                            2             KAKA   29667500      305   \n",
       "James Rodríguez                 2  jamesdrodriguez   18439360      295   \n",
       "Instagram                       2        instagram   36380296       62   \n",
       "Nani                            2         luisnani    3925469      176   \n",
       "Pepe                            2     officialpepe    3564452       42   \n",
       "Radamel Falcao                  2           FALCAO   17433711      619   \n",
       "Dwayne DJ Bravo                 2        DJBravo47    1038819        0   \n",
       "Michael Clarke                  2        MClarke23    2144133      629   \n",
       "Ajinkya Rahane                  2  ajinkyarahane88    5594361       69   \n",
       "\n",
       "                                                Location   Created  \n",
       "FullName                                                            \n",
       "Ashwin Ravichandran                              Chennai  Aug-2011  \n",
       "Gautam Gambhir                                            Dec-2009  \n",
       "Suresh Raina🇮🇳                                     India  Mar-2010  \n",
       "Twitter Sports                                            May-2011  \n",
       "Ishant Sharma                           New Delhi, India  Dec-2009  \n",
       "Rohit Sharma                                              Mar-2010  \n",
       "R P Singh रुद्र प्रताप सिंह            Rae bareli, India  Feb-2010  \n",
       "Amit Mishra                                        Delhi  Apr-2010  \n",
       "Prannoy Roy                                        India  Oct-2009  \n",
       "Chennaiyin FC 🏆🏆                                 Chennai  Sep-2014  \n",
       "zaheer khan                                               Jan-2010  \n",
       "Anirudh Chaudhry                                          Mar-2010  \n",
       "Manchester United                    Manchester, England  Apr-2012  \n",
       "Harbhajan Turbanator                               India  Nov-2009  \n",
       "yuvraj singh                               Mumbai, India  Jul-2009  \n",
       "Sundar Pichai                                             Mar-2008  \n",
       "Kobe Bryant                                               Jan-2013  \n",
       "Rio Ferdinand                 www.neweraglobalsports.com  Jun-2010  \n",
       "JuventusFC                               Allianz Stadium  Feb-2011  \n",
       "Álvaro Arbeloa               Madrid, Comunidad de Madrid  Apr-2009  \n",
       "Xabi Alonso                                               Aug-2010  \n",
       "Kaka                                                      Jul-2009  \n",
       "James Rodríguez              Madrid, Comunidad de Madrid  Apr-2011  \n",
       "Instagram                                                 Aug-2010  \n",
       "Nani                                                      Feb-2011  \n",
       "Pepe                                                      Jul-2011  \n",
       "Radamel Falcao                         Estambul, Turquía  Jul-2009  \n",
       "Dwayne DJ Bravo                                           Sep-2010  \n",
       "Michael Clarke                   Sydney, New South Wales  Nov-2009  \n",
       "Ajinkya Rahane                                    Mumbai  Aug-2011  "
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "favorite_cluster_df = results_df [results_df['spectral'].isin([0])]\n",
    "favorite_cluster_list = list(favorite_cluster_df.index)\n",
    "\n",
    "discover_Friends_toFollow(ids_of_interest = favorite_cluster_list, \n",
    "                          friend_list = friends_of_self, \n",
    "                          prop = .5, count = 30).fillna('').set_index('FullName')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Reco: 4e. After looking at Tech & Media community clustser (Spectral clustering)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'Who-To-Follow' reco after looking at   3 friends' friends:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Freq</th>\n",
       "      <th>UserName</th>\n",
       "      <th>Followers</th>\n",
       "      <th>Friends</th>\n",
       "      <th>Location</th>\n",
       "      <th>Created</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FullName</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>New Scientist</td>\n",
       "      <td>3</td>\n",
       "      <td>newscientist</td>\n",
       "      <td>3528447</td>\n",
       "      <td>145</td>\n",
       "      <td>Worldwide</td>\n",
       "      <td>Jan-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Slashdot</td>\n",
       "      <td>2</td>\n",
       "      <td>slashdot</td>\n",
       "      <td>241209</td>\n",
       "      <td>147</td>\n",
       "      <td>San Francisco</td>\n",
       "      <td>Mar-2007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Ars Technica</td>\n",
       "      <td>2</td>\n",
       "      <td>arstechnica</td>\n",
       "      <td>1168522</td>\n",
       "      <td>1233</td>\n",
       "      <td>NYC - Boston - Chicago - SF</td>\n",
       "      <td>Jan-2007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>GigaOm</td>\n",
       "      <td>2</td>\n",
       "      <td>gigaom</td>\n",
       "      <td>309985</td>\n",
       "      <td>1155</td>\n",
       "      <td>Austin, TX</td>\n",
       "      <td>Mar-2007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Gizmodo</td>\n",
       "      <td>2</td>\n",
       "      <td>Gizmodo</td>\n",
       "      <td>2892136</td>\n",
       "      <td>139</td>\n",
       "      <td></td>\n",
       "      <td>Mar-2007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>TechCrunch</td>\n",
       "      <td>2</td>\n",
       "      <td>TechCrunch</td>\n",
       "      <td>10104486</td>\n",
       "      <td>1050</td>\n",
       "      <td>San Francisco, CA</td>\n",
       "      <td>Mar-2007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>NASA</td>\n",
       "      <td>2</td>\n",
       "      <td>NASA</td>\n",
       "      <td>33248641</td>\n",
       "      <td>298</td>\n",
       "      <td></td>\n",
       "      <td>Dec-2007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Techmeme</td>\n",
       "      <td>2</td>\n",
       "      <td>Techmeme</td>\n",
       "      <td>399577</td>\n",
       "      <td>921</td>\n",
       "      <td>United States</td>\n",
       "      <td>Mar-2007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>The Onion</td>\n",
       "      <td>2</td>\n",
       "      <td>TheOnion</td>\n",
       "      <td>11320535</td>\n",
       "      <td>15</td>\n",
       "      <td></td>\n",
       "      <td>Mar-2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>TeslaRoadTrip</td>\n",
       "      <td>2</td>\n",
       "      <td>TeslaRoadTrip</td>\n",
       "      <td>57198</td>\n",
       "      <td>69</td>\n",
       "      <td>United States</td>\n",
       "      <td>Feb-2013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Neil deGrasse Tyson</td>\n",
       "      <td>2</td>\n",
       "      <td>neiltyson</td>\n",
       "      <td>13475967</td>\n",
       "      <td>39</td>\n",
       "      <td>New York City</td>\n",
       "      <td>Jan-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>CleanTechnica</td>\n",
       "      <td>2</td>\n",
       "      <td>cleantechnica</td>\n",
       "      <td>104221</td>\n",
       "      <td>141</td>\n",
       "      <td></td>\n",
       "      <td>Oct-2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>bill lee</td>\n",
       "      <td>2</td>\n",
       "      <td>westcoastbill</td>\n",
       "      <td>39946</td>\n",
       "      <td>549</td>\n",
       "      <td>Being First</td>\n",
       "      <td>Jan-2007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>SpaceX</td>\n",
       "      <td>2</td>\n",
       "      <td>SpaceX</td>\n",
       "      <td>9051744</td>\n",
       "      <td>107</td>\n",
       "      <td>Hawthorne, CA</td>\n",
       "      <td>Apr-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Rick and Morty</td>\n",
       "      <td>2</td>\n",
       "      <td>RickandMorty</td>\n",
       "      <td>1746681</td>\n",
       "      <td>201</td>\n",
       "      <td></td>\n",
       "      <td>Jul-2012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Hyperloop</td>\n",
       "      <td>2</td>\n",
       "      <td>Hyperloop</td>\n",
       "      <td>180224</td>\n",
       "      <td>95</td>\n",
       "      <td>Hawthorne, CA</td>\n",
       "      <td>Aug-2013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Tim Urban</td>\n",
       "      <td>2</td>\n",
       "      <td>waitbutwhy</td>\n",
       "      <td>208761</td>\n",
       "      <td>578</td>\n",
       "      <td></td>\n",
       "      <td>Mar-2013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>OpenAI</td>\n",
       "      <td>2</td>\n",
       "      <td>OpenAI</td>\n",
       "      <td>379085</td>\n",
       "      <td>0</td>\n",
       "      <td></td>\n",
       "      <td>Dec-2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>The Boring Company</td>\n",
       "      <td>2</td>\n",
       "      <td>boringcompany</td>\n",
       "      <td>327424</td>\n",
       "      <td>5</td>\n",
       "      <td>Los Angeles, CA</td>\n",
       "      <td>May-2017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>TESLARATI</td>\n",
       "      <td>2</td>\n",
       "      <td>Teslarati</td>\n",
       "      <td>127849</td>\n",
       "      <td>71</td>\n",
       "      <td>California</td>\n",
       "      <td>Mar-2013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Backchannel</td>\n",
       "      <td>1</td>\n",
       "      <td>backchnnl</td>\n",
       "      <td>24365</td>\n",
       "      <td>294</td>\n",
       "      <td>San Francisco, CA</td>\n",
       "      <td>Oct-2014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Google Nest</td>\n",
       "      <td>1</td>\n",
       "      <td>googlenest</td>\n",
       "      <td>195530</td>\n",
       "      <td>12</td>\n",
       "      <td>Home</td>\n",
       "      <td>Jul-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Google Australia NZ</td>\n",
       "      <td>1</td>\n",
       "      <td>googledownunder</td>\n",
       "      <td>72566</td>\n",
       "      <td>622</td>\n",
       "      <td>Sydney</td>\n",
       "      <td>Apr-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Google Communications</td>\n",
       "      <td>1</td>\n",
       "      <td>Google_Comms</td>\n",
       "      <td>34452</td>\n",
       "      <td>373</td>\n",
       "      <td>Mountain View, CA</td>\n",
       "      <td>Sep-2017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Matías Duarte</td>\n",
       "      <td>1</td>\n",
       "      <td>MatiasDuarte</td>\n",
       "      <td>32226</td>\n",
       "      <td>391</td>\n",
       "      <td></td>\n",
       "      <td>May-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Google for Nonprofits</td>\n",
       "      <td>1</td>\n",
       "      <td>googlenonprofit</td>\n",
       "      <td>20952</td>\n",
       "      <td>296</td>\n",
       "      <td></td>\n",
       "      <td>Feb-2011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Google.org</td>\n",
       "      <td>1</td>\n",
       "      <td>Googleorg</td>\n",
       "      <td>66519</td>\n",
       "      <td>320</td>\n",
       "      <td>Mountain View, CA</td>\n",
       "      <td>May-2014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>BI Tech</td>\n",
       "      <td>1</td>\n",
       "      <td>SAI</td>\n",
       "      <td>1284518</td>\n",
       "      <td>67</td>\n",
       "      <td>New York, NY</td>\n",
       "      <td>Sep-2007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Eric Schmidt</td>\n",
       "      <td>1</td>\n",
       "      <td>ericschmidt</td>\n",
       "      <td>2311883</td>\n",
       "      <td>187</td>\n",
       "      <td>Mountain View, CA</td>\n",
       "      <td>Dec-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Google Россия</td>\n",
       "      <td>1</td>\n",
       "      <td>GoogleRussia</td>\n",
       "      <td>2088130</td>\n",
       "      <td>90</td>\n",
       "      <td></td>\n",
       "      <td>Jul-2009</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       Freq         UserName  Followers  Friends  \\\n",
       "FullName                                                           \n",
       "New Scientist             3     newscientist    3528447      145   \n",
       "Slashdot                  2         slashdot     241209      147   \n",
       "Ars Technica              2      arstechnica    1168522     1233   \n",
       "GigaOm                    2           gigaom     309985     1155   \n",
       "Gizmodo                   2          Gizmodo    2892136      139   \n",
       "TechCrunch                2       TechCrunch   10104486     1050   \n",
       "NASA                      2             NASA   33248641      298   \n",
       "Techmeme                  2         Techmeme     399577      921   \n",
       "The Onion                 2         TheOnion   11320535       15   \n",
       "TeslaRoadTrip             2    TeslaRoadTrip      57198       69   \n",
       "Neil deGrasse Tyson       2        neiltyson   13475967       39   \n",
       "CleanTechnica             2    cleantechnica     104221      141   \n",
       "bill lee                  2    westcoastbill      39946      549   \n",
       "SpaceX                    2           SpaceX    9051744      107   \n",
       "Rick and Morty            2     RickandMorty    1746681      201   \n",
       "Hyperloop                 2        Hyperloop     180224       95   \n",
       "Tim Urban                 2       waitbutwhy     208761      578   \n",
       "OpenAI                    2           OpenAI     379085        0   \n",
       "The Boring Company        2    boringcompany     327424        5   \n",
       "TESLARATI                 2        Teslarati     127849       71   \n",
       "Backchannel               1        backchnnl      24365      294   \n",
       "Google Nest               1       googlenest     195530       12   \n",
       "Google Australia NZ       1  googledownunder      72566      622   \n",
       "Google Communications     1     Google_Comms      34452      373   \n",
       "Matías Duarte             1     MatiasDuarte      32226      391   \n",
       "Google for Nonprofits     1  googlenonprofit      20952      296   \n",
       "Google.org                1        Googleorg      66519      320   \n",
       "BI Tech                   1              SAI    1284518       67   \n",
       "Eric Schmidt              1      ericschmidt    2311883      187   \n",
       "Google Россия             1     GoogleRussia    2088130       90   \n",
       "\n",
       "                                          Location   Created  \n",
       "FullName                                                      \n",
       "New Scientist                            Worldwide  Jan-2009  \n",
       "Slashdot                             San Francisco  Mar-2007  \n",
       "Ars Technica           NYC - Boston - Chicago - SF  Jan-2007  \n",
       "GigaOm                                  Austin, TX  Mar-2007  \n",
       "Gizmodo                                             Mar-2007  \n",
       "TechCrunch                       San Francisco, CA  Mar-2007  \n",
       "NASA                                                Dec-2007  \n",
       "Techmeme                             United States  Mar-2007  \n",
       "The Onion                                           Mar-2008  \n",
       "TeslaRoadTrip                        United States  Feb-2013  \n",
       "Neil deGrasse Tyson                  New York City  Jan-2009  \n",
       "CleanTechnica                                       Oct-2008  \n",
       "bill lee                               Being First  Jan-2007  \n",
       "SpaceX                               Hawthorne, CA  Apr-2009  \n",
       "Rick and Morty                                      Jul-2012  \n",
       "Hyperloop                            Hawthorne, CA  Aug-2013  \n",
       "Tim Urban                                           Mar-2013  \n",
       "OpenAI                                              Dec-2015  \n",
       "The Boring Company                 Los Angeles, CA  May-2017  \n",
       "TESLARATI                               California  Mar-2013  \n",
       "Backchannel                      San Francisco, CA  Oct-2014  \n",
       "Google Nest                                   Home  Jul-2010  \n",
       "Google Australia NZ                         Sydney  Apr-2009  \n",
       "Google Communications            Mountain View, CA  Sep-2017  \n",
       "Matías Duarte                                       May-2009  \n",
       "Google for Nonprofits                               Feb-2011  \n",
       "Google.org                       Mountain View, CA  May-2014  \n",
       "BI Tech                               New York, NY  Sep-2007  \n",
       "Eric Schmidt                     Mountain View, CA  Dec-2009  \n",
       "Google Россия                                       Jul-2009  "
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "favorite_cluster_df = results_df [results_df['spectral'].isin([4])]\n",
    "favorite_cluster_list = list(favorite_cluster_df.index)\n",
    "\n",
    "discover_Friends_toFollow(ids_of_interest = favorite_cluster_list, \n",
    "                          friend_list = friends_of_self, \n",
    "                          prop = .5, count = 30).fillna('').set_index('FullName')  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'Who-To-Follow' reco after looking at   9 friends' friends:\n",
      "... Cache-Miss for 4 nodes. Updating Cache...\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Freq</th>\n",
       "      <th>UserName</th>\n",
       "      <th>Followers</th>\n",
       "      <th>Friends</th>\n",
       "      <th>Location</th>\n",
       "      <th>Created</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FullName</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>Karan Johar</td>\n",
       "      <td>8</td>\n",
       "      <td>karanjohar</td>\n",
       "      <td>16664147</td>\n",
       "      <td>485</td>\n",
       "      <td>Maharashtra, India</td>\n",
       "      <td>Jul-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Ranveer Singh</td>\n",
       "      <td>7</td>\n",
       "      <td>RanveerOfficial</td>\n",
       "      <td>12789980</td>\n",
       "      <td>862</td>\n",
       "      <td>Mumbai</td>\n",
       "      <td>Aug-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Shahid Kapoor</td>\n",
       "      <td>7</td>\n",
       "      <td>shahidkapoor</td>\n",
       "      <td>14494052</td>\n",
       "      <td>96</td>\n",
       "      <td></td>\n",
       "      <td>Jul-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Abhishek Bachchan</td>\n",
       "      <td>7</td>\n",
       "      <td>juniorbachchan</td>\n",
       "      <td>14707884</td>\n",
       "      <td>410</td>\n",
       "      <td>Mumbai, India</td>\n",
       "      <td>Nov-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Chulbul Pandey</td>\n",
       "      <td>6</td>\n",
       "      <td>BeingSalmanKhan</td>\n",
       "      <td>38222008</td>\n",
       "      <td>24</td>\n",
       "      <td>MUMBAI</td>\n",
       "      <td>Apr-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Barack Obama</td>\n",
       "      <td>6</td>\n",
       "      <td>BarackObama</td>\n",
       "      <td>108917081</td>\n",
       "      <td>611754</td>\n",
       "      <td>Washington, DC</td>\n",
       "      <td>Mar-2007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>taran adarsh</td>\n",
       "      <td>6</td>\n",
       "      <td>taran_adarsh</td>\n",
       "      <td>3680564</td>\n",
       "      <td>179</td>\n",
       "      <td>Mumbai, India</td>\n",
       "      <td>Dec-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Arjun Kapoor</td>\n",
       "      <td>6</td>\n",
       "      <td>arjunk26</td>\n",
       "      <td>6128252</td>\n",
       "      <td>794</td>\n",
       "      <td></td>\n",
       "      <td>Apr-2014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>J</td>\n",
       "      <td>6</td>\n",
       "      <td>jiteshpillaai</td>\n",
       "      <td>721109</td>\n",
       "      <td>1170</td>\n",
       "      <td>Mumbai</td>\n",
       "      <td>May-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Dia Mirza</td>\n",
       "      <td>5</td>\n",
       "      <td>deespeak</td>\n",
       "      <td>3056203</td>\n",
       "      <td>1075</td>\n",
       "      <td>India</td>\n",
       "      <td>Feb-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Twinkle Khanna</td>\n",
       "      <td>5</td>\n",
       "      <td>mrsfunnybones</td>\n",
       "      <td>5316236</td>\n",
       "      <td>88</td>\n",
       "      <td></td>\n",
       "      <td>Aug-2014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Rishi Kapoor</td>\n",
       "      <td>5</td>\n",
       "      <td>chintskap</td>\n",
       "      <td>3354356</td>\n",
       "      <td>27</td>\n",
       "      <td>Home, Mumbai, India</td>\n",
       "      <td>Jan-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Uday Chopra</td>\n",
       "      <td>5</td>\n",
       "      <td>udaychopra</td>\n",
       "      <td>1481809</td>\n",
       "      <td>379</td>\n",
       "      <td>Hell, 7th Fire, Oven 666</td>\n",
       "      <td>Jul-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Jacqueline Fernandez</td>\n",
       "      <td>5</td>\n",
       "      <td>Asli_Jacqueline</td>\n",
       "      <td>14051330</td>\n",
       "      <td>202</td>\n",
       "      <td></td>\n",
       "      <td>Oct-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Azmi Shabana</td>\n",
       "      <td>5</td>\n",
       "      <td>AzmiShabana</td>\n",
       "      <td>3270467</td>\n",
       "      <td>271</td>\n",
       "      <td>ÜT: 19.111431,72.823609</td>\n",
       "      <td>Mar-2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Ronnie Screwvala</td>\n",
       "      <td>5</td>\n",
       "      <td>RonnieScrewvala</td>\n",
       "      <td>2130370</td>\n",
       "      <td>293</td>\n",
       "      <td>Mumbai</td>\n",
       "      <td>May-2012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>barkha dutt</td>\n",
       "      <td>5</td>\n",
       "      <td>BDUTT</td>\n",
       "      <td>6975662</td>\n",
       "      <td>2191</td>\n",
       "      <td>India</td>\n",
       "      <td>Feb-2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Priya Gupta</td>\n",
       "      <td>5</td>\n",
       "      <td>priyagupta999</td>\n",
       "      <td>643397</td>\n",
       "      <td>535</td>\n",
       "      <td>Mumbai</td>\n",
       "      <td>Jun-2012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>MissMalini</td>\n",
       "      <td>5</td>\n",
       "      <td>MissMalini</td>\n",
       "      <td>2757942</td>\n",
       "      <td>4222</td>\n",
       "      <td>Bombay</td>\n",
       "      <td>Oct-2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Farah Khan</td>\n",
       "      <td>5</td>\n",
       "      <td>TheFarahKhan</td>\n",
       "      <td>6738536</td>\n",
       "      <td>74</td>\n",
       "      <td>Mumbai, India</td>\n",
       "      <td>Sep-2010</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      Freq         UserName  Followers  Friends  \\\n",
       "FullName                                                          \n",
       "Karan Johar              8       karanjohar   16664147      485   \n",
       "Ranveer Singh            7  RanveerOfficial   12789980      862   \n",
       "Shahid Kapoor            7     shahidkapoor   14494052       96   \n",
       "Abhishek Bachchan        7   juniorbachchan   14707884      410   \n",
       "Chulbul Pandey           6  BeingSalmanKhan   38222008       24   \n",
       "Barack Obama             6      BarackObama  108917081   611754   \n",
       "taran adarsh             6     taran_adarsh    3680564      179   \n",
       "Arjun Kapoor             6         arjunk26    6128252      794   \n",
       "J                        6    jiteshpillaai     721109     1170   \n",
       "Dia Mirza                5         deespeak    3056203     1075   \n",
       "Twinkle Khanna           5    mrsfunnybones    5316236       88   \n",
       "Rishi Kapoor             5        chintskap    3354356       27   \n",
       "Uday Chopra              5       udaychopra    1481809      379   \n",
       "Jacqueline Fernandez     5  Asli_Jacqueline   14051330      202   \n",
       "Azmi Shabana             5      AzmiShabana    3270467      271   \n",
       "Ronnie Screwvala         5  RonnieScrewvala    2130370      293   \n",
       "barkha dutt              5            BDUTT    6975662     2191   \n",
       "Priya Gupta              5    priyagupta999     643397      535   \n",
       "MissMalini               5       MissMalini    2757942     4222   \n",
       "Farah Khan               5     TheFarahKhan    6738536       74   \n",
       "\n",
       "                                      Location   Created  \n",
       "FullName                                                  \n",
       "Karan Johar                 Maharashtra, India  Jul-2009  \n",
       "Ranveer Singh                           Mumbai  Aug-2009  \n",
       "Shahid Kapoor                                   Jul-2009  \n",
       "Abhishek Bachchan                Mumbai, India  Nov-2009  \n",
       "Chulbul Pandey                          MUMBAI  Apr-2010  \n",
       "Barack Obama                    Washington, DC  Mar-2007  \n",
       "taran adarsh                     Mumbai, India  Dec-2009  \n",
       "Arjun Kapoor                                    Apr-2014  \n",
       "J                                       Mumbai  May-2009  \n",
       "Dia Mirza                                India  Feb-2010  \n",
       "Twinkle Khanna                                  Aug-2014  \n",
       "Rishi Kapoor              Home, Mumbai, India   Jan-2010  \n",
       "Uday Chopra           Hell, 7th Fire, Oven 666  Jul-2009  \n",
       "Jacqueline Fernandez                            Oct-2009  \n",
       "Azmi Shabana           ÜT: 19.111431,72.823609  Mar-2010  \n",
       "Ronnie Screwvala                        Mumbai  May-2012  \n",
       "barkha dutt                              India  Feb-2009  \n",
       "Priya Gupta                             Mumbai  Jun-2012  \n",
       "MissMalini                              Bombay  Oct-2008  \n",
       "Farah Khan                       Mumbai, India  Sep-2010  "
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "discover_Friends_toFollow(ids_of_interest = list(results_df [results_df['spectral'] == 5].index), \n",
    "                          friend_list = friends_of_self, \n",
    "                          prop = 1, count = 20).fillna('').set_index('FullName')  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}