NLP/Models+K-Means/main/K-Means/KMeans+vectorString.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import json\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.cluster import KMeans\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn import preprocessing\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "outputs": [],
   "source": [
    "data = pd.read_json(r'../..//data/SIR_train_set.json')\n",
    "CVE_ID = data[\"CVE_ID\"]"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "outputs": [
    {
     "data": {
      "text/plain": "      baseScore  impactScore  exploitabilityScore  severity\n0           7.5          3.6                  3.9      HIGH\n1           9.8          5.9                  3.9  CRITICAL\n2           7.5          3.6                  3.9      HIGH\n3           8.1          5.9                  2.2      HIGH\n4           8.8          5.9                  2.8      HIGH\n...         ...          ...                  ...       ...\n5619        7.5          3.6                  3.9      HIGH\n5620        6.1          2.7                  2.8    MEDIUM\n5621        6.5          3.6                  2.8    MEDIUM\n5622        6.5          3.6                  2.8    MEDIUM\n5623        5.4          2.7                  2.3    MEDIUM\n\n[5624 rows x 4 columns]",
      "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>baseScore</th>\n      <th>impactScore</th>\n      <th>exploitabilityScore</th>\n      <th>severity</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>7.5</td>\n      <td>3.6</td>\n      <td>3.9</td>\n      <td>HIGH</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>9.8</td>\n      <td>5.9</td>\n      <td>3.9</td>\n      <td>CRITICAL</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>7.5</td>\n      <td>3.6</td>\n      <td>3.9</td>\n      <td>HIGH</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>8.1</td>\n      <td>5.9</td>\n      <td>2.2</td>\n      <td>HIGH</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>8.8</td>\n      <td>5.9</td>\n      <td>2.8</td>\n      <td>HIGH</td>\n    </tr>\n    <tr>\n      <th>...</th>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th>5619</th>\n      <td>7.5</td>\n      <td>3.6</td>\n      <td>3.9</td>\n      <td>HIGH</td>\n    </tr>\n    <tr>\n      <th>5620</th>\n      <td>6.1</td>\n      <td>2.7</td>\n      <td>2.8</td>\n      <td>MEDIUM</td>\n    </tr>\n    <tr>\n      <th>5621</th>\n      <td>6.5</td>\n      <td>3.6</td>\n      <td>2.8</td>\n      <td>MEDIUM</td>\n    </tr>\n    <tr>\n      <th>5622</th>\n      <td>6.5</td>\n      <td>3.6</td>\n      <td>2.8</td>\n      <td>MEDIUM</td>\n    </tr>\n    <tr>\n      <th>5623</th>\n      <td>5.4</td>\n      <td>2.7</td>\n      <td>2.3</td>\n      <td>MEDIUM</td>\n    </tr>\n  </tbody>\n</table>\n<p>5624 rows × 4 columns</p>\n</div>"
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "columns_1 = ['baseScore', 'impactScore', 'exploitabilityScore', 'severity']\n",
    "train_data = data[columns_1]\n",
    "train_data"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "outputs": [
    {
     "data": {
      "text/plain": "0       CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N\n1       CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H\n2       CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N\n3       CVSS:3.0/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H\n4       CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H\n                            ...                     \n5619    CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H\n5620    CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:C/C:L/I:L/A:N\n5621    CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:N/I:H/A:N\n5622    CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:N/I:H/A:N\n5623    CVSS:3.1/AV:N/AC:L/PR:L/UI:R/S:C/C:L/I:L/A:N\nName: vectorString, Length: 5624, dtype: object"
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vectorString = data['vectorString']\n",
    "vectorString"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "outputs": [
    {
     "data": {
      "text/plain": "{'AV': {'network': 2897, 'adjacent': 5516, 'local': 2334},\n 'AC': {'low': 2659, 'medium': 5396, 'high': 2152},\n 'Au': {'none': 3904, 'single': 2309, 'multiple': 3674},\n 'C': {'none': 3904, 'partial': 7704, 'complete': 3143},\n 'I': {'none': 3904, 'partial': 7704, 'complete': 3143},\n 'A': {'none': 3904, 'partial': 7704, 'complete': 3143},\n 'severity': {'low': 2659, 'medium': 5396, 'high': 2152}}"
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 打开并读取JSON文件\n",
    "with open('../..//data/label_word_ids_CVSS2.json', 'r') as file:\n",
    "    # 解析JSON文件\n",
    "    cvss2 = json.load(file)\n",
    "cvss2"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "outputs": [
    {
     "data": {
      "text/plain": "9"
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 打开并读取JSON文件\n",
    "with open('../../data/label_word_ids.json', 'r') as file:\n",
    "    # 解析JSON文件\n",
    "    cvss = json.load(file)\n",
    "cvss['AV']\n",
    "len(cvss)"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "outputs": [
    {
     "data": {
      "text/plain": "{'AV': {'N': 0.20251660258650822,\n  'A': 0.38559944075498076,\n  'L': 0.16315973435861586,\n  'P': 0.24872422229989513},\n 'AC': {'L': 0.5526917480773228, 'H': 0.4473082519226772},\n 'PR': {'N': 0.44796328169822147,\n  'L': 0.3051061388410786,\n  'H': 0.24693057946069993},\n 'UI': {'N': 0.5477760628595482, 'R': 0.4522239371404518},\n 'S': {'U': 0.8439380911435942, 'C': 0.15606190885640583},\n 'C': {'N': 0.44796328169822147,\n  'L': 0.3051061388410786,\n  'H': 0.24693057946069993},\n 'I': {'N': 0.44796328169822147,\n  'L': 0.3051061388410786,\n  'H': 0.24693057946069993},\n 'A': {'N': 0.44796328169822147,\n  'L': 0.3051061388410786,\n  'H': 0.24693057946069993},\n 'severity': {'low': 0.18472974850632207,\n  'medium': 0.37487842156454076,\n  'high': 0.14950673891899402,\n  'critical': 0.2908850910101431}}"
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "columns_2 = ['AV', 'AC', 'PR', 'UI', 'S', 'C', 'I', 'A']\n",
    "for column in columns_2:\n",
    "    cvss[column] = {k[0].upper(): v for k, v in cvss[column].items()}\n",
    "# 计算每一行的总值\n",
    "summ = {key: sum(values.values()) for key, values in cvss.items()}\n",
    "\n",
    "# 计算每个值除以总值\n",
    "cvss = {\n",
    "    key: {subkey: value / summ[key] for subkey, value in values.items()}\n",
    "    for key, values in cvss.items()\n",
    "}\n",
    "cvss"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lx\\AppData\\Local\\Temp\\ipykernel_58136\\555386878.py:10: FutureWarning: DataFrame.applymap has been deprecated. Use DataFrame.map instead.\n",
      "  train_data_temp = train_data_temp.applymap(transform_value)\n"
     ]
    },
    {
     "data": {
      "text/plain": "     AV AC PR UI  S  C  I  A\n0     N  L  N  N  U  H  N  N\n1     N  L  N  N  U  H  H  H\n2     N  L  N  N  U  H  N  N\n3     N  H  N  N  U  H  H  H\n4     N  L  N  R  U  H  H  H\n...  .. .. .. .. .. .. .. ..\n5619  N  L  N  N  U  N  N  H\n5620  N  L  N  R  C  L  L  N\n5621  N  L  N  R  U  N  H  N\n5622  N  L  N  R  U  N  H  N\n5623  N  L  L  R  C  L  L  N\n\n[5624 rows x 8 columns]",
      "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>AV</th>\n      <th>AC</th>\n      <th>PR</th>\n      <th>UI</th>\n      <th>S</th>\n      <th>C</th>\n      <th>I</th>\n      <th>A</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>N</td>\n      <td>L</td>\n      <td>N</td>\n      <td>N</td>\n      <td>U</td>\n      <td>H</td>\n      <td>N</td>\n      <td>N</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>N</td>\n      <td>L</td>\n      <td>N</td>\n      <td>N</td>\n      <td>U</td>\n      <td>H</td>\n      <td>H</td>\n      <td>H</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>N</td>\n      <td>L</td>\n      <td>N</td>\n      <td>N</td>\n      <td>U</td>\n      <td>H</td>\n      <td>N</td>\n      <td>N</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>N</td>\n      <td>H</td>\n      <td>N</td>\n      <td>N</td>\n      <td>U</td>\n      <td>H</td>\n      <td>H</td>\n      <td>H</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>N</td>\n      <td>L</td>\n      <td>N</td>\n      <td>R</td>\n      <td>U</td>\n      <td>H</td>\n      <td>H</td>\n      <td>H</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      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th>5619</th>\n      <td>N</td>\n      <td>L</td>\n      <td>N</td>\n      <td>N</td>\n      <td>U</td>\n      <td>N</td>\n      <td>N</td>\n      <td>H</td>\n    </tr>\n    <tr>\n      <th>5620</th>\n      <td>N</td>\n      <td>L</td>\n      <td>N</td>\n      <td>R</td>\n      <td>C</td>\n      <td>L</td>\n      <td>L</td>\n      <td>N</td>\n    </tr>\n    <tr>\n      <th>5621</th>\n      <td>N</td>\n      <td>L</td>\n      <td>N</td>\n      <td>R</td>\n      <td>U</td>\n      <td>N</td>\n      <td>H</td>\n      <td>N</td>\n    </tr>\n    <tr>\n      <th>5622</th>\n      <td>N</td>\n      <td>L</td>\n      <td>N</td>\n      <td>R</td>\n      <td>U</td>\n      <td>N</td>\n      <td>H</td>\n      <td>N</td>\n    </tr>\n    <tr>\n      <th>5623</th>\n      <td>N</td>\n      <td>L</td>\n      <td>L</td>\n      <td>R</td>\n      <td>C</td>\n      <td>L</td>\n      <td>L</td>\n      <td>N</td>\n    </tr>\n  </tbody>\n</table>\n<p>5624 rows × 8 columns</p>\n</div>"
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#转换数据\n",
    "def transform_value(val):\n",
    "    return val.split(':')[1]\n",
    "temp = []\n",
    "for i in range(vectorString.size):\n",
    "    part =  vectorString[i].split('/')\n",
    "    list_items = part[1::]\n",
    "    temp.append(list_items)\n",
    "train_data_temp = pd.DataFrame(temp, columns=columns_2)\n",
    "train_data_temp = train_data_temp.applymap(transform_value)\n",
    "train_data_temp"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lx\\AppData\\Local\\Temp\\ipykernel_58136\\1422158133.py:2: FutureWarning: Downcasting behavior in `replace` is deprecated and will be removed in a future version. To retain the old behavior, explicitly call `result.infer_objects(copy=False)`. To opt-in to the future behavior, set `pd.set_option('future.no_silent_downcasting', True)`\n",
      "  train_data_temp.replace(cvss, inplace=True)\n"
     ]
    },
    {
     "data": {
      "text/plain": "            AV        AC        PR        UI         S         C         I  \\\n0     0.202517  0.552692  0.447963  0.547776  0.843938  0.246931  0.447963   \n1     0.202517  0.552692  0.447963  0.547776  0.843938  0.246931  0.246931   \n2     0.202517  0.552692  0.447963  0.547776  0.843938  0.246931  0.447963   \n3     0.202517  0.447308  0.447963  0.547776  0.843938  0.246931  0.246931   \n4     0.202517  0.552692  0.447963  0.452224  0.843938  0.246931  0.246931   \n...        ...       ...       ...       ...       ...       ...       ...   \n5619  0.202517  0.552692  0.447963  0.547776  0.843938  0.447963  0.447963   \n5620  0.202517  0.552692  0.447963  0.452224  0.156062  0.305106  0.305106   \n5621  0.202517  0.552692  0.447963  0.452224  0.843938  0.447963  0.246931   \n5622  0.202517  0.552692  0.447963  0.452224  0.843938  0.447963  0.246931   \n5623  0.202517  0.552692  0.305106  0.452224  0.156062  0.305106  0.305106   \n\n             A  \n0     0.447963  \n1     0.246931  \n2     0.447963  \n3     0.246931  \n4     0.246931  \n...        ...  \n5619  0.246931  \n5620  0.447963  \n5621  0.447963  \n5622  0.447963  \n5623  0.447963  \n\n[5624 rows x 8 columns]",
      "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>AV</th>\n      <th>AC</th>\n      <th>PR</th>\n      <th>UI</th>\n      <th>S</th>\n      <th>C</th>\n      <th>I</th>\n      <th>A</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.547776</td>\n      <td>0.843938</td>\n      <td>0.246931</td>\n      <td>0.447963</td>\n      <td>0.447963</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.547776</td>\n      <td>0.843938</td>\n      <td>0.246931</td>\n      <td>0.246931</td>\n      <td>0.246931</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.547776</td>\n      <td>0.843938</td>\n      <td>0.246931</td>\n      <td>0.447963</td>\n      <td>0.447963</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>0.202517</td>\n      <td>0.447308</td>\n      <td>0.447963</td>\n      <td>0.547776</td>\n      <td>0.843938</td>\n      <td>0.246931</td>\n      <td>0.246931</td>\n      <td>0.246931</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.452224</td>\n      <td>0.843938</td>\n      <td>0.246931</td>\n      <td>0.246931</td>\n      <td>0.246931</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      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th>5619</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.547776</td>\n      <td>0.843938</td>\n      <td>0.447963</td>\n      <td>0.447963</td>\n      <td>0.246931</td>\n    </tr>\n    <tr>\n      <th>5620</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.452224</td>\n      <td>0.156062</td>\n      <td>0.305106</td>\n      <td>0.305106</td>\n      <td>0.447963</td>\n    </tr>\n    <tr>\n      <th>5621</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.452224</td>\n      <td>0.843938</td>\n      <td>0.447963</td>\n      <td>0.246931</td>\n      <td>0.447963</td>\n    </tr>\n    <tr>\n      <th>5622</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.452224</td>\n      <td>0.843938</td>\n      <td>0.447963</td>\n      <td>0.246931</td>\n      <td>0.447963</td>\n    </tr>\n    <tr>\n      <th>5623</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.305106</td>\n      <td>0.452224</td>\n      <td>0.156062</td>\n      <td>0.305106</td>\n      <td>0.305106</td>\n      <td>0.447963</td>\n    </tr>\n  </tbody>\n</table>\n<p>5624 rows × 8 columns</p>\n</div>"
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 替换 DataFrame 中的值\n",
    "train_data_temp.replace(cvss, inplace=True)\n",
    "train_data_temp"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "outputs": [
    {
     "data": {
      "text/plain": "            AV        AC        PR        UI         S         C         I  \\\n0     0.202517  0.552692  0.447963  0.547776  0.843938  0.246931  0.447963   \n1     0.202517  0.552692  0.447963  0.547776  0.843938  0.246931  0.246931   \n2     0.202517  0.552692  0.447963  0.547776  0.843938  0.246931  0.447963   \n3     0.202517  0.447308  0.447963  0.547776  0.843938  0.246931  0.246931   \n4     0.202517  0.552692  0.447963  0.452224  0.843938  0.246931  0.246931   \n...        ...       ...       ...       ...       ...       ...       ...   \n5619  0.202517  0.552692  0.447963  0.547776  0.843938  0.447963  0.447963   \n5620  0.202517  0.552692  0.447963  0.452224  0.156062  0.305106  0.305106   \n5621  0.202517  0.552692  0.447963  0.452224  0.843938  0.447963  0.246931   \n5622  0.202517  0.552692  0.447963  0.452224  0.843938  0.447963  0.246931   \n5623  0.202517  0.552692  0.305106  0.452224  0.156062  0.305106  0.305106   \n\n             A  baseScore  impactScore  exploitabilityScore  severity  \n0     0.447963        7.5          3.6                  3.9      HIGH  \n1     0.246931        9.8          5.9                  3.9  CRITICAL  \n2     0.447963        7.5          3.6                  3.9      HIGH  \n3     0.246931        8.1          5.9                  2.2      HIGH  \n4     0.246931        8.8          5.9                  2.8      HIGH  \n...        ...        ...          ...                  ...       ...  \n5619  0.246931        7.5          3.6                  3.9      HIGH  \n5620  0.447963        6.1          2.7                  2.8    MEDIUM  \n5621  0.447963        6.5          3.6                  2.8    MEDIUM  \n5622  0.447963        6.5          3.6                  2.8    MEDIUM  \n5623  0.447963        5.4          2.7                  2.3    MEDIUM  \n\n[5624 rows x 12 columns]",
      "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>AV</th>\n      <th>AC</th>\n      <th>PR</th>\n      <th>UI</th>\n      <th>S</th>\n      <th>C</th>\n      <th>I</th>\n      <th>A</th>\n      <th>baseScore</th>\n      <th>impactScore</th>\n      <th>exploitabilityScore</th>\n      <th>severity</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.547776</td>\n      <td>0.843938</td>\n      <td>0.246931</td>\n      <td>0.447963</td>\n      <td>0.447963</td>\n      <td>7.5</td>\n      <td>3.6</td>\n      <td>3.9</td>\n      <td>HIGH</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.547776</td>\n      <td>0.843938</td>\n      <td>0.246931</td>\n      <td>0.246931</td>\n      <td>0.246931</td>\n      <td>9.8</td>\n      <td>5.9</td>\n      <td>3.9</td>\n      <td>CRITICAL</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.547776</td>\n      <td>0.843938</td>\n      <td>0.246931</td>\n      <td>0.447963</td>\n      <td>0.447963</td>\n      <td>7.5</td>\n      <td>3.6</td>\n      <td>3.9</td>\n      <td>HIGH</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>0.202517</td>\n      <td>0.447308</td>\n      <td>0.447963</td>\n      <td>0.547776</td>\n      <td>0.843938</td>\n      <td>0.246931</td>\n      <td>0.246931</td>\n      <td>0.246931</td>\n      <td>8.1</td>\n      <td>5.9</td>\n      <td>2.2</td>\n      <td>HIGH</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.452224</td>\n      <td>0.843938</td>\n      <td>0.246931</td>\n      <td>0.246931</td>\n      <td>0.246931</td>\n      <td>8.8</td>\n      <td>5.9</td>\n      <td>2.8</td>\n      <td>HIGH</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      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th>5619</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.547776</td>\n      <td>0.843938</td>\n      <td>0.447963</td>\n      <td>0.447963</td>\n      <td>0.246931</td>\n      <td>7.5</td>\n      <td>3.6</td>\n      <td>3.9</td>\n      <td>HIGH</td>\n    </tr>\n    <tr>\n      <th>5620</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.452224</td>\n      <td>0.156062</td>\n      <td>0.305106</td>\n      <td>0.305106</td>\n      <td>0.447963</td>\n      <td>6.1</td>\n      <td>2.7</td>\n      <td>2.8</td>\n      <td>MEDIUM</td>\n    </tr>\n    <tr>\n      <th>5621</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.452224</td>\n      <td>0.843938</td>\n      <td>0.447963</td>\n      <td>0.246931</td>\n      <td>0.447963</td>\n      <td>6.5</td>\n      <td>3.6</td>\n      <td>2.8</td>\n      <td>MEDIUM</td>\n    </tr>\n    <tr>\n      <th>5622</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.452224</td>\n      <td>0.843938</td>\n      <td>0.447963</td>\n      <td>0.246931</td>\n      <td>0.447963</td>\n      <td>6.5</td>\n      <td>3.6</td>\n      <td>2.8</td>\n      <td>MEDIUM</td>\n    </tr>\n    <tr>\n      <th>5623</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.305106</td>\n      <td>0.452224</td>\n      <td>0.156062</td>\n      <td>0.305106</td>\n      <td>0.305106</td>\n      <td>0.447963</td>\n      <td>5.4</td>\n      <td>2.7</td>\n      <td>2.3</td>\n      <td>MEDIUM</td>\n    </tr>\n  </tbody>\n</table>\n<p>5624 rows × 12 columns</p>\n</div>"
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data = pd.concat([train_data_temp, train_data], axis=1)\n",
    "train_data"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lx\\AppData\\Local\\Temp\\ipykernel_58136\\4130220277.py:2: FutureWarning: Downcasting behavior in `replace` is deprecated and will be removed in a future version. To retain the old behavior, explicitly call `result.infer_objects(copy=False)`. To opt-in to the future behavior, set `pd.set_option('future.no_silent_downcasting', True)`\n",
      "  train_data['severity'] = train_data['severity'].replace(category_replacement)\n"
     ]
    },
    {
     "data": {
      "text/plain": "            AV        AC        PR        UI         S         C         I  \\\n0     0.202517  0.552692  0.447963  0.547776  0.843938  0.246931  0.447963   \n1     0.202517  0.552692  0.447963  0.547776  0.843938  0.246931  0.246931   \n2     0.202517  0.552692  0.447963  0.547776  0.843938  0.246931  0.447963   \n3     0.202517  0.447308  0.447963  0.547776  0.843938  0.246931  0.246931   \n4     0.202517  0.552692  0.447963  0.452224  0.843938  0.246931  0.246931   \n...        ...       ...       ...       ...       ...       ...       ...   \n5619  0.202517  0.552692  0.447963  0.547776  0.843938  0.447963  0.447963   \n5620  0.202517  0.552692  0.447963  0.452224  0.156062  0.305106  0.305106   \n5621  0.202517  0.552692  0.447963  0.452224  0.843938  0.447963  0.246931   \n5622  0.202517  0.552692  0.447963  0.452224  0.843938  0.447963  0.246931   \n5623  0.202517  0.552692  0.305106  0.452224  0.156062  0.305106  0.305106   \n\n             A  baseScore  impactScore  exploitabilityScore  severity  \n0     0.447963        7.5          3.6                  3.9  0.149507  \n1     0.246931        9.8          5.9                  3.9  0.290885  \n2     0.447963        7.5          3.6                  3.9  0.149507  \n3     0.246931        8.1          5.9                  2.2  0.149507  \n4     0.246931        8.8          5.9                  2.8  0.149507  \n...        ...        ...          ...                  ...       ...  \n5619  0.246931        7.5          3.6                  3.9  0.149507  \n5620  0.447963        6.1          2.7                  2.8  0.374878  \n5621  0.447963        6.5          3.6                  2.8  0.374878  \n5622  0.447963        6.5          3.6                  2.8  0.374878  \n5623  0.447963        5.4          2.7                  2.3  0.374878  \n\n[5624 rows x 12 columns]",
      "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>AV</th>\n      <th>AC</th>\n      <th>PR</th>\n      <th>UI</th>\n      <th>S</th>\n      <th>C</th>\n      <th>I</th>\n      <th>A</th>\n      <th>baseScore</th>\n      <th>impactScore</th>\n      <th>exploitabilityScore</th>\n      <th>severity</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.547776</td>\n      <td>0.843938</td>\n      <td>0.246931</td>\n      <td>0.447963</td>\n      <td>0.447963</td>\n      <td>7.5</td>\n      <td>3.6</td>\n      <td>3.9</td>\n      <td>0.149507</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.547776</td>\n      <td>0.843938</td>\n      <td>0.246931</td>\n      <td>0.246931</td>\n      <td>0.246931</td>\n      <td>9.8</td>\n      <td>5.9</td>\n      <td>3.9</td>\n      <td>0.290885</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.547776</td>\n      <td>0.843938</td>\n      <td>0.246931</td>\n      <td>0.447963</td>\n      <td>0.447963</td>\n      <td>7.5</td>\n      <td>3.6</td>\n      <td>3.9</td>\n      <td>0.149507</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>0.202517</td>\n      <td>0.447308</td>\n      <td>0.447963</td>\n      <td>0.547776</td>\n      <td>0.843938</td>\n      <td>0.246931</td>\n      <td>0.246931</td>\n      <td>0.246931</td>\n      <td>8.1</td>\n      <td>5.9</td>\n      <td>2.2</td>\n      <td>0.149507</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.452224</td>\n      <td>0.843938</td>\n      <td>0.246931</td>\n      <td>0.246931</td>\n      <td>0.246931</td>\n      <td>8.8</td>\n      <td>5.9</td>\n      <td>2.8</td>\n      <td>0.149507</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      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th>5619</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.547776</td>\n      <td>0.843938</td>\n      <td>0.447963</td>\n      <td>0.447963</td>\n      <td>0.246931</td>\n      <td>7.5</td>\n      <td>3.6</td>\n      <td>3.9</td>\n      <td>0.149507</td>\n    </tr>\n    <tr>\n      <th>5620</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.452224</td>\n      <td>0.156062</td>\n      <td>0.305106</td>\n      <td>0.305106</td>\n      <td>0.447963</td>\n      <td>6.1</td>\n      <td>2.7</td>\n      <td>2.8</td>\n      <td>0.374878</td>\n    </tr>\n    <tr>\n      <th>5621</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.452224</td>\n      <td>0.843938</td>\n      <td>0.447963</td>\n      <td>0.246931</td>\n      <td>0.447963</td>\n      <td>6.5</td>\n      <td>3.6</td>\n      <td>2.8</td>\n      <td>0.374878</td>\n    </tr>\n    <tr>\n      <th>5622</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.447963</td>\n      <td>0.452224</td>\n      <td>0.843938</td>\n      <td>0.447963</td>\n      <td>0.246931</td>\n      <td>0.447963</td>\n      <td>6.5</td>\n      <td>3.6</td>\n      <td>2.8</td>\n      <td>0.374878</td>\n    </tr>\n    <tr>\n      <th>5623</th>\n      <td>0.202517</td>\n      <td>0.552692</td>\n      <td>0.305106</td>\n      <td>0.452224</td>\n      <td>0.156062</td>\n      <td>0.305106</td>\n      <td>0.305106</td>\n      <td>0.447963</td>\n      <td>5.4</td>\n      <td>2.7</td>\n      <td>2.3</td>\n      <td>0.374878</td>\n    </tr>\n  </tbody>\n</table>\n<p>5624 rows × 12 columns</p>\n</div>"
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "category_replacement = {'HIGH':0.14950673891899402, 'MEDIUM':0.37487842156454076, 'CRITICAL':0.2908850910101431, 'LOW':0.18472974850632207}\n",
    "train_data['severity'] = train_data['severity'].replace(category_replacement)\n",
    "train_data"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "outputs": [],
   "source": [
    "# 复制元数据col\n",
    "new_col = train_data.columns"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "outputs": [
    {
     "data": {
      "text/plain": "(5624, 3)"
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# PCA降维\n",
    "pca=PCA(3)\n",
    "pca.fit(train_data)\n",
    "new_data=pca.transform(train_data)\n",
    "new_data.shape"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "outputs": [
    {
     "data": {
      "text/plain": "          0         1         2\n0  0.013773  1.328624  0.647031\n1  1.539486  0.577848 -0.185040\n2  0.013773  1.328624  0.647031\n3  0.763329 -1.328923  0.107140\n4  1.074067 -0.631252 -0.089019\n5  1.073885 -0.633773 -0.097836\n6  0.616217 -1.754719  0.053234\n7  0.013862  1.327285  0.635674\n8 -0.889005  0.561839 -1.746661\n9  1.073885 -0.633773 -0.097836",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>0</th>\n      <th>1</th>\n      <th>2</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>0.013773</td>\n      <td>1.328624</td>\n      <td>0.647031</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>1.539486</td>\n      <td>0.577848</td>\n      <td>-0.185040</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>0.013773</td>\n      <td>1.328624</td>\n      <td>0.647031</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>0.763329</td>\n      <td>-1.328923</td>\n      <td>0.107140</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>1.074067</td>\n      <td>-0.631252</td>\n      <td>-0.089019</td>\n    </tr>\n    <tr>\n      <th>5</th>\n      <td>1.073885</td>\n      <td>-0.633773</td>\n      <td>-0.097836</td>\n    </tr>\n    <tr>\n      <th>6</th>\n      <td>0.616217</td>\n      <td>-1.754719</td>\n      <td>0.053234</td>\n    </tr>\n    <tr>\n      <th>7</th>\n      <td>0.013862</td>\n      <td>1.327285</td>\n      <td>0.635674</td>\n    </tr>\n    <tr>\n      <th>8</th>\n      <td>-0.889005</td>\n      <td>0.561839</td>\n      <td>-1.746661</td>\n    </tr>\n    <tr>\n      <th>9</th>\n      <td>1.073885</td>\n      <td>-0.633773</td>\n      <td>-0.097836</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 对数据进行预处理标准化\n",
    "scaler=preprocessing.StandardScaler().fit(new_data)\n",
    "data_s=pd.DataFrame(scaler.fit_transform(new_data,y=train_data.columns))\n",
    "data_s.head(10)"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 640x480 with 1 Axes>",
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 元数据散点图\n",
    "plt.rcParams['font.sans-serif']=['SimHei']\n",
    "plt.rcParams['axes.unicode_minus'] = False\n",
    "plt.scatter(data_s[0],data_s[1],c='r',label='散点')\n",
    "plt.savefig(\"cluster1.svg\", dpi=300,format=\"svg\")\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.metrics import silhouette_score\n",
    "from sklearn.cluster import KMeans"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "outputs": [
    {
     "data": {
      "text/plain": "Text(0, 0.5, '$J(C_K)$')"
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": "<Figure size 640x480 with 1 Axes>",
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 确定Kmeans K值\n",
    "inertia=[]\n",
    "for k in range(2,10):\n",
    "    kmeans=KMeans(n_clusters=k,random_state=1).fit(data_s)\n",
    "    inertia.append(np.sqrt(kmeans.inertia_))\n",
    "plt.plot(range(2,10),inertia,marker='s')\n",
    "plt.xlabel('$k$') # K\n",
    "plt.ylabel('$J(C_K)$') # 误差平方和"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 640x480 with 1 Axes>",
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjUAAAGxCAYAAACa3EfLAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8hTgPZAAAACXBIWXMAAA9hAAAPYQGoP6dpAABMXklEQVR4nO3deXhU5cH+8e/MZCF7IAkh7JCEAAGCNbgAyiYuqIAiLv3Zvm21dan1VUBZVATFBRS1r22V2rq3KgKCqCgIKqjgSlgCISFsIQmQBZJMSDLJzPn9EUmlLJlAJmdmcn+ui6udYc6Z24dJcucsz2MxDMNARERExMdZzQ4gIiIi0hxUakRERMQvqNSIiIiIX1CpEREREb+gUiMiIiJ+QaVGRERE/IJKjYiIiPgFlRoRERHxCyo1IiIi4hdUakRERMQvmFZqsrOzmTBhAoMGDWLu3Lk0tlqDYRg8/PDDnHfeeaSnpzNt2jSqq6tbKK2IiIh4O1NKjcPh4Pbbbyc1NZXFixeTm5vLkiVLTrvNsmXL2L17N++99x7/+te/yMnJYcGCBS2UWERERLydKaVm7dq12O12pk+fTteuXZk0aRKLFi067TabN2/msssuo1OnTqSkpHDJJZewb9++FkosIiIi3s6UUpOVlUVaWhohISEApKSkkJube9ptkpKSWL58OcXFxeTn5/PRRx8xePDglogrIiIiPiDAjDe12+107ty54bHFYsFqtVJWVkZUVNRJt5k4cSJvvfUWQ4YMAWDEiBFcc801TX7v0tIKGrl8p8ksFmjXLsIj+/Y3Giv3aazcp7Fyn8bKfRor93lyrI7t2x2mlBqbzUZQUNBxzwUHB1NdXX3KUvP6668TGRnJZ599hsViYebMmcybN49p06Y16b1dLjwy4J7at7/RWLlPY+U+jZX7NFbu01i5z5NjdWzf7jDl9FNUVBSlpaXHPVdZWUlgYOApt1m+fDm33HILHTt2JCEhgcmTJzd6HY6IiIi0HqaUmv79+5ORkdHwOC8vD4fDccqjNAAul4uSkpKGx0VFRTidTk/GFBERER9iyumnQYMGYbfbWbx4MRMmTGDBggUMHjwYm81GeXk5YWFh2Gy247ZJT0/npZdewmazUVtby0svvcTIkSPNiC8iIiJeyJRSExAQwJw5c5g8eTLz5s3DarXyxhtvAPWFZ+nSpfTp0+e4be655x7sdjtPPfUUlZWVDB06lAceeMCM+CIiIuKFTCk1AKNGjWLVqlVkZmaSlpZG27ZtAdixY8dJXx8ZGcm8efNaMqKIiIj4ENNKDUBcXBzDhw83M4KIiIj4CS1oKSIiIn5BpUZERET8gkqNiIiI+AVTr6kRERER3+Z0GWTkl1Gzv5xgl4uBnaKwWZswDXAzUqkRERGRM7Imp5j5a3ZyyO5oeK59eBCTRyYxMjm2xfPo9JOIiIg02ZqcYqa+v+24QgNwyO5g6vvbWJNT3OKZVGpERESkSZwug/lrdp72Nc98lovT1bIrgarUiIiISJNk5JedcITmvx2sqCEjv6yFEtVTqREREZEmKW6k0DT1dc1FpUZERESaJDY8qFlf11xUakRERKRJSiobPwITHxHMwE5RLZDmP3RLt4iIiLjFZRj8/eu9/HPDvkZfO2lEYovPV6NSIyIiIo2qdNQxa8UOPt9ZAsD/O7cz/RIiePbz3OMuGo6PCGbSiERT5qlRqREREZHTyi+rYvLSTHKLjxJoszBjdDJXpXYAYERybP2MwlarZhQWERER7/VD3hGmvr+Nsuo6YsKCeGpsX/p3jGz4e5vVQnrXaGJjIygursBo2alpjqNSIyIiIie1KKOAp3+aRK9PfDhPj0ulfUSw2bFOSaVGREREjlPndPH0Z7ks3lQIwGW943jw0l60CbSZnOz0VGpERESkwZGjtUxdvo0f95dhAf54UQ9+PagzFos518k0hUqNiIiIAJBTZGfK0kwKymsIC7Lx6JjeXJQYY3Yst6nUiIiICJ/nFDNzRRZVtS46R7dh/vhUesaEmR2rSVRqREREWjHDMHj5m328+NVeAAZ1jeaJq/oQFRJocrKmU6kRERFppapqnTzy8Q4+zS4G4IZzOnLPsJ4E2HxzFSWVGhERkVboQHk1k5dmkl1USYDVwv2jkrhmQILZsc6KSo2IiEgrk7G/jPvf38bhqlrahgQyb2xfBnZu2cUnPUGlRkREpBVZtqWQJz/dSZ3LoFdcGPPHp9Ihso3ZsZqFSo2IiEgrUOcyeO7zXN7ZWADAqF6xPHx5CiFePqFeU6jUiIiI+LmyqlpmfLCdb/cdAeC2wd245YKuPjGhXlOo1IiIiPixXSWVTF6ayf4j1YQEWpl9RW9GJMeaHcsjVGpERET81LrcEh76KItKh5OEyGDmj08lOS7c7Fgeo1IjIiLiZwzD4PXv9vPXdbsxgHM6RzH36j60DQ0yO5pHqdSIiIj4kepaJ3NWZvNJVhEA1w5IYMrIRAJ9dEK9plCpERER8ROHKmqYsiyT7Qft2KwWpoxI5LqBHc2O1WJUakRERPzA1sJypizbRkmlg6g2ATx5dV/Su0abHatFqdSIiIj4uA8zD/L4qmwcToPE2FCeHpdK5+gQs2O1OJUaERERH+V0Gfxl3W7e/H4/AMMSY5g9JoWwoNb54711/leLiIj4uIrqOh74cDvr9xwG4HcXdOW2wd2w+tmEek2hUiMiIuJj9pYeZfLSTPYeriI4wMrMy3pxae/2ZscynUqNiIiID1m/p5QZH2zHXuOkfXgQ88en0js+wuxYXkGlRkRExAcYhsFbP+bz5y924TJgQMdI5o3tS0yYf0+o1xQqNSIiIl7OUefiiU9z+CDzIABXp8Yz7ZJkggL8f0K9plCpERER8WLFlQ7uX7aNLYXlWC1wz/BEbjyno9+tsN0cVGpERES81PaDFUxZmskhu4OI4ACeuKoP53dva3Ysr6VSIyIi4oVWZh3ikU+yqalz0b1dCPPH96Nr29Y3oV5TqNSIiIh4EZdh8OJXe3jlmzwAhvRox5wrexMerB/ZjdEIiYiIeAl7TR0zP8pi3a5SAH49qDN3Du2BzarrZ9yhUiMiIuIF9h+pYvLSTHaVHCXIZuGBS3sxpm+82bF8ikqNiIiIyb7bd5jpy7dTVl1HbFgQT4/rS2pCpNmxfI5KjYiIiEkMw+DdjEKe+WwnTgP6dojg6XF9iQsPNjuaT1KpERERMUGt08VTa3by3uYDAFzRpz0zRifTJtBmcjLfpVIjIiLSwg4fdTD1/W1szC/HAvzp4h7cnN5ZE+qdJZUaEZFWwukyyMgvo2Z/OcEuFwM7RemuGhNkH7IzZVkmheU1hAXZeOzKPgzp2c7sWH5BpUZEpBVYk1PM/DU7OWR3NDzXPjyIySOTGJkca2Ky1mVNdhEPr9hBdZ2Lrm1DmD8ule4xoWbH8htaCUtExM+tySlm6vvbjis0AIfs9adA1uQUm5Ss9XAZBn//eg9Tl2+nus7F+d2ieeWXA1VompmO1IiI+DGny2D+mp2nfc0zn+UyLDFGp6I85KjDyayPd/DZT+Xxl+d24k8X9yRA493sVGpERPxYRn7ZCUdo/tvBihoy8ss4t0t0y4RqRQrKqpmyLJOcokoCbRamXZLM2H4dzI7lt1RqRET82N7SKrded/+yTNI6RZHSPrz+T3w4HSKCdTfOWdi4v4z739/Gkapa2oUGMm9sX9I6RZkdy6+p1IiI+BnDMNhcUM6iTYWs2nHIrW3Ka5ys21XasOYQQFSbAJLbh5MSF05KfBgp7cPp1jZUp6nc8N7mQuau3onTZdC7fThPjetLh8g2Zsfyeyo1IiJ+otJRx4pth1i8qZCdxZUNzwdYLdS5jFNuFxcexKNX9CanuJIdh+zsOGRnV8lRyqrr+H7fEb7fd6ThtcEBVnrFhdHr2BGd9uEkxoYRHKD7TgDqnC6e/XwXCzMKABidEsfMy3ppQr0WYlqpyc7OZvr06ezbt4/rrruO+++//7SHOadNm8Z77713wvOrV6+mc+fOnowqIuLVsg/ZWbypkI+3H+JorROoLx+XpsQxYWBHDlbUMPX9bafcfsrIJM7tGs25XaMbnnPUudhVcqzk1P9v9iE71XUuthRWsKWwouG1NquFHu1CSWl/fNkJD25dvzcfqapl+gfbG0rgnUO785vzuugUXguyGIZx6vruIQ6Hg8svv5yhQ4dy6623MmfOHC677DImTJhwym2qqqqora1teLxx40Yef/xxPvroI2w29xtwcXEFzf1fbLFAbGyER/btbzRW7tNYua81jlVNnYvV2UUsyihkS2F5w/Pd2oZwbVoCV6XGE9kmsOH5k81TEx8RzKQRiW7PU+N0GeQdrmo4mnPsT1l13Ulf3ymqzX+u0WkfTkr7MGJ9aE2jpnyucosrmbw0k/yyakIDbTwyJoVhSa1n/h9Pfg0e27c7TKnRa9euxW63M336dEJCQpg0aRKzZ88+bakJCQkhJCSk4fFrr73GXXfd1aRCIyLi6/IOV7FkcyHLtx5oKBM2q4URSTFMSOvIuV2iTnpkYGRyLMMSY+pnFLZaz2hGYZvVQveYULrHhHJZn/ZA/fU7Bytq2HGokuyfFZ0DFTXkl1WTX1Z93Dw4MWFBpLQPO67sdIpq49NHM77YWcLMj7I4WuukY1Qb5o9PJSk2zOxYrZIppSYrK4u0tLSGkpKSkkJubq7b22/evJn9+/dz5ZVXNvm9PfF1c2yfPvw12WI0Vu7TWLnP38eqzmWwLreERRkFfLP3SMPz8RHBXDOgA+P7d3DrCEiAzcKgbtHExERQUtI8v1FbLBYSotqQENWG4ckxDc8fqao9ruTsOFjJ3sNHKal08PVuB1/vPtzw2rAg23FHc1Liw+nRLpQAm7nX6TT2uTIMg1e+yeOFL/dgAOldonhybF+iQwJPvoEf8+TXYFP2aUqpsdvtx10HY7FYsFqtlJWVERXV+O1ub775JjfddBNWa9M/8DEx7h3COhOe3Le/0Vi5T2PlPn8bqwNl1bz93T7e/jaPA+XVQP03+GG94rj5/G6M6N3+jO9E8vRYxQJJXdox5mfPHXXUsb2wgm0FZWQWlJNZUM6OAxVUOpz8uL+MH/eXNbw2KMBKSnwEqR0jSe0URWrHSPp0iCQkqOWPzp9srKocTu5btIkPNhcC8OsLu/HQVX0JNLmImc3sr0FTSo3NZiMoKOi454KDg6murm601Bw5coTVq1fzwAMPnNF7N9dvJz9nsdCsv/n4M42V+zRW7vOnsXIZBt/uPcLiTQWs3VmC86f/nrYhgYzt34FrB3SgU3T9Ue7DpfYm79/sseoWFkC35Biu+OmoTp3Txe7So+w4+J8LknccslPpcLIlv4wt+WXwXR4AVgt0axvacHv5sT9RHjoycqqxOlBezZRl28g6aMdmtXD/qEQmpHWk7HDlqXfm5zz5uTq2b3eYUmqioqLIyck57rnKykoCAxv/YK5atYr09HS3juicjGHgsS9kT+7b32is3Kexcp8vj9WRqlqWbz3Ae5sLyTtS3fD8OZ0imZDWkRHJsQT9dNt0c/w3estY2axWkmLDSYoN58rU+udchkFBWfV/XZBcSUmlg92lR9ldepSPtxc17KNDRHBDwen10yms+GacOPDnY7W5oJz7lmVSerSW6JBA5o7twy86R3vFWHoDsz9XppSa/v378+677zY8zsvLw+FwuFVUVqxYwZgxYxp9nYiItzMMgy2FFSzeVMCnO4pw/HRYJizIxpV947kmLaFVXnBqtVjoHB1C5+gQRvWKa3i+2F5z3NGcHYfs5JdVc6CihgMVNXyRW9Lw2uiQQHrFhdE7/j9lp2vbEKxuFh2ny6i/qHp/ecNF1R9tO8gTn+ZQ6zRIjgvj6XGpdIzShHrexJRSM2jQIOx2O4sXL2bChAksWLCAwYMHY7PZKC8vJyws7KR3NVVXV/Ptt98ye/ZsE1KLiDSPSkcdn2w/xKJNheQU/eeURUr7cCakJXBZ7/aEmnDtiLeLDQ8mNjyYIT3bNTxnr6lrKDjZPx3R2V1SyZGqWr7dd4RvfzZxYEigleS4n12Q3D6cnjFhDUfAjjnZ7e+hgVaO1roAGJEcy6zLU/Rv5IVMKTUBAQHMmTOHyZMnM2/ePKxWK2+88QZQX3iWLl1Knz59Tthu48aNREVF0aVLl5aOLCJy1nYWVbJoUwEfbz9EpeM/k+RdkhLHdWkJpHaI8Olbm80QHhzAuV2ij1uMs6bORe7PZkfOPmQnu6iSqloXmwvK2Vzwn3l9AqwWesSENpy+qnTU8eJXe094n2OF5pJesTx2VR+3j/hIyzJl8r1jioqKyMzMJC0tjbZt27bIe2ryPXNprNynsXKfN49VTZ2LNTlFLM4oZNPPfph2bRvChLQEruwb77ELXU/Gm8fKk5wug72HjzbcXr6jqL7slJ9i4sBTiY8IZtmt52n9q//SqiffOyYuLo7hw4ebGUFExCP2H6liyaZClmce5EhV/WzoNgsMS4plQloCg7pG66hMC7JZLfSMCaNnTBhX/HQiwDAMDlTU/HTnlZ1v9h4+bvmHkzlYUUNGftlxR4bEe7SuhTlERDyozmXwZW4JizcXsmHPfyaXax8exDUDEhjXvwNxPrRMgL+zWCwkRLYhIbINw5Nj6d4ulC2FWY1uV/yza23Eu6jUiIicpSJ7DUu3HGDp5sKGi0stwAXd2zIhLYEhPWMI0OkKrxcbHtT4i5rwOml5KjUiImfAZRh8t+8ISzYV8sXO4oZJ8qJDAhnbL55rBiTQOTrk9DsRrzKwUxTtw4OOu+vpv8VHBDOw05nNkyaep1IjItIEZVW1fJB5kCWbC9l3uKrh+bSOkUwYmMCo5LgTbhEW32CzWpg8Momp72875WsmjUjURcJeTKVGRKQRhmGwtbCCxZsL+XRHETV19bf3hgXZuKJPeyakdSQprvVNkuePRibHMnds3xPmqYmPCGbSiERGJseamE4ao1IjInIKRx1OPs46xOKMArJ/NkleclwY16UlcFmf9oQF6duovxmZHMuwxJj6GYWt1oYZhXWExvvpq1FE5L/sLK5kyaZCPtp2sGGSvCCbhdEpcUxI60i/BE2S5+9sVgvpXaNb5Zw+vkylRkQEcNS5WJNTzOJNBWTk/2eSvC7Rbbg2rSNXpcYT3YKT5IlI06nUiEirtv9IFe9tLuT9rcdPknfxzybJ05T4Ir5BpUZEWh2ny+DLXaUs3lTAhj2HOXZmoX14EOP710+S1z5Ck+SJ+BqVGhFpNYrtNSzbeoD3Nh/gYEVNw/MXdKufJG9ooibJE/FlKjUi4tOcLqP+LpX95Se9S8UwDL7PO8LiTYV8vrMEp6v+uExUmwCu7teBawck0KWtJskT8QcqNSLis9bkFJ8wn0j78CAmj0wivUtU/SR5mwrZ+7NJ8gZ0jGRCWgKjesURrEnyRPyKSo2I+KQ1OcUnnfn1kN3B1Pe3EWC1UPfTUZnQQBtX9G3PhLQEkuPCWzqqiLQQlRoR8TlOl8H8NTtP+5o6l0FSbCjXDezI5ZokT6RV0Fe5iPicjPyy0y46eMzkEYmkd23bAolExBvohLKI+JxiNwoNQEllrYeTiIg3UakREZ8TE+bezL6x4UEeTiIi3kSlRkR8Sp3TxSdZhxp9XXxEMAM7RbVAIhHxFrqmRkR8RkV1HdOWb+PbfUcafe2kEYlaVVmklVGpERGfsP9IFfe+t5U9pVWEBFp5dEwfnIZxwjw18RHBTBqRyMjkWBPTiogZVGpExOtl7C9jyrJMyqrraB8exDPX9COlff18M8MSY+pnFLZaTzqjsIi0Hio1IuLVPtp2kDkrs6l1GvSJD2f++FTiwv+z2KTNaiG9azSxsREUF1dgGKfZmYj4NZUaEfFKLsNgwVd7ePmbPABGJMcy+4oUQgJtJicTEW+lUiMiXqe61snsj7P5NLsIgF8P6sIfL+qO1aLTSiJyaio1IuJViisdTFmaSeaBCgKsFqaPTmZsvw5mxxIRH6BSIyJeY2dRJfe+t5UDFTVEtQlg7ti+nNsl2uxYIuIjVGpExCt8tauUGR9s52itk65tQ3j2mn50bRtidiwR8SEqNSJiund+zOeZz3NxGXBulyjmXt2XqBD3lkIQETlGpUZETFPnMnjms1zezSgAYGy/eKZdkkygTSu4iEjTqdSIiCnsNXXM+GA76/ccxgLcdVEPfjWoMxbd4SQiZ0ilRkRaXH5ZFZPey2RXyVGCA6w8OqY3I7SsgYicJZUaEWlRmwvKmbI0k8NVtcSGBfHMNan0iY8wO5aI+AGVGhFpMZ9sP8Qjn+zA4TRIaV+/5EF8RHDjG4qIuEGlRkQ8zjAM/rF+H39fvxeoX4TykTG9CQ3Skgci0nxUakTEo2rqXDz6yQ4+yapf8uDm9M7cdVEPraQtIs1OpUZEPKb0qIP7lm1jc0E5NquFaaOSGD8gwexYIuKnVGpExCN2lVRy75KtFJTXEBEcwJNX9+G8bm3NjiUifkylRkSa3YY9pUxbvp1Kh5PO0W14dnw/useEmh1LRPycSo2INKtFGQU8vWYnTgPO6RTJvLGpRIdqyQMR8TyVGhFpFk6XwbOf5/LOxvolD65MjWfGJckEBWjJAxFpGSo1InLWKh11PPBBFl/tLgXgzqHd+c15XbTkgYi0KJUaETkrB8qrufe9THYWVxIcYGX2FSmM6hVndiwRaYVUakTkjGUWljNpaSalR2uJCQti/vhUUjtoyQMRMYdKjYickU93FDHr4x3U1LlIjgvjmfGpdIhsY3YsEWnFVGpEpEkMw+CVb/J44as9AAzt2Y45V/YmLEjfTkTEXPouJCJuc9S5eHxVNh9uOwTATb/oxP8O66klD0TEK6jUiIhbjhyt5b73M8nIL8dmgftGJTEhraPZsUREGqjUiEij9pQc5Z73tpJfVk1YkI0nr+7DBd3bmR1LROQ4KjUiclrf7D3MtOXbsNc46RjVhmevSaVnTJjZsURETqBSIyKntGRzIfM+zcFpQFrHSJ4a15e2oUFmxxIROSmVGhE5gdNl8H9rd/HvH/IBuLxPex68tBfBWvJARLyYSo2IHOeow8lDH2WxNrcEgNsGd+OWC7pqyQMR8XoqNSLS4GBFDZPe20p2USVBNgsPX57Cpb3bmx1LRMQtKjUiAsD2gxVMei+T4koH7UIDeXpcKv07RpodS0TEbSo1IsKanGJmfpRFTZ2LnjGhPHtNPzpGackDEfEtKjUirZhhGLz2bR5//XIPABd2b8vjV/UhPFjfGkTE9+g7l0grVet08cSqHJZnHgTg+oEduXdEIgFa8kBEfJRp92dmZ2czYcIEBg0axNy5czEMw63tXC4XN954Iy+//LKHE4r4ryNVtdy1aAvLMw9itcB9IxO5b1SSCo2I+DRTSo3D4eD2228nNTWVxYsXk5uby5IlS9za9q233qKiooJf/epXHk4p4p/2lh7llrcy+HF/GWFBNp65ph/Xn9PJ7FgiImfNlFKzdu1a7HY706dPp2vXrkyaNIlFixY1ut3Bgwd59tlnefDBBwkMDGyBpCL+5Ye8I/zurQz2Ha4iITKYf9w0kCE9tIaTiPgHU66pycrKIi0tjZCQEABSUlLIzc1tdLvHH3+cjh07cuDAAX788Ud+8YtfNPm9PTF/2LF9am6yxmms3NfcY7VsywEeX5WD02XQLyGC+eNTiQnzjyUP9Llyn8bKfRor93lyrJqyT1NKjd1up3Pnzg2PLRYLVquVsrIyoqKiTrrNxo0b+fjjjxk2bBj79u3jhRdeYOjQocycObNJ7x0TE3FW2c3at7/RWLnvbMfK5TKY+0kWC77YBcBVAxJ4emIabQJtzRHPq+hz5T6Nlfs0Vu4ze6xMKTU2m42goON/QwwODqa6uvqUpWbhwoWkpaWxYMECLBYL119/PSNGjODmm2+mZ8+ebr93SUkFbl6T7DaLpf4f0hP79jcaK/c1x1hVOZzMXJHFZzn1Sx7cekFX/jCkG/ayo9ibMavZ9Llyn8bKfRor93lyrI7t2x2mlJqoqChycnKOe66ysvK018kcPHiQiy++uGH9mYSEBNq1a0deXl6TSo1h4LEPpyf37W80Vu4707Eqstcw6b1Msg7ZCbRZeOiyXlzRJ75hn/5Inyv3aazcp7Fyn9ljZcqFwv379ycjI6PhcV5eHg6H45RHaQDi4+OpqalpeFxZWUlZWRnx8fGejCrik3YctPObf20k65Cd6JBAXpg4oKHQiIj4K1NKzaBBg7Db7SxevBiABQsWMHjwYGw2G+Xl5TidzhO2ueqqq1i4cCHr168nPz+f2bNn07NnT1JSUlo6vohX+2JnCbe+ncEhu4Me7UJ55ZcDSet06l8YRET8hSmnnwICApgzZw6TJ09m3rx5WK1W3njjDaC+8CxdupQ+ffoct82QIUOYMmUKs2bNorCwkD59+vDnP/+54XSUSGtnGAb/+iGf//tiFwZwfrdonriqLxFtNHG4iLQOFsPdqXw9oKioiMzMTNLS0mjbtm2LvGdxsWcuYoqNjfDIvv2Nxsp9TRmrOqeLuat3snTLAQAmpCUwZUQiATbTJg1vUfpcuU9j5T6Nlfs8OVbH9u0OU3+Fi4uLY/jw4WZGEPF55dW1TFu+ne/2HcEC3DO8Jzf9opOOYopIq6Pj0iI+bP+RKu5ZspW9h6sICbTy2JV9uCgxxuxYIiKmUKkR8VEb95dx37JMyqrriI8I5pnxqfRqH252LBER06jUiPigDzMPMmdlNnUug74dIpg/ri+x4cFmxxIRMZVKjYgPcRkGL361h1e+yQNgVK9YZl2e4pdLHoiINJVKjYiPqK51MuvjHazOLgbgt+d34fYh3bHqgmAREUClRsQrOV0GGfll1OwvJ9jlonN0CPe/v41tByoIsFp44NJkrkrtYHZMERGvolIj4mXW5BQzf81ODtkdDc9ZLeAyIKpNAPPG9eUXnaPNCygi4qVUakS8yJqcYqa+v+2E510/TWZ12+BuKjQiIqfQOqYbFfEBTpfB/DU7T/ua177bj9OlqU1FRE5GpUbES2Tklx13yulkDlbUkJFf1kKJRER8i0qNiJcobqTQNPV1IiKtTZNKTV1dHUuWLDnta7777jscDn3TFWmq2PCgZn2diEhr06RSY7FYePLJJ3nllVf48MMPyczMxOl0Nvx9cXEx9957L2vWrGn2oCL+LqeostHXxEcEM7BTVAukERHxPU26+8lmq5+1tKSkhO3bt/Pmm2+ye/du0tPTufbaa3n22WcZNWoUl19+uUfCivirJZsKmP9ZbqOvmzQiEZtVk+2JiJyMW6UmKyuLbt26ERISQkREBFOmTGn4u4qKCv70pz9x5513ct555zF79myPhRXxR8u2FPLEp/V3Pf0qvTP9EiKY/1nucRcNx0cEM2lEIiOTY82KKSLi9dwqNbNmzWLbtm0MGDCAo0ePsmHDBvbt28ePP/7I+vXrSU9P5/XXX2fevHksW7aMcePGeTq3iF/4IPMAj63MAeCmX3TiTxf3wGKxMCwptn5GYauVYJeLgZ2idIRGRKQRbpWat99+m9LSUjZs2MDnn3/OnXfeSU1NDb/73e/48MMPCQ8PB+D555/n+uuv5/zzz6dDB03hLnI6H207yCMfZ2MA1w/syL3De2L5aR0nm9VCetdoYmMjKC6uwNDUNCIijXLrQuFFixaxbt06nE4nNpuNtWvXcvnll1NaWsrChQs5ePAgAAsWLGDixImsXLnSo6FFfN3KrEPM/ngHBjAhLYEpIxMbCo2IiJwZt47U9OnThzfffBPDMAgKCuLhhx8mLi6OadOmMXDgQN544w1SUlKoqalhxowZBAXpllORU1mdXcTMj7JwGTCufwfuH5WkQiMi0gzcOlLzww8/0L9/f0JDQ9m9ezdhYWF069YNwzBo164dn3zyCXV1dZSVlR13i7eIHO/znGIe+DALpwFXpcYzY3QyVhUaEZFm4VapycrKIisri/z8fL777jv27dtHbm4uN910Ew6HA4fDQWVlJRdccAHz58/3dGYRn7Q2t4TpH2zH6TK4ok97Hry0lwqNiEgzcqvUzJo1i9DQUIYMGcKAAQMoLCzkpptuYsqUKVRUVDBu3DicTidTpkxhw4YNHDhwwNO5RXzKV7tLmbZ8G3Uug0tT4ph5eYruZhIRaWZulZqXXnqJ2tpaIiMjSUxM5LHHHuOee+6hd+/eREVFsWTJEu69916sVitjx45l6dKlHo4t4js27Cnl/mWZ1DoNRvWKZfaY3gSo0IiINDu3LhS+4447sFqtHDhwgMTERPr378/vfvc7bDYbv/71r4mKiuLCCy8E4OqrryYiIsKjoUV8xbd7DzNl2TYcToPhSTHMUaEREfEYt0qN1Vp/QKdDhw4N889cc801ANx6663HvTYhIaE584n4rB/yjjBpaSY1dS4u6tmOx6/qQ4CtScutiYhIE7j9Hdblcp3w3IYNG/j4448BKC0t5cUXX2y+ZCI+LGN/Gfe+t5WaOhdDerTjyav7EqhCIyLiUW59l3344YfZsWPHcc8dPnyYGTNmYBgGTqeTu+++m6+++ora2lqPBBXxFZsLyvnfJVupqnVxQbe2zB3bl6AAFRoREU9z6zvtxo0bqaiooLS0lPXr1wMwZ84camtrueKKK3juueeoqKjgxRdfJDAw0KOBRbxZZmE5dy/ewtFaJ+ldo3lqXF+CVWhERFqEW99tBw4cSHFxMevWreMf//gHtbW1dOnSpaHA5Ofn8+c//5mwsDCPhhXxZtsPVnDX4i1UOpz8onMUz45PpU2gzexYIiKthlulJjExkYMHDzJ27FiqqqrIzs7mnnvuwWKxYBgGzzzzDN27d/dwVBHvteOgnbsWbcFe42Rgp0ievaafCo2ISAtz6+6nLl268O2331JVVcWjjz7Kk08+yUsvvURoaCj3338/AQH1u2nTpg3XXnst/fv392hoEW+SU2Tnj4s2U15dR/+ESJ67th+hQSo0IiItza0jNR06dGDp0qU89NBDJCYmsm7dOn744QdCQkJITU0lLS2NtLQ0ysrKePbZZz2dWcRr5BZXcue7WyirriO1QwT/N6EfYUFu/a4gIiLNzK3vvhEREYSHh7N7925WrFhB165deeuttwgLC2Ps2LG0a9cOgI4dOzJv3jyPBhbxFrtLjnLnu5s5UlVLn/hwnp/Qn/BgFRoREbO4daQmMjKSkJAQXnnlFf7+979jsVjo2bMnOTk51NTUkJOTA0B6ejrvv/++RwOLeIO9pUe5493NlB6tpVdcGM9P6E9EGxUaEREzuVVq2rRpQ1VVFVFRUfz2t7/lyJEjjBkzhpKSEux2O1OnTuWFF14gNDS0YfZhEX+Vd7iKO97dTEmlg6TYMP46cQBRIZrKQETEbG79ahkQEIDdbgfg4osvJjAwkO7duxMeHk5paSnz58/nD3/4A7GxsUycONGjgUXMlF9WX2iK7A56xoTyt4n9iVahERHxCm6VGpvNxj333ANAdHQ0V1xxBQBLly6lU6dOAPzjH//g+uuvZ8iQIXTs2NEzaUVMVFhezR0LN3Owoobu7UL428QBtA0NMjuWiIj8xO1zRTfeeOMJzx0rNADdunVj0aJFKjTilw6UV3P7ws0UltfQtW0IL0wcQEyYCo2IiDdp1gtgunTp0py7E/EKhypquPPdzRSUVdM5ug0vTBxAbHiw2bFEROS/uFVqamtrefPNN9m7dy/33Xcf27Zta/i7d955h+HDh3Prrbd6LKSIWYrtNdzx7mbyjlTTMaq+0LSPUKEREfFGbl1TY7FYeOeddxgyZAgHDhzgiSeeoLa2ltjYWOrq6vjHP/7BHXfcwZdffsnQoUM9nVmkRZRUOrjz3S3sO1xFh4hgXpg4gA6RbcyOJSIip+D23U82W/207+eccw6TJk1i7969ZGZmMmbMGAAmTZpEcnIyNTU1BAfrN1nxbYePOrjz3c3sLj1K+/AgXrh+AB2jVGhERLyZ27OFWSwWLBYLb775JgsXLmxYzPLRRx8FwDAMDMOgtraWH3/80WOBRTztSFUtf1y0hV0lR4kLD+LF69PoHB1idiwREWlEk6ZATUhIYOHChcTFxQH1k/LV1tZisViw2+3Ex8d7JKRISymvruWuRVvIKaokJiyIv00cQJe2KjQiIr6g0QuFd+/eze233w5AUVERr776KrfccguXX345K1euZOjQoXz99dc8+eSTHg8r4kkV1XXctWgLOw7ZaRcayAsTB9C9XajZsURExE2NlpqQkBAuvPBCAFatWkVwcDCGYXD//fdjGAbt2rXjzTffZNu2bSxcuJCtW7d6PLRIc7PX1HH3ki1sP2gnOiSQv04cQI8YFRoREV/SaKnp0KED//M//4NhGGzYsAH4z/U1UH9KasiQIZSUlLBp0yZuu+02zyYWaWaVjjr+d8lWthZWENUmgL9e15+k2DCzY4mISBO5PfmexWJh9uzZDRcEr1mzBpfLRXV1NX/4wx/o2rUr06dP54orrqCqqsqTmUWaTVWtk3uXbGVzQTkRwQH89boB9GofbnYsERE5A026ULi2thaHw8H/+3//j9LSUnr27Em/fv2A+iM2W7du5cEHH/RIUJHmVl3r5N73trIxv5zwYBt/ua4/KfEqNCIivsrtUlNQUMBLL72E1Wpl8+bNWK1WPvjgA6Kjo3n77bcZPXo0vXr18mRWkWZTXetk0tJMfsgrIyzIxvMT+tO3Q4TZsURE5Cy4XWpuvPFGgoKCCAwMJCAgAIvFgsPhoKKigh9++IHs7GweeOABbrjhBmbOnOnJzCJnpabOxX3vb+O7fUcIDbTx52v70S8h0uxYIiJyltwuNffee2/D/9+2bRt9+/Y94TU5OTm6nka8mqPOxdT3t7Fhz2HaBFh57tp+pHWKMjuWiIg0gyav0r1+/XpuvPFGDh06hMPhaHje6XTy2GOPcejQoWYNKNJcap0upn+wna92lxL8U6E5p7MKjYiIv2hSqbHb7cyaNYs//elP1NbWMnbsWMrKygA4evQow4cPZ8aMGR4JKnI26pwuZnywnbW5JQQHWJk/PpVzu0SbHUtERJqR26WmtLSUW2+9leHDh/P73/+eTp06kZiYyMsvvwxAREQEv/nNbwgP190j4l3qXAYPfZTF5ztLCLRZeGpcX87v1tbsWCIi0szcKjUPPfQQEydOZOLEiUyfPr3h+UceeYTly5djt9sbnjs2KZ+IN3C6DGatyOLT7GICrBaeGpvKhd3bmR1LREQ8oNFSc+TIEQ4fPszBgwf59ttvqaiooLS0lOzsbMrLyxk8eDCvv/46u3fvZsuWLQ2no0TM5nQZPPLJDj7JKsJmtfDk1X0Z0lOFRkTEXzV691N0dDR/+ctfKCws5JFHHuHqq6/m0ksv5euvv8Zms2GxWNi6dSsrV67EYrEwbty4lsgtclouw+Cxldl8tO0QNgs8flUfhiXFmB1LREQ8yO1bumNjYzn//PNJSkrirbfe4p133gEgMTGR77//np49e9KunX4LFvO5DIMnVuWwPPMgNgvMubIPI5NjzY4lIiIe5vaFwlarlZdffpnJkyczdepUunXrxp133gnApk2b2LhxY5PeODs7mwkTJjBo0CDmzp2LYRiNbnP11VeTkpLS8OeBBx5o0nuK/zMMg3mrd7J0ywGsFph9RW8uSYkzO5aIiLQAt47UlJWVERUVRWBgIIZh8P7777Ny5UoOHTrE73//ew4cOADAv//9b2pqanjzzTdPuz+Hw8Htt9/O0KFDefbZZ5kzZw5LlixhwoQJp9ymqqqKvLw81q9fT0BAfeygoCB3/zulFTAMg/mf5bJ4UyEW4OHLU7isT3uzY4mISAtptNQYhsGVV15J165dqaqqIj8/n1tvvZWgoCByc3P5/e9/T3Z2NuvXr+fmm2+mtra20Tddu3Ytdrud6dOnExISwqRJk5g9e/ZpS822bdtISUk561Ncnrg569g+deNX4zw1VoZh8NwXu3hnYwEAD13WiytT45v3TVqYPlfu01i5T2PlPo2V+zw5Vk3ZZ6OlxmKxsGbNGn788UfuuOMOfvOb35CcnMxtt92Gy+XivPPOo127dnz55ZdceOGFbr1pVlYWaWlphISEAJCSkkJubu5pt9myZQsHDhzgggsuoK6ujiuvvJIHHnigyUdrYmI8t2ihJ/ftb5pzrAzD4MmPs/j3D/kAPHFtf246r2uz7d9s+ly5T2PlPo2V+zRW7jN7rNw6/RQUFMQ555xDZGQkq1atYsWKFUyZMoWioiKcTifR0dHs2rXL7Te12+107ty54bHFYsFqtTac5jqZXbt2ce6553LXXXdRUVHBlClTePXVV/nDH/7g9vsClJRU4MblO01isdT/Q3pi3/6mucfKMAz+9uUeXvkmD4BplyQxumdbiosrzn7nJtPnyn0aK/dprNynsXKfJ8fq2L7d4fbdTzabjalTp2KxWBgzZgwXXXQRpaWl2Gw22rZty5w5c9wOaLPZTjjCEhwcTHV19SlLzSOPPHLc4z/+8Y+8/vrrTS41hoHHPpye3Le/aa6x+vvXexsKzZQRiUxI6+h3/wb6XLlPY+U+jZX7NFbuM3us3L77KSAggDFjxjQ8joiIoFu3bkB9STnvvPPcftOoqChKS0uPe66yspLAwEC399GuXTstntnK/XPDXl5avw+Ae4f35IZfdDI5kYiImKnJq3Q3h/79+5ORkdHwOC8vD4fDccqjNAA33HADhYWFDY8zMjLo2LGjJ2OKF3vt2zxe/GovAHdf3INfntu5kS1ERMTfmVJqBg0ahN1uZ/HixQAsWLCAwYMHY7PZKC8vx+l0nrBNUlISM2fOZNOmTbz33nu88sor3HTTTS0dXbzAm9/v5y/rdgNw59Du/GpQF5MTiYiINzCl1AQEBDBnzhweffRRzj//fFavXs2UKVOA+sKTnZ19wjZTp04lKCiIX//61zz//PPcd999XHPNNS0dXUz21o/5/PmL+ovS/zC4G78933/uchIRkbPj9oXCzW3UqFGsWrWKzMxM0tLSaNu2LQA7duw46esjIyP561//2pIRxcss3FjAM5/V3/r/uwu68vsLu5mcSEREvIlppQYgLi6O4cOHmxlBfMSSzYU8tWYnAP9zXhduH6xCIyIixzPl9JNIU7y/5QBPrMoB4P+d25k/Du2ORVN8iojIf1GpEa/2QeYB5qysv8bqxl904n+H9VChERGRk1KpEa+1YvtBHvk4GwOYOLAjk4b3VKEREZFTUqkRr7Qy6xCzVuzAAK4dkMCUkYkqNCIicloqNeJ11mQXMfOjLFwGjOvXgamXJGFVoRERkUao1IhX+WJnMTM+zMJpwJWp8cy4NFmFRkRE3KJSI15jXW4J05Zvx+kyuLxPex66tJcKjYiIuE2lRrzC17tLmbp8G3Uug9EpcTx8eQo2qwqNiIi4T6VGTPfNnsPctyyTWqfByORYHhnTmwAVGhERaSKVGjHVd/sOM3lZJg6nwfCkGB67UoVGRETOjKnLJEjr4nQZZOSXUbO/nGCXC5fLYNLSTGrqXAzt2Y7Hr+pDgE09W0REzoxKjbSINTnFzF+zk0N2xwl/d2H3tsy9ui+BKjQiInIWVGrE49bkFDP1/W2n/Psr+8YTFKBCIyIiZ0c/ScSjnC6D+T+trn0qz6/bjdNltFAiERHxVyo14lEZ+WUnPeX0cwcrasjIL2uhRCIi4q9UasSjihspNE19nYiIyKmo1IhHxYYHNevrRERETkWlRjxqYKcoYsNOX1jiI4IZ2CmqhRKJiIi/UqkRj7JZLXSMDD7tayaNSNSSCCIictZUasSjvthZwubCCqxAu9DA4/4uPiKYuWP7MjI51pxwIiLiVzRPjXiMvaaOeatzALh5UGfuHNqjfkZhq5Vgl4uBnaJ0hEZERJqNSo14zN++3MMhu4PO0W34/YXdsFktpHeNJjY2guLiCgxNTSMiIs1Ip5/EIzYXlLMoowCA6Zck0ybQZnIiERHxdyo10uxqnS7mrMzGAK5Kjee8bm3NjiQiIq2ASo00u1e/zWN3yVHahQbyv8N6mh1HRERaCZUaaVa7S47yyjf7AJg8IpHokMBGthAREWkeKjXSbFyGweOrsql1Ggzp0Y7RKXFmRxIRkVZEpUaazXubC8nILyck0Mq0S5KwWHS7toiItByVGmkWhypqeH7tbgDuHNqDDpFtTE4kIiKtjUqNNIun1uyk0uGkX0IEEwd2NDuOiIi0Qio1ctbW5BTz+c4SbFYLD4zupVmCRUTEFCo1clYqqut4avVOAP5nUGeS4sJMTiQiIq2VSo2clb+s201xpYOubUP43QXdzI4jIiKtmEqNnLGN+8tYsrkQgBmjkwkO0MdJRETMo59CckZq6lw8tjIbgPH9O3Bul2hzA4mISKunUiNn5JVv9rH3cBUxYUHcfbGWQhAREfOp1EiT7Syu5LVv8wC4b2QiEW0CTE4kIiKiUiNN5HQZPL4ymzqXwcWJMYxMjjU7koiICKBSI020eFMBWworCAuycf8oLYUgIiLeQ6VG3HagvJq/rtsDwB8v6kF8RLC5gURERH5GpUbcYhgGc1fv5GitkwEdI5mQlmB2JBERkeOo1IhbPs0u5stdpQRYLTxwaTJWnXYSEREvo1IjjSqvruXpNfVLIfz2/C70jNFSCCIi4n1UaqRR//fFbkqP1tKjXSi/Oa+r2XFEREROSqVGTuv7fUdYtvUAUL8UQpCWQhARES+ln1ByStW1Th5fVb8UwoS0BAZ2jjI5kYiIyKmp1Mgp/XPDPvKOVBMXHsRdF/UwO46IiMhpqdTISWUfsvPG9/sBuH9kEuHBWgpBRES8m0qNnMDpMnhsVQ5Ol8HI5FiGaykEERHxASo1coJ3Nuaz7UAF4cE27huZaHYcERERt6jUyHEKyqp58as9APzp4p7EhmspBBER8Q0qNdLAMAye/DSHqloX53SOYnz/DmZHEhERcZtKjTT4JKuI9XsOE2SzMGO0lkIQERHfolIjABypquWZz3IB+N0FXeneLtTkRCIiIk2jUiMAPPfFLg5X1ZIYG8qvB3UxO46IiEiTqdQI3+w5zIeZB7EAD4zuRaBNHwsREfE9+unVylXXOnni0xwArj+nI/07RpqcSERE5Myo1LRyf/96L/ll1cRHBHPH0O5mxxERETljppWa7OxsJkyYwKBBg5g7dy6GYbi9bXl5OUOHDmX//v0eTOj/sg5W8K8f6sdw6qgkwoK0FIKIiPguU0qNw+Hg9ttvJzU1lcWLF5Obm8uSJUvc3n7evHkUFRV5MKH/q3MZzFmZg8uA0SlxXJQYY3YkERGRs2JKqVm7di12u53p06fTtWtXJk2axKJFi9za9rvvvmPNmjVER0d7NqSfe+uH/ew4ZCeyTQCTR2gpBBER8X2mnG/IysoiLS2NkJAQAFJSUsjNzW10O4fDwcyZM3nwwQd5+umnz+i9PTGf3LF9+spcdfuPVLHg670A/O+wnsSGB7XYe/vaWJlJY+U+jZX7NFbu01i5z5Nj1ZR9mlJq7HY7nTt3bnhssViwWq2UlZURFRV1yu1efPFFunfvzpgxY8641MTERJzRdmbvu7kYhsE9S7dRU+dicGIMvxuehMWEr1hfGCtvobFyn8bKfRor92ms3Gf2WJlSamw2G0FBxx8dCA4Oprq6+pSlJjc3l7fffpv33nvvrN67pKSCJlyT7BaLpf4f0hP7bm4fZB7ky53FBAdYuW94T0pK7C36/r40VmbTWLlPY+U+jZX7NFbu8+RYHdu3O0wpNVFRUeTk5Bz3XGVlJYGBgSd9vWEYPPTQQ9xzzz3Ex8ef1XsbBh77cHpy382h9KiDZ39aCuHWC7rSOTrEtLzePlbeRGPlPo2V+zRW7tNYuc/ssTLlQuH+/fuTkZHR8DgvLw+Hw3HKozQFBQX88MMPzJs3j/T0dNLT0ykoKGDs2LEsX768hVL7vmc+y6Wsuo7kuDBuTu/c+AYiIiI+xJQjNYMGDcJut7N48WImTJjAggULGDx4MDabjfLycsLCwrDZbA2vj4+PZ/Xq1cft45e//CXPPPMMffr0aen4Pumr3aV8klWE1QIPXtqLAC2FICIifsaUUhMQEMCcOXOYPHky8+bNw2q18sYbbwD1hWfp0qXHlZWAgIDjLiw+9lyHDh0ICwtr0ey+6KjDydyflkK48Red6NtBF72JiIj/MW0K2VGjRrFq1SoyMzNJS0ujbdu2AOzYscOt7desWePJeH7lxa/2UFheQ0JkMLcN7m52HBEREY8wdV78uLg4hg8fbmYEv5d5oIJ3NuYDMO2SZEKDbI1sISIi4pt0YYUfq3O6eGxlNi4DLu/TnsE92pkdSURExGNUavzYm9/vJ6eokqg2AUwa3tPsOCIiIh6lUuOn9h2u4qX19UshTBqRSNvQllsKQURExAwqNX7IMAyeWJWNw2lwfrdorujT3uxIIiIiHqdS44eWbz3I93llBAdYmXZJsilrO4mIiLQ0lRo/U1zp4LkvdgFw2+BudI4OMTmRiIhIy1Cp8TPz1+RSUVNH7/bh3HSulkIQEZHWQ6XGj6zLLeHT7CJsx5ZCsOq0k4iItB4qNX6i0lHHkz8thfDLczuTEh9uciIREZGWpVLjJ/62bg+H7A46RbXhD4O7mR1HRESkxanU+IHNBeW8m1EAwPTRybQJ1FIIIiLS+qjU+Ljan5ZCMIArU+M5v1tbsyOJiIiYQqXGx732bR67So7SNiSQe4ZpKQQREWm9VGp82J6So7z8zT4AJo9IJDok0OREIiIi5lGp8VEuw+DxVdnUOg0G92jLpb3jzI4kIiJiKpUaH7V0cyEb88sJCdRSCCIiIqBS45OK7DX839rdANwxtAcJkW1MTiQiImI+lRof9NSaXCodTlI7RHD9wI5mxxEREfEKKjU+5rOcYj7LKcZmtfDApcnYtBSCiIgIoFLjU+w1dcxbvROAXw/qTHKclkIQERE5RqXGh/xl3W6KKx10bRvCLRdoKQQREZGfU6nxERn7y1i8qRCAGaOTCQ7QP52IiMjP6SejD3DUuXhsVTYA4/p14Nwu0eYGEhER8UIqNT7glW/2sae0inahgdw9rIfZcURERLySSo2Xyy2u5NVv8wC4b2QSkW20FIKIiMjJqNR4MZdh8NjKHOpcBhf1bMeoXrFmRxIREfFaKjVebFFGIVsKywkLsjFVSyGIiIiclkqNlzpYUcPfvqxfCuHOoT2Ijwg2OZGIiIh3U6nxQoZhMPfTHCodTvonRHLdwASzI4mIiHg9lRovtDq7mHW7Sgn4aSkEq047iYiINEqlxsuUV9fy1Jr6pRB+c14XEmPDTE4kIiLiG1RqvMz/rd1N6dFaurcL4bfndzU7joiIiM9QqfEiP+QdYdmWAwA8MLoXQVoKQURExG36qeklqmudPL4qB4AJaQkM7BxlciIRERHfolLjJV7+Zh/7DlcRFx7EXRdpKQQREZGmUqnxAjlFdl7/bj9QvxRCeHCAyYlERER8j0qNyZwugzkrc3C6DIYnxTAiWUshiIiInAmVGpMtzChg24EKwoJs3D8qyew4IiIiPkulxkSF5dW88NNSCHdf3IO4cC2FICIicqZUakxSvxTCTqpqXZzTKZLxA7QUgoiIyNlQqTHJyqwivtpdSqDNwozRvbQUgoiIyFlSqTHBkapa5n+WC8Dvzu9K95hQkxOJiIj4PpUaE/z5i10crqqlZ0wo/3NeF7PjiIiI+AWVmhb2zd7DfJB5EAvwwKW9CLTpn0BERKQ56CdqC6qudfLET0shTBzYkQEdI01OJCIi4j9UalrQS+v3kl9WTfvwIO68qLvZcURERPyKSk0L2XHQzr++r18KYeolyYQFaSkEERGR5qRS0wLqXAZzVmbjNOCSXnFcnBhjdiQRERG/o1LTAt7+MZ+sQ3YiggOYPDLR7DgiIiJ+SaXGw/YfqeLFr/YA8L/DehAbFmRuIBERET+lUuNBhmHw5Kc51NS5SO8Sxdh+HcyOJCIi4rdUajxoxfZDfLP3CEE2C9NH98KipRBEREQ8RqXGQw4fdfDMT0sh3HphN7q2DTE5kYiIiH9TqfGQZz7fRVl1HclxYfwqvbPZcURERPyeSo0HrN9TysfbD2G1wAOjkwnQUggiIiIep5+2zayq1smTPy2FcMM5nUhN0FIIIiIiLUGlppm9+NUeCsprSIgM5vYh3c2OIyIi0mqo1DSjzMIK3v4xH6hfCiE0yGZyIhERkdbD5xYgKigooKSkhJSUFIKCzJ/IzukyyMgv4+i+Mp75ZAcuAy7rHceQHu3MjiYiItKqmFZqsrOzmT59Ovv27eO6667j/vvvb3QelyeeeIKlS5cSHR3N0aNHefXVV0lMNG/ZgTU5xcxfs5NDdkfDcxbgvG5tTcskIiLSWply+snhcHD77beTmprK4sWLyc3NZcmSJafd5ptvvuHzzz9n9erVfPLJJwwdOpSXXnqphRKfaE1OMVPf33ZcoQEwgEc/yWZNTrE5wURERFopU47UrF27FrvdzvTp0wkJCWHSpEnMnj2bCRMmnHKboKAgHn30UcLDwwHo06cPX3/9dZPfuzkm9XW6DOav2Xna1zzzWS7Dk2KwWTWL8M8dG39Nrtw4jZX7NFbu01i5T2PlPk+OVVP2aUqpycrKIi0tjZCQ+ll2U1JSyM3NPe0255xzTsP/Ly0tZfHixfzqV79q8nvHxEQ0eZv/tj635IQjNP/tYEUNuypquTAx5qzfzx81x79Da6Gxcp/Gyn0aK/dprNxn9liZUmrsdjudO/9nll2LxYLVaqWsrIyoqKjTbrtw4UIee+wx0tPTue6665r83iUlFRhGkzc7Tm7BEbdflxxl/sXM3sRiqf/QN8e/g7/TWLlPY+U+jZX7NFbu8+RYHdu3O0wpNTab7YQ7l4KDg6murm601IwfP564uDhmzZrFm2++yc0339yk9zYMznrAY8PcKyqxYUH6QjiF5vh3aC00Vu7TWLlPY+U+jZX7zB4rUy4UjoqKorS09LjnKisrCQwMbHTboKAgRowYwd13382iRYs8FfG0BnaKon346YtNfEQwAzudvqCJiIhI8zGl1PTv35+MjIyGx3l5eTgcjtMepXnttddYvnx5w+PAwECsVnPmDrRZLUwemXTa10wakaiLhEVERFqQKa1g0KBB2O12Fi9eDMCCBQsYPHgwNpuN8vJynE7nCdt06dKFxx9/nA0bNrBr1y7++c9/cvnll7d09AYjk2OZO7bvCUds4iOCmTu2LyOTY01KJiIi0jpZDMOcs1+rV69m8uTJBAcHY7VaeeONN0hKSiIlJYWlS5fSp0+fE7Z55ZVX+Oc//0ldXR0TJ07k3nvvbfLRmuLi5r2I6diMwjVWK8EuFwM7RekIzWlYLBAbG9Hs/w7+SGPlPo2V+zRW7tNYuc+TY3Vs32691qxSA1BUVERmZiZpaWm0bdsys/B6csD1wW+cxsp9Giv3aazcp7Fyn8bKfd5Sakxd+ykuLo7hw4ebGUFERET8hFbpFhEREb+gUiMiIiJ+QaVGRERE/IJKjYiIiPgFlRoRERHxCyo1IiIi4hdUakRERMQvqNSIiIiIXzB18j0zWDywgsGxfXpi3/5GY+U+jZX7NFbu01i5T2PlPk+OVVP2aeoyCSIiIiLNRaefRERExC+o1IiIiIhfUKkRERERv6BSIyIiIn5BpUZERET8gkqNiIiI+AWVGhEREfELKjUiIiLiF1RqpEWVl5ezadMmysrKzI4iIiJ+RqXmLHz66aeMGjWKvn37Mm7cOHJzc82O5NVWrFjByJEjefDBBxk2bBgrVqwwO5JPuOWWW1iyZInZMbzWnDlzSElJafgzevRosyN5vaeeeorbb7/d7BhebcmSJcd9ro790dfiid59912GDRtGWloav/rVr8jLyzMvjCFnZO/evcagQYOMDz/80CgqKjLuvvtu44YbbjA7ltcqLy83zj//fGP79u2GYRjG4sWLjREjRpicyvstW7bM6NWrl7F48WKzo3itG264wfj888+NsrIyo6yszKioqDA7klfbvn27MXDgQGPfvn1mR/FqNTU1DZ+psrIyo7Cw0Dj//PONvXv3mh3Nq+zdu9cYNmyYsXXrViM/P9+YPn268ctf/tK0PDpSc4Zyc3OZPHkyY8aMITY2lptuuont27ebHctr2e12ZsyYQe/evQHo27cvhw8fNjmVdzty5Ahz586lR48eZkfxWnV1deTk5JCenk5kZCSRkZGEh4ebHctruVwuZs6cyW9+8xu6dOlidhyvFhQU1PCZioyMZOnSpYwePZquXbuaHc2rbNu2jbS0NFJTU+nYsSMTJkxg7969puVRqTlDI0aM4IYbbmh4vHv3brp162ZiIu+WkJDA2LFjAaitreW1117TaYJGzJ07l0suuYSBAweaHcVrZWdn43K5GD9+PAMGDOCWW26hoKDA7Fhe66233iI7O5tOnTqxevVqHA6H2ZF8Qk1NDa+//jq33Xab2VG8TlJSEhs2bGD79u1UVFTw73//myFDhpiWR6WmGTgcDl555RVuvPFGs6N4vaysLIYOHcq6det48MEHzY7jtTZs2MD69eu57777zI7i1Xbu3EmPHj2YN28e77//PgEBATz00ENmx/JKlZWVPP/883Tp0oWCggJeffVVfvnLX1JdXW12NK+3fPlyBgwYQOfOnc2O4nWSkpK47LLLGD9+POnp6WRkZDB16lTT8qjUNIPnn3+ekJAQJk6caHYUr5eSksI///lPunXrplJzCjU1NTz88MPMmjVLp1IaMXbsWJYsWcI555xD9+7defjhh/n666+x2+1mR/M6q1atoqqqitdee427776bV155hcrKSpYtW2Z2NK/39ttvc9NNN5kdwytt3ryZzz77jIULF/L9999z1VVX8fvf/x7DMEzJo1JzltavX8+//vUv5s+fT2BgoNlxvJ7FYqFfv348+eSTrFy5kvLycrMjeZ2//e1v9OvXj+HDh5sdxefExMTgcrk4dOiQ2VG8zoEDB0hLS6Ndu3YABAQEkJKSYur1D75g79697Nu3j8GDB5sdxSt98MEHXHnllaSlpREREcE999xDXl4eWVlZpuQJMOVd/UReXh6TJ09m5syZJCUlmR3Hq3377bd89tlnDYclg4KCsFgsWK3q1f9t+fLlHD58mPT0dACqq6tZsWIFmzdvZtasWeaG8zJz586lb9++XH311QBs3LgRq9VKQkKCycm8T4cOHaipqTnuuYKCAs455xyTEvmGFStWMHz4cP3Segoul+u4mz4qKyupqqrC6XSakkel5gxVV1dz++23M2rUKEaPHk1lZSUAoaGhWCwWk9N5n+7du7Nw4UK6d+/OxRdfzHPPPceQIUN0euUk/v3vf1NXV9fweN68eaSlpXHNNdeYmMo79e7dm+eee47Y2FicTiePPvoo48ePJyQkxOxoXmfYsGE8+uijvPXWW4wYMYKVK1eSlZXFn//8Z7OjebV169bpa+800tPTmTZtGq+++ioxMTG8++67xMXFkZKSYkoei2HWiS8f9+mnn/LHP/7xhOdXr16ti8lO4auvvuLxxx+nsLCQoUOHMmvWrIZD4XJq06ZN47zzzuPaa681O4pXmj9/Pm+99RY2m42rr76aSZMmERoaanYsr/TDDz8wb948srKyiIuLY8aMGYwcOdLsWF6rurqa9PR0li1bRmJiotlxvJJhGPztb39j0aJFFBUVkZyczGOPPUbfvn1NyaNSIyIiIn5BFzSIiIiIX1CpEREREb+gUiMiIiJ+QaVGRERE/IJKjYiIiPgFlRoRERHxCyo1IiIi4hdUakTEZy1ZsoRx48Y1PH755ZcZPHgwRUVFJqYSEbNomQQR8Qvbt2/nueee469//StxcXFmxxERE+hIjYj4vKqqKiZNmsTNN9/MRRddZHYcETGJSo2I+LwnnniCsLAw7r33XrOjiIiJdPpJRHzanj17yMrKon///gQGBpodR0RMpCM1IuLTqquruf/++9mxYwdr1qwxO46ImEilRkR8WlJSErfccgsTJ07k+eefxzAMsyOJiElUakTEpwUE1J9Fv+2229i5cyerVq0yOZGImEWlRkT8Qnx8vI7WiLRyKjUi4jduu+029uzZw4oVK8yOIiImsBj6lUZERET8gI7UiIiIiF9QqRERERG/oFIjIiIifkGlRkRERPyCSo2IiIj4BZUaERER8QsqNSIiIuIXVGpERETEL6jUiIiIiF9QqRERERG/8P8Bo5ci8uO0ijQAAAAASUVORK5CYII="
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.ylabel('$J(C_K)$')  # 误差平方和\n",
    "# 存放轮廓系数\n",
    "Scores = []\n",
    "plt.rcParams['font.sans-serif'] = ['SimHei']  #显示中文标签\n",
    "plt.rcParams['axes.unicode_minus'] = False  #用来正常显示负号\n",
    "for k in range(2, 9):\n",
    "    estimator = KMeans(n_clusters=k)  #构造聚类器\n",
    "    estimator.fit(data_s)\n",
    "    Scores.append(silhouette_score(data_s, estimator.labels_, metric='euclidean'))\n",
    "X = range(2, 9)\n",
    "plt.xlabel('K')\n",
    "plt.ylabel('轮廓系数')\n",
    "plt.plot(X, Scores, 'o-')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 640x480 with 1 Axes>",
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# calinski_haarbaz指数\n",
    "from sklearn.metrics import calinski_harabasz_score\n",
    "haraba=[]\n",
    "plt.rcParams['font.sans-serif']=['SimHei'] #显示中文标签\n",
    "plt.rcParams['axes.unicode_minus']=False #用来正常显示负号\n",
    "for k in range(2,9):\n",
    "    estimator=KMeans(n_clusters=k) #构造聚类器\n",
    "    estimator.fit(data_s)\n",
    "    haraba.append(calinski_harabasz_score(data_s,estimator.labels_))\n",
    "X=range(2,9)\n",
    "plt.xlabel('K')\n",
    "plt.ylabel('calinski_harabaz指数')\n",
    "plt.plot(X,haraba,'o-')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "outputs": [],
   "source": [
    "#建立模型\n",
    "cluster=KMeans(n_clusters=4,random_state=5).fit(data_s)"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "outputs": [
    {
     "data": {
      "text/plain": "array([[ 0.36355751, -1.81817227,  0.18306699],\n       [-0.95490222,  0.17679215, -1.69155755],\n       [ 1.31656935,  0.10349958, -0.16019117],\n       [-0.51047021,  0.36281987,  0.90438576]])"
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#聚类中心\n",
    "centers=cluster.cluster_centers_ #聚类中心\n",
    "centers"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "outputs": [
    {
     "data": {
      "text/plain": "4749.265326507992"
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 查看总距离平方和\n",
    "inertia = cluster.inertia_\n",
    "inertia"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "outputs": [
    {
     "data": {
      "text/plain": "0.5513427055662409"
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 轮廓系数均值\n",
    "c_preds1 = cluster.labels_\n",
    "silhouette_score(data_s,c_preds1)"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 640x480 with 1 Axes>",
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "plt.figure()\n",
    "plt.scatter(data_s.values[:, 0], data_s.values[:, 1], c=c_preds1)#原始数据散点图，按照分类查看\n",
    "plt.scatter(centers[:, 0], centers[:, 1],\n",
    "            marker='x', s=169, linewidths=3,\n",
    "            color='r', zorder=10) # 重心红色X进行突出\n",
    "plt.savefig(\"cluster.svg\", dpi=300,format=\"svg\")"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "outputs": [],
   "source": [
    "# 层次聚类\n",
    "from sklearn.cluster import AgglomerativeClustering\n",
    "from sklearn.metrics import confusion_matrix\n",
    "clustering = AgglomerativeClustering(linkage='ward', n_clusters=3)\n",
    "res = clustering.fit(data_s)"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 640x480 with 1 Axes>",
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 存放轮廓系数\n",
    "Scores = []\n",
    "plt.rcParams['font.sans-serif']=['SimHei'] #显示中文标签\n",
    "plt.rcParams['axes.unicode_minus']=False #用来正常显示负号\n",
    "for k in range(2, 9):\n",
    "    estimator = AgglomerativeClustering(n_clusters=k)  # 构造聚类器\n",
    "    estimator.fit(data_s)\n",
    "    Scores.append(silhouette_score(data_s, estimator.labels_, metric='euclidean'))\n",
    "X = range(2, 9)\n",
    "plt.xlabel('k')\n",
    "plt.ylabel('轮廓系数')\n",
    "plt.plot(X, Scores, 'o-')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 1000x1000 with 2 Axes>",