aml_project_full_dataset.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "e5c9004a",
   "metadata": {},
   "source": [
    "Amanda Rodgers\n",
    "July 18, 2024\n",
    "Anti-money laundering Project"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bda4cfe3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Dataset: https://www.kaggle.com/datasets/anshankul/ibm-amlsim-example-dataset?select=transactions.csv\n",
    "# Tutorial: https://medium.com/@jasonclwu/network-analysis-for-anti-money-laundering-with-python-ad981792a947\n",
    "# Github: https://github.com/jasonclwu/202209_AML_Network_Analysis/blob/main/AML_Network_Analysis.ipynb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6f24be2a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import libraries\n",
    "import pandas as pd\n",
    "import networkx as nx\n",
    "from pyvis.network import Network as net\n",
    "# from pyvis import Network as net\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "pd.options.display.float_format = '{:,.0f}'.format"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9cc52f4a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create df with dataset\n",
    "txn_small = pd.read_csv('transactions.csv', usecols=['SENDER_ACCOUNT_ID','RECEIVER_ACCOUNT_ID','TX_AMOUNT'], dtype='object')\n",
    "txn_small['TX_AMOUNT'] = txn_small['TX_AMOUNT'].astype(float)\n",
    "txn_small"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b2b215fb",
   "metadata": {},
   "source": [
    "<h2>Sum transaction amounts for each node pair</h2>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "beb30e66",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Form nodes and edges by grouping by account numbers with a count of transactions and sum of amounts\n",
    "# Money came from source account and target is the account it went to\n",
    "# Create new columns agg_txn_count(number of transactions) and agg_txn_amt (sum of all transactions)\n",
    "edge_small = txn_small.groupby(['SENDER_ACCOUNT_ID', 'RECEIVER_ACCOUNT_ID'],as_index=False).agg({'TX_AMOUNT':['count','sum']})\n",
    "edge_small.columns = ['source','target','agg_txn_count','agg_txn_amt']\n",
    "edge_small"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "36a0d29f",
   "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>source</th>\n",
       "      <th>target</th>\n",
       "      <th>agg_txn_count</th>\n",
       "      <th>agg_txn_amt</th>\n",
       "      <th>value</th>\n",
       "      <th>title</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>884</td>\n",
       "      <td>24</td>\n",
       "      <td>4,219</td>\n",
       "      <td>4,219</td>\n",
       "      <td>from: 1\\nto: 884\\nagg_count: 24\\nagg_amount: 4...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10</td>\n",
       "      <td>8317</td>\n",
       "      <td>21</td>\n",
       "      <td>2,003</td>\n",
       "      <td>2,003</td>\n",
       "      <td>from: 10\\nto: 8317\\nagg_count: 21\\nagg_amount:...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10</td>\n",
       "      <td>8520</td>\n",
       "      <td>21</td>\n",
       "      <td>2,003</td>\n",
       "      <td>2,003</td>\n",
       "      <td>from: 10\\nto: 8520\\nagg_count: 21\\nagg_amount:...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100</td>\n",
       "      <td>3472</td>\n",
       "      <td>18</td>\n",
       "      <td>978</td>\n",
       "      <td>978</td>\n",
       "      <td>from: 100\\nto: 3472\\nagg_count: 18\\nagg_amount...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100</td>\n",
       "      <td>9954</td>\n",
       "      <td>18</td>\n",
       "      <td>978</td>\n",
       "      <td>978</td>\n",
       "      <td>from: 100\\nto: 9954\\nagg_count: 18\\nagg_amount...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>95</th>\n",
       "      <td>1017</td>\n",
       "      <td>4185</td>\n",
       "      <td>24</td>\n",
       "      <td>607</td>\n",
       "      <td>607</td>\n",
       "      <td>from: 1017\\nto: 4185\\nagg_count: 24\\nagg_amoun...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>1017</td>\n",
       "      <td>7378</td>\n",
       "      <td>24</td>\n",
       "      <td>607</td>\n",
       "      <td>607</td>\n",
       "      <td>from: 1017\\nto: 7378\\nagg_count: 24\\nagg_amoun...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>1017</td>\n",
       "      <td>7990</td>\n",
       "      <td>24</td>\n",
       "      <td>607</td>\n",
       "      <td>607</td>\n",
       "      <td>from: 1017\\nto: 7990\\nagg_count: 24\\nagg_amoun...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>1017</td>\n",
       "      <td>9896</td>\n",
       "      <td>24</td>\n",
       "      <td>607</td>\n",
       "      <td>607</td>\n",
       "      <td>from: 1017\\nto: 9896\\nagg_count: 24\\nagg_amoun...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>1018</td>\n",
       "      <td>1071</td>\n",
       "      <td>22</td>\n",
       "      <td>565</td>\n",
       "      <td>565</td>\n",
       "      <td>from: 1018\\nto: 1071\\nagg_count: 22\\nagg_amoun...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   source target  agg_txn_count  agg_txn_amt  value  \\\n",
       "0       1    884             24        4,219  4,219   \n",
       "1      10   8317             21        2,003  2,003   \n",
       "2      10   8520             21        2,003  2,003   \n",
       "3     100   3472             18          978    978   \n",
       "4     100   9954             18          978    978   \n",
       "..    ...    ...            ...          ...    ...   \n",
       "95   1017   4185             24          607    607   \n",
       "96   1017   7378             24          607    607   \n",
       "97   1017   7990             24          607    607   \n",
       "98   1017   9896             24          607    607   \n",
       "99   1018   1071             22          565    565   \n",
       "\n",
       "                                                title  \n",
       "0   from: 1\\nto: 884\\nagg_count: 24\\nagg_amount: 4...  \n",
       "1   from: 10\\nto: 8317\\nagg_count: 21\\nagg_amount:...  \n",
       "2   from: 10\\nto: 8520\\nagg_count: 21\\nagg_amount:...  \n",
       "3   from: 100\\nto: 3472\\nagg_count: 18\\nagg_amount...  \n",
       "4   from: 100\\nto: 9954\\nagg_count: 18\\nagg_amount...  \n",
       "..                                                ...  \n",
       "95  from: 1017\\nto: 4185\\nagg_count: 24\\nagg_amoun...  \n",
       "96  from: 1017\\nto: 7378\\nagg_count: 24\\nagg_amoun...  \n",
       "97  from: 1017\\nto: 7990\\nagg_count: 24\\nagg_amoun...  \n",
       "98  from: 1017\\nto: 9896\\nagg_count: 24\\nagg_amoun...  \n",
       "99  from: 1018\\nto: 1071\\nagg_count: 22\\nagg_amoun...  \n",
       "\n",
       "[100 rows x 6 columns]"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Make df smaller for faster processing\n",
    "first_5000_rows_df = edge_small.head(100)\n",
    "first_5000_rows_df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9116b6d3",
   "metadata": {},
   "source": [
    "<h2>Construct Network with NetworkX</h2>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "f6b17c05",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create graph network with networkX\n",
    "G = nx.from_pandas_edgelist(first_5000_rows_df, source='source', target='target')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "cd46e90e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['1', '884', '10', '8317', '8520']"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Access node and edge information from G, variable that stores graph\n",
    "# Show first five nodes of graph, pattern is source, target, source, target, source, target\n",
    "list(G.nodes)[0:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "5d57fa08",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('1', '884'),\n",
       " ('10', '8317'),\n",
       " ('10', '8520'),\n",
       " ('100', '3472'),\n",
       " ('100', '9954')]"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Show the first 5 edge of G\n",
    "# Shows ('target', 'source')\n",
    "list(G.edges)[0:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "19ed07ae",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('1', '884', {}),\n",
       " ('10', '8317', {}),\n",
       " ('10', '8520', {}),\n",
       " ('100', '3472', {}),\n",
       " ('100', '9954', {})]"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Show the first 5 edge of G with all attributes with data=True, there are no attributes in this case\n",
    "# [('source', 'target' {attributes})]\n",
    "list(G.edges(data=True))[0:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "0560b5dd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('1', {}), ('884', {}), ('10', {}), ('8317', {}), ('8520', {})]"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Show the first 5 nodes of G with all attributes with data=True, there are no attributes in this case\n",
    "list(G.nodes(data=True))[0:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "07ccfa55",
   "metadata": {},
   "source": [
    "<h2>Various network layouts</h2>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "18cd9a7e",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAzMAAAJrCAYAAADUAc2YAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy81sbWrAAAACXBIWXMAAA9hAAAPYQGoP6dpAABcW0lEQVR4nO3df3Dc933f+df3u2tCXoimhYVBmTEXDI0fjREkUVoWPNQUsw1F34wv6IGZG7Eh2WnK69y5Yq7GDTXlsFPxx7Uc6cI5KjGnnrmcJppCvIHmZogJLm7PNBIUokMLgRNNA6EOAZQlFjUtolhINLhrL73YvT+opUAQP/bH97vf7+e7z8eM/zCX+u6H5AL4vr6fz/v9tvL5fF4AAAAAYBjb6wUAAAAAQDkIMwAAAACMRJgBAAAAYCTCDAAAAAAjEWYAAAAAGIkwAwAAAMBIhBkAAAAARiLMAAAAADASYQYAAACAkQgzAAAAAIxEmAEAAABgJMIMAAAAACMRZgAAAAAYiTADAAAAwEiEGQAAAABGIswAAAAAMBJhBgAAAICRCDMAAAAAjESYAQAAAGAkwgwAAAAAIxFmAAAAABiJMAMAAADASIQZAAAAAEYizAAAAAAwEmEGAAAAgJEIMwAAAACMRJgBAAAAYCTCDAAAAAAjEWYAAAAAGIkwAwAAAMBIhBkAAAAARiLMAAAAADASYQYAAACAkQgzAAAAAIxEmAEAAABgJMIMAAAAACMRZgAAAAAYiTADAAAAwEiEGQAAAABGIswAAAAAMBJhBgAAAICRCDMAAAAAjESYAQAAAGAkwgwAAAAAIxFmAAAAABiJMAMAAADASIQZAAAAAEYKe70AAABQHalMVreTKT3I5rQlbGtXtF71ddwKADAX38EAAAiw6btLujKW0MjNeSUW08qveM2SFGuIKN7epCNdMbVu3+rVMgGgLFY+n89v/tsAAIBJ5hbTOj04oeszCwrZlpZz6/+4L7y+r6VRF3o7tbMhUsWVAkD5CDMAAosjNQi69T7jA+MJnRmaVDaX3zDErBayLYVtS+d6OnR4T8zFlQOAMwgzAAKFIzUIus0+49s+/Sl99JOfVfw+Jw+26US8teLrAICbCDMAAoEjNQi6Uj7jTnntUKdeZIcGgI8RZgAYjyM1CLpyP+OV+lTI0h+f+LLan/1M1d4TAEpBmAFgtMsj07p4bari63CkBn7l1Ge8Es0czwTgU4QZAMYaGE/o1NUJx67HkRr4jdOf8UpwPBOAHxFmABhpbjGtA5dGlcnmHLtmXdjWcN9+btLgC258xp3A8UwAfmJ7vQAAKMfpwQllHa4dyObyOj3oj6fggBufcScs5/LKZHM6dXVCl0emvV4OgBpHmAFgnOm7S7o+s+B4IfRyLq/rMwuamV9y9LpAqdz6jDvt4rUpvT2e8HoZAGoYYQaAca6MJRSyLVeuHbItvfXu2jdnqUxWk3fu6b3Eh5q8c0+pTNaVNQBufsad9srQpOYW014vA0CNYhQ2AOOM3Jx37Yn1ci6vkal5nVWHJIZwwhtufsadVjie2X+8y+ulAKhBhBkARrmfySrh8lPgRDKtmx/8WP/yWz/YcEBhXtLsYlr9Y7N683u36fIER1TjM+6klcczW5oI9ACqi2NmAIwym0zJ7efVeUm/cfnPdONWUpI2fUJeeP3GraQOXBrVADUEqEA1PuNO2+h4JgC4iTADwCgPqtSm9sFyruRjPnR5ghOq9Rl3UuF4JgBUG2EGgFG2hM34tkWXJ5TLlM/4aolkmqYYAKrOzO+YAGrWrmi9zOjxRJcnlMekz/hKeUm3kymvlwGgxhBmABilvi6smCEF9gzhRDlM+oyvZuIROQBmI8wAME68vcmIGRwM4US5TPmMr2bqETkA5uK7DgDjHOmKGTODgy5PKIdJn/ECSw+PyAFANRFmABindftW7WtpNOLJNV2eUA6TPuMFsWhE9XWMrwNQXYQZAEa60NupsCE3enR5QjlM+oyHbEvxtiavlwGgBhFmABhpZ0NE53o6vF5GUejyhHKY9BlfzuV1dG/M62UAqEGEGQDGOrwnppMH27xeRlHo8oRymPAZD9mW9rU0qqVpq9dLAVCDONwKwEjTd5d0ZSyhkZtm1KPQ5QnlOhFvVePTdfrng+9rOe+/pgBh29KF3k6vlwGgRhFmABhlbjGt04MTuj6zoJBtGdHxiS5PqNThPTH9/p9M6869n3q9lCec7+nQTkPn4gAwH2EGgDEGxhM6MzSp7McBxoQgI0nP1G+hyxMqcj+T1Y98GGRePtiuF/dQKwPAO/x0BWCEyyPTunhtyutllOVe+meaW0zz9Bplm02m5JfoHrIthW1L53s6CDIAPMchbgC+NzCeMDbIPJTX6cEJrxcBg/mhgURh5k337qiG+/YTZAD4AjszAHxtbjGtM0OTXi+jIst56frMgmbml+j4hLJUq4HE57c9pQ/u/fSxXSBLDwdixtuadHRvjM8wAF8hzADwtdODE49qZEwWsi299W5CZw2ZGwJ/2RWtlyW5etTMkjTct1/Sw7lID7I5bQnb2hWtp+YLgG9xzAyAb03fXdL1mQVjCv03spzLa2TKjDbS8J/6urBiLtdcxaIR1deFVV8XVseObXou9ow6dmwjyADwNcIMAN+6MpZ4dE4/CBLJtFKZrNfLgKHi7U2yXPpyCNmW4m1N7lwcAFzE4xYAvjVycz4QuzIFeT08vtOxY5vXS6m6VCbL0aVKWZJbMzOXc3kd3UtBPwDz8JMEgC/dz2SVWEx7vQzH+aErVbVM313SlbGERm7OK7GYfrKovCGieHuTjnTF1LqdovKNDIwn9OaN265cO2Rb6t4dpbAfgJEIMwB8yU9zNZxUra5UXppbTOv04ISuzywoZFtr7q7lJc0uptU/Nqs3v3db+1oadaG3k1k8a3C7o1/YtnSht9O16wOAm4L/UxWAkYK4g2HpYVeqIBsYT+jApVHduJWUpE2PCRZev3ErqQOXRjUwnnj0WiqT1eSde3ov8aEm79yr2Xojtzv6ne/pIEQCMBY7MwB8KYg7GIVuUUF1eWS67OGmy7m8lnN5nbo6obe/P6fF+w84mqZPOvq55R9172L4JQCjBe9uAUAgFOZqBEXQu0UNjCfKDjKrvZf4SLOrgoz0+NG0F15/R8feGNNcAOuqCqbvLul/GXjPtevblhSg/hoAahRhBoAvVWOuRjUFuVuU2zUdq210NC0I5hbTOvbGmF54/R394IMl194nlxezjwAYjzADlIhz/NUTb28KxJyZkG1pX0tjYLtFuV3TsZ7lXF6ZbE6nrk7o8sh01d/fDatrjtzG7CMApgvu4W3AQU62mGXeRvGOdMX05vdue72MigW5W5TbNR3FunhtSp97us7o+o9Kao7KVcuzjwAEA3dQwAacajHLvI3ytG7fqn0tjbpxK2n08MygdotKZbL6/T+d9k3txStDk+r+YqORf9dO1hyVKoidAwHUDiufd2ueMGC2gfGEzgxNKvtxl6VihWxLYdvSuZ4O/Z0vNm4ahlb+d8u5PPM2VplbTOvApVFlDL3hevlgu16Kt3i9DMdsFMy9Vhj+2H+8y+ullMTrz/i3fufL7MwAMBZhBliDU8c9CvUe5YahwwYfmXHSwHhCp65OeL2MohX+Dc/3dBh97GmlYnYp/WK473mj6pOOvTHm2e6jJen9s1/hqCsAY9EAAFjFyeMeyyXu6hT+m6AVNVfq8J6YTh5s83oZmyqE1+7dUQ337Q9MkCl1EKaXQralt941p7tZoebIq7/ToM8+AhB8fAcDVqh2i9nNBKGo2Skn4q1qfLqurKN/Tgp/vCvxRN1TNKJ4W5OO7o0ZtSuwGS+K0iuxnMtrZGpeZ9Xh9VKKcmUs4dlOV9BnHwGoDYQZYAWvWsxuxOSiZqcd3hMrqQ7JaSHb0tGuZr38lfaa6EjnZVF6JQrthk34Nxm5Oe9ZMA/y7CMAtYNjZsDHvD7usZ5sLq/Tg+bUi7htZ0NE/ce79J2vP69jXc1qjkZUrUk0hZu/+rqwOnZs03OxZ9SxY5sRN82l8tsuZSkK7Yb97n4mq8Ri2pP3DvrsIwC1I3g/gYEyeXncYyPLubyuzyzo373/Iz3f+rlA3jiXo3X7Vp3t6dBZdTwxu+dffesHGru96Oi/ZaFTVq3c/Plxl7IUJrQbnk2mPOsGF+TZRwBqC3dFwMe8PO5RjK9d+Utm0qyjsFNS8Npv/pIOXBp19N+zlm7+/DIIsxJbwv4/eOBl4Arq7CMAtcf/3+2BKrj5wY8169Fxj1KsHND5wuvv6NgbY5ozYN3VtrMhonM9zhaA19LNX2GX0lSWpF3Req+XsSmvAtfLB9tpKgIgMAgzqHkD4wn9d5e/6/UySlLYcbhxK6kDl0Y1MG5OK9pqcbKdc63d/Pl9l3IzprQb3hWtr1q9V8i2VBe29dqhzkANcQUA/3+3B1xkWtvZ1QpzbE5dndDC/YxOxFu9XpKvVNLOOYiDL4vhZVG6E0xqN1xfF1asIVKVXeHu3VFd6O2smd1FALWDnRnULFPbzq7n4rUpvc0OzRMO74lpuG+/undHJWnT41NBHXxZLC+L0p1gWrvheHuTq0f6fuHZrRrue179x7sIMgACiZ0Z1CST285uhJk0ayu0c56+u6QrYwmNTM0rkUzXzODLUpjQBWwje5qfMerf7khXTG9+77Zr1//G33/OqL8PACgVYQY1yfS2s+spzKTpP97l9VJ8aaN2zkEdfFkqE7qAbeQ//PCeBsYTOmzIjlrr9q3a19KoG7eStBIHgDKY/VMLKINfh2M6oTCTZmZ+yeul+F4tDL4sRzWL0t3wIJvTqasTujwy7fVSinaht1Nhh4+a1VIrcQC1jTCDmmN629nNhGxLb71L7QzKUyhKN51JNWS0EgeA8hFmUHNMbzu7meVcXiNT814vAwZzuyi9Wl4ZmjRmDhOtxKsrlclq8s49vZf4UJN37imVyXq9JABl4lwFaorpbWeLlUimlcpkOTqFsrhdlF4tptWQ0UrcXY8agNycV2JxjQYgDRHF25t0pCum1u3UGgGmsPL5fHAfUQOrTN65p69+w6wBmeX61u98WR07tnm9DBjq2Btjjhele2W473mjCuHnFtM6PTih6zMLCtnWhv8Ghdf3tTQyR2Ydpfx92paUy0u//IVt+qd/t1Vdu6M8FAJ8jjCDmvJe4kP1fvOG18uoisGvdeu52DNeLwOGmltM68ClUWUMb9Ucsi0d62rWWYdrUqqBVuKVGxhPlLXTtVIzOzaArxFmUFPYmQGKNzCe0KmrE14vo2LN0YhGT8a9XkZFaCVeussj044NRmYHDPAvGgCgppjedrZYlh7+WYFKOFmU7qVCDZnJaCVemoHxhGNBRtKjXZ0bt5I6cGlUA4Z0ygNqAWEGNSUobWc3E4tGuNmBI07EW/XqoU7VhW1jO5zlJd1OprxeBqpkbjGtM0OTrlx7OZdXxsBZRkCQEWZQc4LSdnY9IdtSvK3J62UgQA7viWm4b79++QvmHlt8YHjtD4p3enBC2So0rjBplhEQZIQZ1JwjXbFAdGhaz3Iur6N7ac0KZ+1siOiXfu6zMvU5wJYwP+5qwfTdJV2fWaja93iTZhkBQcV3d9Sc1u1bta+lMZC7MyHb0r6WRjobwRUjN+dl4nMAashqx5WxRFW/tz/I5vRP336vau8H4EmEGdSkC72dCgcwzIRtSxd6O71eBgLI5IGz1JDVjpGb81Xdec9L+svER/rNb/4ZOzSARwgzqEk7GyI6Z+Dcic2c7+mgZShcMZtMycBNGWrIaoiXgfsvEx/R5QzwCGEGNSsobWcLXj7Yrhf3UCsDd5haQE8NWe3wMnDnJbqcAR4hzKCmVdJ2NmRbqgvbeu1Qpz776U+5tMLi1/BSvMWTNaA2mFhATw1ZbfFL4KbLGVBd5v10AhxWaDvbvTsqSZuGmsLr3bujGu7br1+NPaOPfvIz19e50RrYkYHbTBw4Sw1ZbfFT4KbLGVA9VEQCelhD03+8S9N3l3RlLKGRqXklkunHjixYelhIHG9r0tG9sUdPe88OTSpkW1UpOl1vDYDbCgNnZ128QasL28o4+HSdGrLaUgjcfqjtyubyOj04of7jXV4vBQg8wgywQuv2rTrb06Gz6lAqk9XtZEoPsjltCdvaFa1fsyOS291zPr/tKf3r3/rVDdcAVEO8vUn9Y7OufN5DtqW/vyemxq1bdPHaVMXXo4as9lQjcBdrOZfX9ZkFzcwv8dAJcJl/9mQBn6mvC6tjxzY9F3tGHTu2rRkiqtE954N7P1Xb9q3rrgGoFjcHzhYK9Z2qY6OGrDbF25t8M0MsZFt6611qZwC3EWaAClSje05e0u1kyuV3ATbn1sDZ1YX6ldaxsSNTu9wM3KVazuU1MjXv9TKAwOMxL1CBanXP8UuXHuBCb6cOXBp19IZxrUL9SurYULsKgfvGraQvQk0imVYqk2VXHXARX11ABarVPcdPXXpQ2woDZ09dnXDsmhsV6pdTx4ba5kbgLldhZ71jxzavlwIEFndIQAWq0a7W+vh9AL9wcuBsKYX6xdSxAYXA7RfsrAPuIswAFSh0z3FTLBrhpg2+Q6E+/MzJwF0pdtYBd/EVBlTIze45IdtSvK3JlWsDlaJQH35WSeB2CjvrgPusfD7v/aFSwGDTd5f0wuvvuHb94b7nKWyG71GoD7+aW0zr9OCErs8sVG3AcUFzNKLRk/GqvR9QiwgzgAOOvTHmePeckG2pe3eUCdIwDoX68KOVgXs26f5gzZBt6VhXs876qH4HCCLCDOCAucW0DlwaVcbBQs+6sK3hvv3rdnkCAJQnlcnq3VtJ/e/f/mvdvHvftfdhZx1wHzUzgAPc6J6zUbtaAED56uvC+vVf2K5vf32/fjX2Wce7Uq4eBAvAPYQZwCFetasFAJTv9158zvGOY2sNggXgDsIM4CDa1QKAWdhZB8xGzQzgglK65xRe39fSqAu9nfwABAAPXB6Z1sVrUxVf5+WD7TyQAqqIMAO4iHa1AGCOgfGEzgxNKpvLl9SdMmRbCtuWzvd0cEQYqDLCDFAltKsFAP9jZx0wC2EGAABgFXbWATMQZgAAADbAzjrgX4QZAAAAAEaiNTMAAAAAIxFmAAAAABiJMAMAAADASIQZAAAAAEYizAAAAAAwEmEGAAAAgJEIMwAAAACMRJgBAAAAYCTCDAAAAAAjEWYAAAAAGCns9QIQHKlMVreTKT3I5rQlbGtXtF71dXzEAAAA4A7uNFGR6btLujKW0MjNeSUW08qveM2SFGuIKN7epCNdMbVu3+rVMgEAABBAVj6fz2/+24DHzS2mdXpwQtdnFhSyLS3n1v8YFV7f19KoC72d2tkQqeJKAQAwF6cegI0RZlCygfGEzgxNKpvLbxhiVgvZlsK2pXM9HTq8J+biCgEAMBenHoDiEWZQkssj07p4bari65w82KYT8VYHVgQAQDA4deqB3RzUEsIMijYwntCpqxOOXe+1Q516kR0aAAAqPvXwT37ti/ow9TN2c1BzCDMoytxiWgcujSqTzTl2zbqwreG+/dTQAABqmlOnHixL2uiujhpWBBFzZlCU04MTypbwpKgY2Vxepwed2+kBAMA0A+MJR4KMtHGQkfRox+fGraQOXBrVwHjCkfcFvESYwaam7y7p+sxCSdvexVjO5XV9ZkEz80uOXhcAABPMLaZ1Zmiy6u+7nMsrk83p1NUJXR6Zrvr7A04izGBTV8YSCtmWK9cO2ZbeeveTJ0OpTFaTd+7pvcSHmrxzT6lM1pX3BQDAa26ceijVxWtTepsdGhiM1hbY1MjNecd3ZQqWc3l9+z9+8Oh9KFoEANSCwqkHP3hlaFLdX2ykhgZGogEANnQ/k1Xn2W/L7Q9JyLK0vMFHkaJFAECQnB2aVP/YrGsPC0sRsi11746q/3iX10sBSsYxM2xoNplyPchI2jDISBQtAgCCxc1TD6WihhUmI8xgQw8cbMXsBIoWAQCmu5/JKrGY9noZj1ldwwqYgjCDDW0J+/cjQtEiAMBE1Tr1UIrlXF4jU/NeLwMomX/vVOELu6L1cqePmTNeGZrUnM+ebgEAsBG/nXooSCTTdBGFcQgz2FB9XVgxHxfbM3gTAGAav556yEu6nUx5vQygJP78aoKvxNubXJszUymKFgEApvHzqQe/7hpVghl2wcacGWzqSFdMb37vttfLWFehaPFsT4fXSwEAYFOFUw+zPjwm7dddo1JN313SlbEEM+xqQDA+sXBV6/at2tfS6OvdGYoWAQAmibc3yY8/Vv9idtHrJVRkbjGtY2+M6YXX31H/2KxmVwUZ6eFxutnFtPrHZvXC6+/o2Btj1N8ajDCDolzo7VTYj991P0bRIgDAKFZePhkz85h/9W//2tgb+4HxhA5cGtWNW0lJ2nSODzPsgoEwg6LsbIjonI+PcVG0CAAwxcB4Qm/emPV6GWsytbHO5ZFpnbo6oUw2V/IwUmbYmY0wg6Id3hPTyYNtXi9jXUEsWgQABMvcYlpnhia9Xsa6TGysMzCe0MVrU45cixl25iHMoCQn4q169VCn6sK272poglK0CAAIrtODE8r68XzZCoXGOiZwIxwyw84s3P2hZIf3xDTct1/du6OStGmoqUbmsfSw1SUAAH41fXdJ12cWSj4GVW0mNdZxIxyaetSuVhFmUJadDRH1H+/Sd77+vI51Nas5GnmiZ74lqTka0T/Yu0s7tj3l6npi0Yjq6+g0DgDwrytjCd+daliPl411ip0L41Y4NPGoXS3j7g8Vad2+VWd7OnRWHUplsrqdTOlBNqctYVu7ovWPBYz+sVlXnkaFbEvxtibHrwsAgJNGbs77flemoNBYp2PHtqq8XzlzYQrh0K17C2bYmYEwA8fU14XX/abn5uDN5VxeR/fGXLk2AADF2OyB3v1MVgnD6jCq0VhnbjGt04MTuj6zsG4wWTkX5s3v3da+lkZd6O10NRwWjtqdFWHG7wgzqIrC4M0bt5KOfuMJ2Za6d0fV0sT0XgBAdZWym/BgOffE8Ea/c7uxzsB4QmeGJh/VvJQ6FybjctgqHLXjGLu/Wfl83rSvLRhqbjHt+DefurCt4b792tkQceyaAABspJjdhILC67/8hW36D//lXhVXWRlL0vtnv+LajfzlkWnH2im76Vu/8+WqHbVDeWgAgKpxY/Dm+Z4OggwAoGrKnTI/8UNzgozkbmMdJ+fCuI0Zdv7Hvhmq6vCemBbuZxz5JvbywXa9uIdaGQBAdVSym2BI3b8kdxvr+H1o6GrMsPM/wgyq7kS8VY1P1z06J1tKDU3IthS2LZ3v6SDIAACqxqTdhEq52VjHhKGhBcywMwNxE54odfBm4fXu3VEN9+0nyAAAqsa03YRKhGxL+1oaXWmsY8rQ0AJm2JmBfyF4pjB481E3mKl5JZJrdIOJRhRva9LRvTG6lgEAqs6k3YRKhW1LF3o7Xbm2m3NhnMYMO3MQZuC5UgZvAgBQTYXdhFrhZmMdk4aGMsPOHNwlwlc2GrwJAEC1mbCbYEmOzLBxs7GOSUNDmWFnFmpmAAAA1mHCbkJD/Ra9eqhTdWF70xrU1UK2pbqwrdcOdeqleItLK5Rmkyljhoa6edQOziPMAAAArMGU3YTF1AP9xi/t8HVjHZPmtTDDziwcMwMAAFiDKbsJeUm3kyl17Njm28Y6psxrYYadeQgzAAAAazBpN2HlWv3YWGdXtN6x2p6NbAlZWs6LGXY1xIyYDAAAUGWm7CZI66+10Fjnudgz6tixzbMOofV1YcVcPrrVHI3oT/7XX/PtUTu4g50ZAACANVRrN6FSpkyqj7c3qX9s1pWGCoW5MMywqz1WPp/3+9coAACAJ/b/7ohmfd4EoDka0ejJuNfL2NT03SW98Po7rl1/uO/5NYOJX47awR3m7J8CAABUWby9qeR2x9Vk0qT61u1bta+l0fG/z5BtaV9L47o7LH45agd3EGYAAADWcaQr5us5M6ZNqr/Q26mww2GGuTC1jTADAACwDrd2E5yw2Y6EH+1siOhcT4ej12QuTG0jzAAAAGzAjd0EJ5i6I3F4T0wnD7Y5ci3mwoAwAwAAsAE3dhOcYPKOxIl4q1491Km6sF3yrlfItlQXtvXaoU69FG9xaYUwBd3MAAAAinB5ZFoXr015vQxJD3ckgnAjP7eY1unBCV2fWVDItjasTyq8vq+lURd6O40NcnAWYQYAAKBIA+MJnRmaVDaXL3vKfF6q+BpBO1rFXBiUizADAABQAid2E9iRWB9zYVAKwgwAAEAZnNhNYEcCqAxhBgAAoEJO7CawIwGUjjADAAAAwEjEfbiOJ00AAABwA3eUcMWjM8A355VYXOMMcENE8fYmHemKqXU7Z4ABAABQOo6ZwVF0ZwEAAEC1EGbgmEp775/r6dDhgPXNBwAAgHsIM3CEU1ORTx5s04l4qwMrAgAAQNDZXi8A5hsYTzgSZCTp4rUpvT2ecORaAAAACDbCDMqWymT1Jz+4q1f+aNLR674yNKm5xbSj1wQAAEDwcMwMJdmoS5lTQral7t1R9R/vcuHqAAAACApaM6MopXQpq9RyLq/rMwuamV9SSxNtmwEAALA2jplhUwPjCR24NKobt5KS5GqQKQjZlt56l9oZAAAArI8wgw1dHpnWqasTymRzVQkxBcu5vEam5qv2fgAAADAPYQbrcrJLWTkSybRSmaxn7w8AAAB/I8xgTXOLaZ0ZcrZLWanykm4nU56uAQAAAP5FNzOs6dgbY7pxK1nVo2VrefYzdfpvOz6vI10xtW6nGQAAAAA+QZjBE6bvLumF19/xehmPFLqn7Wtp1IXeTu1siHi9JAAAAPgAx8zwhCtjCYVsy+tlPFLYHbpxK6kDl0Y1MF56l7NUJqvJO/f0XuJDTd65Ry0OAABAADBnBk8YuTnv+fGytSzn8lrO5XXq6oQW7md0It664e/faMCnJSnWEFG8vYkjbAAAAIbimBkecz+TVefZb8uED8Vrhzr14p7YE79eyoBPjrABAACYi2NmeMxsMmVEkJGkV4YmNbeYfuzXSh3w6cQRNgAAAHiDMIPHPMjmvF5C0bK5vE4PTjz6/5UM+FzO5ZXJ5nTq6oQuj0w7vVQAAAC4gJoZPGZL2Jx8u5zL6/rMgmbml/T92Q8dG/B58dqUPvd03ZpH2AAAAOAf5ty5oip2Revlnz5mmwvZlr757/+T4wM+1zrCBgAAAH8hzOAx9XVhxQwqgl/O5fXHEz/Sz5adPR63+ggbAAAA/Icwgyc8t/OzXi+hJJlsTk53kl55hA0AAAD+RJjBk0w6Z+aikG3prXfpbgYAAOBXhBk84b3ER14vwReWc3mNTM17vQwAAACsgzCDx9zPZJWg8P2RRDKtVCbr9TIAAACwBsIMHmPS0MxqyEu6nUx5vQwAAACsgTCDx5g0NLNa+DsBAADwJ4Zm4jEmDc2sFv5OAABwXyqT1e1kSg+yOW0J29oVrVd9Hbeq2BifEDymMDSTo2YPWXr4dwIAAJw3fXdJV8YSGrk5r8Ri+rH7D0tSrCGieHuTjnTF1Lp9q1fLhI9Z+Xye+1Y8Zv/vjmiWJgCSpOZoRKMn414vAwCAQJlbTOv04ISuzywoZFta3mBgXOH1fS2NutDbqZ0GDfeG+zg/gyfE25sUshk2E7ItxduavF4GAACBMjCe0IFLo7pxKylJGwaZla/fuJXUgUujGhhnBhw+QZjBE450xTb9xlILlnN5Hd0b83oZAAAExuWRaZ26OqFMNlfyvcZyLq9MNqdTVyd0eWTapRXCNIQZPKF1+1bta2l0fHcmZFtqjpqxNRyyLe1raVRLE+dzAQBIZbKavHNP7yU+1OSde2XNYBsYT+jitSlH1nPx2pTeZocGomYG65hbTOvApVFlHGxLXBe2Ndy3X3/2nxZ0ZmhS2VzetztAhbVyLhcAUKucLM53876Cn9W1jTCDdQ2MJ3Tq6oRj13vtUKde3PPw2FYphX9eWLlWAABqiRvF+cfeGNONW0lHf96HbEvdu6PqP97l2DVhHsIMNnR5ZNqRLeGXD7brpXjLE7/+6KnP1LwSybQvWkKvt1YAAIJuYDxR1umJkG0pbFs619Ohw6seBk7fXdILr7/j9FIfGe57nmPhNYwwg01V+o3tfE9HUbscq4dl/c9v/YXmPvxJJUsvWqlrBQAgaJx6gHnyYJtOxFsf/f+zQ5PqH5t17RTG1rqwfvNXv8AsmhpFmEFRvOgH7/Y3P4ne9QAASO4eLa/G/Dp+ntcuwgxKstGxMEtSLBpRvK1JR/fGKt7ydXtbesdnn9LBX3jWkbUCAGAqt4rz//Vv/ao+/amQfuuNMceuu5mNjrshmAgzKNvqY2G7ovWqrws7+h5uFAzalrT356P6v//xXseuCQCAqdz4WesHq4+7IZgIM/A1WjkCAOAet09BeI3upMHH0Ez42s6GiM71dDh6zfM9HQQZAAAkXRlLOD4k209eGZrUnMv1OvAWYQa+d3hPTCcPtjlyrZcPtvOEBgCAj43cnA/c8bKVsrm8Tg8619gA/kOYgRFOxFv16qFO1YXtkp8ghWxLdWFbrx3qZH4MAAAfu5/JKhHwXYvlXF7XZxY0M7/k9VLgEsIMjHF4T0zDffvVvTsqSZuGmsLr3bujGu7bz44MAAArzCZTvhhW7baQbemtdxNeLwMucbb1FOCynQ0R9R/vqmqLaAAAguiBg811/Gw5l9fI1LzOytkaXPgD3cxgvGq0iAYAIGgm79zTV7/xXa+XURWWpPfPfoX7gwDiXxTGq68Lq2PHNq+XAQQWDwyAYNoVrZcl1cRRs7yk28kU9wsBxE8jAMATHh3lvDmvxOIaRzkbIoq3N+lIV0yt2znKCZiovi6sWENEswFvAlBQK8fqag1hBgBqQLG7K3OLaZ0enND1mQWFbGvNlq15SbOLafWPzerN793WvpZGXejtZH4TYKB4e5P6x2YD3Z65YEuYvldBRM0MAARUqbsrA+MJnRmaVDaXL+nGJmRbCtuWzvV06DBdAwGjTN9d0guvv+P1MlxHzUxwEWYAIGCK2V0pKLzeHI1oNln5UZOTB9t0It5a8XUAVM+xN8Z041Yy0LszzdGIRk/GvV4GXMB+GwAEyMB4QgcujerGraQkbXpzUnjdiSAjSRevTentceY5ACa50NupUGnzqI0Ssi3F25q8XgZcQpgBgIC4PDKtU1cnlMnmPH3C+srQpOZqpKAYCIKdDRHFovVeL8M1y7m8ju7lCGxQEWYAIAAGxhO6eG3K62VIkrK5vE4PTni9DABFmr67pJn5+14vwxUh29K+lkYGaAcYYQYADDe3mNaZoUmvl/HIci6v6zMLmplf8nopAIpwZSyhkB3Mc2Zh29KF3k6vlwEXEWYAwHCnByeU9Vnhrm1Jf/hnt71eBoAijNycD2zx//meDtrGBxz96QDAYNN3l3R9ZsHrZTwhl5eu/HlC351ZYLgm4GP3M1klAlrj9vLBdr1Iu/jAY2cGAAzm9+MhheGaL7z+jo69MUZjAMBnZpMpBWlPJmRbqgvbeu1Qp16Kt3i9HFQBYQYADGbC8ZDC+m7cSurApVEN0LoZ8I0H2ZzXS3BE4aFO9+6ohvv2syNTQzhmBgCGMu14yHIur+VcXqeuTmjhfobhmoAPbAmb/VzbkhSLRhRva9LRvTG6ltUgwgwAGMrk4yEXr03pc0/X8fQU8NiuaL0sycjvJb9/+Ff0639ju+rruJ2tZWbHcQCoYaYfD2G4JuC9+rqwYgZ2+7IkggwkEWYAwFimHw9huCbgD/H2Jl83EllLLBohyEASYQYAjFU4HmIqhmsC/nCkK+b7RiIrhWxL8bYmr5cBnyDMAIChTD0eslLItvTWu3Q3A7zUun2r9rU0GrM7s5zL6+he6u3wEGEGAAxm4vGQlZZzeY1MzXu9DKDmXejtVNjh7yW29fB/TgrZlva1NNK1DI8QZgDAYKYdD1lLIplWKpP1ehlATdvZENG5ng5Hr/nywXZ9KuTsrWbYtnSht9PRa8JshBkAMJhpx0PWkpd0O5nyehlAzTu8J6aTB9scudbLB9v1tV9rcTwgne/p0E7Dj9fCWYQZADCcG8dDqs30NtNAUJyIt+rVQ52qC9slPyQJ2ZbqwrZeO9Spl+ItkpwPSMymwmqEGQAwnBvHQ6rN9DbTQJAc3hPTcN9+de+OStKmoabwevfuqIb79j8ROJwOSMBKVj6fN/uwNQDUiFQmq9vJlB5kc9oStrUrWv/YnIXLI9O6eG3KwxWWx5L0/tmvMDMC8KHpu0u6MpbQyNS8Esm0Vt40Wno47yXe1qSje2ObFuXPLaZ1enBC12cWFLKtDev9Cq/va2nUhd5OjpZhXYQZAPCxRzcSN+eVWFzjRqIhonh7k450xdS6fasGxhM6MzSpbC5vTGOA5mhEoyfjXi8DwCY2e6BSLCcDEkCYAYANOPXDu1SVPMGU9Oi/9buQbelYV7POGn5MDkB5vPoei+AgzADAKqXuhjit3N2VkG0pbFs619Ohw3ti+nfv/0hfu/KXjq/PacN9z/P0FQBQFsIMAHzMD+e5nap7OXmwTb/d/fP6xbPfll+/yYdsS927o+o/3uX1UgAAhiLMoCxsCyNonNoNqXQNp65OVHSNlV471Kl//e//k2YX045d00l1YVvDffsp7AUAlI0wg6J5ffQGcIuTuyEn4q1l/bdzi2kduDSqjIPzVurCtn7xMxn9xYIt2SHHruuU1w51MjMCAFARwgw25YejN4Bb3NgNKecG/dgbY7pxK+lsB7Lcsn5656ae+sKXnLumQ14+2M7MCABAxZhShg0NjCd04NKobtxKStKmN1qF12/cSurApVENjCdcXyNQrrnFtM4MTTp6zVeGJjVX4rGu6btLuj6z4HwrZTukp77wJf3N2GdLHlTntn/YvcvrJQAAAoAwg3VdHpnWqasTymRzJd9kLefyymRzOnV1QpdHpl1aIVCZ04MTyjocILK5vE4PlrbTc2Us4VrYCNmWdjXWK+yzMHM7mfJ6CQCAACDMYE0D4wnHJolfvDalt9mhgc+4tRuynMvr+syCZuaXiv5vRm7OuzbgcjmX1/dnP9Q5n81xeeBgbRAAVFMqk9XknXt6L/GhJu/cUyqT9XpJNY32UzWk2A5kbh296f5iIzU08I3CbogbIcK2pEvDU/onv9ayaae/+5msEi53G0sk0/qNX9qhhfsZxx5SVGpLmGdpAMxBEyT/IswEXDlffG4evWGeBPzCzd2QXF761sQH+tbEB5v+kJtNplyfA5PXw2NdJ+Ktany6rqwW1E6yJO2K1nvy3gBQimKaIOUlzS6m1T82qze/d5smSFVGN7OAKrcD2T/e9/P6B3847tq6mPQNP7ifyaqzysMk1+v0917iQ/V+84br7z/4tW49F3tGUmnfH9zQHI1o9GS8qu8JAKXyw/wxbI59/gCqpAPZb7/5fVku1QmHbEtvvUvtDLz357eSVQ0y0vqd/qp13Grl++xsiKj/eJe+8/XndayrWc3RiKrVHiBkW4q3NVXp3QCgPDRBMgfHzAKmkuF/bj+dXc7lNTI1r7PyVyEyasvAeEL/4o/e9+z9lz9+wnfq6oQW7mf0290/L0tyNVytd6yrdftWne3p0Fl1PFZT98GPf6qvXflLV9aynMvr6F6eVALwL6ebIH3u6ToGBLuIMBMgTn7xuSWRTCuVyW5YEA24pZKw74bCD7lYQ0SzLjYBiEUjm37N1deF1bFj26P/v6+l0fEhniHbUvfuKEdNAfgWTZDMwzGzgHDji88NhULkYtH+EE7xa9h/ZWhSz4bTUt6dVsXlHuu60Nvp+GyasG3pQm+no9cEACf5Zf4Yisfj8YBw44vPLZvNl6D9IZzm57CfefAzjf7lTT31hS+5cv1yj3XtbIjoXE+HTl117gfw+Z4OnkwC8K3C/DGnrZw/VtiZLnZcBjbH31oAuPXF55b1Cp5pfwi3+Drs2yE99YUv6Vdjn9V/+C/3fHWs6/CemGOzaV4+2M6ZcQCe2yhEuDl/LGRbujwyo89+egsPax1Ga+YAODs0qf6xWc9mRpTCkvT+2a888fSB9odwy/TdJb3w+jteL2NDIdvSf/8rO/THf/UjZTbZuSxFXdjWcN/+isN+pV+f53s6CDIAPFPsiY9v/8cP9KN7P3V1LaWMy+BhbXEIMwGw/3dHXC0edtJa8yWcKso+ebBNJ+KtFV8HwWJK2G+ORvS1/V909FjXa4c6HQsR5c6u4ocxAK+U9H3LsrTso1tiHtYWj2NmhrufySphSJCRpL/V/Mxj/5/2h3DbyM153wcZ6WGnv9/4pR2+PdZVmE3z6Ann1LwSyTWecEYjirc16ejeGF3LAHhm5Y6yVMTMPR8FGenJNv48rF0fYcZws8lU1Yf/VeL2wiedzGh/CLeZFPYLnf5OxFvV+HSdb491rTebhgJWAH7htzb8leJh7cZozWy4zTqD+c1fJD7SzPySJNofwn2mhf3C1/PhPTEN9+1X9+6opIchZSOF17t3RzXct79qP/AKs2meiz2jjh3bCDIAPOfXNvyVemVoUnOGPJyrNsKM4dbrDOZXIdvSW+8mHnVgc/r4z8r2h4BpYX/l13PhWNd3vv68jnU1qzka0epIY+lhrc2xrmYN9z2v/uNd7EoCqFl+bsNfKR7Wro/HaIbbFa2XJRnz9Hk5l9fI1LykzTt6lKsQmM72dDh+bZjFpLBv6eHX82oc6wKAjRVq+QbGE452hPSTtWbV4CF+Chquvi6sWEPEmG5m0sNC5z/567uuFWUXAtNZEWZqnUlhPxaNbBpMCse6AACPdyuzLcmAXi8V4WHt2sx5bIl1xdubNj1T7yd5SXMf/sTV90gk00plsq6+B/yvEPZN0BDZ4vUSAMAYA+MJHbg0qhu3kpKCH2Skx0+34BOEmQA40hUzovVsNRU6QwGmhP335j7S2+MJr5cBAL53eWRap65OKJPN1dz9Dw9rn0SYCYDW7Vu1r6XR62X4jmnF33CHSWGfbjUAsLGgdisrFg9rn0SYCYh//tVf8HoJvmNS8TfcUwj7JuzO0K0GANYX5G5lpeBh7eO42wsIU548S9LOhk8/0WLWaet1hkJtutDbqbABYYbW4gCwPjfm05mIh7WP428jIExJ6SHb0q+3b3e9KLuYzlCoHTsbIjpnSPeXQrcaAMAn3JpPZ6K/mP3Q6yX4CmEmIExJ6cu5vI7ujblalB2yLcXbmly5Nsx1eE9MJw+2eb2MTdGtBgCedGUsYcRx4Wr4V//2B9RXrmDGHTA2VZin4Wch29K+lka1NG11tSi7EJiA1U7EW/VPfu2LXi9jU3SrAYDHjdycZ1fmY9RXPo4wExAmzNMI25Yu9HZKcq8oe2VgAtby1c7Pe72ETdGtBgA+cT+TVYKdiEeor3wcYSZA/D5P43xPh3auCFxuFGWvDEzAWkzYxZTMqYMDALfNJlNiT+Zx1Fd+gjATIH6ep/HywXa9uOfxo19uFGWvDkzAaibsYkrm1MEBgNt4uPMk6is/wU/LAPHbPA3bkurCtl471KmX4i1r/h4ni7LXCkzAWvy+i0lrcQD4BA931kZ95UN8OgLGT/M0On9um4b79m8aME7EW/XqoU7Vhe2SbzBDtrVpYAJW8/MupkRrcQBYyZTjwdVGfeVDhJmA8dM8jQu9nUUf+Tq8J6bhvv3q3h2VpE1DTeH17t3RogITsJLfdjFXorU4ADzOlOPBXuAIHmEmkPwwT6OcYzI7GyLqP96l73z9eR3ralZzNPLEkxhLUnM0omNdzRrue179x7uokUFZ/LSLuRKtxQHgSX4/HuwVjuBJVj6f9+9ZC1RkYDyhM0OTyubyVT9S0xyNaPRkvOLrpDJZ3U6m9CCb05awrV3Reo7fwDED4wmduuqfXv0h21L37qj6j3d5vRQA8JXpu0t64fV3XLt+yJKWDbsjtiS9f/YrNX9fRJwLsFKPbjnFyWMy9XVhdezYpudiz6hjx7aa/4KFs/ywi7kSrcUBYG1uHg9+6lO2ju3dpeG+5x1tStTs8skR6isf4m8g4ApHt6bvLunKWEIjU/NKJNOu9mvnmAxMciLeqsan68raxbQsycm9bVqLA8D6LvR26sClUUdPm2wJWfrO1/c/+t57omlr2T8TQralsG3pfE+HXtwT039dyqh/bNaV0zHUV36CY2Y1qHB06/TghCZ+eE9Ofo1xTAammltM6/TghK7PLChkWxv+8Cm8vq+lUX/j2a36g+/+54rf/+WD7XTkA4BNOH08+LVDnWs2ESr3Z8LK5kduH40b7nteLU1bXbu+KQgzNWxuMa0Dl0aVcbATRl3Y1nDffp4uw1gb7WJaeritH29r0tG9sUc/RMqtT1v9FA8AsLnLI9O6eG2q4usU8xCpnJ8JKx17Y0w3biUd3Z3hwfHjCDM1rlpPOAATldKAwomneACA4njxEKmcpkQ8OHYfYQZVfcIBBF2lT/EAAMUx5SESD47dRZiBJI7JAG6gtTgAuM+Eh0g8OHYPYQaPmPKEAwAAYC1+fojEg2N3EGbwBBOecAAAAJiGB8fOI8xgQ35+wgEAAGAiHhw7hzADAAAAeIQHx5UhzAAAAAAwku31AgAAAACgHIQZAAAAAEYizAAAAAAwEmEGAAAAgJEIMwAAAACMRJgBAAAAYCTCDAAAAAAjEWYAAAAAGIkwAwAAAMBIhBkAAAAARiLMAAAAADASYQYAAACAkQgzAAAAAIxEmAEAAABgJMIMAAAAACMRZgAAAAAYiTADAAAAwEiEGQAAAABGIswAAAAAMBJhBgAAAICRCDMAAAAAjESYAQAAAGAkwgwAAAAAI4W9XkCtSGWyup1M6UE2py1hW7ui9aqv468fAAAAKBd30y6avrukK2MJjdycV2IxrfyK1yxJsYaI4u1NOtIVU+v2rV4tEwAAADCSlc/n85v/NpRibjGt04MTuj6zoJBtaTm3/l9x4fV9LY260NupnQ2RKq4UAAAAMBdhxmED4wmdGZpUNpffMMSsFrIthW1L53o6dHhPzMUVAgAAAMFAmHHQ5ZFpXbw2VfF1Th5s04l4qwMrAgAAAIKLbmYOGRhPOBJkJOnitSm9PZ5w5FoAAABAUBFmHDC3mNaZoUlHr/nK0KTmFtOOXhMAAAAIEsKMA04PTihbQn1MMbK5vE4PTlR0jVQmq8k79/Re4kNN3rmnVCbr0OoAAAAA79GauUJ/NfeRrs8sOH7d5Vxe12cWNDO/pJam4ts20w4aAAAAtYIGAGVYGRhmXTwKFrItHetq1tmejk1/L+2gAQAAUGsIMyUoJTA4pTka0ejJ+Ia/h3bQAAAAqEUcMyvSysAgqSpBRpISybRSmazq69b+p6qkHfTyx+Hn1NUJLdzP0A4aAAAARiHMFMGp+THlyEu6nUypY8e2J15zuh30556u04vs0AAAgCKlMlndTqb0IJvTlrCtXdH6dR/AAm7g07YJJwNDueYW00+EGbfaQXd/sZEaGgAAsC6aDcFPqJnZwNxiWgcujSqTzXm6jk+FLP1vf+8XH6trOfbGmG7cSjp63C1kW+reHVX/8S7HrgkAAIKBZkPwI+bMbMCN+THl+Nnyw7qWyyPTkh4+Ebk+s+B43c7KdtAAAAAFA+MJHbg0qhu3kpI2rx0uvH7jVlIHLo1qYDzh+hpRmwgz63ArMFTi4rUpvT2e0JWxhEK25cp7hGxLb73LNxwAAPDQ5ZFpnbo6oUw2V/J90XIur0w299hDWcBJhJl1uBkYKvHK0KSu/eAD10LWci6vkal5V64NAADM4nSzobfZoYHDCDPrGLk576tdmYKfLed056OfuvoehXbQAACgdrnVbGjOxYHjqD2EmTXcz2SV8OkXWjXyVaEdNAAAqF1u1A5nc3mdHpxw9JqobbRmXsNsMiX/7clU1wOPO7gBAADvFGqHnbay2VBLk3ttm5l/Uzv4V10DN/LSljCbdgAA1KorYwnZljsnQgrNhs72dDh6Xebf1CbCzBpq/UbekrQrWu/1MgAAgAfmFtMaGE+4drS90GzorD4JM5XspBQz/yYvaXYxrf6xWb35vdvMvwkQwswadkXrZUk1e9QsFo2wFQsAQA0aGE/olT96Xw+W3b0LSiTT+qu5j3T1vR9WtJMyMJ7QmaHJR7U9pc6/OdfT8dhQcpjHyufztXrPvqFfOX9NH/3kZ14vo+pCtqVjXc2Ob/0CAAB/uzwy7Vgb5mKtt5Oy+vW1dlKcWu/Jg206EW+t+DrwRm2fp1rHwHiiJoOM9PCJxdG9PKEAAKCWODlPphSl7qQMfDynhvk3KGBnZpW5xbQOXBpVxoAmAJs9zSjnet27o+o/3uXYNQEAgL+ZdO8jSf/jl39e/e/OOrreurCt4b791NAYiJ2ZVdzoqe6GHZ99SmHbcvSaYdvShd5OR68JAAD8zZR7n4L/67v/WT9bdjZ4Mf/GXISZFQo91Z3c7XCDZUkHf+FZnXO4ruV8TwdPJAAAqCGm3Pus5vRyV86/gVkIMytcGUso5PBuhxvyeenXf+FzOrwnppMH2xy55ssH2/Ui3TwAAKgpptz7VENh/g3MQphZYeTmvDFPJv7Pd/6zJOlEvFWvHupUXdgu+ZtRyLZUF7b12qFOvRRvcWOZAADAx0y693FbYf4NzEKY+dj9TFaJxbTXyyjayq3Qw3tiGu7br+7dUUnaNNQUXu/eHdVw3352ZAAAqEGm3ftUQyKZViqT9XoZKAGTET82m0wZNSSzsBVamAezsyGi/uNdmr67pCtjCY1MzSuRXGMAVTSieFuTju6NqaVp7QFUAAAg+Ey796mGvKTbyZQ6dmzzeikoEmHmYw8MaUdYUNgKPavHmwC0bt+qsz0dOqsOpTJZ3U6m9CCb05awrV3RetXX8U8OAADMu/epFv5ezMKd7ce2hM07cVfYCl0voNTXhXmyAAAA1mTivU818PdiFv61PrYrWi/TenkUtkIBAABKZeK9TzU4PccP7iLMfKy+LqyYgTNW2AoFAADlMPXex23/8ls/8HoJKAFhZoV4e5NxvdbZCgUAAOVy897HsFuqRxieaRbuhFc40hUzqte6pYdbxAAAAOVw894nl5d2bHvKlWu7ieGZZiHMrNC6fav2tTQaszsTi0boTgYAAMrm1r1PyLa0r6VRB7/0rDH3VQUMzzQLYWaVC72dRhR+hWxL8bYmr5cBAAAM58a9T9i2dKG307hTLwUMzzQHYWaVnQ0Rnevp2Pw3emw5l9fRvTGvlwEAAAznxr3P+Z4O7WyIGHfqpYCOseYgzKzh8J6YTh5sc+RazQ0R17ZuW5q2OnpdAABQm5y893n5YLte3PPJA1dTTr2sRsdYMxBm1nEi3qpXD3WqLmyXHEZCtqW6sK3XDnXqreNdrm3dAgAAOMWpe5+X4i2PvWbKqZfV6BhrBv6VNnB4T0zDffvVvTsqSZt+YRde794d1XDffr24J+bq1i0AAICTnLj3We+6Tu38VAMdY81h5fN586qyPDB9d0lXxhIamZpXIpnWyr80Sw87i8XbmnR0b2zN41+XR6Z18dpUxet4+WD7E088AAAAnFbpvc9aBsYTOjM0qWwu7+vGAM3RiEZPxr1eBopAmClDKpPV7WRKD7I5bQnb2hWtL6pFcrlfwCHbUti2dL6nY90nHgAAAG4p995nLXOLaZ0enND1mQWFbMt3oSZkWzrW1ayzBh6Nq0WEmSor5Qu48Pq+lkZd6O3kaBkAAAiMjXZ+vDbc9zyNlgxBmPGIG1u3AAAAJirs/Lz/w3v6Z1cnPFtHyLbUvTuq/uNdnq0BpSHM+ICTW7cAAACmmrxzT1/9xnc9e/+6sK3hvv2chjEId8w+UF8XVseObV4vAwAAwFO7ovWyJM+OnNEx1jy0ZgYAAIAv1NeFFfMoTKwe9gkzEGYAAADgG/H2ppKHdpZro2GfMANhBgAAAL5xpCvmervmYod9wv+omQEAAIBvtG7fqn0tjbpxK+lKqGmmY2yg0M0MAAAAvjK3mNaBS6PKZHOOXfNTIUt/fOLLan/2M45dE97jmBkAAAB8ZWdDROd6Ohy95r/8e79IkAkgwgwAAAB85/CemE4ebHPkWnQqCy6OmQEAAMC3BsYTOjM0qWwuX1INTci2FLYtne/pIMgEGGEGAAAAvja3mNbpwQldn1lQyLY2DDWF1/e1NOpCbydDMAOOMAMAAAAjTN9d0pWxhEam5pVIprXyJtaSFKNTWc0hzAAAAMA4qUxWt5MpPcjmtCVsa1e0XvV1TB2pNYQZAAAAAEYivgIAAFSAHQLAO3ylAQAAlOhR7cbNeSUW16jdaIgo3t6kI10xtW6ndgNwC8fMAAAAikRXLcBfCDMAAABFqHTeybmeDh0uc94JR9mAtRFmAAAANnF5ZFoXr01VfJ2TB9t0It5a1O/lKBuwOcIMAADABgbGEzp1dcKx6712qHPDifQcZQOKR5gBAABYx9xiWgcujSqTzTl2zbqwreG+/WsGDy+PsgEmsr1eAAAAgF+dHpxQtoRQUYxsLq/Tg0/u9FwemdapqxPKZHMlBRlJWs7llcnmdOrqhC6PTDu1VMD3CDMAAABrmL67pOszCyUHi80s5/K6PrOgmfmlR782MJ5wpCZHki5em9Lb4wlHrgX4HWEGAABgDVfGEgrZlivXDtmW3nr3YeCYW0zrzNCko9d/ZWhSc4tpR68J+BFhBgAAYA0jN+cd35UpWM7lNTI1L6m6R9mAoKFBOQAALmAuiNnuZ7JKuLyzkUim9VdzH+n6zILj1155lK2libbNCC6+qwIA4BDmggTHbDIlt9u95iX9wXdvbdp+uVyFo2xnezocvzbgF7RmBgCgQswFCZ73Eh+q95s3XH+fZz9Tpw9+nHHt+s3RiEZPxl27PuA1amYAAKjAwHhCBy6N6satpCRt+oS98PqNW0kduDSqAbpO+dKWcHVukdwMMtLDo2ypTNbV9wC8RJgBAKBMzAUJrl3RernTx6y68pJuJ1NeLwNwDTUzAICaV06xvtNzQT73dJ1eZHK7b9TXhRVriGg2AO2NH2RzXi8BcA1hBgBQkyop1ndrLkj3FxupofGReHuT+sdmXWvPXC1zHz4MZHTVQxDRAAAAUFOcKNY/9saYbtxKOnqTG7Itde+Oqv94l2PXRGWm7y7phdff8XoZjqKrHoKGMAMAqBkD4wmdGZpUNpcvKYiEbEth29K5ng79zdgzrt7gDvc9z1wQH3EjuPoBXfUQFIQZAEBNuDwy7UiNy3M7P6u/+uE91+aCHOtqZi6Ij8wtpnXg0qgyAa07WRnUD1OzBQPRzQwAEHhOFuu/N/eRa0/pl3N5jUzNu3JtlGdnQ0TnAhwu6aoH0xFmAACB5kaxvpuYC+I/h/fEdPJgm9fLcN3Fa1N6m7lHMAxhBgAQaKcHJ5Q1qN6BuSD+dCLeqlcPdaoubCtklzaBptTf76VXhiY1F4B21KgdhBkAQGBN313S9ZkF44q3mQviT4f3xDTct1/du6OSNg8phdd/6ee2ub42p2RzeZ0enPB6GUDRaDQOAAisK2OJTdsv+9GWMM8a/WpnQ0T9x7s+mVM0Na9Eco05RdGI4m1NOro3pqWfZtX7zRteLbkky7m8rs8saGZ+ia56MALdzAAAgbX/d0eMnOB+5G/H9A+7dzEDxBCpTFa3kyk9yObWHEw5eeeevvqN73q4wtLQVQ8mIcwAAALpfiarzrPflok/5JgBEiypTFa/aNhnsTka0ejJuNfLADbFPjYAIJBmkymjbh5XKhyLu3ErqQOXRjVAhymj1deFFTMskNJVD6YgzAAAAikIRfTMAAmOeHuTUV3N6KoHUxBmAACBFLQiemaAmO1IV8y4RhRBeCCA4AvWd3oAAD62K1ovc56DF4cZIOZ66lMhGbQxIyl4DwQQTHxKAQCBZGKdwmaYAWKu04MTRoVrSw8fCAB+R5gBAASWaXUKm1k5AwTmeDS81aBTZrFo5LH20oBfEWYAAIFlYp3CZkK2pbfepXbGJIXhraYI2ZbibU1eLwMoCmEGABBYrdu3al9Lo1E3kptZzuU1MjXv9TJQgpGb80aF6uVcXkf3xrxeBlAUwgwAINAu9HYqHKAwIzEDxCT3M1klDGraELIt7WtpVEvTVq+XAhSFMANjpDJZTd65p/cSH2ryzj1+kAMoys6GiM71dHi9DEcxA8Qcpg1vDduWLvR2er0MoGhUdsHXpu8u6cpYQiM355VYTD/2A8GSFGuIKN7epCNdMbVu5ykSgLUd3hPTX3+wpDdv3PZ6KY5hBogZTPt3Ot/ToZ0B6wKIYCPMwJfmFtM6PTih6zMLCtnWmmeN85JmF9PqH5vVm9+7rX0tjbrQ28k3YQBry0uWJeVNeky+AWaAmMGkf6eXD7brxT3UysAs5nyFoWYMjCd04NKobtxKStKmRZOF12/cSurApVENMCEbwBpGbs4HJsgwA8QcJgxv/VTI0muHOvVSvMXrpQAlY2cGvnJ5ZFoXr02V9d8u5/JazuV16uqEFu5ndCLe6vDqAJjKtCLszTADxByF4a2zPv78/fGJL6v92c94vQygLOzMwDcGxhNlB5nVLl6b0tvs0AD4mGlF2BthBoh5/Dy8tTkaIcjAaIQZ+MLcYlpnhiYdvea/+KNJ3fzgx45eE4CZTCvC3ggzQMzj1+GtBGMEAXvU8IXTgxPKOvyN/sFyTl/5vetqpuMZUPNMKsLeSMi21L07ygwQwxSGt964lfRVqCEYIwiC8d0dRpu+u6TrMwuufYMvdDx74fV3dOyNMc0VeW6ZuTZAcJhQhF0MZoCYy2/DWxmOiaBgZwaeuzKWWLf9slNWdzw719Ohw2u0n2SuDRBMJhRhF4MZIOYqDG89dXXC66VIIhgjOKx8PiiNKmGq/b874skNxsmDbY86nhUz16ag8DpzbQCznB2aVP/YrK+O+ZTi5YPttM4NgEq6djrptUOdzJRBIBBm4Kn7maw6z37bsy5Drx3qVF7SmaFJZT9u7VyskG0pbFvr7vIA8Jfpu0t64fV3vF5GSQrfZ873dHDjGSAD44myfu44NfSVYIwgIczAU5N37umr3/iuZ+/v1PG2lbs8APzr2BtjvivCXgs7wMFX7omA/+aLUf3en0yX/QCOYIygIczAU+8lPlTvN294vQxHsGUP+N/cYloHLo0q4+NWzc3RiOJtTTq6N0Zxdg14VKs5Na9Eco1azTU+DxyNBj5BmIGn/vQHd/WP/s33vV6GI+rCtob79vODAvC5gfGEb4qwV/r8tqc03Ldf9XX05qlVqUxWt5MpPcjmtCVsa1e0fsPPQzlBCAgawgw89Vt/8K5u3Ep6vQxHFOY/9B/v8nopADbhlyLsgpBt6VhXs872dHi9FBiq1CAEBAWfcnhm+u5SYIKM9LD98/WZBc3ML/EEDPC5E/FWNT5dV1YRthsYXohK1deF1bFjm9fLAKqOoZnwTGG+TJCEbEtvvZvwehkAinB4T0zDffvVvTsqSZt+P3Lr+xXDCwGgfIQZeGbk5rznT0OdtpzLa2Rq3utlACjSzoaI+o936Ttff17HuprVHI1odWSx9LAo/1hXs/r/0d9WXdjZH50MLwSA8lEzA094PV/GTZak989+hbPKgKE2qz1wuoEAnRABoHzcbcETs8lUIIOMJOUl3U6mOLsMGGqz2oPDe2JauJ9xpIHAywfbCTIAUAHCDDzxwMczHpwQ9D8fUOsqaSDA8EIAcA41M/DEFofPnPtN0P98AMpvINC9O6rhvv0EGQBwADsz8MSuaL0sKZBHzSw9/PMBCL5CAwGGFwKAN2gAAM/s/90RzS6mvV6G45qjEY2ejHu9DAAeYXghAFQP313hmXh7k/rHZgPVnjlkW4q3NXm9DAAeYnghAFQPB/vhmSNdsUAFGYkp3gAAANXEzgw807p9q/a1NOrGrWQgQk3IttS9O8p5eKyJo0cAADiPmhl4am4xrQOXRpUJQCvjurCt4b792tkQ8XopNcXPIeFRUfjNeSUW1ygKb4go3t6kI10xtW4nBAMAUCrCDDzn9DRtrzDFu3qcDglOB6K5xbROD07o+syCQra14c5j4fV9LY260NtJGAYAoASEGXju8si0I5O0vfTywXa9FG/xehmB52RIcGvXZGA8UdEgxXM9HTpMKAYAoCiEGXjKD7sylvXw5tWyNr45Xo0p3tXlVEhwc9fEqWB+8mCbTsRbK74OAABBR5iBZ/xSL1MXtvVvfvtv6/LIDMeCfMqpkHDwS9s1OvVfXdk1cTqYc2wRAIDNEWbgmWNvjPmik9nKm0amePuPH3bvVlpr18SNYF5qQwk/N0IA3MLnHgBhBp6YvrukF15/x+tlbFjrwg9J7/ll92611bsmbgTzQqvv/uNd6/4euqWhFvG5B7ASYQaeODs0qf6xWU92Zah1MYdfdu9WW7lr4nYwH+57/oldQLqloRbxuQewFtvrBaA2jdycr/oNasi2JEndu6Ma7ttPkPG56btLuj6z4LsgI0nZXF6nBx8efbsylnj02XJayLb01ruJx35tYDyhA5dGdeNWUpI2/fspvH7jVlIHLo1qYDyx4e8H/IjPPYD1cGYGVXc/k1ViMV2196PWxUyFkODHMLOcy+v6zIJm5pdcDebLubxGpuZ1Vh2SKmuEsPxx04NTVye0cD9DtzQYg889gI0QZlB1s8mUqnF7+n/8D7+s9me3UutiKC9270oRsi394Z/ddj2YJ5JppTJZ/b9/dcexeUwXr03pc0/XPbY7SY0Y/GhgPOHq5x6A+fhJhap7UKVi7p9vrFfHjm1VeS84q9q7d+VYzuX1pzfnXQ/meUnv3krqzNCko9d9ZWhSn9/2af3pX89TSA1fmltMu/K57/5iIzU0QIBQM4Oq2xKuzseuWu8D51Vr965Sdz76SVXe59X/76/1s2VnHwJksjn9gz/8c/WPzWp2VZCRHoao2cW0+sdm9cLr7+jYG2Oa83nARLCcHpxQ1uHd2ZX1bgCCgbs9VN2uaL3cKZf+hPXx+8BM1dq9q5Rluf1Jfmh6/r7cOnFHITX8yK0GICvr3QAEA2EGVVdfF1bM5S3+WDTCeX+DsavmP8u5vDLZnE5dndDlkWmvl4OAq3aXQADm4o4Bnoi3N7n6gyre1uTKtVEd1di9c8rntz3l9RKq7uK1Kb3NDg1cVI0ugQCCgTADTxzpirn6g+roXrrVmKwau3dOsCT9XReDuZ+9MjRJDQ0qlspkNXnnnt5LfKjJO/eUymSr0gCk0CUQgPk4hwNPtG7fqn0tjY5Pdw/Zlrp3R5klEwDx9ib1j836uj1zLBrRP+zepSt/Xnu7FIVC6v7jXV4vBYaZvrukK2OJdbvoPbvtqap0CbydTNHxEggAdmbgmQu9nQo7/EQ7bFu60Nvp6DXhDTd375xQOM5YCOa1tjtDITVKNbeY1rE3xvTC6+9s2EXvR/d+WpX1mNJoBMDGCDPwzM6GiM71dDh6zfM9HcwPCAi/h4SVxxndCOYmoJAaxRoYT+jApVHduJWUtHkXvWqg0QgQDHwlw1OH98R08mCbI9d6+WA7k50Dxq8hIWRb2tfS+Og4oxvB3AQUUqMYl0emderqhDLZnC9CjET7fiBICDPw3Il4q1491Km6sF3yU/iQbakubOu1Q516Kd7i0grhFb+GhLWOMzoZzE1CITU2MjCe0MVrU14v4wm07weCgzADXzi8J6bhvv3q3h2VpE1DTeH17t1RDfftZ0cmwA7viWnbpz/l9TIes95xxkqCuakKhdTAanOLaZ0ZmvR6GU+gfT8QLDyWgG/sbIio/3jXJ51upuaVSD7Z6SYWjSje1qSje2N0LasB9zNZ/fgnP/N6GY9sdpzx8J6Y/s4XG3V6cELXZxYUsi3fHK1xC4XUWMvpwQllffjZp30/ECyEGfhO6/atOtvTobPqUCqT1e1kSg+yOW0J29oVredoQI2ZTaZcb9O6mZBtKWxbOt/TUdQuYDHBPEgopMZq03eXdH1mwetlrGlP8zOPPQjj5wxgNr5a4Wv1dWHmANQ4L5/6F3ZVundHdaG3s+ROeWsF8/d/eE//7OqESyuuPgqpsZYrYwlf70puNusm1hBRvL1JR7piat3OCQDAzwgzAHzNi6f+bhxnDGowp5Aaaxm5Oe/bIDM++6FeeP2ddcNWXtLsYlr9Y7N683u3ta+lsayHGQCqg59AAHxtV7ReluTqES1L0v/zP/03CtmW68dMqvHnqRYKqbGW+5msEotpr5exqc3CVuH1G7eSOnBpVOd6OnSYZjOA7xBmAPhafV1YsYaIZl28OYpFI/pbuxpcu/5K1fjzVAuF1FiLH+rcnLScy2s5l9epqxNauJ/RiXir10sCsAJVmwB8L97e5FqrYy92F9z881TL6sGhQEGQu9tdvDalt8cTXi8DwAqEGQC+d6Qr5tr5ey92F9z881TLWoNDASn43e1eGZrUXAB2VoGgCPZ3HACB0Lp9q/a1NDq+m+HV7oJbf55qWm9wKFCoCwuqbC6v04PB6UgImI4wA8AIF3o7FXb45t/L3QU3/jzVstngUNS2Ql1YUC3n8ro+s6CZ+SWvlwJAhBkAhtjZENG5ng5Hr+nl7oIbfx43hWxLdWFbrx3q1EvxFq+XA58LQl3YRkK2pbfepXYG8APCDABjHN4T08mDbY5cy43dhVQmq8k79/Re4kNN3rmnVCa74e938s/jlsL9aPfuqIb79rMjg6IEoS5sI8u5vEam5r1eBgDRmhmAYU7EW9X4dJ3ODE0q+3HL1GKFbEth29L5ng7HbsornSReyZ/HbWHb0tGuZscGh6J2FOrCbtxK+uoz7aREMq1UJsvQWMBjVj6fD+Z3GQCBNreY1unBCV2fWVh3kndB4XUnJ3k7/f43P/ixTl2d0HtzH8mWtFFz28L1dmx7Snd//FMtu/BdPGRbOtbVrLMGHYWDv8wtpnXg0qgyAW7V/K3f+bI6dmzzehlATSPMADDao52RqXklkmvsjEQjirc1Obq7MDCeqGhnqDBJfKNdnXw+L8t6vOZg9Z8nn5deeP0dR/5Maxnue54dGVRkYDyhU1eD2/lr8Gvdei72jNfLAGoaYQZAYKQyWd1OpvQgm9OWsK1d0XrHj4BcHpnWxWtTFV+nuSGi2cX0prs6tiXl8tJzOz+rVw91qv3Zzzz2+rE3xhw/yhOyLXXvjqr/eJdj10Ttcuprxo/YmQG8R5gBgCJ5+ZR59a5OgRtHeerCtob79jNHBo4pdzfTzyxJ75/9CjUzgMfoZgYARZhbTOvM0KRn77+cyyuTzenU1QldHpl+9OtBa1mNYDq8J6bhvv3q3h2VpEC0bY5FIwQZwAcIMwBQhNODE8r65InyxWtTenv8kxkXfm9ZDUgPg3f/8S595+vP61hXs9GDNUO2pXhbk9fLACCOmQHApqbvLrlaaF+OurCtoZf+jrK5/KMaoe/f/lAX/t0PfNGyGtjM5J17+uo3vuv1MspGgwzAH9gfBYBNXBlLbFqoX22ZbE5f+b3rj/2aJenz256SJN2599OiW0Z374461rIaKNYDQ1s2FxpkEGQAfyDMAMAmRm7O+yrIrCevT0KMJG3fWifLsnTno59UpWU1UIotYTNPuodtSxd6O71eBoCPEWYAYAP3M1klFtNeL6MkheB1dymjsG3pbM+X9LeaG1xtWQ2Uale0XpYk/z8meBwNMgB/MfOxCABUyWwyZdzNVkGhA9qZof+okZvzei72jDp2bCPIwBfq68LGNQGgQQbgP4QZANiAqef6V1vdAQ3wg3h7k+/bNIdsS3VhW68d6tRL8RavlwNgFcIMAGzA1HP9a3llaFJzhh2ZQ7Ad6Yr5vh6te3dUw3372ZEBfCo4P6UBwAWFc/1BkM3ldXpwwutlAI+0bt+qfS2Nvtyd2fpUWMN9z6v/eBc1MoCPEWYAYAMmnutfz3Iur+szC5qZX/J6KcAjF3o7FfZZmAnZln7zuS/Q6Q8wAGEGADZhwrn+YoVsS2+9S+0M/GNnQ0Tnejq8XsZjlnN5Hd3LsTLABIQZANiECef6i7Wcy2tkat7rZQCPObwnppMH27xehqSHgX9fSyO7MoAhCDMAsAk/n+svRyKZViqT9XoZwGNOxFv16qFO1YVtT7/WGIoJmIUwAwBF8OO5/nLlJd1OprxeBvCEw3tiGu7br+7dUUnaNNS4EXoYigmYhTADAEXw47n+SgRlfg6CZ2dDRP3Hu/Sdrz+vY13Nao5GnugoaElqjkZ0rKtZw33PO3ZEjaGYgHmsfD4fjIPgAFAFl0emdfHalNfLqNi3fufL6tixzetlAEVJZbK6nUzpQTanLWFbu6L1qq8LP/Z7BsYTOjM0qWwuX1KNW8i2FLYtne/pIMgABiLMAECJCjdNP1vOycS+AJak989+5YmbQcB0c4tpnR6c0PWZBYVsa8NQU3h9X0ujLvR2crQMMBRhBgDKMHYrqcN/8K5M/AbaHI1o9GTc62UArpm+u6QrYwmNTM0rkUw/9nVqSYpFI4q3Neno3hhdywDDEWYAoAzH3hjTjVtJ41o2h2xLx7qadTZA9T/ARoo5ogbAXHw1A0CJpu8u6frMgtfLKAvDAFFr6uvC1IcBAUY3MwAo0ZWxhJEzZxgGCAAIGsIMAJRo5Oa8ccfLJIYBAgCChzADACW4n8kqsZj2ehllYRggACBoCDMAUILZZMrIDmYMAwQABBENAACgBA+yOa+XUDSGAQIAgo4wAwAl2BL2/4Z2YRhg9+4owwABAIFGmAGAEuyK1suSfHnUjGGAAIBaQ5gBgBLU14UVa4ho1mdNAH7/8K/o1//GdoYBAgBqiv/PSwCAz8Tbm3w1Z8aSCDIAgJpEmAGAEh3pivlqzkwsGiHIAABqEmEGAErUun2r9rU0+mJ3JmRbirc1eb0MAAA8QZgBgDJc6O1U2AdhZjmX19G9tF0GANQmwgwAlGFnQ0Tnejo8XUPItrSvpZGuZQCAmkWYAYAyHd4T08mDbZ69f9i2dKG307P3BwDAa4QZAKjAiXirXj3UqbqwXfUamvM9HQzEBADUNCufz/unJQ8AGGpuMa3TgxO6PrOgkG253u3s5YPteine4up7AADgd4QZAHDQ9N0lXRlLaGRqXolkWiu/wVp62Eb557Z9Wt+f/VDL+XxJoSdkWwrbls73dOjFPRT9AwBAmAEAl6QyWd1OpvQgm9OWsK1d0fpH82BK2ckpvL6vpVEXejs5WgYAwMcIMwDgoWJ2cuJtTTq6N0bXMgAAViHMAIBPbLSTAwAAnkSYAQAAAGAkWjMDAAAAMBJhBgAAAICRCDMAAAAAjESYAQAAAGAkwgwAAAAAIxFmAAAAABiJMAMAAADASIQZAAAAAEYizAAAAAAwEmEGAAAAgJEIMwAAAACMRJgBAAAAYCTCDAAAAAAjEWYAAAAAGIkwAwAAAMBIhBkAAAAARiLMAAAAADASYQYAAACAkQgzAAAAAIxEmAEAAABgJMIMAAAAACMRZgAAAAAYiTADAAAAwEiEGQAAAABGIswAAAAAMBJhBgAAAICRCDMAAAAAjESYAQAAAGAkwgwAAAAAIxFmAAAAABiJMAMAAADASIQZAAAAAEYizAAAAAAwEmEGAAAAgJEIMwAAAACMRJgBAAAAYCTCDAAAAAAjEWYAAAAAGIkwAwAAAMBIhBkAAAAARiLMAAAAADASYQYAAACAkQgzAAAAAIxEmAEAAABgJMIMAAAAACMRZgAAAAAYiTADAAAAwEj/P9KUb+FgAp9JAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 800x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Networkx Default Layout\n",
    "# Create a network graph using networkx\n",
    "plt.figure(figsize=(8,6))\n",
    "nx.draw(G)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d0be0930",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# Debug: Print the graph information\n",
    "# print(f\"Nodes: {G.nodes()}\")\n",
    "# print(f\"Edges: {G.edges()}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4b752e2c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create a Pyvis network and visualize the NetworkX graph\n",
    "# Pyvis Default Layout\n",
    "# nt = net.Network(width='100%', height='100%', directed=True)\n",
    "# nt.from_nx(G)\n",
    "# nt.show('01_pyvis_default.html')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c57a34f4",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Debug: Check the Pyvis network\n",
    "# print(f\"Pyvis Nodes: {nt.nodes}\")\n",
    "# print(f\"Pyvis Edges: {nt.edges}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "826fac6a",
   "metadata": {},
   "source": [
    "<h3>? can't get plot to render below?</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "3ea9a485",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "pyvis.network.Network"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "net"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "28538121",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Help on method show in module pyvis.network:\n",
      "\n",
      "show(name, local=True, notebook=True) method of pyvis.network.Network instance\n",
      "    Writes a static HTML file and saves it locally before opening.\n",
      "    \n",
      "    :param: name: the name of the html file to save as\n",
      "    :type name: str\n",
      "\n"
     ]
    }
   ],
   "source": [
    "help(nt.show)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "bdc5fdbe",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "01_pyvis_default.html\n"
     ]
    }
   ],
   "source": [
    "# Is not working, check to see if file was created and if browser and cloud are compatible\n",
    "# see if versions of pyvis and networkX are compatible\n",
    "# ERROR AttributeError: module 'pyvis.network' has no attribute 'network'(pyvis object net.network is not\n",
    "# being created)\n",
    "\n",
    "nt = net(width='100%', height='100%', directed=True)\n",
    "nt.from_nx(G)\n",
    "nt.show('01_pyvis_default.html', notebook=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9834e830",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Debug\n",
    "# Ensure that the Pyvis network object nt is correctly initialized \n",
    "# and that the from_nx method is successfully converting the NetworkX graph.\n",
    "from pyvis.network import Network\n",
    "\n",
    "nt = Network(width='100%', height='100%', directed=True)\n",
    "nt.from_nx(G)\n",
    "\n",
    "# Check the Pyvis network\n",
    "print(nt.nodes)\n",
    "print(nt.edges)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "530fe1dc",
   "metadata": {},
   "outputs": [],
   "source": [
    "pip install --upgrade networkx"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ee107e51",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "pip install --upgrade pyvis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5820b664",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Debug\n",
    "# Create a Pyvis network and visualize the NetworkX graph\n",
    "nt = Network(width='100%', height='100%', directed=True)\n",
    "nt.from_nx(G)\n",
    "\n",
    "# Debug: Check the Pyvis network\n",
    "print(f\"Pyvis Nodes: {nt.nodes}\")\n",
    "print(f\"Pyvis Edges: {nt.edges}\")\n",
    "\n",
    "# Show the network\n",
    "nt.show('1_pyvis_default.html')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3dfbf05c",
   "metadata": {},
   "source": [
    "<h3>Create function for drawing network graph</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "id": "fbac68ed",
   "metadata": {},
   "outputs": [],
   "source": [
    "def show_graph(graph, filename):\n",
    "    nt = net(width='100%', height='100%', directed=True)\n",
    "    nt.from_nx(graph)\n",
    "    nt.show(filename, notebook=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "id": "c7ffff88",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "pyvis.network.Network"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "net # This is a class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "92756437",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<class 'pyvis.network.Network'> |N|=0 |E|=0"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "net() # Calling a class calls the constructor function, which returns an instance of Network"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3800ace1",
   "metadata": {},
   "source": [
    "<h3>Set Node Attributes</h3>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "10564caf",
   "metadata": {},
   "source": [
    "<h3>Set the node size according to the degree (number of connections) of node</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "db2730ec",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('1', 1), ('884', 1), ('10', 2), ('8317', 1), ('8520', 1)]"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# List source, \n",
    "# ?? follow on graph what these mean('source1', has 1 connection), ('targetamt884', 'is connected with source1'), ('source10', 'has 2 connections')\n",
    "list(G.degree)[0:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "d83afba0",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creates a dictionary of key = node id and value = node degree (how many connections that node has)\n",
    "# Nodes with a bigger degree will appear bigger in the graph\n",
    "degree_dict_G = dict(G.degree)\n",
    "nx.set_node_attributes(G, degree_dict_G, 'value')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "id": "5efdde4c",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('1', {'size': 10, 'value': 1}),\n",
       " ('884', {'size': 10, 'value': 1}),\n",
       " ('10', {'size': 10, 'value': 2}),\n",
       " ('8317', {'size': 10, 'value': 1}),\n",
       " ('8520', {'size': 10, 'value': 1})]"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# List graph dictionary of nodes with attributes size and value being number of connections (same as size but pyvis \n",
    "# recognizes it different)\n",
    "# size = number of connections to that node\n",
    "# Below shows that the new attribute value has been added\n",
    "list(G.nodes(data=True))[0:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7eabe46c",
   "metadata": {},
   "source": [
    "<h3>Graph won't render, pyvis net.network object never created</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "id": "c0fb2f74",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "02_node_size.html\n"
     ]
    }
   ],
   "source": [
    "show_graph(G, '02_node_size.html')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a6371d17",
   "metadata": {},
   "source": [
    "<h3> ?? graph won't render</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "3f154271",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "ename": "AttributeError",
     "evalue": "'Network' object has no attribute 'show_graph'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[29], line 2\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[38;5;66;03m# Show graph to that nodes with higher degrees are larger\u001b[39;00m\n\u001b[0;32m----> 2\u001b[0m nt\u001b[38;5;241m.\u001b[39mshow_graph()\n",
      "\u001b[0;31mAttributeError\u001b[0m: 'Network' object has no attribute 'show_graph'"
     ]
    }
   ],
   "source": [
    "# Show graph to that nodes with higher degrees are larger\n",
    "# Besides using degree as the centrality measurement, \n",
    "# you can also use other measurements under graph theory (link) such as eigenvector, closeness, betweenness etc\n",
    "# For any above, you have to obtain the measurement for each node in a dictionary form, then pass the dictionary \n",
    "# to the node attribute with attribute name of value.\n",
    "nt.show_graph()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5fe313e0",
   "metadata": {},
   "source": [
    "<h3>Display Node Information on Hover</h3>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ae653cc5",
   "metadata": {},
   "source": [
    "<h3> graph will not render???</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "265e2c85",
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'show_graph' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[30], line 6\u001b[0m\n\u001b[1;32m      4\u001b[0m     title_dict_G[node] \u001b[38;5;241m=\u001b[39m \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mid: \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mnode\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m'\u001b[39m\n\u001b[1;32m      5\u001b[0m nx\u001b[38;5;241m.\u001b[39mset_node_attributes(G, title_dict_G, \u001b[38;5;124m'\u001b[39m\u001b[38;5;124mtitle\u001b[39m\u001b[38;5;124m'\u001b[39m)\n\u001b[0;32m----> 6\u001b[0m show_graph(G, \u001b[38;5;124m'\u001b[39m\u001b[38;5;124m03_node_title_1.html\u001b[39m\u001b[38;5;124m'\u001b[39m)\n",
      "\u001b[0;31mNameError\u001b[0m: name 'show_graph' is not defined"
     ]
    }
   ],
   "source": [
    "# Use title attribute to show info of node when you hover over it\n",
    "# if you want to add more info in addition to account number, pass more information to the title attribute\n",
    "title_dict_G = {}\n",
    "for node in list(G.nodes):\n",
    "    title_dict_G[node] = f'id: {node}'\n",
    "nx.set_node_attributes(G, title_dict_G, 'title')\n",
    "show_graph(G, '03_node_title_1.html')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0469c759",
   "metadata": {},
   "source": [
    "<h3>Retrieve neighbors for each node</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "03af9543",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['3674', '6042', '6735', '7028', '8634']"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# List neighbors of node 2121\n",
    "list(G.neighbors('2121'))[0:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "b5c2f1b0",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create a dictionary of neighbor list for each node\n",
    "neighbor_dict_G = {}\n",
    "for node in list(G.nodes):\n",
    "    neighbor_dict_G[node] = ','.join(G.neighbors(node))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1f3e393e",
   "metadata": {},
   "source": [
    "<h3> ? graph won't show, error: NameError: name 'show_graph' is not defined</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "1820cc22",
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'show_graph' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[36], line 8\u001b[0m\n\u001b[1;32m      5\u001b[0m     title_dict_G[node] \u001b[38;5;241m=\u001b[39m \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mid: \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mnode\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;130;01m\\n\u001b[39;00m\u001b[38;5;124mdegree: \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mdegree_dict_G[node]\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;130;01m\\n\u001b[39;00m\u001b[38;5;124mneighbor: \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mneighbor_dict_G[node]\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m'\u001b[39m\n\u001b[1;32m      6\u001b[0m nx\u001b[38;5;241m.\u001b[39mset_node_attributes(G, title_dict_G, \u001b[38;5;124m'\u001b[39m\u001b[38;5;124mtitle\u001b[39m\u001b[38;5;124m'\u001b[39m)\n\u001b[0;32m----> 8\u001b[0m show_graph(G, \u001b[38;5;124m'\u001b[39m\u001b[38;5;124m04_node_title_2.html\u001b[39m\u001b[38;5;124m'\u001b[39m)\n",
      "\u001b[0;31mNameError\u001b[0m: name 'show_graph' is not defined"
     ]
    }
   ],
   "source": [
    "# Incorporate the dictionary to the attribute title to display: \n",
    "# Display ID, Degree and Neighbors in Graph (Hover on Node)\n",
    "title_dict_G = {}\n",
    "for node in list(G.nodes):\n",
    "    title_dict_G[node] = f'id: {node}\\ndegree: {degree_dict_G[node]}\\nneighbor: {neighbor_dict_G[node]}'\n",
    "nx.set_node_attributes(G, title_dict_G, 'title')\n",
    "\n",
    "show_graph(G, '04_node_title_2.html')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a7139a6b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# NOTE: you can update the node attribute dictionary under G which the attribute name (dictionary key) \n",
    "# needs to match with pyvis’s parameter name (label, title, size, value and etc.)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2b0644a2",
   "metadata": {},
   "source": [
    "<h3>Setting Edge Attributes</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c470c697",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set the edge attributes dictionary under G\n",
    "# you can adjust the edge width according to the aggregated transaction amount between accounts, \n",
    "# and show detailed edge information on hover such as payer, payee, aggregated transaction amount \n",
    "# and aggregated transaction count"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "e20cfa03",
   "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>source</th>\n",
       "      <th>target</th>\n",
       "      <th>agg_txn_count</th>\n",
       "      <th>agg_txn_amt</th>\n",
       "      <th>value</th>\n",
       "      <th>title</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>884</td>\n",
       "      <td>24</td>\n",
       "      <td>4,219</td>\n",
       "      <td>4,219</td>\n",
       "      <td>from: 1\\nto: 884\\nagg_count: 24\\nagg_amount: 4...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10</td>\n",
       "      <td>8317</td>\n",
       "      <td>21</td>\n",
       "      <td>2,003</td>\n",
       "      <td>2,003</td>\n",
       "      <td>from: 10\\nto: 8317\\nagg_count: 21\\nagg_amount:...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10</td>\n",
       "      <td>8520</td>\n",
       "      <td>21</td>\n",
       "      <td>2,003</td>\n",
       "      <td>2,003</td>\n",
       "      <td>from: 10\\nto: 8520\\nagg_count: 21\\nagg_amount:...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100</td>\n",
       "      <td>3472</td>\n",
       "      <td>18</td>\n",
       "      <td>978</td>\n",
       "      <td>978</td>\n",
       "      <td>from: 100\\nto: 3472\\nagg_count: 18\\nagg_amount...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100</td>\n",
       "      <td>9954</td>\n",
       "      <td>18</td>\n",
       "      <td>978</td>\n",
       "      <td>978</td>\n",
       "      <td>from: 100\\nto: 9954\\nagg_count: 18\\nagg_amount...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  source target  agg_txn_count  agg_txn_amt  value  \\\n",
       "0      1    884             24        4,219  4,219   \n",
       "1     10   8317             21        2,003  2,003   \n",
       "2     10   8520             21        2,003  2,003   \n",
       "3    100   3472             18          978    978   \n",
       "4    100   9954             18          978    978   \n",
       "\n",
       "                                               title  \n",
       "0  from: 1\\nto: 884\\nagg_count: 24\\nagg_amount: 4...  \n",
       "1  from: 10\\nto: 8317\\nagg_count: 21\\nagg_amount:...  \n",
       "2  from: 10\\nto: 8520\\nagg_count: 21\\nagg_amount:...  \n",
       "3  from: 100\\nto: 3472\\nagg_count: 18\\nagg_amount...  \n",
       "4  from: 100\\nto: 9954\\nagg_count: 18\\nagg_amount...  "
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Obtain the edge attribute dictionary by updating the source dataframe edge_small\n",
    "edge_small['value'] = edge_small['agg_txn_amt']\n",
    "edge_small['title'] = edge_small.apply(lambda df: f'from: {df.source}\\nto: {df.target}\\nagg_count: {df.agg_txn_count:,.0f}\\nagg_amount: {df.agg_txn_amt:,.0f}', axis=1)\n",
    "edge_small.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e0c0f92b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Value controls the size (width) of edge and title controls the information to be shown on hover"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4570e2d7",
   "metadata": {},
   "source": [
    "<h3>Setting edge size and information to display on hover</h3>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "27608058",
   "metadata": {},
   "source": [
    "<h3> ?? graph won't render</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "c9a7b5c6",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'show_graph' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[39], line 8\u001b[0m\n\u001b[1;32m      5\u001b[0m     edge_dict_G[\u001b[38;5;28mtuple\u001b[39m([row[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124msource\u001b[39m\u001b[38;5;124m'\u001b[39m], row[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mtarget\u001b[39m\u001b[38;5;124m'\u001b[39m]])] \u001b[38;5;241m=\u001b[39m {\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mvalue\u001b[39m\u001b[38;5;124m'\u001b[39m:row[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mvalue\u001b[39m\u001b[38;5;124m'\u001b[39m], \u001b[38;5;124m'\u001b[39m\u001b[38;5;124mtitle\u001b[39m\u001b[38;5;124m'\u001b[39m:row[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mtitle\u001b[39m\u001b[38;5;124m'\u001b[39m]}\n\u001b[1;32m      6\u001b[0m nx\u001b[38;5;241m.\u001b[39mset_edge_attributes(G, edge_dict_G)\n\u001b[0;32m----> 8\u001b[0m show_graph(G, \u001b[38;5;124m'\u001b[39m\u001b[38;5;124m05_edge_title_and_value.html\u001b[39m\u001b[38;5;124m'\u001b[39m)\n",
      "\u001b[0;31mNameError\u001b[0m: name 'show_graph' is not defined"
     ]
    }
   ],
   "source": [
    "# create an edge attribute dictionary for value and title and pass it to G so pyvis will update the edge \n",
    "# layout accordingly\n",
    "# graph is now able to show more information with differentiated edge size and hover information\n",
    "edge_dict_G = {}\n",
    "for index, row in edge_small.iterrows():\n",
    "    edge_dict_G[tuple([row['source'], row['target']])] = {'value':row['value'], 'title':row['title']}\n",
    "nx.set_edge_attributes(G, edge_dict_G)\n",
    "\n",
    "show_graph(G, '05_edge_title_and_value.html')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "09f3e046",
   "metadata": {},
   "source": [
    "<h3>Community Detection with Large Dataset</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9deebc50",
   "metadata": {},
   "outputs": [],
   "source": [
    "# explore the network structure of a larger dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "f246708d",
   "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>source</th>\n",
       "      <th>target</th>\n",
       "      <th>agg_edge_large_count</th>\n",
       "      <th>agg_edge_large_amt</th>\n",
       "      <th>value</th>\n",
       "      <th>title</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>63958</th>\n",
       "      <td>9703</td>\n",
       "      <td>6059</td>\n",
       "      <td>20</td>\n",
       "      <td>2,506</td>\n",
       "      <td>2,506</td>\n",
       "      <td>from: 9703\\nto: 6059\\nagg_count: 20\\nagg_amoun...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38025</th>\n",
       "      <td>7369</td>\n",
       "      <td>9795</td>\n",
       "      <td>2</td>\n",
       "      <td>1,338,379</td>\n",
       "      <td>1,338,379</td>\n",
       "      <td>from: 7369\\nto: 9795\\nagg_count: 2\\nagg_amount...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64700</th>\n",
       "      <td>9722</td>\n",
       "      <td>9962</td>\n",
       "      <td>20</td>\n",
       "      <td>2,973</td>\n",
       "      <td>2,973</td>\n",
       "      <td>from: 9722\\nto: 9962\\nagg_count: 20\\nagg_amoun...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41653</th>\n",
       "      <td>7881</td>\n",
       "      <td>9751</td>\n",
       "      <td>20</td>\n",
       "      <td>5,923</td>\n",
       "      <td>5,923</td>\n",
       "      <td>from: 7881\\nto: 9751\\nagg_count: 20\\nagg_amoun...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23451</th>\n",
       "      <td>5376</td>\n",
       "      <td>4205</td>\n",
       "      <td>2</td>\n",
       "      <td>1,755,996</td>\n",
       "      <td>1,755,996</td>\n",
       "      <td>from: 5376\\nto: 4205\\nagg_count: 2\\nagg_amount...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61987</th>\n",
       "      <td>9647</td>\n",
       "      <td>9206</td>\n",
       "      <td>20</td>\n",
       "      <td>3,453</td>\n",
       "      <td>3,453</td>\n",
       "      <td>from: 9647\\nto: 9206\\nagg_count: 20\\nagg_amoun...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5068</th>\n",
       "      <td>2132</td>\n",
       "      <td>8921</td>\n",
       "      <td>23</td>\n",
       "      <td>701</td>\n",
       "      <td>701</td>\n",
       "      <td>from: 2132\\nto: 8921\\nagg_count: 23\\nagg_amoun...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21665</th>\n",
       "      <td>5185</td>\n",
       "      <td>5233</td>\n",
       "      <td>21</td>\n",
       "      <td>448</td>\n",
       "      <td>448</td>\n",
       "      <td>from: 5185\\nto: 5233\\nagg_count: 21\\nagg_amoun...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68701</th>\n",
       "      <td>998</td>\n",
       "      <td>275</td>\n",
       "      <td>15</td>\n",
       "      <td>5,846</td>\n",
       "      <td>5,846</td>\n",
       "      <td>from: 998\\nto: 275\\nagg_count: 15\\nagg_amount:...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43896</th>\n",
       "      <td>8141</td>\n",
       "      <td>7693</td>\n",
       "      <td>23</td>\n",
       "      <td>240</td>\n",
       "      <td>240</td>\n",
       "      <td>from: 8141\\nto: 7693\\nagg_count: 23\\nagg_amoun...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2000 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      source target  agg_edge_large_count  agg_edge_large_amt     value  \\\n",
       "63958   9703   6059                    20               2,506     2,506   \n",
       "38025   7369   9795                     2           1,338,379 1,338,379   \n",
       "64700   9722   9962                    20               2,973     2,973   \n",
       "41653   7881   9751                    20               5,923     5,923   \n",
       "23451   5376   4205                     2           1,755,996 1,755,996   \n",
       "...      ...    ...                   ...                 ...       ...   \n",
       "61987   9647   9206                    20               3,453     3,453   \n",
       "5068    2132   8921                    23                 701       701   \n",
       "21665   5185   5233                    21                 448       448   \n",
       "68701    998    275                    15               5,846     5,846   \n",
       "43896   8141   7693                    23                 240       240   \n",
       "\n",
       "                                                   title  \n",
       "63958  from: 9703\\nto: 6059\\nagg_count: 20\\nagg_amoun...  \n",
       "38025  from: 7369\\nto: 9795\\nagg_count: 2\\nagg_amount...  \n",
       "64700  from: 9722\\nto: 9962\\nagg_count: 20\\nagg_amoun...  \n",
       "41653  from: 7881\\nto: 9751\\nagg_count: 20\\nagg_amoun...  \n",
       "23451  from: 5376\\nto: 4205\\nagg_count: 2\\nagg_amount...  \n",
       "...                                                  ...  \n",
       "61987  from: 9647\\nto: 9206\\nagg_count: 20\\nagg_amoun...  \n",
       "5068   from: 2132\\nto: 8921\\nagg_count: 23\\nagg_amoun...  \n",
       "21665  from: 5185\\nto: 5233\\nagg_count: 21\\nagg_amoun...  \n",
       "68701  from: 998\\nto: 275\\nagg_count: 15\\nagg_amount:...  \n",
       "43896  from: 8141\\nto: 7693\\nagg_count: 23\\nagg_amoun...  \n",
       "\n",
       "[2000 rows x 6 columns]"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "txn_full = pd.read_csv('transactions.csv', usecols=['SENDER_ACCOUNT_ID','RECEIVER_ACCOUNT_ID','TX_AMOUNT'], dtype='object')\n",
    "txn_full['TX_AMOUNT'] = txn_full['TX_AMOUNT'].astype(float)\n",
    "\n",
    "edge_large = txn_full.groupby(['SENDER_ACCOUNT_ID', 'RECEIVER_ACCOUNT_ID'],as_index=False).agg({'TX_AMOUNT':['count','sum']}).sample(2000, random_state=721)\n",
    "edge_large.columns = ['source','target','agg_edge_large_count','agg_edge_large_amt']\n",
    "edge_large['value'] = edge_large['agg_edge_large_amt']\n",
    "edge_large['title'] = edge_large.apply(lambda df: f'from: {df.source}\\nto: {df.target}\\nagg_count: {df.agg_edge_large_count:,.0f}\\nagg_amount: {df.agg_edge_large_amt:,.0f}', axis=1)\n",
    "\n",
    "edge_large"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "664f2eb4",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import the dataframe to networkx\n",
    "# #Note that we add the argument \"edge_attr\" in \"from_pandas_edgelist\" so that we don't have to add edge attribute seperately.\n",
    "H = nx.from_pandas_edgelist(edge_large, source='source', target='target', edge_attr=['title','value'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "6428beca",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('9703',\n",
       "  '6059',\n",
       "  {'title': 'from: 9703\\nto: 6059\\nagg_count: 20\\nagg_amount: 2,506',\n",
       "   'value': 2505.8}),\n",
       " ('6059',\n",
       "  '7182',\n",
       "  {'title': 'from: 6059\\nto: 7182\\nagg_count: 22\\nagg_amount: 228',\n",
       "   'value': 228.14})]"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Verify if the edge attributes are successfully imported to H\n",
    "list(H.edges(data=True))[0:2]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ea8bb573",
   "metadata": {},
   "source": [
    "<h3>? graph will not render Still getting this error:AttributeError: type object 'Network' has no attribute 'Network'</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "bd706216",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "ename": "AttributeError",
     "evalue": "type object 'Network' has no attribute 'Network'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[43], line 3\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[38;5;66;03m# Generate the graph with pyvis \u001b[39;00m\n\u001b[1;32m      2\u001b[0m \u001b[38;5;66;03m# Add a line nt.toggle_physics(False) beforehand otherwise the graph will take a very long time to load\u001b[39;00m\n\u001b[0;32m----> 3\u001b[0m nt \u001b[38;5;241m=\u001b[39m net\u001b[38;5;241m.\u001b[39mNetwork(width\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124m100\u001b[39m\u001b[38;5;124m%\u001b[39m\u001b[38;5;124m'\u001b[39m, height\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124m100\u001b[39m\u001b[38;5;124m%\u001b[39m\u001b[38;5;124m'\u001b[39m, directed\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mFalse\u001b[39;00m)\n\u001b[1;32m      4\u001b[0m nt\u001b[38;5;241m.\u001b[39mtoggle_physics(\u001b[38;5;28;01mFalse\u001b[39;00m)\n\u001b[1;32m      5\u001b[0m nt\u001b[38;5;241m.\u001b[39mfrom_nx(H)\n",
      "\u001b[0;31mAttributeError\u001b[0m: type object 'Network' has no attribute 'Network'"
     ]
    }
   ],
   "source": [
    "# Generate the graph with pyvis \n",
    "# Add a line nt.toggle_physics(False) beforehand otherwise the graph will take a very long time to load\n",
    "nt = net.Network(width='100%', height='100%', directed=False)\n",
    "nt.toggle_physics(False)\n",
    "nt.from_nx(H)\n",
    "nt.show('06_H_full.html')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4b772249",
   "metadata": {},
   "outputs": [],
   "source": [
    "# The graph has too much info to interpret so we need to separate it into groups with community detection\n",
    "# from fund flow analysis perspective, a reasonable approach is to group linked nodes in a single group and \n",
    "# those non-linked nodes into another group so that each group are unconnected, which can be assessed by the \n",
    "# function nx.connected_components(H)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "547a71f2",
   "metadata": {},
   "source": [
    "<h3>Identify mutually exclusive communities</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "9ceb31eb",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[{'4333', '6059', '7182', '9703'},\n",
       " {'4297', '5573', '7266', '7369', '9795'},\n",
       " {'5788', '8914', '9722', '9962'},\n",
       " {'6333', '7881', '7911', '9751', '9797'},\n",
       " {'4205', '474', '5366', '5376', '8708', '9789'}]"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Separate nodes that are connected and not connect\n",
    "list(nx.connected_components(H))[0:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "36734dd5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1041"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check how many unconnected groups there are in H: \n",
    "len(list(nx.connected_components(H)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "c1441d17",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Assign the group number (from 1 to 1,041) to each node in the form of dictionary \n",
    "# In other words: Assign a community to each group\n",
    "group_dict_H = {}\n",
    "\n",
    "for group, nodes in enumerate(sorted(list(nx.connected_components(H)), key=len, reverse=True), start=1):\n",
    "    for node in nodes:\n",
    "        group_dict_H[node] = group"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "19be9a45",
   "metadata": {},
   "outputs": [],
   "source": [
    "# NOTE: ## Other algorithms: https://networkx.org/documentation/stable/reference/algorithms/community.html"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "994c2977",
   "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>source</th>\n",
       "      <th>group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>62</th>\n",
       "      <td>1087</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>1609</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>1890</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>1905</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>2369</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3035</th>\n",
       "      <td>5233</td>\n",
       "      <td>1039</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3036</th>\n",
       "      <td>275</td>\n",
       "      <td>1040</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3037</th>\n",
       "      <td>998</td>\n",
       "      <td>1040</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3038</th>\n",
       "      <td>7693</td>\n",
       "      <td>1041</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3039</th>\n",
       "      <td>8141</td>\n",
       "      <td>1041</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3040 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     source  group\n",
       "62     1087      1\n",
       "36     1609      1\n",
       "49     1890      1\n",
       "32     1905      1\n",
       "20     2369      1\n",
       "...     ...    ...\n",
       "3035   5233   1039\n",
       "3036    275   1040\n",
       "3037    998   1040\n",
       "3038   7693   1041\n",
       "3039   8141   1041\n",
       "\n",
       "[3040 rows x 2 columns]"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get a summary of group size \n",
    "groups = pd.DataFrame({'source':group_dict_H.keys(), 'group':group_dict_H.values()}).sort_values(by=['group','source'])\n",
    "groups"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f1deaa19",
   "metadata": {},
   "source": [
    "<h3>Display ID, group, degree and neighbors in graph (hover on node)</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "57b7720c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set the node attribute value and title for each node under H, add one more element which is the group attribute\n",
    "# Group attribute will assign different node and edge colors for each group\n",
    "nx.set_node_attributes(H, group_dict_H, 'group')\n",
    "\n",
    "degree_dict_H = dict(H.degree)\n",
    "nx.set_node_attributes(H, degree_dict_H, 'value')\n",
    "\n",
    "neighbor_dict_H = {}\n",
    "for node in list(H.nodes):\n",
    "    neighbor_dict_H[node] = ','.join(H.neighbors(node))\n",
    "\n",
    "title_dict_H = {}\n",
    "for node in list(H.nodes):\n",
    "    title_dict_H[node] = f'id: {node}\\ngroup: {group_dict_H[node]}\\ndegree: {degree_dict_H[node]}\\nneighbor: {neighbor_dict_H[node]}'\n",
    "nx.set_node_attributes(H, title_dict_H, 'title')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e2d8284e",
   "metadata": {},
   "outputs": [],
   "source": [
    "# With group attribute we can use it to filter the nodes for graph generation with H.subgraph(). "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "368ab661",
   "metadata": {},
   "source": [
    "<h3>Create subgraph to display selected groups only</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "9eaa7797",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Generate the network graph for group 1 to 4:target_groups = [1,2,3,4]\n",
    "# Display group 1-4 with pyvis default layout\n",
    "target_groups = [1,2,3,4]\n",
    "target_nodes = [node for node, group in group_dict_H.items() if group in target_groups]\n",
    "H_sub = H.subgraph(target_nodes)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "00b6cb6f",
   "metadata": {},
   "source": [
    "<h3> ?? graph wil not render</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "e1d4c5de",
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'show_graph' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[55], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m show_graph(H_sub, \u001b[38;5;124m'\u001b[39m\u001b[38;5;124m7_communities_pyvis_default.html\u001b[39m\u001b[38;5;124m'\u001b[39m)\n",
      "\u001b[0;31mNameError\u001b[0m: name 'show_graph' is not defined"
     ]
    }
   ],
   "source": [
    "# ERROR: NameError: name 'show_graph' is not defined\n",
    "# should have a network graph with diff colors\n",
    "show_graph(H_sub, '7_communities_pyvis_default.html')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "754a6b04",
   "metadata": {},
   "source": [
    "<h3>?? graph won't render Display group 1-4 with networkx kamada kawai layout</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "4d05f99b",
   "metadata": {},
   "outputs": [
    {
     "ename": "KeyError",
     "evalue": "'1'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyError\u001b[0m                                  Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[57], line 6\u001b[0m\n\u001b[1;32m      4\u001b[0m pos \u001b[38;5;241m=\u001b[39m nx\u001b[38;5;241m.\u001b[39mkamada_kawai_layout(H_sub, scale\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m1000\u001b[39m)\n\u001b[1;32m      5\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m node \u001b[38;5;129;01min\u001b[39;00m nt\u001b[38;5;241m.\u001b[39mnodes:\n\u001b[0;32m----> 6\u001b[0m     node[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mx\u001b[39m\u001b[38;5;124m'\u001b[39m] \u001b[38;5;241m=\u001b[39m pos[node[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mid\u001b[39m\u001b[38;5;124m'\u001b[39m]][\u001b[38;5;241m0\u001b[39m]\n\u001b[1;32m      7\u001b[0m     node[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124my\u001b[39m\u001b[38;5;124m'\u001b[39m] \u001b[38;5;241m=\u001b[39m pos[node[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mid\u001b[39m\u001b[38;5;124m'\u001b[39m]][\u001b[38;5;241m1\u001b[39m]\n\u001b[1;32m      9\u001b[0m nt\u001b[38;5;241m.\u001b[39mshow(\u001b[38;5;124m'\u001b[39m\u001b[38;5;124m08_communities_kamada_kawai.html\u001b[39m\u001b[38;5;124m'\u001b[39m)\n",
      "\u001b[0;31mKeyError\u001b[0m: '1'"
     ]
    }
   ],
   "source": [
    "# ??? get ERROR nt = net.Network(width='100%', height='100%', directed=True)\n",
    "nt.from_nx(H_sub)\n",
    "\n",
    "pos = nx.kamada_kawai_layout(H_sub, scale=1000)\n",
    "for node in nt.nodes:\n",
    "    node['x'] = pos[node['id']][0]\n",
    "    node['y'] = pos[node['id']][1]\n",
    "\n",
    "nt.show('08_communities_kamada_kawai.html')\n",
    "nt = net.Network(width='100%', height='100%', directed=True)\n",
    "nt.from_nx(H_sub)\n",
    "\n",
    "pos = nx.kamada_kawai_layout(H_sub, scale=1000)\n",
    "for node in nt.nodes:\n",
    "    node['x'] = pos[node['id']][0]\n",
    "    node['y'] = pos[node['id']][1]\n",
    "\n",
    "nt.show('08_communities_kamada_kawai.html')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "26b2b8e5",
   "metadata": {},
   "source": [
    "<h3> ?? graph won't render</h3>\n",
    "<h3>Display group 1-4 with networkx spring layout</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "47353934",
   "metadata": {},
   "outputs": [
    {
     "ename": "AttributeError",
     "evalue": "type object 'Network' has no attribute 'Network'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[58], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m nt \u001b[38;5;241m=\u001b[39m net\u001b[38;5;241m.\u001b[39mNetwork(width\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124m100\u001b[39m\u001b[38;5;124m%\u001b[39m\u001b[38;5;124m'\u001b[39m, height\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124m100\u001b[39m\u001b[38;5;124m%\u001b[39m\u001b[38;5;124m'\u001b[39m, directed\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mTrue\u001b[39;00m)\n\u001b[1;32m      2\u001b[0m nt\u001b[38;5;241m.\u001b[39mfrom_nx(H_sub)\n\u001b[1;32m      4\u001b[0m pos \u001b[38;5;241m=\u001b[39m nx\u001b[38;5;241m.\u001b[39mspring_layout(H_sub, scale\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m1000\u001b[39m, seed\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m721\u001b[39m, k\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m0.01\u001b[39m)\n",
      "\u001b[0;31mAttributeError\u001b[0m: type object 'Network' has no attribute 'Network'"
     ]
    }
   ],
   "source": [
    "nt = net.Network(width='100%', height='100%', directed=True)\n",
    "nt.from_nx(H_sub)\n",
    "\n",
    "pos = nx.spring_layout(H_sub, scale=1000, seed=721, k=0.01)\n",
    "for node in nt.nodes:\n",
    "    node['x'] = pos[node['id']][0]\n",
    "    node['y'] = pos[node['id']][1]\n",
    "\n",
    "nt.show('09_communities_spring.html')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8e3ce9ab",
   "metadata": {},
   "source": [
    "<h3>Identify end-to-end fund flow of any given node</h3>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "78147c07",
   "metadata": {},
   "source": [
    "<h3> ?? graph won't render</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "09066498",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "ename": "AttributeError",
     "evalue": "type object 'Network' has no attribute 'Network'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[60], line 9\u001b[0m\n\u001b[1;32m      5\u001b[0m target_nodes \u001b[38;5;241m=\u001b[39m [node \u001b[38;5;28;01mfor\u001b[39;00m node, group \u001b[38;5;129;01min\u001b[39;00m group_dict_H\u001b[38;5;241m.\u001b[39mitems() \u001b[38;5;28;01mif\u001b[39;00m group \u001b[38;5;129;01min\u001b[39;00m target_groups]\n\u001b[1;32m      7\u001b[0m H_sub \u001b[38;5;241m=\u001b[39m H\u001b[38;5;241m.\u001b[39msubgraph(target_nodes)\n\u001b[0;32m----> 9\u001b[0m nt \u001b[38;5;241m=\u001b[39m net\u001b[38;5;241m.\u001b[39mNetwork(width\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124m100\u001b[39m\u001b[38;5;124m%\u001b[39m\u001b[38;5;124m'\u001b[39m, height\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124m100\u001b[39m\u001b[38;5;124m%\u001b[39m\u001b[38;5;124m'\u001b[39m, directed\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mTrue\u001b[39;00m)\n\u001b[1;32m     10\u001b[0m nt\u001b[38;5;241m.\u001b[39mfrom_nx(H_sub)\n\u001b[1;32m     12\u001b[0m pos \u001b[38;5;241m=\u001b[39m nx\u001b[38;5;241m.\u001b[39mspring_layout(H_sub, scale\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m1000\u001b[39m, seed\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m721\u001b[39m, k\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m0.01\u001b[39m)\n",
      "\u001b[0;31mAttributeError\u001b[0m: type object 'Network' has no attribute 'Network'"
     ]
    }
   ],
   "source": [
    "# Extend the code further to show the network graph of any given node and its connected nodes. \n",
    "# Find the end-to-end fund flow of account number 9989\n",
    "# Most powerful use case in AML investigation which allow you to identify all related accounts \n",
    "# from an end-to-end perspective, staring with any given node\n",
    "# Highlight the input node 9989 by changing its shape, image and color attribute\n",
    "target_node = '9989'\n",
    "target_groups = [dict(H.nodes(data=True))[target_node]['group']]\n",
    "target_nodes = [node for node, group in group_dict_H.items() if group in target_groups]\n",
    "\n",
    "H_sub = H.subgraph(target_nodes)\n",
    "\n",
    "nt = net.Network(width='100%', height='100%', directed=True)\n",
    "nt.from_nx(H_sub)\n",
    "\n",
    "pos = nx.spring_layout(H_sub, scale=1000, seed=721, k=0.01)\n",
    "for node in nt.nodes:\n",
    "    node['x'] = pos[node['id']][0]\n",
    "    node['y'] = pos[node['id']][1]\n",
    "\n",
    "[node for node in nt.nodes if node['id']==target_node][0]['shape'] = 'image'\n",
    "[node for node in nt.nodes if node['id']==target_node][0]['image'] = 'https://openmoji.org/data/color/svg/1F608.svg'\n",
    "[node for node in nt.nodes if node['id']==target_node][0]['color'] = '#ff4f4f'\n",
    "\n",
    "nt.show('10_show_group_with_given_node.html')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "d9b6c40d",
   "metadata": {
    "scrolled": true
   },
   "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>sample_size</th>\n",
       "      <th>max_group_size</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1000</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1500</td>\n",
       "      <td>28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2000</td>\n",
       "      <td>64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2500</td>\n",
       "      <td>121</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3000</td>\n",
       "      <td>406</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>131</th>\n",
       "      <td>66500</td>\n",
       "      <td>9995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>132</th>\n",
       "      <td>67000</td>\n",
       "      <td>9997</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>133</th>\n",
       "      <td>67500</td>\n",
       "      <td>9998</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>134</th>\n",
       "      <td>68000</td>\n",
       "      <td>9999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>135</th>\n",
       "      <td>68500</td>\n",
       "      <td>9999</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>136 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     sample_size  max_group_size\n",
       "0           1000              11\n",
       "1           1500              28\n",
       "2           2000              64\n",
       "3           2500             121\n",
       "4           3000             406\n",
       "..           ...             ...\n",
       "131        66500            9995\n",
       "132        67000            9997\n",
       "133        67500            9998\n",
       "134        68000            9999\n",
       "135        68500            9999\n",
       "\n",
       "[136 rows x 2 columns]"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Determine the manageable timeframe coverage (or sample size in our case), \n",
    "# check the maximum group size under different sample size\n",
    "# Replace random sampling with different range of timeframe (e.g. week, month, quarter, year etc.) \n",
    "# For example, there are 11 nodes in the biggest group if we have sampled 1,000 edges, \n",
    "# 28 nodes in the biggest group for a 1,500 edge sample, so forth and so on.\n",
    "\n",
    "edge_full = txn_full.groupby(['SENDER_ACCOUNT_ID', 'RECEIVER_ACCOUNT_ID'],as_index=False).agg({'TX_AMOUNT':'count'}).drop(columns=['TX_AMOUNT']).rename(columns={'SENDER_ACCOUNT_ID':'source','RECEIVER_ACCOUNT_ID':'target'})\n",
    "\n",
    "max_dict = {}\n",
    "for i in range(1000, len(edge_full)+1, 500):\n",
    "    G = nx.from_pandas_edgelist(edge_full.sample(i, random_state=721), source='source', target='target')\n",
    "    max_dict[i] = max(sorted([len(x) for x in list(nx.connected_components(G))]))\n",
    "\n",
    "max_group_size_per_sample = pd.DataFrame({'sample_size':max_dict.keys(), 'max_group_size':max_dict.values()})\n",
    "max_group_size_per_sample"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f42cffc7",
   "metadata": {},
   "source": [
    "<h3>?? has error Export node and edge information for further analysis</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "73295a3d",
   "metadata": {},
   "outputs": [
    {
     "ename": "AttributeError",
     "evalue": "'float' object has no attribute 'split'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[62], line 2\u001b[0m\n\u001b[1;32m      1\u001b[0m H_sub_nodes \u001b[38;5;241m=\u001b[39m pd\u001b[38;5;241m.\u001b[39mDataFrame(nt\u001b[38;5;241m.\u001b[39mnodes)[[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mid\u001b[39m\u001b[38;5;124m'\u001b[39m,\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mgroup\u001b[39m\u001b[38;5;124m'\u001b[39m,\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mvalue\u001b[39m\u001b[38;5;124m'\u001b[39m,\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mtitle\u001b[39m\u001b[38;5;124m'\u001b[39m]]\u001b[38;5;241m.\u001b[39mrename(columns\u001b[38;5;241m=\u001b[39m{\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mid\u001b[39m\u001b[38;5;124m'\u001b[39m:\u001b[38;5;124m'\u001b[39m\u001b[38;5;124msource\u001b[39m\u001b[38;5;124m'\u001b[39m,\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mvalue\u001b[39m\u001b[38;5;124m'\u001b[39m:\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mdegree\u001b[39m\u001b[38;5;124m'\u001b[39m,\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mtitle\u001b[39m\u001b[38;5;124m'\u001b[39m:\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mneighbor\u001b[39m\u001b[38;5;124m'\u001b[39m})\u001b[38;5;241m.\u001b[39msort_values(by\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124msource\u001b[39m\u001b[38;5;124m'\u001b[39m)\n\u001b[0;32m----> 2\u001b[0m H_sub_nodes[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mneighbor\u001b[39m\u001b[38;5;124m'\u001b[39m] \u001b[38;5;241m=\u001b[39m H_sub_nodes[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mneighbor\u001b[39m\u001b[38;5;124m'\u001b[39m]\u001b[38;5;241m.\u001b[39mapply(\u001b[38;5;28;01mlambda\u001b[39;00m x:x\u001b[38;5;241m.\u001b[39msplit(\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mneighbor: \u001b[39m\u001b[38;5;124m'\u001b[39m)[\u001b[38;5;241m-\u001b[39m\u001b[38;5;241m1\u001b[39m])\n\u001b[1;32m      3\u001b[0m H_sub_nodes\u001b[38;5;241m.\u001b[39mhead(\u001b[38;5;241m5\u001b[39m)\n",
      "File \u001b[0;32m/opt/anaconda3/lib/python3.11/site-packages/pandas/core/series.py:4764\u001b[0m, in \u001b[0;36mSeries.apply\u001b[0;34m(self, func, convert_dtype, args, by_row, **kwargs)\u001b[0m\n\u001b[1;32m   4629\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mapply\u001b[39m(\n\u001b[1;32m   4630\u001b[0m     \u001b[38;5;28mself\u001b[39m,\n\u001b[1;32m   4631\u001b[0m     func: AggFuncType,\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m   4636\u001b[0m     \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs,\n\u001b[1;32m   4637\u001b[0m ) \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m>\u001b[39m DataFrame \u001b[38;5;241m|\u001b[39m Series:\n\u001b[1;32m   4638\u001b[0m \u001b[38;5;250m    \u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m   4639\u001b[0m \u001b[38;5;124;03m    Invoke function on values of Series.\u001b[39;00m\n\u001b[1;32m   4640\u001b[0m \n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m   4755\u001b[0m \u001b[38;5;124;03m    dtype: float64\u001b[39;00m\n\u001b[1;32m   4756\u001b[0m \u001b[38;5;124;03m    \"\"\"\u001b[39;00m\n\u001b[1;32m   4757\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m SeriesApply(\n\u001b[1;32m   4758\u001b[0m         \u001b[38;5;28mself\u001b[39m,\n\u001b[1;32m   4759\u001b[0m         func,\n\u001b[1;32m   4760\u001b[0m         convert_dtype\u001b[38;5;241m=\u001b[39mconvert_dtype,\n\u001b[1;32m   4761\u001b[0m         by_row\u001b[38;5;241m=\u001b[39mby_row,\n\u001b[1;32m   4762\u001b[0m         args\u001b[38;5;241m=\u001b[39margs,\n\u001b[1;32m   4763\u001b[0m         kwargs\u001b[38;5;241m=\u001b[39mkwargs,\n\u001b[0;32m-> 4764\u001b[0m     )\u001b[38;5;241m.\u001b[39mapply()\n",
      "File \u001b[0;32m/opt/anaconda3/lib/python3.11/site-packages/pandas/core/apply.py:1209\u001b[0m, in \u001b[0;36mSeriesApply.apply\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m   1206\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mapply_compat()\n\u001b[1;32m   1208\u001b[0m \u001b[38;5;66;03m# self.func is Callable\u001b[39;00m\n\u001b[0;32m-> 1209\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mapply_standard()\n",
      "File \u001b[0;32m/opt/anaconda3/lib/python3.11/site-packages/pandas/core/apply.py:1289\u001b[0m, in \u001b[0;36mSeriesApply.apply_standard\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m   1283\u001b[0m \u001b[38;5;66;03m# row-wise access\u001b[39;00m\n\u001b[1;32m   1284\u001b[0m \u001b[38;5;66;03m# apply doesn't have a `na_action` keyword and for backward compat reasons\u001b[39;00m\n\u001b[1;32m   1285\u001b[0m \u001b[38;5;66;03m# we need to give `na_action=\"ignore\"` for categorical data.\u001b[39;00m\n\u001b[1;32m   1286\u001b[0m \u001b[38;5;66;03m# TODO: remove the `na_action=\"ignore\"` when that default has been changed in\u001b[39;00m\n\u001b[1;32m   1287\u001b[0m \u001b[38;5;66;03m#  Categorical (GH51645).\u001b[39;00m\n\u001b[1;32m   1288\u001b[0m action \u001b[38;5;241m=\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mignore\u001b[39m\u001b[38;5;124m\"\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28misinstance\u001b[39m(obj\u001b[38;5;241m.\u001b[39mdtype, CategoricalDtype) \u001b[38;5;28;01melse\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[0;32m-> 1289\u001b[0m mapped \u001b[38;5;241m=\u001b[39m obj\u001b[38;5;241m.\u001b[39m_map_values(\n\u001b[1;32m   1290\u001b[0m     mapper\u001b[38;5;241m=\u001b[39mcurried, na_action\u001b[38;5;241m=\u001b[39maction, convert\u001b[38;5;241m=\u001b[39m\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mconvert_dtype\n\u001b[1;32m   1291\u001b[0m )\n\u001b[1;32m   1293\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mlen\u001b[39m(mapped) \u001b[38;5;129;01mand\u001b[39;00m \u001b[38;5;28misinstance\u001b[39m(mapped[\u001b[38;5;241m0\u001b[39m], ABCSeries):\n\u001b[1;32m   1294\u001b[0m     \u001b[38;5;66;03m# GH#43986 Need to do list(mapped) in order to get treated as nested\u001b[39;00m\n\u001b[1;32m   1295\u001b[0m     \u001b[38;5;66;03m#  See also GH#25959 regarding EA support\u001b[39;00m\n\u001b[1;32m   1296\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m obj\u001b[38;5;241m.\u001b[39m_constructor_expanddim(\u001b[38;5;28mlist\u001b[39m(mapped), index\u001b[38;5;241m=\u001b[39mobj\u001b[38;5;241m.\u001b[39mindex)\n",
      "File \u001b[0;32m/opt/anaconda3/lib/python3.11/site-packages/pandas/core/base.py:921\u001b[0m, in \u001b[0;36mIndexOpsMixin._map_values\u001b[0;34m(self, mapper, na_action, convert)\u001b[0m\n\u001b[1;32m    918\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28misinstance\u001b[39m(arr, ExtensionArray):\n\u001b[1;32m    919\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m arr\u001b[38;5;241m.\u001b[39mmap(mapper, na_action\u001b[38;5;241m=\u001b[39mna_action)\n\u001b[0;32m--> 921\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m algorithms\u001b[38;5;241m.\u001b[39mmap_array(arr, mapper, na_action\u001b[38;5;241m=\u001b[39mna_action, convert\u001b[38;5;241m=\u001b[39mconvert)\n",
      "File \u001b[0;32m/opt/anaconda3/lib/python3.11/site-packages/pandas/core/algorithms.py:1814\u001b[0m, in \u001b[0;36mmap_array\u001b[0;34m(arr, mapper, na_action, convert)\u001b[0m\n\u001b[1;32m   1812\u001b[0m values \u001b[38;5;241m=\u001b[39m arr\u001b[38;5;241m.\u001b[39mastype(\u001b[38;5;28mobject\u001b[39m, copy\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mFalse\u001b[39;00m)\n\u001b[1;32m   1813\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m na_action \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[0;32m-> 1814\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m lib\u001b[38;5;241m.\u001b[39mmap_infer(values, mapper, convert\u001b[38;5;241m=\u001b[39mconvert)\n\u001b[1;32m   1815\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m   1816\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m lib\u001b[38;5;241m.\u001b[39mmap_infer_mask(\n\u001b[1;32m   1817\u001b[0m         values, mapper, mask\u001b[38;5;241m=\u001b[39misna(values)\u001b[38;5;241m.\u001b[39mview(np\u001b[38;5;241m.\u001b[39muint8), convert\u001b[38;5;241m=\u001b[39mconvert\n\u001b[1;32m   1818\u001b[0m     )\n",
      "File \u001b[0;32mlib.pyx:2926\u001b[0m, in \u001b[0;36mpandas._libs.lib.map_infer\u001b[0;34m()\u001b[0m\n",
      "Cell \u001b[0;32mIn[62], line 2\u001b[0m, in \u001b[0;36m<lambda>\u001b[0;34m(x)\u001b[0m\n\u001b[1;32m      1\u001b[0m H_sub_nodes \u001b[38;5;241m=\u001b[39m pd\u001b[38;5;241m.\u001b[39mDataFrame(nt\u001b[38;5;241m.\u001b[39mnodes)[[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mid\u001b[39m\u001b[38;5;124m'\u001b[39m,\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mgroup\u001b[39m\u001b[38;5;124m'\u001b[39m,\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mvalue\u001b[39m\u001b[38;5;124m'\u001b[39m,\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mtitle\u001b[39m\u001b[38;5;124m'\u001b[39m]]\u001b[38;5;241m.\u001b[39mrename(columns\u001b[38;5;241m=\u001b[39m{\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mid\u001b[39m\u001b[38;5;124m'\u001b[39m:\u001b[38;5;124m'\u001b[39m\u001b[38;5;124msource\u001b[39m\u001b[38;5;124m'\u001b[39m,\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mvalue\u001b[39m\u001b[38;5;124m'\u001b[39m:\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mdegree\u001b[39m\u001b[38;5;124m'\u001b[39m,\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mtitle\u001b[39m\u001b[38;5;124m'\u001b[39m:\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mneighbor\u001b[39m\u001b[38;5;124m'\u001b[39m})\u001b[38;5;241m.\u001b[39msort_values(by\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124msource\u001b[39m\u001b[38;5;124m'\u001b[39m)\n\u001b[0;32m----> 2\u001b[0m H_sub_nodes[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mneighbor\u001b[39m\u001b[38;5;124m'\u001b[39m] \u001b[38;5;241m=\u001b[39m H_sub_nodes[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mneighbor\u001b[39m\u001b[38;5;124m'\u001b[39m]\u001b[38;5;241m.\u001b[39mapply(\u001b[38;5;28;01mlambda\u001b[39;00m x:x\u001b[38;5;241m.\u001b[39msplit(\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mneighbor: \u001b[39m\u001b[38;5;124m'\u001b[39m)[\u001b[38;5;241m-\u001b[39m\u001b[38;5;241m1\u001b[39m])\n\u001b[1;32m      3\u001b[0m H_sub_nodes\u001b[38;5;241m.\u001b[39mhead(\u001b[38;5;241m5\u001b[39m)\n",
      "\u001b[0;31mAttributeError\u001b[0m: 'float' object has no attribute 'split'"
     ]
    }
   ],
   "source": [
    "H_sub_nodes = pd.DataFrame(nt.nodes)[['id','group','value','title']].rename(columns={'id':'source','value':'degree','title':'neighbor'}).sort_values(by='source')\n",
    "H_sub_nodes['neighbor'] = H_sub_nodes['neighbor'].apply(lambda x:x.split('neighbor: ')[-1])\n",
    "H_sub_nodes.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "05ad5283",
   "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>source</th>\n",
       "      <th>target</th>\n",
       "      <th>agg_edge_large_count</th>\n",
       "      <th>agg_edge_large_amt</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6847</th>\n",
       "      <td>2394</td>\n",
       "      <td>9158</td>\n",
       "      <td>3</td>\n",
       "      <td>22,831,488</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20399</th>\n",
       "      <td>5049</td>\n",
       "      <td>9989</td>\n",
       "      <td>24</td>\n",
       "      <td>371</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31225</th>\n",
       "      <td>6642</td>\n",
       "      <td>2394</td>\n",
       "      <td>20</td>\n",
       "      <td>3,114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32421</th>\n",
       "      <td>6743</td>\n",
       "      <td>9912</td>\n",
       "      <td>20</td>\n",
       "      <td>2,012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34844</th>\n",
       "      <td>7119</td>\n",
       "      <td>9587</td>\n",
       "      <td>24</td>\n",
       "      <td>363</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      source target  agg_edge_large_count  agg_edge_large_amt\n",
       "6847    2394   9158                     3          22,831,488\n",
       "20399   5049   9989                    24                 371\n",
       "31225   6642   2394                    20               3,114\n",
       "32421   6743   9912                    20               2,012\n",
       "34844   7119   9587                    24                 363"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "H_sub_edges = edge_large[(edge_large['source'].isin(target_nodes)) | (edge_large['target'].isin(target_nodes))].drop(columns=['value','title']).sort_values(by=['source','target'])\n",
    "H_sub_edges.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "0d3ce478",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Show maximum group size per sample\n",
    "max_group_size_per_sample.plot(x='sample_size', y='max_group_size', figsize=(10,8))\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.11.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}