Network_analysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "import networkx as nx\n",
    "import pandas as pd\n",
    "from tqdm import tqdm_notebook,tqdm\n",
    "import numpy as np\n",
    "import parmap\n",
    "import json\n",
    "parent_path='Data/'\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(parent_path+'users_information.json', 'r') as fp:\n",
    "    dict_users=json.load(fp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "import json\n",
    "with open(parent_path+'fear_speech_data.json', 'r') as fp:\n",
    "    fear_speech_data=json.load(fp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'message_text': '*प्रशासक समिति*✊🚩  ●●●●●●●●●●● ● ● ● 😎🚩 *आंतकवादी संगठनों का💣🔪 इस्लामिक नाम और उनका इस्लाम.....* *धर्म से जुड़ा हुआ अर्थ...* 🐖🐖🐖 *1.लश्करे तैयबा-फरिश्तो की सेना* *2.अल कायदा-अल्लाह का कायदा...* *3.जेश ए मोहम्द-मोहम्मद साहेब का दल...* *4.तहरिक ए तालिबान-पवित्र योद्धाओ का दल...* *5.हिजबुल मुजादिन-इस्लामी बलिदानियो का समूह...* *6.बोको हराम -पैगम्बर मुहम्मद की शिक्षा को फैलाने के लिए प्रतिबद्ध..* *सभी मुस्लमान अज्ञानि अल्लाह की बताई हुई रूहानी किताब क़ुरआन की बताई राह* *(पूरी दुनिया को इस्लाम बनाना)पर ही चल रहे है कोई ज्यादा बच👨\\u200d👨\\u200d👦* *पैदा करके तो कोई लव जिहाद👫 करके तो कोई काफ़िर(गेर मुसलमान)को मारकर..* *धरती पर आंतक🔫💣💣💣🔫फेला रहे है\\ufeff।।* 😡😡😡😎😡😡😡  *जय सनातन धर्म की*🚩🚩🚩  🙏🚩🇮🇳🔱🏹🐚🕉',\n",
       " 'translated_text': '* Administrator ✊ 🚩   Committee * ● ●●●●●●●●●● 🚩  😎  ● ● ● नाम 🔪  💣  * Islamic name of terrorist organizations and their meaning 🐖  🐖  🐖  in Islam… .. * Religion… * ... 1. * 1. Army of Lashkar-e-Taiba-Farishto * * 2. Al Qaeda-Qaeda of Allah ... * * 3. Team of Jesh-e-Mohammad-Mohammed Saheb ... * * 4. Tahrik-e-Taliban-Holy Warriors Party ... * * 5. Hijbul Mujadin-group of Islamic sacrifices ... * * 6. Boko Haram - Committed to spread the teachings of Prophet Muhammad .. * * All the Muslim ignorance told the Quran of Allah the 👨 👦   👨  spiritual book The path 👫  is going on * (making the whole world Islam), by creating 🔫  💣  💣  💣  🔫 😡 😡 😡 😡  😡   😎   😡    some 🚩 🚩  🙏 🇮 🔱 🐚 🕉   🏹   🇳   🚩   🚩   more children * * *, by creating some love jihad and by killing a kafir (Ger Muslim) .. * * Terror on Earth 💣💣💣🔫Fella is there. * 😡😡😡😎😡😡😡 * Jai of Sanatan Dharma * 🚩🚩🚩 🙏🚩🇮🇳🔱🏹🐚🕉  ',\n",
       " 'annotation_list': ['Fear speech', 'Fear speech', 'Normal'],\n",
       " 'propagation': [{'group_id': 9087,\n",
       "   'user_id': 229869,\n",
       "   'timestamp': 1538130086000},\n",
       "  {'group_id': 7, 'user_id': 215, 'timestamp': 1550186113000}]}"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fear_speech_data['0']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_user_matched=pd.read_csv('Data/final_user_selection_matched.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2932"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df_user_matched[df_user_matched['annotation']=='Non fear speech'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 109542/109542 [2:03:31<00:00, 14.78it/s]  \n"
     ]
    }
   ],
   "source": [
    "tuples_user_info=[]\n",
    "\n",
    "list_users= list(dict_users.keys())\n",
    "for i in tqdm(range(len(list_users))):\n",
    "    for j in range(i,len(list_users)):\n",
    "        if(i==j):\n",
    "            continue\n",
    "        temp_user1_groups= set(dict_users[str(list_users[i])]['group_the_user_participate'])\n",
    "        temp_user2_groups= set(dict_users[str(list_users[j])]['group_the_user_participate'])\n",
    "        \n",
    "        \n",
    "        common_groups=len(temp_user1_groups.intersection(temp_user2_groups))\n",
    "        if(common_groups>0):\n",
    "            tuples_user_info.append([list_users[i],list_users[j],common_groups])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "import networkx as nx"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "G = nx.Graph()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "G.add_weighted_edges_from(tuples_user_info)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {},
   "outputs": [],
   "source": [
    "nx.write_weighted_edgelist(G, \"Networks/users_network.edgelist.gz\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load precomputed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import json\n",
    "\n",
    "# with open(\"Networks/user_dict.json\", 'r') as fp:\n",
    "#     dict_users_ids=json.load(fp)\n",
    "# with open(\"Networks/user_labels_dict.json\", 'r') as fp:\n",
    "#     dict_users_names=json.load(fp)\n",
    "\n",
    "    \n",
    "G_new=nx.read_weighted_edgelist(\"Networks/users_network.edgelist.gz\")\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 109542/109542 [00:00<00:00, 624287.43it/s]\n"
     ]
    }
   ],
   "source": [
    "dict_users_ids={}\n",
    "list_users= list(dict_users.keys())\n",
    "for i in tqdm(range(len(list_users))):\n",
    "    dict_users_ids[dict_users[str(i)]['user_id']]=str(i)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [],
   "source": [
    "k_cores=nx.algorithms.core.core_number(G_new)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/punyajoy/.conda/envs/punyajoy_gpu/lib/python3.7/site-packages/ipykernel_launcher.py:4: TqdmDeprecationWarning: This function will be removed in tqdm==5.0.0\n",
      "Please use `tqdm.notebook.tqdm` instead of `tqdm.tqdm_notebook`\n",
      "  after removing the cwd from sys.path.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "738a0c42d69d4743aa4d61c3a129a6ed",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value=''), FloatProgress(value=0.0, max=4782.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "dict_user_num_messages={}\n",
    "dict_group_num_messages={}\n",
    "\n",
    "for key in tqdm_notebook(fear_speech_data.keys(),total=len(fear_speech_data)):\n",
    "    element = fear_speech_data[key]\n",
    "    \n",
    "    count_fearspeech=element['annotation_list'].count('Fear speech')\n",
    "    count_normal=element['annotation_list'].count('Normal')\n",
    "    \n",
    "    if(count_fearspeech>count_normal):\n",
    "        for ele in element['propagation']:\n",
    "            try:\n",
    "                dict_user_num_messages[ele['user_id']]+=1\n",
    "            except KeyError:\n",
    "                dict_user_num_messages[ele['user_id']]=1\n",
    "            try:\n",
    "                dict_group_num_messages[ele['group_id']]+=1\n",
    "            except KeyError:\n",
    "                dict_group_num_messages[ele['group_id']]=1\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "dict_user_num_messages_sorted={k: v for k, v in sorted(dict_user_num_messages.items(), key=lambda item: item[1],reverse=True)}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib\n",
    "import seaborn as sns\n",
    "import matplotlib.ticker as mtick\n",
    "matplotlib.rcParams['figure.figsize'] = (5, 5)\n",
    "sns.set(context='paper',style='white',font_scale=1.5)\n",
    "\n",
    "font = {\n",
    "    'weight' : 'normal',\n",
    "    'size'   : 11}\n",
    "\n",
    "ax = sns.histplot(x=list(dict_user_num_messages.values()),bins=100)\n",
    "plt.xlabel('amount of fear speech')\n",
    "plt.ylabel('number of users')\n",
    "plt.yscale('log')\n",
    "\n",
    "\n",
    "\n",
    "sns.set(context='paper',style='white',font_scale=1.0)\n",
    "\n",
    "ax2 = plt.axes([0.55, 0.53, .3, .3])\n",
    "sns.ecdfplot(x=list(dict_user_num_messages_sorted.values()),ax=ax2)\n",
    "ax2.xaxis.set_major_formatter(mtick.PercentFormatter())\n",
    "\n",
    "plt.xlim(left=0,right=30)\n",
    "plt.ylim(bottom=0.5,top=1)\n",
    "\n",
    "\n",
    "plt.xlabel('% users')\n",
    "plt.ylabel('% messages')\n",
    "\n",
    "\n",
    "plt.savefig('../Fear_speech_results/fear_speech_per_users.pdf',bbox_inches='tight',dpi=400)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [],
   "source": [
    "list_cores_fs=[]\n",
    "list_cores_nfs=[]\n",
    "for index,row in df_user_matched.iterrows():\n",
    "    if(row['annotation']==\"Fear speech\"):\n",
    "        try:\n",
    "            list_cores_fs.append(k_cores[dict_users_ids[row['user_id']]])\n",
    "        except KeyError:\n",
    "            list_cores_fs.append(0)\n",
    "    else:\n",
    "        try:\n",
    "            list_cores_nfs.append(k_cores[dict_users_ids[row['user_id']]])\n",
    "        except KeyError:\n",
    "            list_cores_nfs.append(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.18155984866426092\n"
     ]
    }
   ],
   "source": [
    "from statistics import mean, stdev\n",
    "from math import sqrt\n",
    "\n",
    "# test conditions\n",
    "c0 = list_cores_fs\n",
    "c1 = list_cores_nfs\n",
    "\n",
    "cohens_d = (mean(c0) - mean(c1)) / (sqrt((stdev(c0) ** 2 + stdev(c1) ** 2) / 2))\n",
    "\n",
    "print(cohens_d)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2932"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(list_cores_fs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAVQAAAFKCAYAAABCTqdeAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjAsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+17YcXAAAgAElEQVR4nOzdeVyU1f7A8c8s7DAsLiiiIi64I4IbluWKuZTatdTyWle92Z6Zla22qaXZ1Uq7N1s1LU3btLTs55KWCy4RCu6KgCjINsM6y/P7gxxFQAdkZli+79erV855nmfmC8jX8zznnO9RKYqiIIQQ4oapnR2AEELUFZJQhRCimkhCFUKIaiIJVQghqokkVCGEqCaSUIUQoppIQhVCiGri0IS6YcMGJkyYQPfu3QkLC7vu+WfPnmXy5MlEREQQHR3NO++8g0ybFULUVA5NqDqdjgkTJvDcc89d91yz2cy0adMICgpix44drFixgh9++IGPP/7YAZEKIUTlOTSh3nzzzYwYMYLmzZtf99zY2FjOnDnDzJkz8fLyIjQ0lClTprBy5UoHRCqEEJVXY5+hJiYm0rJlS3Q6nbWtc+fOJCcnYzAYnBiZEEKUT+vsACpiMBjw8fEp1XYpuRoMBry9va/7HrY8pxVCiKo4cuRImbYam1C9vb3L9ERzc3Otx2xV3hctRH0377O97IxLdcpnB6qzmdu/4JrnKMYiPMN6okJl52hUeLTqgtrNs1JXVdRZq7EJtX379pw5cwa9Xm/tqR46dIjg4OBKJVQhajtFUbhycssve87wf7FnUaurnmziT1xkkO4EPVTx6NSF1RCljRQLPk1DaDRsnuM+04EcmlDNZjMmkwmj0QhAUVERAC4uLqjVpR/nRkVF0aJFC+bPn8+zzz7L+fPnWbZsGePHj3dkyELYRX6hkfSsintpFkVh4cr9GE0WUtLLjhn4u5oY1Cy3yp/fxuMEPd3TCBl6Lz7hA6r8PqI0lSProa5bt45Zs2aVaf/8889p3rw5w4cP58MPPyQqKgoomYc6e/Zs9u3bh4eHB3fddRdPPPEEKpVt/zKHhYXJLb+okeaviGX7gZTrnjfc5xA+lhz81Xl4qYqs7UG+avw79qry56tUKnQ9huPi17jK71GfVZRbHJpQHU0SqnAGxWREUSzXPOfVxT/RoCiZkYGXk6pJn4nFkGl9rUKh8dApaHUNcG/WrtRzPpXWpfoDFzarKLfU2GeoQtR0Xy3/nl1Hc+CKOybFZEQxG6977UlTILcFmPAP71eqXeOpw7NN95IXag1qrWu1xizsSxKqEOU4d/oUa9Zux2iyYCrMw6KosKg0XDS6Y0GFBoWjBf509Ciie7urbpvValz8m17z/W8B+vToj28j3TXPE7WLJFQhAHOBgUtD6UWpR9n9Syy/nWtAVLAKjacfrirQurnjr4ZiEzTWqYkEbr1pBEHNmzg3eFFjSEIV9V7e0b2cX1N6Gk+RuT0B3oE888QoJ0UlaiNJqKLeMhfoyfptNfoDm9H6NqLpva8C8Pvhi+zfn4Emv9jJEYraRhKqqLcubv4MQ9wWXJu0xq/vaOsUos1xxwEYP6S9M8MTtZAkVFHnGLPSsBSXrP7Jjf0Ji7HQuoSxMDkR1BpUWheM6Wfx7X07uR1G8eb3h7Bs2QlA3PEMJg3vyM3dmjntaxC1kyRUUasoZhOFKUdQTEawmFEslpL/KxZy9/5I4dmEMtd4hIbj2rCkZKRn2yg03gFoffxLjoV04eCRHE6fy+WOW1oDEBHWmH6STEUVSEIVtYZiMnL+23fIP7IbtYc3qNSo1Brr/y1FeXiEhuMXfSfuwX8Xr1CpSs65piR03q78Y0Bbu38Nom6ThCpqPEVRyNr2JQUnD1J07ji67jE0vO3fzg5LiDIkoYoaqzDlKOe/no9iKsZSaMCjVVf8b74b/353VdtnKIpC8gUDRtO1l4oKYQtJqKLGSv/hXcyGTALHPovGwwe34DCbC+PYauHK/Wzdn0xoM99qfV9RP0lCFTWSpSgf1FoCBkzEq10Pu3zG6s1H2bo/mTv7t2HisI52+QxRv0hCFTVK3tG9FKUeJ3vn1wBofRtV6/srioLZomCxKBw8mk6vTk0YO7Admhso1izEJZJQRY1yYd3baH0b4t68A4H/eBqN540VD5mxaBunU3Ox/J1Iry5WOWFIGF4eUgpPVA9JqKJGUVBoNOIR3JtXzyqllPQ8Jg7rSPuW/qjVKtRqFZq//++iVdO0gVe1fI4QIAlV1BAmQxbn1y4As+mG3yu/0Eh+oQmjyYLFohDS1If2IQHVEKUQ1yYJVTiFuTCPs0sfAYsZlYsbZn1JpXrfniNwbdKqSu9ZWGTiwNELzPl0r7VNrQIfTynSLBxDEqpwuLwjezj/9ZsANBgyGa2uIQBav8a4BYZc89q0i3l8uv4wZkvZeaN7DqVhUUCrUbP0mQF4e7jg6qLB1eV6K6WEqB6SUIVD5B3dS86u7zBmncdsyETt4U3jkY/h2TbS5vfIMRQx6/0dZOqLyl0m2qKJji6tG9Dt6gr6QjiIJFThEHkJv2MpysO39+1ovHzx6dzv+hddYXf8OdZuOU5GTiHTx0cwIKqFnSIVouokoQqHMMRvxzOsF369Rtp+TYGRzXuSMBQU89UvR2kVpKNfRDNJpqLGkoQq7M5cYADAr0/lthP581g6n/94mPC2jegfGcxD/wjH3VX+yoqaS/52CrtRFAsXf/6E3NgfQaVG7V65OZ+ZOYV4ubvw8pTedopQiOolCVXYTfr372KI345rYCuCJr6K2s2zUtdv2Hmy3NF8IWoqSajCLrJ2fI3h0A78okfjf+sEVCq1zdcaTRZW/ZyIocDIPTGyr5OoPWz/Wy6EDRSLmcKUo2RtW4VX+17433x3pZIpwIWsfNb8eozenZvSvX2gnSIVovpJD1VUq9z9v3Bx04cABNx6Dypt5QqPTJv3KynpJYNYU0d1wU0m5YtaRBKqqBa5BzaTl7iLgpMHcGvahmb/erNS1yuKQvyJi6SkG3jwzq50Cm0gyVTUOpJQxQ0z5+WQ8eNStL6N8b91Al5hvWy6bsGKfWw7kFymvXtYY5pIFahqs2rVKr788ksAsrKy6Nu3L82bN+fHH39EpVIxffp0BgwYgMVi4fXXX2fPnj24urry0ksv0a1bN9atW8e2bdvIzMwkJSWF5557jj/++IPff/+dbt26MXfuXCd/hTWIUoe1a9fO2SHUC9l7NignXh+jGHMzK3XdHU99pyz6cr8SdzxdOZqUqZzLMNgpQqEoimIwGJTbb79dOXjwoHLnnXcqRUVFyvnz55X+/fsrBoNB2bBhg/Lggw8qFotFSUxMVGJiYhRFUZS1a9cqI0aMUAoLC5WEhASla9euyr59+xSLxaKMHj1aOXHihJO/MserKLdID1XcMMNfW1G7e6Hx9qv0tUN6t6R9Symt5wgvvfQSd911FwcOHCAmJgZXV1caN25Mx44dOX78OAcOHGDEiBGoVCrCwsLw8PAgIyMDgOjoaNzc3GjXrh1ubm50794dgDZt2nDu3DlCQ0Od+aXVGJJQxQ0xGbIoOncC/353X3MDvWff38Ghkxdx1V4e8TdbFFw0MtHEEb744gsA7rnnHj799NNKX+/iUjK4qFarrX++9NpsNldLjHWBJFRxQ3L++BYAXdSwa5536ORF+nRpyh39WlvbXLRq2W3UAeLi4lizZg0rV64EoHv37rzxxhvcd999ZGdnc/jwYdq0aUNERAQbN27ktttu4/jx4xQWFtKwYUMnR1+7SEIVVaYoCjl71uPROgKNh/c1z1Wr4B8D2tKuhb+DohOXrFq1iszMTMaPHw/AgAEDuOWWWxg9ejQqlYoXXngBLy8vhg4dSmxsLCNHjsTFxUUGm6pApShXb1tWd4SFhXHkyBFnh1FnGRL/4MLaBQRNmoN7cNg1z73jqe+Y/1g/SaiiTqgot0gPVVRJ1o6vydq2CpWbJ27NyhZ7vuTseT0PvfV/QMktvhB1mfwNF5VmLsz7e2lpH4ImvnbNpaUnkrMBeOPBaEKa3tiW0DfCaDQybtw4oqKi2LhxY5njW7Zs4e6772bcuHHExcU5IUJRF0gPVVRaUXLJrY5/v7twbXTtYs8/7DhJgM6drm0aOSK0Cmm1WhYvXsxXX31V5pjZbGbRokWsWLGCvLw8nnjiCVatWuWEKEVtJwlVVFrWzq8BrptMAQqKzPTq3MTeIV2XSqWicePy95o6ffo0ISEheHt74+3tjclkoqioCDc3NwdHKWo7ueUXlWIxFlGUfARdj+HXPTc24Tw5hiJaN6v8hH9HysnJQae7/DhCp9ORnZ3txIhEbSU9VFEpxowUALw73XTdc7/ffoKmDbwIb1uz5zL6+vqi1+utr/V6PX5+tv0jEBZ27dkNouax58wfSajCZorFTOa2kmeLbkEVj+xfqUenwBpf6KRly5acPn2a/Px88vLy0Gg0lbrdl6l5tYe9/wGUhCpsZji0g4IT+/HpNui65/55LJ2z5/V0at3AAZHZ5vHHHyc+Ph5PT0/i4uLw8/Nj0KBBhIaG8sgjj3D//fejUqmYNWuWs0MVtZQkVGGTonMnSP9+MSoXNxoOe+Ca6/YBft2bhM7LjT6dmzoowutbtGhRhccGDhzIwIEDHRiNqIscOihlsVhYuHAh0dHRREREMHnyZFJSUio8//vvv2fkyJF0796dfv368cYbb1BcXOzAiMUlKR8/DUCTu2bZvKVJp9YNaNHEeXNPhXA0hybUZcuWsX79elasWMGOHTsICgpi2rRpWMrZ2TIxMZFnnnmGhx9+mH379rFq1Sp27NjBkiVLHBmyAIxZaQA0uuNxPEK62HTNuYw8e4YkxDUZ8ou5mFNQ5j97c+gt/5dffsmUKVOstRNnzpxJdHQ0+/bto0ePHqXOPXv2LL6+vgwdOhSAZs2aceutt5KYmOjIkAVQcOYQAJ6twis851xGHrviz7F5bxIq4EyanvB2zp3ML+onQ34x97y8EYvF8WVKHJZQ9Xo9KSkpdO7c2dqm0+lo2bIlCQkJZRLqTTfdRHBwMBs2bGDo0KGkpKTwf//3f/z73/92VMgCUMwmjBnJuDZuicar4lJ7H3wTx5HTmXi4abk1sjm392vN4J7Xn/gvRHUrMpqxWBQW/DsCf+/Sm0TevN6+n+2whGowlOxkeeUEagAfHx/rsSt5eHjwj3/8g5dffpmZM2diNpsZPXo0o0aNcki8okRO7I/k7P4eXeTQa55XUGjizgFtGTuwnYMiE7VZTmYWRQWFdnnvzKySOcX5q5/DRW3/2/wrOSyhenuX1Mu8cgL1pdeXjl3pm2++YeHChXzwwQd0796djIwMXnzxRZ555hkWLFjgkJgF5O7bhMbLj4ZDp1Z4TtrFPBJOZ3JTtyAHRiZqAnOBAcVUjKHASLGx7FhIeQyGPKb/7y8sdhzCUWOh5X2vo/O7alD08z52+0xwYEL18fGhWbNmxMfH06VLycCGXq8nKSmJDh06lDk/Pj6eXr16ERUVBUDjxo256667eOqppxwVcr2X/fs6TFlpNBrxcIXnmMwWHv67PN+w6FaOCk3UAOYCAwlvP4De4sq8nNsrlSDVwKJ/d8Lb2z6LPtw83PENcHztXYcOSo0bN46PPvqI3r17ExgYyPz58wkJCSEyMrLMuZGRkbzyyiscOHCAbt26kZmZyerVq0s9gxX2VZR2Eq2uIT7hAyo85/MfEyg2Wbj3tvZoZX+oesNcYCA77RzPZ/0DC2rUKnj3oUg83TQ2Xe+shGdvDk2oU6ZMQa/XM2HCBAoKCoiMjGTp0qWo1WpiY2OZOnUqGzZsICgoiGHDhpGens6sWbM4f/48Hh4e9OzZk9mzZzsy5HpLsZjJO7IX357X3itKURR6dWrC3YNkTXtddOmW/kqWonyS/zedHLMbFsby9kM9CGraEG9PVydFWXM4NKGq1WpmzJjBjBkzyhyLioriwIEDpdomTZrEpEmTHBWe+FtxRjLJ/30cAPfgso9jLjl08iLfbjtB367y7LSuMRcYMOdlk/y/6aCUfTaar7jjMfol+CSBBg39JZn+TZaeilLMhXnWZOp/y3i8wnpWeG7i6UwA7hvR0SGxCccw5aST9P5DJYlUpSb4gUWo3Tytxw0FRqbP343lkwTUahVuLrbd5tcHklCFVf7xfaR9NQeAZv96C7emrSs898+j6Xy64TAR7RrV+GpSwnbmAkNJMgWCH1iExsuvzI62ZksBFovCO9NvoUmAp/ROryAJVVhlbFoGQMsnP7vuttArf07E38eNx8dFOCI0cQPKew5a4bmGbFAsBD+wCNeGwRjyiym6aslmlr4IAH8fN0mmV5GEKlAsZtK/fxdT9gUaDn/wuskUSn6purdvTANfDwdEKK6noqR5aQCpvOegFVKp0Xj5XXMJp9zql08SqsCkv4jh0G/49rodn3DbSti5aNV0aV2zK/HXdrb2LK+bNMt5DnotKq0rGg9vinIu39r7+5QuuO3mopHeaTkkoQoKz5YUnPGLHn3dOqfCMUoNDNniGknzUoKsKn8fN7kTsZEkVEFR8hFUWlfUNv7SnUzJISlNLxP5q9mlHqm1xwk29yxvNGmK6iEJtZ7Lif2J3IOb8elyq82FozNzS4paRMv80xty5S19mdt2lZoWDy9B6yslEGsTSaj1mGIxc3HTMtRunvjfMr5S1+q8XHHRSg+1qswFBs68c3/pW/orbtud3eM05BdbR/OF7SSh1lPm/FzOffEKAE0nvobW27Ztk0X1UEzFoFho9q/5aP7+3js7iV5y5ei+jOZXjiTUeqrwbCLGiykEDJiIW2CIzddZLArHk7PtF1g9o/H2Q+sT4OwwSsnSF8nE/SqSe7Z6ypyfg8rFDb8+lSvYfSYtly82JtK9fWM7RVa3mQsMmPSZJRPoayBDfjGPLNiCWq2SZFoF0kOtp7J3rkMxm2w+/2hSFkvW/snF7EI0ahVPju9ux+jqnnKLjajUqLT2T1iG/GKKjGabzr3UO13y9ABJplUgCbUeKk5PwpRzgYbDH7L5mj+PpZNXYOTOAW3oFNpA5qtWQqkBKAcPPFVlwzq1WlVmIr+wjSTUeshSVLI2+1qFo6+25tdjhDTVMeqWNvYKq866cgBK69fYoQNPlzasK2+1U0VkFVTVSUKth3L2lGz9aGsvM+1iHgVFJm6JaGbPsOqcS/NMLz0v1XiXrdzkKLLayTEkodZDisWMV/veNp//1/EMAPqGS0KtyNXr7subqO+M56Uyl9SxJKHWUy4BlVvlFNJUh588VytXhevuHfi89NJE/EcWbCnzvFTmkjqOTQk1MTERjUZD27ZtAdi2bRtr166lbdu2PPjgg2i1kpdrC1NOOvlHduMaPdrZoTjc6tWrWbduHS4uLsyZM4fmzZtbj33xxRd8//33uLi4MGvWLDp16mTz+15ZkPnKdfc3mkRtHZ3PLzRZE6larWLJ0wPwdL/8OynPRB3Hpkz40ksvMWnSJNq2bUtaWhqPPfYYPXr0YPXq1RQVFcnWzrVI1o6vAWwu01dXZGdns2bNGlatWsXhw4dZsGABixYtAiAjI4MffviBlStXkpWVxfTp01m+fHml3r+6191fyMpn6pzNNo/OX0qkUvTZuWxKqKdOnaJDh5LN2n7++We6dOnCsmXL+OOPP3j++eclodYiKrUGn/ABuAQ0ten848nZLF59kJCmOjtHZl9xcXH07NkTrVZL165dOXXqlPVYSkoKbdq0QaPR0LBhQ86fP09xcTGurrYlppbTP6nW23lDfjFT52wGKNPbrIj0QmsGmxKq0WjEza3k+dmePXvo168fACEhIWRkZNgvOlGtFEXBkPgHnm2ibL5mf+IFtBoVL/yrlx0js7+cnBx8fX2trxXlcs+vRYsWHDp0iIKCAtLS0khNTSUnJ4dGjWzrcVZXMr10i3/l5PrmgT7V8t7CMWxKqK1atWLTpk0MHTqUnTt38sADDwCQnp6OTle7ey71iTHjLJb8XNyatLLp/BxDEct/SqCBrzuBAbZVe6+pdDodR44csb5Wqy+vuvb392fq1KlMnTqVoKAg2rVrh7+/v91juvIZ6ZXPQUvik8n1tZFNCfXhhx/miSeeYP78+fTt25cuXboAsGPHDjp2lC2Ea4uMjR8C4NP11uueqygKTy7aDsBbj95sz7AcIjw8nCVLlmA2m0lMTKRly5aljg8bNoxhw4aRlJTE4sWL7T7QWt4KpisHlOQWvnay6W/NoEGD2Lp1K+np6YSFhVnb+/Tpw5AhQ+wWnKg+itlIYdJhfMIH2lQB/mJOIRcy8xlxUysa+9fu3imAn58fo0aN4p577kGr1fLGG2+wbt06goOD6dmzJzNmzCA9PR1fX19eeuklu8dT3gomSaK133UTqtFopH///nzyySfWgalLIiJkC+HaQv/XNgB8e4206fz9Ry4AcPegsOucWXuMHz+e8eMvF9K+spf69ttvOyMkWcFUx1w3obq4uAClnzmJ2ifz1+Vo/Zvg2qh5hefk5hWTmm5g6do4Tqbm4O6qwddbekxC2MqmW/6xY8eyfPlyZs+ebedwhD0Ycy5gKTQQcGvF25zo84u556WfrK/v6NeaCTFhUlWqml05ki/qHpsS6oULF9i4cSO7du2iU6dOeHqWfqb22muv2SU4UT3SvpoDgHuL8lf/HD+bzfT/lDwSmPfwTXQKbeCw2OqTqweiZElo3WNTQk1KSrKO5l+4cKHUMenB1GzGzFSM6WdpOHRqhbf7P+85A8BHLwyuEwNQNZEhv5i0zPxSA1EyCFX32JRQK7sMT9QMisVM6ucvAuBdQe3TIqOZM+dyGRDVXJKpnVy96Z1sLVJ3VWqynV6v58yZM7Rr187mZXnCeXL2bsCcl41f9GjUV5WOyyswsmJjAut3lCzBlNt8+7lyipQk07rNpoRaUFDA7Nmz+e6771CpVPz88880b96cl156iSZNmvDQQ7ZvpSEcx3gxFY1PAwL631vmWPyJDH78/TShzXx57r6etX4llLNdqzLUpQEoKVxS99mUUBctWkRiYiIrVqxg6tSp1vZ+/frx/vvvS0KtgRSzCf2BX/Du2r/c47vi0/Bw07LoyVsdG1gdZMu+TTIAVT/YlFB//vln3nrrLaKiShfVaN26NWfPnrVLYOLGGOJLlo36hJdNqPmFRjbvTaJDSM3aD762smXfJhmAqh9sSqgZGRk0bVq23JvJZMJstm17WuFYxuzzaLz8cG9ettZCUpoegIfHhjs6rDpNVj0Jm5Y/hYaGsn///jLtv/76a6m1/aLmKDj5J25NW5c7re2vEyUlF1s2kUph1UEm6YtLbOqhTp06lVdffZX8/HwAdu/ezapVq1i+fDn/+c9/7BqgqLzijGSKUo/h071s4Rqz2cIXGxMJ0ElpuOoy/Z1t8oxUADYm1OHDh1NUVMR7771HQUEBL7zwAk2bNuX1119n4MD6tZVGbZCz+weAckf3L2QVYLYoPHin3O5Xl09fGiLPSAVQiXmoY8aMYcyYMWRmZqIoCg0ayLzFmsiUexH9wc24BbdH4+5V5vj3v50AoFvb6tv/qL6T56bikkqXkAoICMDd3Z2tW7dy+vRpO4QkbsS5L0vqKgSOeqLc44YCI1EdAnF3k51qhahuNiXUGTNm8PnnnwMl9VHvuusupk2bxogRI9iyZYtdAxSVo1Jr8bvpH+XuwJmRXcDWfcnovOTWVAh7sCmh7t69m+7duwOwZcsWDAYDO3bs4JFHHmHJkiV2DVDYzmTIpvj8KTRefuUef3vlPgDuH2H7nvNCCNvZlFBzcnJo2LAhADt37mTIkCE0bNiQESNGcOLECbsGKGyX/sO7AHi161Hu8fgTFxncswV+svmbEHZhU0INCAggOTkZgN9//51evUq2FC4sLKxUJX+LxcLChQuJjo4mIiKCyZMnk5KSUuH5hYWFzJs3j379+tGtWzcGDx7Mtm3bbP68+sSUe5GCkwfx6tAHra5hmeOFRSYA+kdVXLFfCHFjbBqZGDp0KE899RQhISEYDAb69u0LQEJCQpndI69l2bJlrF+/nhUrVhAYGMi8efOYNm0a3333XZnErCgKDz/8MABffPEFzZs3Jy0tDZPJZPPn1SeW4gIAGg0vv67C8x/sBKCBzt1hMQlR39iUUGfOnEmTJk1ITU3l2WefxcOjZJrIhQsXuOuuu2z+sC+//JIpU6YQGhpqfd/o6Gj27dtHjx6lb1N37tzJ3r172bp1KwEBJWvOmzRpYvNn1Tc5u74v+YOm7I/07Hk9R5OyGd63FUGNvB0cmRD1h00JVavVcv/995dpnzx5ss0fpNfrSUlJoXPnztY2nU5Hy5YtSUhIKJNQd+3aRXBwMEuXLuXHH3/Ezc2N/v378+STT+LlVXZ+ZX1nLsjFq2PfMnVPAd78fC8AE2LaOzosIeoVmxLq3r17r3n86mRYHoPBAJQk0Sv5+PhYj10pKyuLEydO0LdvXzZv3kxWVhaPPPIIb775Jq+++qotYdcbFmMR+Uf3ouseU+5xVxcN9w5tL9OlhLAzmxLqxIkTUalUKMrleo9XFt1ISEi47nt4e5fcaur1+lLter3eeuxKXl5eaDQannrqKdzc3PDw8GDq1Km89tprklCvkn8sFgDfXiPLPa5Rq/B0d3FkSELUSzYl1KtH1o1GI4cOHWLJkiU8/fTTNn2Qj48PzZo1Iz4+ni5dugAlyTQpKYkOHTqUOf/SpoBXJm7ZELB8xRfOoHJ1R+sfWKrdbFFYuvZPEs9kMahnCydFJ0T9YdOcp8DAwFL/BQcHExMTw4wZM3j//fdt/rBx48bx0UcfcerUKfLz85k/fz4hISFERkaWOXfw4ME0aNCAd955h+LiYs6fP8+yZcuIiSn/tra+8wjpikpV+se5/MfDbNp1hpvCgxjYQxKqEPZW6bX8V2rZsiWHDx+2+fwpU6Zw2223MWHCBKKjo0lJSWHp0qWo1WpiY2OJiIggNTUVKLnl//jjj4mPj6dXr16MHTuW7t2729wjrk+yd65FMRWXaTcUGOnXrRkz741Cq7mhH7UQwgZVrpCRmZnJf//7X5o1a2bzNWq1mhkzZjBjxqj39DAAACAASURBVIwyx6Kiojhw4ECptrZt28oW1tdRfLHkHyDdVbVPjSYLm3adYWCP5qjV8qhECEewKaF26tSpzPNLs9mMp6cnCxcutEtgwjbn174FgHuL0uvzjyZlATC0d4ijQ6qxVq9ezbp163BxcWHOnDk0b3551dg333zDihUrUKvVjBgxgkmTJjkxUlFb2ZRQX3vttTKDQw0aNKBr1674+vraLThxfWZDFr69RqLxKD1T4sCRCwC0aiY/H4Ds7GzWrFnDqlWrOHz4MAsWLGDRokXW4//9739Zt24d7u7ujBgxgvHjx+PqKtPMROXYlFDHjBlj7zhEFRSmHMNSYCjTOwWIO55Bj46Bsi3H3+Li4ujZsydarZauXbty6tSpUsdDQ0OtW/y4u7uj0cj3TVSeVBmuxUw5Jb1Qz9BupdpzDEUknM4kprftdRbqupycnFJ3U1fOqQaIiYlh1KhRaDQaJk+eLAlVVIkM/dZyWr9AVNrLk/ZzDEXc+/JGAO4eJDvSXqLT6cjNzbW+vrIYj8Fg4L///S8bN27kl19+4eeff7bONhGiMiSh1mL5x/eXmS41a8kOAF6c3ItG/rLX0SXh4eHs3bsXs9nMoUOHSlVJU6vVuLi44OnpiaurK+7u7uUuhxbieuSWv5Yy5+di+Gsr7s0vrzIzmS2cPW/g/hGd6NlRKnNdyc/Pj1GjRnHPPfeg1Wp54403WLduHcHBwfTs2ZPRo0dz9913o1Kp6NatG+3atXN2yKIWqjChdujQgR07dtCgQQNmzZrF888/X+6ae+EcubElt/WNRj5ibTuWlA1A62AZ2S/P+PHjGT9+vPX1lb3U++67j/vuu88JUYm6pMJbfnd3d+uo57fffktxcdmVOMI5FEUhd/8mtLqGuPhf7olmGwoBCJctooVwigp7qF27duXhhx+ma9euKIrC3LlzcXcvv9r7a6+9ZrcARVnFaacw52WX6p0WFpuY8+m1yywKIeyrwoT65ptv8sEHH3DmzBlUKhVnz57FxaVsCTipAOV4OXvXA+Dd6WZr2697zwIw/9Gby71GCGF/FSbUJk2aMHv2bADat2/P+++/T4MGDRwVl6iAoijkH9+HT/chqK7Y7uREcjY+nq6EtfR3YnRC1G82jfInJibaOw5ho+K0U1gKDLgFtrK2KYrC3sPn6RseJHcMQjiRzdOmTp06xbJlyzh+/DhQUglq8uTJtGrV6jpXiuqUl/g7AN6dL9/aGwqMZBuKaBHo46ywhBDYOLF/586djBw5koSEBMLDwwkPD+fw4cPcfvvt/PHHH/aOUVxBH7cNz7BeqF0vT9pf/9tJAG4KD3JWWEIIbOyhLly4kPHjx/P888+Xan/99dd5++23+frrr+0SnCjLbMjErWmbUm0mi0K3do3w15U/C0MI4Rg29VCPHj1aakL0JRMmTODo0aPVHpQoX8HpvwDwCOlS5pi7qxTzEMLZbEqo3t7epKWllWlPTU2V1VMOVHzhDKDCrUlIqfYffjuJ2aKUe40QwnFsSqiDBw/mxRdf5LfffqOgoICCggK2b9/Oyy+/zODBg+0do7iCW3AYKs3l+cBGk4WCIhORYY2dGJUQAmx8hvrMM88wa9Yspk6dWmpaTkxMjGya50BZO9ei9Q4o1fZ/sUkAhLeT5aZCOJtNCdXLy4vFixeTlJRUatrUlXvyCPsy6bOw5Ofi0/fOUu0rNyUS0lRHcGOZMiWEs1WqfF+LFi1o0UL2d3eGwqR4ALzCepVqN5os9I8MdkZIQoirSIHpWqLgVBwanwC0vqVv7d1ctTRt6OWkqIQQV5KEWksUph5Dq2tYqi3hVCYZ2QVoNfJjFKImkN/EWkKlccGna/9SbRk5BQBEyAi/EDWCJNRawJR7keK0k2Xa07PyCdC5SQ9ViBpCfhNrgfQNSwHwbBtVqv3r/zvmjHCEEBWwaZQ/MzOTN998k99//52LFy+W2dM8ISHBLsGJSxS8u/ZH63N5DurRpCz0+Uaenhh1jeuEEI5kU0J9/vnnOXr0KP/85z9p3Lix1Nx0IEVRKE4/iy64vbXNYlH4z5cHALi5WzNnhSaEuIpNCXXPnj18/PHHhIeH2zsecZWcP77FrL+I1u/ydKnUDANnz+sZ0qvlNa4UQjiaTc9QfX198fKSuY7OUJyRjNYvEJ8ut1rbLj1xmXpHZ+cEJYQol00Jddq0aSxduhSTyWTveMTVVODZOqJU0697S9bvq9Ty6EWImsSmW/6ffvqJv/76i1tuuYXQ0NAyu59+/PHHdgmuvlPMRgxxW9F1jynVvuPPVNo298PNRWqgClGT2JRQmzRpQpMmTewdi7iKMTsdAM+wnta2FT8lcD4zn9v7hTorLCFEBWxKqHPnzrV3HOIqiqKQ/MGjALgHtbW2f7X5KK2CdNx+c2tnhSaEqEClqk2dO3euVPk+6bXajyFuCwCN73wKtXvpAcF/jyq7BYoQwvlsSqgFBQXMnj2b77//3jqpX61Wc/vttzN79mzc3WVzuOpmLjTg0qgF3u37WNuMJjMAWq0scKuK1atXs27dOlxcXJgzZ06per7/+te/MBqNABw4cICdO3fi6+vrrFBFLWXTb+Zbb73Fnj17eO+999i7dy979+5l8eLF7N69m7feesveMdZLhr+2o3bzKNX2y56S0X2dl6szQqrVsrOzWbNmDStWrGDmzJksWLCg1PGPP/6Y5cuX89JLL9GzZ09JpqJKbEqomzZt4tVXX2XgwIH4+Pjg4+PDoEGDeOWVV9i4caO9Y6yXjFlpuDUpPfB06MRFwlr4E9RQNkasrLi4OHr27IlWq6Vr166cOnWq3PPWr1/P8OHDHRydqCtsSqh6vZ7g4LJV4YODgzEYDNUelAC1qzserS6vTDt7Xs/2gyl4ebhc4ypRkZycnFK9zqvrUVzyyy+/MGTIEEeFJeoYm56htm3bltWrV/PMM8+Uav/qq69o27ZtBVeJqjJmX8BsyCrVdjIlB4CZ90Y6I6RaT6fTceTIEetrtbpsX+LgwYO0atUKH5/K7c8VFhZ2w/GJusGmhPr444/z0EMPsW/fPqKiSqobxcbGcvjwYZYsWWLXAOujgpMHAXALamNty9IX0sDXHW9PeX5aFeHh4SxZsgSz2UxiYiItW5atg7B+/XpGjBhRqfe9MkkLYdMt/y233MLatWsJCQlh165d7Nq1i5CQENauXUu/fv3sHWO9k71zLWpPHVpvf2vb+h2nMFvKv00V1+fn58eoUaO45557mDt3LjNmzGDdunXs2bMHALPZzPbt2+nfv/913kmIiqmUih4m1QFhYWG1sgdx9n9PoOseg2/Ubda2J97ZytDeIQztE+K8wIQQQMW5xaETGi0WCwsXLiQ6OpqIiAgmT55MSkrKda+Lj4+nU6dOTJw40QFROpcp9yLG9LOoNDL4JERtU+Ez1M6dO7N9+3YCAgLo1KnTNYtKx8fH2/Rhy5YtY/369axYsYLAwEDmzZvHtGnT+O6778odJAAoKipi1qxZ9OjRA7PZbNPn1GbFF5MB8GxzefBp5aZETiTn4CIT+oWo0SpMqK+99hre3t7WP1dHlf4vv/ySKVOmEBpaMr9y5syZREdHs2/fPnr06FHuNe+88w69e/dGp9NZn3fVdWp3L7Q+l5+fHjmThb+PG/0iyk5dE0LUHBUm1NGjR1v/PGbMmBv+IL1eT0pKCp07Xy6KrNPpaNmyJQkJCeUm1L1797Jlyxa+/fZbli1bdsMx1AamnHQsRQWl2lQqGNa3lfRQhajhbJo2NXDgQL7++mv8/f1Ltefm5jJ69Gh+/fXX677HpQUAOp2uVLuPj0+5iwPy8vJ47rnnmDNnDh4eHmWO11WGuK2o1KXrnGblFjkpGnHgwAHmzZuHi4sLnp6eLFiwAJPJxNNPP01eXh7R0dE8+mhJVbAtW7bwwQcfoFKpeO655+jatatD4vnss8/YuHEjAQEBNGzYkHfeecdh8WRkZPDII4+g1Woxm8288sortGjRgmeffZYLFy7Qtm1bXn75ZdRqNXFxccyZMwdFUZg2bZpdZlSUF098fDxLliyhadOmACxfvhzAPvEoNggLC1MyMjLKtKenpyudOnWy5S2U3NxcpV27dkpcXFyp9mHDhimfffZZmfNffPFF5bXXXrO+Xrx4sXLvvffa9FmXtGvXrlLn1wSpK19RMn9bY339v2/ilBFPfqts2HnSiVHVX2lpaUp+fr6iKIqycuVKZcmSJcq8efOUH3/8UVEURZk6dapy7NgxxWQyKXfccYei1+uVtLQ0Zdy4cQ6LZ/HixcpPP/1U6jxHxWMymRSz2awoiqLs2rVLmTFjhrJixQrlww8/VBRFUWbPnq1s3bpVURRFGTdunJKWlqYYDAbljjvuUEwmk0PiWbt2rbJs2bIy595IPBXllmv2UPfu3Wv984EDB0ot3TObzezcudPmEn4+Pj40a9aM+Ph4unQpKT+n1+tJSkqiQ4cOZc7fsWMHubm5/PDDDwAUFhZiMpno1asXX3/9dalKQXWFpSifgpN/4t68o7Xtz2PpRLRrxLDoVk6MrP4KDAy0/tnFxQWNRsP+/ft5/PHHAbj11lvZu3cvKpWKkJAQvL298fb2xmQyUVRUhJubm93jMZlMLF26lM8//5wJEyYwYsQITp8+7ZB4NJrLd1N6vZ727dsTGxvLI488Alz+/vTu3Ruz2WyNPyQkhNOnT9O6dfXW9S0vHoA1a9awefNmYmJiuO+++ygqKrJLPNdMqBMnTkSlUqFSqazfoCt5enry8ssv2/xh48aN46OPPqJ3794EBgYyf/58QkJCiIwsu5zyq6++KjWq/8knn3Dw4EEWLVpEo0aNypxfF5hySir0e/1dod+QX8yZNL1sFV0DZGVlsXLlSpYtW8YPP/xgLVmp0+lITk4mJyen1OMsnU5HdnZ2qQRor3hUKhWPPvooer2eSZMmERkZ6dB4jh8/zgsvvMC5c+d49913+f33362frdPpyMnJITs7u9SS3kvt9nB1PCEhIdxxxx2YzWamTZtGt27daNq0qV3iuWZC3bZtG4qicOutt/LNN98QEBBgPebi4oK/v3+lRv+nTJmCXq9nwoQJFBQUEBkZydKlS1Gr1cTGxjJ16lQ2bNhAUFBQmaTp7e2Nq6trnS5qXZCUAIBLQMmzno9/OARA/8i61xuvTQoKCnj88cd54YUXCAgIwMPDw9rb0+v1+Pr64uvri16vt16j1+vx8/NzSDyX+Pj40Lt3b44ePUpwcLDD4mnTpg1ffvkliYmJvPjiizRr1ozc3FwaNWp0ze+PvUokXh3PmjVrgJLe68CBAzl8+DDt27e3SzzXTKiX/jVLTEy84Q+CkoIUM2bMYMaMGWWORUVFceDAgQqvvfTgvy4rSjlSUp1fXfJj+XVvEjG9W9LAVwp4O4vJZGL69OlMnDiR7t27AxAZGcm2bdsYMmQI27dv58knn6Rly5acPn2a/Px88vLy0Gg01X57XVE8er0eHx8fTCYTBw8eZOzYsTRv3twh8RQXF+PqWlJfwsfHB3d3d3r06MH27dtp3bo127dv56abbsLd3R2NRsOFCxfw9vbmzJkz5dZTsEc8l74/iqIQGxvL2LFj7RaPTaP8//vf//D392fs2LGl2tesWUN2djZTp0694UDqO0txIYb47eiibkOlUhF3PB2LAjG9W6LRyHQpZ1m/fj2xsbHk5eXx+eefc8sttzB16lSefvppPvnkE3r37m2tuPbII49w//33o1KpmDVrlsPiOXXqFCdOnMBsNjNixAhatWrlsHgOHTrE22+/bb1TffbZZwkNDeXZZ5/lnnvuoXXr1tZ6H88++yyPPfYYiqLw8MMPo9VWagemKsfz8ccfs3PnTlQqFVFRUURHR9svHltGtAYMGKDs3bu3TPu+ffuUgQMH2jwy5mi1aZS/IPmocuL1MUrRhSRFURTlm63HlNuf+k4xmsxOjkwIcbWKcotNXZ8LFy6U++yycePGnD9//sazuuDizx8BoPVrDEBGdiFhLfzRSu9UiFrDpt/WBg0acPTo0TLtR44csduD7vrEnJ9LUeoxdFHDULuUPOf6/rcTmMwWJ0cmhKgMmxLq4MGDmTt3LocPH7a2HTp0iDfffJOYmBi7BVdf5OwumWvrF315ia+ftxt3D2rnrJCEEFVg01PY6dOnk5iYyJ133mmdX5abm0tkZCRPPvmkXQOsD7J/X4dbUFtrQZSTKTlk6YtQqW+8II0QwnFsSqienp4sX76cP/74g0OHSuZGdurUiT59+lznSmELtYc3fn3vtL5evbnk8UqX1g2dFZIQogoqNU+gT58+kkTtQKXWoFKVPH0pNprZGZdKt3aN8HCr/mklQgj7sfk3Nicnh99++43U1FSKi4tLHStvWaqwTWHqccx5OSU1+oCE05kATBjS3plhCSGqwKaEGhcXx9SpU1EUBYPBQEBAABcvXsTd3Z3GjRtLQr0Bxr8r9Lu37ASU7Bfv5eFCh1YB17pMCFED2TTKP3/+fIYMGcKuXbtwc3Nj1apVbNmyhY4dO/LUU0/ZO8Y6z6VhsHW61OY9Z8krMDo5IiFEVdiUUBMSErjvvvtQq9Wo1WqKi4tp0qQJTz31FAsXLrR3jPVKjqGIbm3rZjUtIeo6mxKqRqOxrnNt0KABaWlpAPj7+5Oammq/6OqB3L0/YikutL5WqaBzmwZOjEgIUVU2PUNt164dCQkJtGzZkvDwcD744AMsFgtr1qyxFmIQVWMpKsCrbRQAhUUmDhxNp1OoJFQhaiObeqgPPvigtYf6+OOPk5mZydSpU4mNjeWFF16wa4B1WXFGMsbMVNyalVQrSsvMByCyg32KEgsh7MumHmpUVJQ1oQYHB7Nhwways7Px9fWtlu2l6ytTbgYAXmG9AfhlzxkAQprqKrxGCFFzXbeHajKZ6NatGydOnCjV7ufnJ8n0Bhn+2gYqNWrXkgLSuXnFtG8pFaaEqK2u+5ur1Wpp0qQJFotUPqpuJn0mHn/PPwU4cjqLti38r3GFEKIms6kr9K9//Yv33nuPwsLC658sbGLSZ1J4Jh6P0G5AyXSpcxfzaOTn4eTIhBBVZdMz1F9//ZW4uDj69etHaGgonp6epY5//PHHdgmuLrv0/NQnfCAA+YUmAG4Klx1OhaitbEqogYGBDBkyxN6x1Cs5u74HQO3hRdrFPP49dzMAnu5SEEWI2qrC395vv/2WYcOG4erqyrx58xwZU72gmI34dB+CSqXm8YVbAXjh/p54ebg4NzAhRJVV+Ax11qxZGAwGADp06MDFixcdFlRdZy7MI/9YLGqXktF9s0XhwTu70qtzUydHJoS4ERUm1ICAAP7880+gpAKSTJGqPvlH9wDg23MEZ9JyKSo2y2CUEHVAhbf8Y8aM4cEHH0Sj0aBSqax7a5cnPj7eLsHVVUXnTqBy80TjE8CJoyXl+zq0kuWmQtR2FSbUGTNmMHjwYE6dOsUzzzzD008/jY+PjyNjq5MURSE39ic8w3qhUqnYtj+ZwABPvOXZqRC13jWHlLt27UrXrl3ZtWsXY8aMwdvb21Fx1VmGuC0A+N/0DwDOZ+bRJli24haiLrBpYv/cuXMlmVYDRbGQueULXAKa4hrYCrPZwvnMAqI6NHZ2aEKIaiCLxh0oL+EPzHnZeHe+BZVKxaFTFzGZLTTy87z+xUKIGk8SqgMpxiJcGgThf/NYAExmBR9PF8LbSYV+IeoCSagOlLNnA4rp8n5R5zPzrUtOhRC1nyRUBzFmn6f4wmk8w3pZ27bEnsVFKz8CIeoK+W12EGNGCgD+0WMAMOQXk3A6k7sGtXNmWEKIaiQJ1YHUnjo0Xr4AXMgqAKBv1yBnhiSEqEaSUB0kd99GLAUG6+s9h0t2jg0MkBF+IeoKSagOYikuxKt9b+vr5PMGGgd4opHtToSoM+S32UEUUzFuTVtbX6tU0LtzEydGJISobpJQHcCYeY6i1GOo/i7XZzRZ2Lo/GRQnByaEqFaSUB0g/+RBAHTdBwNw7GwWAJEdAp0WkxCi+klCdYDCswmo3TxBpUZRFF5dtguATqFSsk+IukQSqiOoVPh0G4hKpcKiQF6hiQdGd8HNRePsyIQQ1UgSqp0pFjP5R/ZYX2/adRqA7u2lwpQQdY0kVDszG7L/HuFvA8B3204Q2syXoIZSDlGIukYSqp3lHizZHtojNJzjZ7NJzcgjUnqnQtRJklDtSFEsZP+2GtfAVmg8fDhw9AIAdw2U9ftC1EUOTagWi4WFCxcSHR1NREQEkydPJiUlpdxzDx48yL///W+io6Pp3r07o0eP5ueff3ZkuDcs7/BOABoMvh+AH347SSN/D9zdrrnzjBCilnJoQl22bBnr169nxYoV7Nixg6CgIKZNm4bFYilzbk5ODsOGDWP9+vXExsYybdo0ZsyYQVxcnCNDrjJLUT4Xvv0PKjdP3Ft0BMDHy5V7Yto7OTIhhL04NKF++eWXTJkyhdDQULy8vJg5cyanTp1i3759Zc695ZZbGDVqFAEBAajVamJiYmjbtm2559ZEeUd2A9B03POoVCpS0w0kpelRqVROjkwIYS8OS6h6vZ6UlBQ6d+5sbdPpdLRs2ZKEhITrXn/+/HlOnjxJ+/a1o4eXves7tH6BuAeXxHv4VCYAPTvK6igh6iqHJVSDoaR0nU6nK9Xu4+NjPVaRvLw8Hn30Ufr370+fPn3sFmN1MWalYUw/i1eHy7H++PsptBoV3p6uToxMCGFPDkuol7ah1uv1pdr1ev01t6jW6/VMmTKFRo0a8eabb9o1xupyqTq/X+9R1jYPNy33Du3grJCEEA7gsITq4+NDs2bNiI+Pt7bp9XqSkpLo0KH8RJOVlcWkSZNo2rQpixYtwtW1dvTu8o7uQeXihsbTB4AsfSFxxzPk+akQdZxDB6XGjRvHRx99xKlTp8jPz2f+/PmEhIQQGRlZ5tz09HQmTpxIWFgYCxYsQKutHVONLEUF6A9uxrVRC2vb/sSS+af9I4OdFZYQwgEcmqWmTJmCXq9nwoQJFBQUEBkZydKlS1Gr1cTGxjJ16lQ2bNhAUFAQX331FceOHSM5OZmNGzda32PkyJG8+uqrjgy7UvKOlFSSajT8IWtbbMJ5NGoVfj5uzgpLOJjFYiE1NRWj0Xj9k0WN4+fnh5+fX6XvKlWKotTZMsdhYWEcOXLEoZ+Z9vVb5B/bR+isr6xtryzbRXBjbybf3vkaV4q6JDk5GZ1OV2YQVtR8FouFCxcuoCgKTZs2LfecinKLLD2tRubCPPKP7Ma353BrmyG/mNiE86jl+Wm9YjQaJZnWUmq1msDAQAoKCip/rR3iqbcKju8HQBc11Nr2+1/nALi9X6hTYhJCVJ5KparSILIk1GpUmHIUldYVra4hAIqisPzHBLw8XAjQuTs5OiGEvUlCrSaKoqCP24JXhz6oVCXf1gNH08k2FDFucJhMmRKiHpCEWk0M8dtQigvwaFky8GS2KLy1PBa1Ckbc1MrJ0QkhHEESajXJ/v0bNN4BeHftD8DXvx4lr8DIxGEd0Wrk2yxqrkvzvbdv316m/d1337W+DgsLo2vXrkRERFj/++CDDwAoKChg7ty59O/fn4iICPr06cM///lPh8+ycbbaMVu+hjNmpmLMSMa39+3WW/sffz9Fs0be/GNAWydHJ8T1+fv789Zbb9G3b180moo3j/zwww/p1atXmfa5c+dy7NgxPv/8c5o3b05ubi5//PFHjVmQYzKZ0Gg0dn/0Jl2napB3LBYAvz5jADhzLpfM3CLGDQlzZliihjEX5mHSZ9r9P3NhXqVju/POO8nLy2P16tVV+toOHDjAbbfdRvPmzYGSIkgxMTG0bt26wmvCwsLYvXu39XVycjJhYWEkJycDkJiYyL333ktUVBQ9evRgzJgxnDx50nr+t99+y+23305kZCTDhw9nw4YN1mO7d+8mLCyMDRs2MHjwYLp160Z+fj4rVqxg0KBBREREEB0dzbPPPlulr7ciNeOfj1rMXJhH5ubP0Po2tq7dv5CVD0DvTk2cGZqoQRSLmaT3pqEU5dv9s1RunoQ8+Skqte3blLu7uzNjxgzeeOMNRo4cec2CReWJioriww8/xGKxEBERQYcOHW649sbs2bO56aab+PTTTwE4duyYdW7vunXreO+991i8eDEdO3Zk//79PPDAAwQGBhIVFWV9j02bNrFmzRo8PDw4d+4c8+fPZ82aNbRr1468vDwOHz58QzFeTRLqDSpMKvmBBI59xtr2f7FncXXRyFYnwkql1tDikQ9QjEX2/ywXt0ol00uGDx/O8uXLWbp0KTNnziz3nGnTppV6JLBp0yYaNGjAc889R+vWrdm0aROLFi3CYrEQExPD888/j6+vb5W+DhcXF86dO0dqaiotWrQoVUTpk08+4cEHH7TWV46KimLEiBF88803pRLqU089hZ+fHwAajQZFUTh+/DhBQUF4e3vTo0ePKsVWEfmNv0E5e35ApXHBLTDE2nYiJYdOrQKcF5SokTTuXuDu5ewwKqRSqZg1axb//Oc/GTduXLnnfPDBB+U+Q3VxceHee+/l3nvvxWw2ExsbyzPPPMMbb7zBW2+9VaV45s2bx9KlS5k0aZI1QT/xxBN4enpy5swZ5syZU6qkp9lsLpVMAYKDLxckat68OQsXLmTVqlW89NJLtGrVivvvv59hw4ZVKb7ySEK9ARZjEYVnDuETPsDalpSWy7mMPO4eJDubitqnW7duDB48mAULFlT5PTQaDb169WLo0KHs3LmzwvM8PT1LLe+8cOFCqePNmjXj9ddfB+DMmTM89NBDeHt789hjj9GwYUOmT5/OyJEjrxmLWl16mGjQoEEMGjQIk8nEzz//zIwZM+jcuTMtWrSo4B0qRwalbkDxhSQAfLrHWNuKjGYABvaonh+QEI42Y8YMtm7dytGjR22+ZvHixezdu5e8vDwUGf5gwgAAFrpJREFUReHw4cP88ssvZXqMV+rcuTPr1q2jqKiIjIwM3n///VLH161bR1paGoqi4O3tjUajsSbISZMm8d577/HXX39hsVgoLi4mLi6uVL3lq508eZJt27aRl5eHVqvFx8cHRVHKJN0bIT3UG2C8WDIa6R7Uxtr2w28nKzpdiFohKCiI++67zzrH1Baurq7MmTOHs2fPYjabadiwITExMTz22GMVXvPyyy/z/PPP07t3b4KDg5kyZQo7duywHt+9ezcLFy7EYDDg7e3NgAEDmDJlClCSUP39/Xn55ZdJSkpCo9HQrl27a36e0Whk6dKlHDt2DEVRCAoK4q233ir1WOBGSfm+KlJMRk69OQ7UGkJnlUw1MRQYGf/Cj/ToGMhLk3vb5XNF7XDq1ClatZIVcrXZtX6GUr6vmpkMJbuYNrn7OWtbbl7JCO600V2dEpMQwrkkoVZR1rYvAXBvdnny/jdbTwDg7enilJiEEM4lCbUKFLMRQ/x2PNtEonbzAEqKoWz84zR9ujTF010SqhD1kSTUKihMLnl2EtD/Xmvbz7tOAzAhpr0zQhJC1ACSUKvg3MrXAHAJuLzfTG5+MS2b+BDSVLa9EKK+koRaSYrJCBYTjUY8jEpbcmtvNJlZ8VMivt6yq6kQ9Zkk1EoyHPoNALegy2X5fjuYAshkfiHqO0molWQuzMO1cQtcGzW3tm3ec5bmgT4MiGp+jSuFEHWdJNRKUMwmMjd/isVYbG07npzNXycyCGvh78TIhBA1gSw9rYSc3T8A0GjYNAAsFoUXlpYUf3joHzKZX9ROEydO5MCBA7i4XJ7uFxkZybJlywBYuXIlK1euJDU1FY1GQ/PmzZkyZUq1VmmqKyShVoIh4Xc0Pg3wCOkCwOlzueQVmvj3qC64aCtff1KImuKBBx7g0UcfLdO+fv163n33XZYsWUK3bt0oKioiPj6ewsJCJ0RZPqPRWOofA2eSW/5KUEzF+PW+3fp6x58lg1Ejbw51VkhC2NWBAweIjIwkIiIClUqFu7s7UVFR3HTTTRVec/XmfgADBgxg3bp1AOTm5jJ9+nR69epF9+7diYmJYePGjdZz//zzTyZOnEivXr3o378///nPfzCZTNbjYWFhfPbZZ4wdO5bw8HB27NjBrl27GDNmDJGRkfTq1Ytx48aRk5NTzd+N65Meqo3yj+/HmJGM+u8CwYYCI2t+PUZw48ptFSHqL0OBkaJi0/VPvEFurlq8PaqnxxYVFcXMmTNZuHAhvXv3plOnTlWuwH/JRx99RF5eHr/++iteXl6kpqZa66KePHmS++67jzlz5jB48GDOnz/PQw89hJubGw8++KD1Pb766iveffddQkNDKSoqYvDgwTzxxBOMGTMGo9HIoUOHnNJrlYRqA0WxkPZ1SWVw7y63oigK09/ZCvx/e3cfFVX57QH8i2ZIvGgG4Uumhg2oDCMwIoKpo4maZYpl6lUTEAErpHdDQdBKU3xDChSsMGFpSGkGXP2lsUxLjdIMkExelSuiyOgMMOAM+/7B5VwPIErOwJD7s9YsmfM858w+Z8H2zJw5ewOrFnt0YGSss9Dp6uH34SFUawyfUB/p/hCSV01B1za0L9++fTsSExOF5x9//DG8vLwwZcoUdO/eHampqdi7dy8qKyshl8sRFhYGieSfFVHv1q0blEolCgoK4OjoiH79+gljycnJePbZZzFlyhQADUWmAwICsGnTJlFC9fHxERoAdu/eHd26dUNJSQnKy8tha2sLZ2fnfxTb/eKEeg9q8s8AOi0en/4mbmnr4bP6EG5W1WH6WDvYPGrW0eGxTqBr1y7YscKr3c5Q25JMAWDx4sUtfoYKAAqFAgqFAgCQn5+PyMhIBAYG4vDhw/+oLbOfnx90Oh1WrFiBsrIyjBo1Cu+88w769++PoqIinDx5Ej/++KMwv76+Hk2rjDatYRobG4vt27fD29sb5ubmmDZtGoKCglptiW0InFDvwY2T3wEAzB3ccfi3UtysqsPsifb4r8l83z67dxZm3fT2Vryj2NnZYeHChQgKCsKNGzeEBni3Mzc3R3X1/3d31Wq1qKioEJ6bmZlh6dKlWLp0KZRKJVatWoUPPvgAu3btgo2NDaZPn47Vq1e3GkfTKvv29vbYsGEDAODcuXPw8/NDnz59MHPmzPvZ3Tbji1J3oSk9j5qiP2HlOhkmXR/Cf04VY0BvS06m7IGwd+9eZGRk4Pr1hvq/ZWVl2L17NwYPHtxiMgUaWpscPnwY5eXl0Gg02LBhg+ii0uHDh/H3339Dq9XCzMwMpqamQoKcM2cOMjIycPDgQdTV1UGn06G4uBhHjx69Y4x1dXVITU0VYrSwsECXLl3a/ewU4DPUVlG9Dpd3rQQA9JqwAJpaLXILr/MdUeyB0aNHD3z11VeIjIyERqOBpaUl3NzcWm2PsnDhQpw/fx6TJ0+GlZUVAgICYGtrK4xfunQJn3zyCa5evYqHH34YMplMaMbn5OSEHTt2YMuWLVi5ciW0Wi369euHOXPmtBrnwYMHERUVhZqaGvTs2RPe3t6YNm1aq+sYArdAaYXqjyO4+v2neMzLFz1GTMXxs/+DtYm/YvObY2H3RMv/OzMGcAuUfwNugaJnWtV1PPz4QPQYMRUAsPvQX3ioqwknU8ZYizihtkJ5PBWka7hv/8r1ahRdvskFpBljd8QJ9Q5qinNA2jr0UswDESH0s4b2ts958Ns4xljLOKHegfJYCgDgEYkbSq+qUV5Zgzle9jDv5F97YYwZDifUFpBOi5qibFg4jgEALPu04ex05vinW1uNMfaA44TaAtWZHwAQLJzGobZOhxvqOvhNGwbTblxRit27+vr6jg6B/UO3bt1qdvPAveCE2oRWXYlr/x2PLt3N8cggGd6ObvhCsYdT3w6OjHUmPXv2RHl5ebNbJpnxu3XrFi5dugQbG5s2r8tf7G+iMjMZANA/KAaXr1WhpEyFqZ6D8Pijj3RwZKwz6dmzJ8rKylBYWPiP7ndnHadLly6wtbXFI4+0/W+eE+ptSHsLqj+OwHyoJypqH8LiNf8BAMzxsu/gyFhnY2Jigj59+tx9IvtX4YR6m4ofdwEAejzzCuavOwIAWLnIndtDM8buSbt+hlpfX4+NGzfCw8MDzs7O8PPzQ2lp6R3n5+bmYvbs2ZDJZBg3bhx27txp0Phunvoepv0kSPjxKjR1OgR6O0E+xPbuKzLGGNo5oSYkJOD777/Hrl27cOzYMfTt2xeBgYEtXg1Vq9VYtGgRRo8ejVOnTmHz5s2IiYkRtUrQp9qyQgBAmWQmDp4ohtTOGlM9+Uv8jLF7164Jdffu3Vi0aBGeeuopmJub491330VhYSF+++23ZnMPHTqELl26CO0Phg8fjpdffhnJyckGiU1dcg6pVSMQkdpwxrzS390gr8MY+/dqt4SqUqlQWloKR0dHYZmVlRUGDBiAc+fONZufl5eHoUOHir4L5ujoiLy8PIPEd/aaKY7WDoF8iC12f/gcf+eUMdZm7XZRSq1WA2hIoreztLQUxprOt7S0FC2zsrJqcW5r7O3bdoX+/PdA8vo2rcIYYwDaMaFaWDR0B1WpVKLlKpVKGGs6//a2CUBD+9mW5t7J/dRCZYyxtmq3t/yWlpbo168fsrOzhWUqlQolJSUYMmRIs/kODg7Izc0VXbDKycmBgwOXz2OMGad2vSg1e/Zs7NixA4WFhaiursb69esxcOBAuLq6Npvr5eUFnU6H2NhY1NXV4ezZs0hJSblrKwTGGOso7doCpb6+Hps2bcLevXtRU1MDV1dXREZG4oknnkBWVhb8/f2RlpaGvn0b7pvPzc1FZGQkzp07h0cffRR+fn5YsGBBe4XLGGNt8q/uKcUYY+2Jq00xxpiecEJljDE94YTKGGN6wgkVbS/a0h7S0tIwd+5cuLi4tHhzwt0Kx2g0GoSHh8PNzQ0uLi4ICQmBUqk0aMzr16/H1KlT4eLigtGjRyM0NBSVlZVGHfdnn32GZ599Fq6urhg5ciT8/PxEd+4ZW7wtee2112Bvb4+TJ08Ky37++WdMmzYNMpkMkyZNQnp6umidyspKhISEwMXFBW5ubggPD0ddXZ1B49y6dSuGDBkCZ2dn4fHWW28J453hWN8VMdq2bRspFArKz88ntVpNK1asoOeff550Ol2HxXT06FE6cOAApaSkkEQiEY2pVCoaNWoUbd26lTQaDZ0+fZpGjBhBGRkZwpywsDCaMWMGlZWVkVKpJH9/f1q8eLFBY96wYQPl5ORQXV0dXbt2jXx8fCggIMCo4y4oKCClUklERLW1tbRjxw7y9PQknU5nlPE29e2335Kvry9JJBI6ceIEERFdvHiRnJyc6Ouvv6ba2lo6cuQIOTk50ZkzZ4T1fH19yd/fn5RKJZWVldGMGTMoMjLSoLFGR0fTvHnzWhzrDMf6XnBCJSKFQkFJSUnC8xs3btCwYcPo1KlTHRhVgxMnTjRLqKmpqcIffaN169bR/PnziYiopqaGpFIpZWZmCuMXLlwgiURCpaWl7RM4ER05coScnZ07Tdy1tbX0xRdfkEQiIaVSafTxXr58mcaOHUulpaWihBodHU0vvfSSaO7SpUtp2bJlRNSQcCUSCV24cEEYz8zMJJlMRhqNxmDxtpZQjf1Y36sH/i1/W4u2GIO7FY4pKipCbW0tpFKpMG5nZwczM7N23adffvlFdGebscadmZkJuVwOqVSKtWvXwsfHBz169DDaeAGAiBAaGoqgoCDhe9uN8vLyRL/PTePOy8uDmZkZ7OzshHGpVIqamhoUFhYaNO7s7Gy4u7tDoVDg7bffxsWLF4WYjPVYt8UDX7G/rUVbjMHdCsc0/tt0TnvuU3p6OlJSUrBr1y5hmbHGPW7cOGRlZUGpVGLfvn1C6xJjjRcAkpOTQUR45ZVXmo2p1WoMHjxYtKxp3C3F3DhmKJMmTYK3tzf69u2L8vJybNiwAT4+Pti/f79RH+u2eOATaluLthiDuxWOuX2fevXqJcxpr31KS0tDREQEYmNjMWzYsE4Td8+ePbFgwQKMGDECTz31lNHGW1JSgtjYWOzZs6fFcQsLi2a/z03jbpqEGucbMm6JRCL8bGtri48++ghyuRynT5822mPdVg/8W/62Fm0xBncrHDNw4ECYmpqK9ik/Px81NTUGLy6TkpKCyMhIxMXFwd1dXKTbmONuVF9fD61Wi+LiYqONt/Fs2tvbGyNHjsTIkSMBAEuWLMHKlSvh4OAgiqlp3A4ODqiurkZ+fr4wnp2dje7du2PQoPbrUmFiYgITExMQkdEe6zbr4M9wjcK2bdtowoQJVFBQQFVVVRQWFtbhV/m1Wi1pNBr66aefSCKRkEajIY1GI7r6HBMTQ7W1tfTHH3+Qm5sbpaenC+uHhYXRzJkz6cqVK6RUKikgIID8/f0NGnNiYiK5ubnR2bNnWxw3xrgTExOpvLyciIgqKipoxYoVJJfL6erVq0YZLxFRdXU1Xb58WfSQSCSUnp5OSqWSSkpKyMnJifbu3Ut1dXXCBaemV/kDAgJIqVTSlStXaObMmRQREWHQuNPS0qiiooKIiK5du0bLli0jhUJBKpXKaI91W3FCJSKdTkdRUVHk7u5OMpmMfH196eLFix0aU2pqKkkkkmaPxiu5OTk5NGvWLJJKpTRmzBhKTEwUrV9TUyMkB2dnZwoODqbKykqDxiyRSGjo0KE0fPhw0eP2q7DGFndQUBB5eHiQTCYjT09PCgwMpOzsbKON905u/90gIjp+/Dg9//zzJJVKaeLEiZSWliaaX1FRQcHBweTs7ExyuZzCwsIMeoWfiCggIIBGjhxJTk5ONHr0aHrzzTepqKhIGO8sx7o1XByFMcb05IH/DJUxxvSFEypjjOkJJ1TGGNMTTqiMMaYnnFAZY0xPOKEyxpiecEJlzMCWLVuGhQsXdnQYrB1wQmWMMT3hhMo6BUNXk++M+JgYH06ozCCSkpLw3HPPwdHREaNGjcIbb7whjKnVaoSHh8Pd3R2Ojo7w9vbGsWPHhPFLly7B3t4e3333Hfz9/TF8+HBs2bIFAFBcXIw33ngDcrkcI0aMgK+vL/76669WY5k/fz6WL1+OTz/9FJ6ennBzc8N7772HqqoqYU5Lb8v3798vaj+zdetWTJw4Eenp6fDy8oJMJsOSJUugVqtx6NAhTJo0Cc7OzggODm5W7QkAvvzySzzzzDOQyWQIDg5u1r4jLS0NL774IqRSKcaPH481a9agurpatB+hoaHYvHkzRo8eDYVC0ep+s/bHCZXpXXR0NKKiojB37lwcOHAACQkJGDp0qDAeGhqKY8eOYf369di/fz9cXFwQGBgoqn4EAFFRUXjhhRdw4MABzJ49G9euXcPcuXPRq1cvJCUlYc+ePRg0aBAWLFiA69evtxrTwYMHcePGDezcuRMbN25EZmYm4uPj27xvV69exb59+xAdHY34+Hj8/vvvCA4ORkpKCrZs2YL4+HhkZWUhLi5OtN7Zs2dx4sQJJCQkYPv27cjLy8Py5cuF8W+++QYRERHw8fFBeno6PvnkE/z8889YuXKlaDsZGRm4fv06vvzyS3z++edtjp8ZWEcXE2D/LlVVVSSVSikhIaHF8aKiIpJIJKJWFkRE06dPb9aiIyYmRjQnOjqaXn75ZdGy+vp6mjBhAn3xxRd3jGnevHn0wgsviJaFh4fTrFmzhOfvv/8+vfrqq6I5+/btE7WfiY6OpiFDhggVk4iIIiIiyMHBQbRs9erVNGPGDNG2hw8fTjdv3hSWNVYRaywOolAoKDk5WfT6p06dEtqxNO6Hl5dXh1ZBY6174AtMM/26cOECamtr4enpecdxAJDL5aLlcrkcZ86cES1zcnISPf/zzz+Rk5MDZ2dn0XKNRoPi4uJW42paM/Pxxx8Xfcxwr2xtbUUFjq2trWFtbS1aZmNj0+yM2c7OTlRt3sXFBUDD8bC0tERpaSnWrl2LdevWCXPo/+oWFRcXC8di2LBhojYhzLhwQmVGy8zMTPS8vr4e7u7uCA8Pbza3aWuMprp16yZ63ljY+E7PAUCr1TbbzkMPif9kTExMWtz27YWS76Zx7vLly4Vi0bfr3bu38HPTY8KMCydUpld2dnYwNTXF8ePHW6yk/vTTTwNoqDo/duxYYXlWVtZdOyQ4Ojri22+/Re/evWFqaqrXuB977LFmZ8i5ubl6235+fj7UarXQruP06dMAgMGDB8Pa2hp9+vRBYWEhZs2apbfXZO2P3zswvTI3N4ePjw9iYmKQlJSEwsJC5OXlYdu2bQCAJ598EpMnT0ZkZCR++ukn5Ofn48MPP8Tff/8NPz+/Vrc9b9486HQ6LFmyBFlZWbh06RKysrKwadMm/P777/cVt4eHBwoKCpCUlISSkhJ8/fXXyMjIuK9t3s7ExATvvfcezp8/j19//RWrVq3C+PHjMWDAAABASEgIvvrqK8TGxuL8+fMoKCjADz/80OLZODNefIbK9C4kJAS9evXCzp07sWbNGlhZWYk+M/3oo4+wbt06vPvuu1Cr1ZBIJIiLixO1NW6JtbU19uzZg40bN+L111+HWq2GjY0NXF1dYWNjc18xe3h4ICQkBHFxcYiKioJCocBrr72GVatW3dd2Gzk5OcHV1RW+vr5QqVQYM2aMaNvTp0+HhYUF4uPjERcXh65du6J///6YOHGiXl6ftQ+u2M8YY3rCb/kZY0xPOKEyxpiecEJljDE94YTKGGN6wgmVMcb0hBMqY4zpCSdUxhjTE06ojDGmJ5xQGWNMT/4Xm+GA2h1CaRcAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 360x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import matplotlib\n",
    "matplotlib.rcParams['figure.figsize'] = (5, 5)\n",
    "\n",
    "#     'weight' : 'normal',\n",
    "#     'size'   : 15}\n",
    "# matplotlib.rc('font', **font)\n",
    "#data_timeline.plot()\n",
    "sns.set(context='paper',style='white',font_scale=1.5)\n",
    "\n",
    "font = {\n",
    "    'weight' : 'normal',\n",
    "    'size'   : 11}\n",
    "#ax = sns.histplot(x=list_cores_nfs,bins=50,color='#DE8F05')\n",
    "ax=sns.ecdfplot(x=list_cores_nfs,color=sns.color_palette()[1])\n",
    "ax=sns.ecdfplot(x=list_cores_fs,color=sns.color_palette()[0])\n",
    "\n",
    "plt.xlabel('core number')\n",
    "plt.ylabel('fraction of users')\n",
    "#ax.set(ylim=(0, 1000))\n",
    "plt.legend(['NFS users','FS users'])\n",
    "\n",
    "sns.set(context='paper',style='white',font_scale=1.0)\n",
    "\n",
    "ax2 = plt.axes([0.55, 0.45, .3, .3])\n",
    "sns.ecdfplot(x=list_cores_nfs,color=sns.color_palette()[1], ax=ax2)\n",
    "sns.ecdfplot(x=list_cores_fs,color=sns.color_palette()[0],ax=ax2)\n",
    "\n",
    "ax2.set_title('zoom')\n",
    "ax2.set_xlim([200,350])\n",
    "ax2.set_ylim([0.7,1])\n",
    "plt.ylabel('')\n",
    "#plt.yscale('log')\n",
    "plt.savefig('../Fear_speech_results/kcore_cdf.pdf',bbox_inches='tight',dpi=400)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [],
   "source": [
    "from numpy.random import seed\n",
    "from numpy.random import randn\n",
    "from scipy.stats import mannwhitneyu\n",
    "\n",
    "seed(1)\n",
    "stat, p = mannwhitneyu(list_cores_fs, list_cores_nfs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Statistics=3794736.500, p=0.000000\n"
     ]
    }
   ],
   "source": [
    "print('Statistics=%.3f, p=%.6f' % (stat, p))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:.conda-punyajoy_gpu] *",
   "language": "python",
   "name": "conda-env-.conda-punyajoy_gpu-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}