P3_ap2_test.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "c8799379",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:52:34.481686Z",
     "start_time": "2024-09-24T12:52:34.000346Z"
    }
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "34871c03",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:09:09.869162Z",
     "start_time": "2024-09-24T12:09:09.837254Z"
    }
   },
   "outputs": [],
   "source": [
    "part1=pd.read_csv('RAWDATA/training_set_2ap_loc0_nav82.csv')\n",
    "part2=pd.read_csv('RAWDATA/training_set_2ap_loc0_nav86.csv')\n",
    "part3=pd.read_csv('RAWDATA/training_set_2ap_loc1_nav82.csv')\n",
    "part4=pd.read_csv('RAWDATA/training_set_2ap_loc1_nav86_new.csv')\n",
    "part5=pd.read_csv('RAWDATA/training_set_2ap_loc2_nav82.csv')\n",
    "\n",
    "resource=pd.concat([part1,part2,part3,part4,part5],axis=0)\n",
    "resource=resource.iloc[:390,:]    #删掉异常的行值"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "fe2e4952",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:09:10.317333Z",
     "start_time": "2024-09-24T12:09:10.313100Z"
    }
   },
   "outputs": [],
   "source": [
    "columns=[0,1,2,4,6,7,8,12,19,20,22,23,25,26,28,29,31,32,40]\n",
    "resource=resource.drop(resource.columns[columns],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "159e7665",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:09:10.842323Z",
     "start_time": "2024-09-24T12:09:10.842323Z"
    }
   },
   "outputs": [],
   "source": [
    "resource.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e6d05847",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:09:11.081619Z",
     "start_time": "2024-09-24T12:09:11.027751Z"
    }
   },
   "outputs": [],
   "source": [
    "for i in range(resource.shape[0]):\n",
    "    if resource.iloc[i,list(resource.columns).index('bss_id')]==0:\n",
    "        resource.iloc[i,list(resource.columns).index('ap_from_ap_0_sum_ant_rssi')]=resource.iloc[i,list(resource.columns).index('ap_from_ap_1_sum_ant_rssi')]\n",
    "        resource.iloc[i,list(resource.columns).index('ap_from_ap_0_max_ant_rssi')]=resource.iloc[i,list(resource.columns).index('ap_from_ap_1_max_ant_rssi')]\n",
    "        resource.iloc[i,list(resource.columns).index('ap_from_ap_0_mean_ant_rssi')]=resource.iloc[i,list(resource.columns).index('ap_from_ap_1_mean_ant_rssi')]\n",
    "        resource.iloc[i,list(resource.columns).index('sta_from_sta_0_rssi')]=resource.iloc[i,list(resource.columns).index('sta_from_sta_1_rssi')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "2d961879",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:09:14.518088Z",
     "start_time": "2024-09-24T12:09:14.514641Z"
    }
   },
   "outputs": [],
   "source": [
    "resource=resource.drop(['ap_from_ap_1_sum_ant_rssi','ap_from_ap_1_max_ant_rssi',\n",
    "                       'ap_from_ap_1_mean_ant_rssi','sta_from_sta_1_rssi'],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "7e00f80e",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:09:15.922691Z",
     "start_time": "2024-09-24T12:09:15.409098Z"
    }
   },
   "outputs": [],
   "source": [
    "resource.index=range(resource.shape[0])\n",
    "# resource.to_excel('programe3_2ap.xlsx')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f06f10ba",
   "metadata": {},
   "source": [
    "# 读入人工筛选的数据集"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "084b0f3f",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:52:41.073943Z",
     "start_time": "2024-09-24T12:52:40.597260Z"
    }
   },
   "outputs": [],
   "source": [
    "resource=pd.read_excel('programe3_2ap.xlsx') #手动删除per为1的数据"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "b80b242a",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:52:44.986243Z",
     "start_time": "2024-09-24T12:52:44.978339Z"
    }
   },
   "outputs": [],
   "source": [
    "# 定义管理rssi中异常值数据，并创造出新特征值\n",
    "def cal_pd(apmax,pd):\n",
    "    apmax = eval(apmax)\n",
    "    comparison_result = [1 if x >= pd else 0 for x in apmax]\n",
    "    proportion_of_ones = sum(comparison_result) / len(apmax)\n",
    "    return proportion_of_ones\n",
    "\n",
    "def cal_ed(apmax,ed):\n",
    "    apmax = eval(apmax)\n",
    "    comparison_result = [1 if x >= ed else 0 for x in apmax]\n",
    "    proportion_of_ones = sum(comparison_result) / len(apmax)\n",
    "    return proportion_of_ones\n",
    "\n",
    "def cal_nav(apmean,nav):\n",
    "    apmean = eval(apmean)\n",
    "    comparison_result = [1 if x >= nav else 0 for x in apmean]\n",
    "    proportion_of_ones = sum(comparison_result) / len(apmean)\n",
    "    return proportion_of_ones    \n",
    "\n",
    "\n",
    "def cal_amsdu(seq_time, ppdu_dur, ampdu, per, throughput):\n",
    "    # 计量单位：微秒 us\n",
    "    # 一次发包成功总耗时为RTS+SIFS+CTS+SIFS+ppdu_dur\n",
    "    # ppdu为一个数据帧的时长，即从RTS-CTS后发出数据帧到收到该帧的ACK\n",
    "    success_dur = ppdu_dur\n",
    "    # 一次发包失败总耗时为RTS+ACK_timeout = 85us\n",
    "    fail_dur = 85.0\n",
    "\n",
    "    # 假设成功发出x个ppdu包， 则有 per/(1-per) * x 个失败包\n",
    "    # 总共耗时seq_time = per/(1-per) * x *fail_dur + x * success_dur\n",
    "    x = seq_time / (success_dur + (per / (1-per) * fail_dur))\n",
    "    \"\"\"\n",
    "    在802.11 WLAN中，最常见的PHY头部大小通常是160位（20字节），适用于802.11a、802.11g、802.11n、802.11ac和802.11ax等标准。这是因为这些标准大多数都基于OFDM（正交频分复用）调制，采用相同的PHY头部结构。\n",
    "    对于802.11b（DSSS和CCK），虽然它的PHY头部是192位（24字节），但在现代无线网络中，802.11a/g/n/ac/ax的使用更为广泛，因此160位的PHY头部是最常见的。\n",
    "    \"\"\"\n",
    "    pkg_len = 1500\n",
    "    # 有效吞吐量（Mbps）为 x*effective_ppdu_len*8 / (60 * 1e6)\n",
    "    # throughput = (x * effective_ppdu_len * 8.0) / (60 * 1e6)\n",
    "    ppdu_len = throughput * 60 * 1e6 / (x*8)\n",
    "    amsdu = ((ppdu_len-20) / ampdu - 40) / 1524\n",
    "    return amsdu"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "91b5535a",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:52:46.364897Z",
     "start_time": "2024-09-24T12:52:46.183217Z"
    }
   },
   "outputs": [],
   "source": [
    "resource['pro_pd'] = resource.apply(lambda row: cal_pd(row['ap_from_ap_0_max_ant_rssi'], row['pd']), axis=1)\n",
    "resource['pro_ed'] = resource.apply(lambda row: cal_ed(row['ap_from_ap_0_max_ant_rssi'], row['ed']), axis=1)\n",
    "resource['pro_nav'] = resource.apply(lambda row: cal_nav(row['ap_from_ap_0_mean_ant_rssi'], row['nav']), axis=1)\n",
    "\n",
    "resource['num_amsdu'] = resource.apply(lambda row: cal_amsdu(row['seq_time']*1000000, \n",
    "                                                     row['ppdu_dur']*1000000,\n",
    "                                                     row['num_ampdu'],\n",
    "                                                     row['per'], row['throughput']), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "7fcaca61",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:52:46.944317Z",
     "start_time": "2024-09-24T12:52:46.934739Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 389 entries, 0 to 388\n",
      "Data columns (total 23 columns):\n",
      " #   Column                      Non-Null Count  Dtype  \n",
      "---  ------                      --------------  -----  \n",
      " 0   protocol                    389 non-null    object \n",
      " 1   pd                          389 non-null    int64  \n",
      " 2   ed                          389 non-null    int64  \n",
      " 3   nav                         389 non-null    int64  \n",
      " 4   ap_from_ap_0_sum_ant_rssi   389 non-null    object \n",
      " 5   ap_from_ap_0_max_ant_rssi   389 non-null    object \n",
      " 6   ap_from_ap_0_mean_ant_rssi  389 non-null    object \n",
      " 7   sta_to_ap_0_sum_ant_rssi    389 non-null    object \n",
      " 8   sta_to_ap_1_sum_ant_rssi    389 non-null    object \n",
      " 9   sta_from_ap_0_sum_ant_rssi  389 non-null    object \n",
      " 10  sta_from_ap_1_sum_ant_rssi  389 non-null    object \n",
      " 11  sta_from_sta_0_rssi         389 non-null    int64  \n",
      " 12  nss                         389 non-null    int64  \n",
      " 13  mcs                         389 non-null    int64  \n",
      " 14  per                         389 non-null    float64\n",
      " 15  num_ampdu                   389 non-null    int64  \n",
      " 16  ppdu_dur                    389 non-null    float64\n",
      " 17  seq_time                    389 non-null    float64\n",
      " 18  throughput                  389 non-null    float64\n",
      " 19  pro_pd                      389 non-null    float64\n",
      " 20  pro_ed                      389 non-null    float64\n",
      " 21  pro_nav                     389 non-null    float64\n",
      " 22  num_amsdu                   389 non-null    float64\n",
      "dtypes: float64(8), int64(7), object(8)\n",
      "memory usage: 70.0+ KB\n"
     ]
    }
   ],
   "source": [
    "resource.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "3769ab64",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:52:48.388396Z",
     "start_time": "2024-09-24T12:52:47.781341Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "num_amsdu std is 1.108420263672169\n",
      "num_amsdu mean is 3.4488594574078184\n",
      "pearson between num_amsdu and num_ampdu is 0.6327102984327106\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjMAAAHFCAYAAAAHcXhbAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABa70lEQVR4nO3deXwTZf4H8M+kV3qmB7ZNF2gjq2gphxUKRW5FQS0g+lNBEE9YFRXxQNS17S5awNXVXRYQXEVlEf3tigurIohcCshlhVLUFVsOaalQSFqgLW3m90d/iU1zTSaTY5rP+/XKSzPzzPN8n+eZmXxJZqaCKIoiiIiIiFRKE+gAiIiIiLzBZIaIiIhUjckMERERqRqTGSIiIlI1JjNERESkakxmiIiISNWYzBAREZGqMZkhIiIiVWMyQ0RERKrGZCbEff3117jpppvQtWtXREVFIS0tDfn5+Xj88cd91ua2bdtQVFSEM2fO2K1buHAhli1b5rO2vVFZWQlBEGTFV15ejqKiIlRWVioe14YNG9C3b1/ExsZCEAR89NFHirdBri1btgyCIPhkfkmdioqKIAhCoMMIGUxmQtjHH3+MgQMHwmQyYf78+Vi3bh1ee+01XHXVVXj//fd91u62bdtQXFysumTGG+Xl5SguLlb8w04URdx6662IiIjA6tWrsX37dgwdOlTRNoiIgl14oAOgwJk/fz4MBgM+++wzhIf/uivcfvvtmD9/fgAjU5YoimhoaEB0dHSgQ1Hc8ePHUVtbi5tuuglXX311oMMhIgoIfjMTwk6dOoVOnTrZJDIWGo39rrFixQrk5+cjLi4OcXFx6NOnD/7+979b169fvx5jx45F586dodVq8dvf/hbTpk3DyZMnrWWKiorw5JNPAgAMBgMEQYAgCNi0aROysrJw4MABbN682bo8KyvLuq3JZMITTzwBg8GAyMhI/OY3v8GMGTNw9uxZmzgFQcD06dOxePFiXH755YiKisLbb7/tdByysrJw4403YtWqVejVqxe0Wi0uvvhi/OUvf5E0jl9++SWuvvpqxMfHIyYmBgMHDsTHH39sXb9s2TL8z//8DwBg+PDh1r65+wbKXb1FRUXo3LkzAGDWrFl249Xepk2bIAgC3nvvPTz77LPIyMhAQkICrrnmGnz//fd2Y3LXXXfZ1TFs2DAMGzbMrs4VK1Zg1qxZ0Ov1iIuLQ0FBAU6cOIG6ujpMnToVnTp1QqdOnXD33Xejvr7eZb/b2717N26//XZkZWUhOjoaWVlZmDBhAg4fPmxTzvJTzxdffIH7778fKSkpSEhIwJ133omzZ8+iuroat956KxITE6HX6/HEE0/gwoULNnUsWrQIvXv3RlxcHOLj43HZZZfhmWeesSmzY8cOXHXVVdBqtcjIyMDs2bPt6gFa98OioiK75c7G1uLChQtITU3F5MmT7dadOXMG0dHRmDlzJgDAbDZjzpw56N69O6Kjo5GYmIhevXrhtddec1o/wH3B3b5g+Ul5/vz5eOGFF9C1a1dotVr07dsXGzZssIvr448/Rp8+fRAVFQWDwYA//elPdmVc/UztbF8h6fjNTAjLz8/HG2+8gUceeQR33HEHcnNzERER4bDs888/jz/+8Y8YP348Hn/8ceh0OpSVldmcRA4dOoT8/Hzcd9990Ol0qKysxCuvvIJBgwZh//79iIiIwH333Yfa2lr89a9/xYcffgi9Xg8AyM7OxqpVq3DLLbdAp9Nh4cKFAICoqCgAwLlz5zB06FAcO3YMzzzzDHr16oUDBw7g+eefx/79+/H555/b/D790UcfYevWrXj++eeRnp6O1NRUl2NRWlqKGTNmoKioCOnp6fjHP/6BRx99FE1NTXjiiSecbrd582aMHDkSvXr1wt///ndERUVh4cKFKCgowHvvvYfbbrsNN9xwA1588UU888wz+Nvf/obc3FwAQLdu3byq97777kPv3r0xfvx4PPzww5g4caJ1vFx55plncNVVV+GNN96AyWTCrFmzUFBQgIMHDyIsLMzt9s7qHD58OJYtW4bKyko88cQTmDBhAsLDw9G7d2+89957+Oabb/DMM88gPj5ecqIItH4IdO/eHbfffjuSk5NRVVWFRYsWoV+/figvL0enTp1syt93330YP348Vq5caW2zubkZ33//PcaPH4+pU6fi888/x7x585CRkWFNDFauXIkHH3wQDz/8MP70pz9Bo9Hgxx9/RHl5ubXu8vJyXH311cjKysKyZcsQExODhQsXYsWKFbLGzZGIiAhMmjQJixcvxt/+9jckJCRY17333ntoaGjA3XffDaD129WioiI899xzGDJkCC5cuIDvvvvO4U+4jnBfcLwvWCxYsACZmZl49dVXYTabMX/+fIwePRqbN29Gfn4+gNZr1saOHYv8/HysXLkSLS0tmD9/Pk6cOCFr/EgmkULWyZMnxUGDBokARABiRESEOHDgQLGkpESsq6uzlvvpp5/EsLAw8Y477pBct9lsFi9cuCAePnxYBCD++9//tq576aWXRABiRUWF3XY9evQQhw4dare8pKRE1Gg04q5du2yW//Of/xQBiJ988ol1GQBRp9OJtbW1kmLNzMwUBUEQS0tLbZaPHDlSTEhIEM+ePSuKoihWVFSIAMS33nrLWmbAgAFiamqqzXg1NzeLOTk5YufOnUWz2SyKoij+7//+rwhA3Lhxo6SYpNZriemll15yW+fGjRtFAOL1119vs/yDDz4QAYjbt2+3GZMpU6bY1TF06FCb+bHUWVBQYFNuxowZIgDxkUcesVk+btw4MTk52W2srjQ3N4v19fVibGys+Nprr1mXv/XWWyIA8eGHH7ZrE4D4yiuv2Czv06ePmJuba30/ffp0MTEx0WXbt912mxgdHS1WV1fbxHPZZZfZ7dMAxMLCQrs6nI1tW/v27RMBiEuWLLFZnpeXJ1555ZXW9zfeeKPYp08fl3U5wn3B9b5gOa4yMjLE8+fPW5ebTCYxOTlZvOaaa6zL+vfv77Rc249YR+cPC2f7CknHn5lCWEpKCrZu3Ypdu3Zh7ty5GDt2LH744QfMnj0bPXv2tP48tH79erS0tOChhx5yWV9NTQ1+97vfoUuXLggPD0dERAQyMzMBAAcPHvQq1v/85z/IyclBnz590NzcbH1dd9111p+p2hoxYgSSkpIk19+jRw/07t3bZtnEiRNhMpmwd+9eh9ucPXsWX3/9NW655RbExcVZl4eFhWHy5Mk4duyY3Vf2UviqXosxY8bYvO/VqxcA2H1V74kbb7zR5v3ll18OALjhhhvsltfW1nr080J9fT1mzZqF3/72twgPD0d4eDji4uJw9uxZh/uVJ7G07XNeXh7OnDmDCRMm4N///rfNz6MWGzduxNVXX420tDTrsrCwMNx2222S+yNFz549ceWVV+Ktt96yLjt48CB27tyJe+65xybmb7/9Fg8++CA+++wzmEwmj9rhvvDrckd9Hj9+PLRarfV9fHw8CgoKsGXLFrS0tODs2bPYtWuX03LkP/yZidC3b1/07dsXQOvv9bNmzcKf//xnzJ8/H/Pnz8cvv/wCANbrMxwxm8249tprcfz4cfz+979Hz549ERsbC7PZjAEDBuD8+fNexXjixAn8+OOPTn8Ga//BY/n5Sqr09HSny06dOuVwm9OnT0MURYdtZWRkuNzWFV/Va5GSkmLz3vLTlDdzlJycbPM+MjLS5fKGhgabRM2ViRMnYsOGDfj973+Pfv36ISEhAYIg4Prrr3cYsyexNDQ0WN9PnjwZzc3NWLp0KW6++WaYzWb069cPc+bMwciRIwG0jrurfUVJ99xzDx566CF89913uOyyy/DWW28hKioKEyZMsJaZPXs2YmNjsXz5cixevBhhYWEYMmQI5s2bZz2mXeG+8OvytvuChbO5bmpqQn19Perq6mA2m/22T5BzTGbIRkREBAoLC/HnP/8ZZWVlAICLLroIAHDs2DF06dLF4XZlZWX49ttvsWzZMkyZMsW6/Mcff1Qkrk6dOiE6Ohpvvvmm0/Vtefp8h+rqaqfL2p/wLZKSkqDRaFBVVWW37vjx4w7jksJX9XpCq9WisbHRbvnJkyd93nZbRqMR//nPf1BYWIinn37auryxsRG1tbWKt3f33Xfj7rvvxtmzZ7FlyxYUFhbixhtvxA8//IDMzEykpKS43FfaioqKcjiGUhPRCRMmYObMmVi2bBleeOEFvPvuuxg3bpzNN47h4eGYOXMmZs6ciTNnzuDzzz/HM888g+uuuw5Hjx5FTEyMB713LFT3BcD5eSEyMhJxcXEIDw+HIAiS9gnLNzftx9Kbf5jQr/gzUwhz9GEJ/PqTkOVbgGuvvRZhYWFYtGiR07osyUP7C1Bff/11u7Ku/vUXFRXlcPmNN96IQ4cOISUlxfpNUtuXq7t4pDhw4AC+/fZbm2UrVqxAfHy89YLd9mJjY9G/f398+OGHNjGbzWYsX74cnTt3xqWXXmrtFyDtX7ye1OsrWVlZ2Ldvn82yH374wauft+QQBAGiKNrtV2+88QZaWlp81m5sbCxGjx6NZ599Fk1NTThw4ACA1rvRNmzYYHNxZ0tLi8PnMjkawy+++ELyzypJSUkYN24c3nnnHfznP/9BdXW1zU9M7SUmJuKWW27BQw89hNraWsWeaRTK+8KHH35o841NXV0d1qxZg8GDByMsLAyxsbHIy8tzWq6ttLQ0aLVau7H897//7ZPYQw2/mQlh1113HTp37oyCggJcdtllMJvNKC0txcsvv4y4uDg8+uijAFpPZs888wz++Mc/4vz585gwYQJ0Oh3Ky8tx8uRJFBcX47LLLkO3bt3w9NNPQxRFJCcnY82aNVi/fr1duz179gQAvPbaa5gyZQoiIiLQvXt3xMfHo2fPnli5ciXef/99XHzxxdBqtejZsydmzJiBf/3rXxgyZAgee+wx9OrVC2azGUeOHMG6devw+OOPo3///rLHIiMjA2PGjEFRURH0ej2WL1+O9evXY968eS7/dVtSUoKRI0di+PDheOKJJxAZGYmFCxeirKwM7733njXJy8nJAQAsWbIE8fHx0Gq1MBgMTr/1kVqvr0yePBmTJk3Cgw8+iJtvvhmHDx/G/Pnzrd/S+UtCQgKGDBmCl156CZ06dUJWVhY2b96Mv//970hMTFS0rfvvvx/R0dG46qqroNfrUV1djZKSEuh0OvTr1w8A8Nxzz2H16tUYMWIEnn/+ecTExOBvf/ub3eMBgNYx/P3vf4/nn38eQ4cORXl5ORYsWACdTic5pnvuuQfvv/8+pk+fjs6dO+Oaa66xWV9QUICcnBz07dsXF110EQ4fPoxXX30VmZmZuOSSS7wbkDb9CLV9wSIsLAwjR47EzJkzYTabMW/ePJhMJhQXF1vL/PGPf8SoUaMwcuRIPP7442hpacG8efMQGxtr842RIAiYNGkS3nzzTXTr1g29e/fGzp07Fb0TLqQF9PJjCqj3339fnDhxonjJJZeIcXFxYkREhNi1a1dx8uTJYnl5uV35d955R+zXr5+o1WrFuLg48YorrrC5Mr+8vFwcOXKkGB8fLyYlJYn/8z//Ix45csThlfqzZ88WMzIyRI1GY3OXT2VlpXjttdeK8fHxIgAxMzPTuk19fb343HPPid27dxcjIyNFnU4n9uzZU3zsscds7i4BID700EOSxyEzM1O84YYbxH/+859ijx49xMjISDErK8vujgdndyNs3bpVHDFihBgbGytGR0eLAwYMENesWWPXzquvvioaDAYxLCzM6V0NntYr526m//3f/3XbL7PZLM6fP1+8+OKLRa1WK/bt21f84osvnN7B0r5Oy90k7e8+KywsFAGIv/zyi9t4LY4dOybefPPNYlJSkhgfHy+OGjVKLCsrs7vLxtM2p0yZIsbGxlrfv/322+Lw4cPFtLQ0MTIyUszIyBBvvfVWcd++fTbbffXVV+KAAQPEqKgoMT09XXzyySfFJUuW2N3N1NjYKD711FNily5dxOjoaHHo0KFiaWmppLuZLFpaWsQuXbqIAMRnn33Wbv3LL78sDhw4UOzUqZMYGRkpdu3aVbz33nvFyspKl/VyX3C9L1jGYd68eWJxcbHYuXNnMTIyUrziiivEzz77zC6u1atXi7169bLOwdy5c61ttWU0GsX77rtPTEtLE2NjY8WCggKxsrKSdzMpQBBFUfRX4kQUjLKyspCTk4P//Oc/gQ6FiIJAZWUlDAYDXnrpJZfPmaLgwWtmiIiISNV4zQwRBYTZbIbZbHZZxtGf2qCOh/sCeYs/MxFRQNx1110u/2YW0PpHQqnj475A3mIyQ0QBUVlZ6fApu21JefAbqR/3BfIWkxkiIiJSNV4ATERERKrW4a+oMpvNOH78OOLj433+oDEiIiJShiiKqKurQ0ZGBjQa19+9dPhk5vjx407/nhAREREFt6NHj7r8Q8dACCQz8fHxAFoHIyEhIcDREBERkRQmkwldunSxfo670uGTGctPSwkJCUxmiIiIVEbKJSK8AJiIiIhUjckMERERqRqTGSIiIlI1JjNERESkakxmiIiISNWYzBAREZGqMZkhIiIiVWMyQ0RERKrGZIaIiIhUrcM/AZiIKNS1mEXsrKhFTV0DUuO1yDMkI0wj2C3v0yURK74+jMpT5yCKIuK04QjXCOiflQIzRKzaewzHzjQgI1GLnAwdUuKiUFPXgK3//QVVxgak66Iw5JJUpOuikZ7gvJ0rM5Ow5/Bpu3gCNS7VpgacrGvAyfpGlP1sQkxUGK7smgQRInZV1uLQiXo0tIhI0Iajb2YS4qMjUG1swKn6RsRERuDKzES0iCI2fvcLABHXZqfjrqsMiAzX2LVVU9eATrFRgACcrG+U1H9n8ye1f462k7suWAmiKIqBDsKXTCYTdDodjEYj/5wBEYWctWVVKF5Tjipjg3WZXqfFmN56rP62yma50py1oxEAs2hbrrAgG6Ny9D6LpT1H46IkAcDUIQbMvj7bbVuu+u9s/tyNl6vtAMha58/5ATz7/GYyQ0TUQa0tq8IDy/ci2E/yln/zL5qU65cPTH+Oy8jsVHxeXuOyLWf9dxanu/FytZ2zONytc9Wer3jy+c1rZoiIOqAWs4jiNeVBn8gAv36IFq8pR4vZtxH7e1zWu0lkAMf9dxWnq/GSsp2rGKTGF2yYzBARdUA7K2p9+hOS0kQAVcYG7Kyo9Wk7wTou7fvvLk5n4+Wr/vlrfuRiMkNE1AHV1AXfB7YUvo472MfFEp/UONuXC9XxYzJDRNQBpcZrAx2CLL6OO9jHxRKf1DjblwvV8WMyQ0TUAeUZkqHXaRHcN9T+SkDrXTN5hmSfthOs49K+/+7idDZevuqfv+ZHLiYzREQdUJhGsN5qG2wf3O1Z4issyPb580zajos/jMxOhQDXc+Co/67mz9V4SdlO7jp/zI9cTGaIiDqoUTl6LJqUi3Sd7U8Dep0W04YYoNf59icDZ+20/zxM12n9etuvZVx82X8BwLQhBiy9s5/DOWjLWf+dzZ+78XK13eJJuVgsY52/b8v2FJ8zQ0TUwfEJwK7HhU8ADs4nAPOheW0wmSEiIlIfPjSPiIiIQgaTGSIiIlI1JjNERESkakxmiIiISNWYzBAREZGqMZkhIiIiVWMyQ0RERKrGZIaIiIhUjckMERERqRqTGSIiIlI1JjNERESkakxmiIiISNWYzBAREZGqMZkhIiIiVWMyQ0RERKrGZIaIiIhULTzQARCRvRaziJ0Vtaipa0BqvBZ5hmSEaQS369SqbZ86xUYBAnCyvtFt/zwdC2fl2y+/MjMJew6fdluvu/ZbzCJ2HDqF7T+dBCAgv1sKBlyc4tF8NTWb8fa2SuyqPIWYyHDcnNsZA3/bya6OFrOIbT+exL/2HsO5pmb0y0rBhLyu+MfXh7G2rApHa88jLioM+d06YVR2OmrPN6H2bBOS46KQGhcFsyji64paACKu7JyEz78/gX0/G5GgjcDUwRdj0KUX4XxTCx57/xscrj2H+KjWWA5Wm2AWRQgQEK8Nh0YjIC4qDAeP1+G48TzCNQJM5y8AAKLCNRAEDTQa4NrL03HXIAMiwzU2Y1ltakBtfSOSYyORrouWPaeezn3bdlPjtdbxEEURiTER6BSvRXqCFn26JGLF14dxuPYcuiTF4NLUOOysrMXPp8/DLIo4ebYJ5xsv4IJZRESYBo3NLWi60ILzzWYIEJAaH4VL0uJxvqkFDc0t6Ns1GZdnJKD2XJN136+pa8TJugacrG/CgZ+NiI4MQ9+sZGTrW8u52kfl7stSOBpDAEFxPhJEURT93ur/KykpwYcffojvvvsO0dHRGDhwIObNm4fu3btby4iiiOLiYixZsgSnT59G//798be//Q09evSQ1IbJZIJOp4PRaERCQoKvukKkmLVlVSheU44qY4N1mV6nRWFBNgA4XTcqR+/3WJXgqL9tOeufq3FyNBbOyo/prcfqb6tslmsEwNzmzOioXnftry2rwtMf7seZcxds4kiMicDc8T0lzVfJJ+VYsrUC7c/SMZFheOXW3tY61pZVYeYH3+JcU4vbOoOJAGDqEAOu6JrkdB+QM6dS58rR3KuFo31U7r4shaMxTIyJAACbfVzJ85Enn98BTWZGjRqF22+/Hf369UNzczOeffZZ7N+/H+Xl5YiNjQUAzJs3Dy+88AKWLVuGSy+9FHPmzMGWLVvw/fffIz4+3m0bTGZITdaWVeGB5XvR/qAUALtlbdcBwKJJuapLaJz1ty1H/XM1Tu3LSm3Hlfb1umt/6hADXt9S4bLOxW7mq+STckl1AMDvlu91WU7tBEifU6lzFarknC88GUMlz0eqSWba++WXX5CamorNmzdjyJAhEEURGRkZmDFjBmbNmgUAaGxsRFpaGubNm4dp06a5rZPJDKlFi1nEoHlfyPpXogAgXafFl7NGqOYnJ0/627Z/AFxu134svBlXR/VufnI4hr600WV9rpJPC72L+WpqNuOy339q8y9qR9LiIwEIOFHX6C581dN7MKeezFUo8uR8Ief4Uep85Mnnd1BdAGw0GgEAycmtv8NVVFSguroa1157rbVMVFQUhg4dim3btjmso7GxESaTyeZFpAY7K2pln3RFAFXGBuysqFU2KB/ypL9t++duu/Zj4c24Oqr33e2VbuuT8i9EV/P17vZKt4kMAJyoawqJRAbwbE49matQ5Mn5Qs7xE4jzUdAkM6IoYubMmRg0aBBycnIAANXV1QCAtLQ0m7JpaWnWde2VlJRAp9NZX126dPFt4EQKqanz/qSrRB3+IifWmroGydtZyik9JodrzylWl7PYlGyjI/F0TjmOrkkZR2+OH3+ej4ImmZk+fTr27duH9957z26dINh+TSWKot0yi9mzZ8NoNFpfR48e9Um8REpLjdcGRR3+IifW1Hit5O0s5ZQek8zkGMXqchabkm10JJ7OKcfRNSnj6M3x48/zUVAkMw8//DBWr16NjRs3onPnztbl6enpAGD3LUxNTY3dtzUWUVFRSEhIsHkRqUGeIRl6nRZyfmEW0HpNgeVWSTXwpL9t++duu/Zj4c24Oqp3cn6W2/qktOVqvibnZ0HKpQZp8ZFIi4+S0Jr6eTKnnsxVKPLkfCHn+AnE+SigyYwoipg+fTo+/PBDfPHFFzAYDDbrDQYD0tPTsX79euuypqYmbN68GQMHDvR3uEQ+FaYRrLdftz9xCE7+v+37woJs1Vz8C7jub1vt+ydlnNqOhdR2XGlbb2S4xmX7ltuN3XE1X5HhGtw/2H0dxWNzUDxW2mMq1EyA9DmVOlehytPzhafHT6DORwFNZh566CEsX74cK1asQHx8PKqrq1FdXY3z588DaP15acaMGXjxxRexatUqlJWV4a677kJMTAwmTpwYyNCJfGJUjh6LJuUiXWf79Wy6TovFk3Kx2Mk6Nd6WDTjvb1uO+udqnByNhbPyep0W04YYoG+3vP05uH297tqffX02Fk/KtT6Ho62kmAi3t2UDwOzrszFtiAGOflGPjQyz1jEqR4/Fk3IRExnmsr5gJACYNsSAxZNy7ebAQu/hnEqdK2dzrxbt91G5+7IUzsYwKSbCbh8P1PkooLdmO7vu5a233sJdd90F4NeH5r3++us2D82zXCTsDm/NJjXiE4D5BGALPgGYTwAO1ScAq/Y5M77AZIaIiEh9VPucGSIiIiJPMZkhIiIiVWMyQ0RERKrGZIaIiIhUjckMERERqRqTGSIiIlI1JjNERESkakxmiIiISNWYzBAREZGqMZkhIiIiVWMyQ0RERKrGZIaIiIhUjckMERERqRqTGSIiIlI1JjNERESkakxmiIiISNXCAx0AEQWfFrOInRW1qKlrQGq8FnmGZIRpBJfrAWBnRS2qTQ2orW9Ecmwk0nXR1m3d1dnUbMa72ytxuPYcMpNjMLF/JnZX1uLDvcdwtqkF/bKSMGlAFkqPnkG18TxO1jfh9NkmHDeeR7ouCimxWnSKj0J6gtZhm51iowABOFnf6LB9X4yTp9u1mEXs+OkUtv14EsdOnwMgICNJi6ToSJw5dwHHjechiiJ+kxSNq7pdhAHdUhCmEaxjV3HyLKqN52GGiKZmETkZOiTFRsLUcAGiKCIhKgLfnahDfWMz0hKicGXXZKQlaGEWRWw/dArfHjuDmMgw5BlSMGVgFiLDNdZY285raoIW5hYRX1eeAiAgv1sKBlyc4nIfsKzzdKx8PReBaLep2Yy3t1ViZ8VJnGsyo+dvdEiJi0KnuNbx6tMlESu+Pmw9Fibn/zoXO346he2HTgEQkX9xJ+s+IDW+KzOTsOfwab+Pk68JoiiKgQ7Cl0wmE3Q6HYxGIxISEgIdDlHQW1tWheI15agyNliX6XVaFBZkY1SO3uH6xJgIAMCZcxfs6tPrtBjTW4/V31Y5rbPkk3Is3VoBs0JnI2dtti9jaV8Od+Pk6XZjeuvx/u5jDsfQmcSYCPTLSsKGgzWKjZ2FIADXXJ6Ksp9NTsewrdjIMESEaxzG72j/8Hb825I7F4Fot+STcizZWgFPPnk1AnD15anYVXnabnwTYyIwd3xPh+05ik8jwGZf8cc4yeXJ5zeTGSKyWltWhQeW70X7k4Ll321ThxiwZEuF3Xo5LHVek52K9eU1CtQor/1Fk3I9PpG7GydndTrbLhR5M/5tyZ0Lb8lpt+STcry+pULxWABgcbv2pO5rvh4nb3jy+c1rZogIQOvX0cVryh2e/CzLlm5VJpGx1CkCAUlkLO0DQPGacrR48LWGlHFyVKer7UKR3PFvS+5ceEtOu03NZizd6ptEBgCKVh+wtufJvubLcfInJjNEBKD1egZXPyeIgOI/ZQSaCKDK2ICdFbWSt5EyTo7qdLddKJIz/m3JnQtvyWn33e2VPj1+qk2N1vY83dd8NU7+xGSGiAAANXWh+0HrSd+llm1fLpTH1x25YyN3Lrwlp93DtecUjcFVe74ez2DEZIaIAACp8dpAhxAwnvRdatn25UJ5fN2ROzZy58JbctrNTI5RNAZX7fl6PIMRkxkiAgDkGZKh12nh7CZNAa13QnQkAlrv5rDcOiyFlHFyVKe77UKRnPFvS+5ceEtOu5Pzs3x6/KQnRFnb83Rf89U4+ROTGSICAIRpBBQWZAOA3UnQ8v7+wQYIDtbLYalnZHaqArXJax8ACguyPXrOhpRxclSnq+1Ckdzxb0vuXHhLTruR4RrcP9igaBxtFY3pYW3Pk33Nl+PkT0xmiMhqVI4eiyblIl1n+3Vzuk6LRZNyMfv6bIfrE2MirM8SaU+v02LaEAP0Tupcemc/TBtiUPRfrc7adNS+nNtR3Y2TszqdbWeJ19kYOpMYE4GR2ak++Re/ILQmmq7GsK3YqDCn8Sc52D+8Gf+25M6Ft+S0O/v6bEwbYoDg4Xxp/n8uHI1vYkyE3W3ZruJrv6/4epz8hc+ZISI7fAKwMuPk6XZ8ArD/5yIQ7fIJwNLwoXltMJkhIiJSHz40j4iIiEIGkxkiIiJSNSYzREREpGpMZoiIiEjVmMwQERGRqjGZISIiIlVjMkNERESqxmSGiIiIVI3JDBEREakakxkiIiJSNSYzREREpGpMZoiIiEjVmMwQERGRqjGZISIiIlVjMkNERESqxmSGiIiIVC080AEQ0a9azCJ2VtSipq4BqfFa5BmSEaYRvC6rRNuerr8yMwl7Dp+2Ky817haziB2HTmH7TycBCOhvSIZGEHDybKN1OwCyY3a2zttx8GR8LfFXG8+j9mwTEmMiceZcE5LjopCe4DimTrFRMIsivq6oBSAi/+JOGNAtxVpux6FT+PK/v2D/cSOiI8KRZ0jClIEGRIY7/7erlD63nYv8bikYcHGK3VhWnTmPb46ehgjAkBKLyflZLtv1hhL7v6d1NDWb8fa2CuyqPI3YyDCMz+2Mgb/t5LZdJY9VpY/7jkIQRVEMVONbtmzBSy+9hD179qCqqgqrVq3CuHHjrOvr6+vx9NNP46OPPsKpU6eQlZWFRx55BA888IDkNkwmE3Q6HYxGIxISEnzQCyJlrC2rQvGaclQZG6zL9DotCguyMSpHL7usEm3LWa8RAHObs4tep8WY3nqs/rbKbdxry6rw9If7cebcBacxJ8ZEAIBNGakxA3C4zl18csfd0XaO4m/PWUyOxuK2vp3x/u5jDusTAEwdYsDs67Mlxda+z47mIjEmAnPH9wRgP5YWGgG4f7Djdr2hxP7vaR0ln5RjyZYKtP/AjI0Mw8u39nbarpLHqtLHfbDz5PM7oMnMp59+iq+++gq5ubm4+eab7ZKZ+++/Hxs3bsQbb7yBrKwsrFu3Dg8++CD+9a9/YezYsZLaYDJDarC2rAoPLN9rd6K0/Htr0aRc68nKk7JKtD11iMHhSdzdeqnax722rAq/W77Xq7pcxexpnFLHwdm4OxvfQJjWLqGRMvevb6lQvF1vKLH/e1pHySflbsdhsYN2lTxWlT7u1cCTz++AXjMzevRozJkzB+PHj3e4fvv27ZgyZQqGDRuGrKwsTJ06Fb1798bu3bv9HCmR77SYRRSvKXf4YWdZVrymHC1m0aOySrQtAli61XGi4m69VG3jbmo2o2j1Aa/qchezr+p0NO6uxjcQlmypQFOzGYC0/W7pVu8TGUs9lna9ocT+72kdTc1mLJGQ0LVvV8ljVenjviMK6guABw0ahNWrV+Pnn3+GKIrYuHEjfvjhB1x33XVOt2lsbITJZLJ5EQWznRW1Ln9CEAFUGRuws6LWo7JKtA3Y/lQkZ70Ulrjf3V6JalOj1/X54pzuqk5n4y5lfP1JBPDu9koA0vY7pcbRLP7arjeU2P89rePd7ZWSktH27Sp5rCp93HdEQZ3M/OUvf0F2djY6d+6MyMhIjBo1CgsXLsSgQYOcblNSUgKdTmd9denSxY8RE3mupk7ah11NXYNHZZUs5y+Ha88FOgSvtB/PYBtf4Ncx9ndsSsytEvu1p3V4EnfbupU8BtV6PPtT0CczO3bswOrVq7Fnzx68/PLLePDBB/H555873Wb27NkwGo3W19GjR/0YMZHnUuO1kst5UlbJcv6SmRwT6BC80n48g218gV/H2N+xKTG3SuzXntbhSdxt61byGFTr8exPQZvMnD9/Hs888wxeeeUVFBQUoFevXpg+fTpuu+02/OlPf3K6XVRUFBISEmxeRMEsz5AMvU4LZzdXCmi9YyHPkOxRWSXaBlrvSPFmvRSWuCfnZyE9IcrL2pSJyZM6nY27lPH1JwHA5PwsANL2O6Xu+NUIv7brDSX2f0/rmJyfJWn+2rer5LGq9HHfEQVtMnPhwgVcuHABGo1tiGFhYTCbvb+QjChYhGkE6+3C7U9WlveFBdkI0wgelVWibQGtt9bKWS9V27gjwzUoGtNDZk3SYnb0/0rU6WjcXY1vIEwd8uvzZqTsS5Y+e+v+wa6fcyOVEvu/p3VEhmswdYj7cWjfrpLHqtLHfUcU0GSmvr4epaWlKC0tBQBUVFSgtLQUR44cQUJCAoYOHYonn3wSmzZtQkVFBZYtW4Z33nkHN910UyDDJlLcqBw9Fk3KRbrO9mvidJ3W7pZLT8oq0fbs67NlrW9/XtXrtJg2xAC9m7hH5eixeFKu9TksziTFRNiVkRLz4km5WOxgnbv43I2Ds3F3Nr6O4m/PWUztJcZEYNoQg9P6BDi+PVrK3Dubi6SYCOtYOotPIyh7W7aUmKXs/57WMfv6bEwbYnCYkMZGhTm8LVupWH1RV0cU0OfMbNq0CcOHD7dbPmXKFCxbtgzV1dWYPXs21q1bh9raWmRmZmLq1Kl47LHHIAjSMlA+Z4bUhE8Ato2HTwDmE4DlxuyLOvgEYP9SzUPz/IHJDBERkfqo5qF5RERERN5iMkNERESqxmSGiIiIVI3JDBEREakakxkiIiJSNSYzREREpGpMZoiIiEjVmMwQERGRqjGZISIiIlVjMkNERESqxmSGiIiIVI3JDBEREakakxkiIiJSNSYzREREpGpMZoiIiEjVmMwQERGRqoUHOgBSrxaziJ0Vtaipa0BqvBZ5hmSEaQS361vMInb8dArbDp3Ez6fPIyNRi6u6XYQB3VIAwLpNp7goQASqTQ3Ye6QWNaZGxEWFY9wVv0G4RoOTZxutZU6ebfQohkDzJK62ZTvFRgECcLLecX+DgZS+WcpUmxpQW9+I5NhIpOuiJY9DsPTdVzG1r/fKzCTsOXza5v3Xh07hX98cw7mmFvTLSsaUgVmIDNe4rCfPkAwATo9Lb/tiOba3HzoFQET/rBRAAL6uOAVAQH63FAy4OCXg8wa02QeN51F7tgnJcVFIT7Afa0/HR045NRzXwU4QRVEMdBC+ZDKZoNPpYDQakZCQEOhwOoy1ZVUoXlOOKmODdZlep0VhQTZG5eidrh/TW4/3dx/DmXMX7OqMiQxDZLjG4TqppMRgWR8onsTlqGxbwdCftqT0zVWfPBmHQPfdVzE5qlcjAGY3Z2pBAKYONmD29dlO60mMiQAAm2PMclyu/rbKq76sLavC0x/ud3v8JsZEYO74nkF3DFq0H2tPxkfqPqG24zpQPPn8ZjJDHltbVoUHlu9F+x3H8u+IqUMMWLKlwm69P7iLwbJ+0aTcgJwo3I1d27iclXW3XaBI6RsASX2SMg6B7LuvYpIy5+5MG2LAFV2TvK7Hk76sLavC75bv9aj+xUF2DHqq/fhI3SfUdlwHkief37xmhjzSYhZRvKbc4YFoWbZ0a2ASGUsMoosYLMuK15Sjxd0/dRUmZewscbkq62q7QJHSt6LVB1C0WlqfpIxDoPruq5ikzrk7S7dWoPDfZV7XI7UvLWYRRavLPa4/2I5BT7Udn6Zms6R9wlU5Z9sE8rhWEyYz5JGdFbVOvxoFWg/CYDj2XMUgAqgyNmBnRa3f4gGkjZ0lLndlnW0XKFL6Vm1qRLVJWp+kjkMg+u6rmDyZc1fMInCirsnregBpfbFc++SpYDwGPWUZn3e3V0raJ9yVc7RNII9rNeEFwOSRmjrlTgSB5u++SG1PblyBnBtftO1Jnf7su6/mMZiPLVexeRN3sB6Dnjpce07Rcm0F834RTJjMkEdS47WBDkEx/u6L1PbkxhXIufFF257U6c+++2oeg/nYchWbN3EH6zHoqczkGEXLtRXM+0Uw4c9M5JE8QzL0Oi2c3TQooPVugEDTCHAZo173622q/iJl7CxxuSvrbLtAkdK39ITWW1+l9EnqOASi776KyZM5d0UjAGnxkV7XA0jrS54hGekJnn/gBvIYVIplfCbnZ0naJ9yVc7RNII9rNWEyQx4J0wgoLGi99bP9AWl5f/9gAwQH6/3B0u79gw02MbVdDwCFBdl+f46DlLGzxOWqrKvtAkVK34rG9EDRmGy3dQmQNg6B6ruvYpI65+7cP9iA4rE5XtcjtS9hGkHSvLYXyGNQqUQPaO1HZLhG0j7hqpyzbfi8GWmYzJDHRuXosWhSLtLb/QsnXafFokm5mH19tsP1ep0W04YYrM+6aC8mMszpOqncxWBZH6jbHd2NXdu4nJV1t12gSOmbpYyzfx3rPRiHQPbdVzE5q1fK55kgtN6WPfv6bKf1JMZE2B1jluOy/Zx40pdROXosnpQr6fhNiokI2G3ZANzug+3HWur4SN0n1HZcqwWfM0Oy8QnA8vEJwHwCsCf18gnAyuMTgIMfH5rXBpMZIiIi9eFD84iIiChkMJkhIiIiVWMyQ0RERKrGZIaIiIhUjckMERERqRqTGSIiIlI1JjNERESkakxmiIiISNWYzBAREZGqhcvZ6J133nG5/s4775QVDBEREZGnZP05g6SkJJv3Fy5cwLlz5xAZGYmYmBjU1tYqFqC3+OcMiIiI1Mfnf87g9OnTNq/6+np8//33GDRoEN577z1ZQRMRERHJodg1M5dccgnmzp2LRx99VKkqiYiIiNxS9ALgsLAwHD9+XMkqiYiIiFySdQHw6tWrbd6LooiqqiosWLAAV111lSKBEREREUkhK5kZN26czXtBEHDRRRdhxIgRePnll5WIi4iIiEgSWcmM2WxWOg7VaTGL2FlRi5q6BqTGa5FnSEaYRgh0WF5rMYvY9uNJ/GvvMZxraka/rBRMGZiFyPDWXySbms1Y9mUF1h2sBiBgZHYa7r7KAAB4d3slDteeQ2ZyDCbn/7qNI03NZqflW8widhw6he0/nUSzWYTp/AX8Ut+I2Mhw3NTnNwgP06CmvhEn6xpx6mwDqo2N+E1SNAZ264R+WcnYc/g0qk0NqK1vRHJsJNJ10cgzJAOArDlzN9eO1stti0gqueegjnruChYc38CQdWu2UrZs2YKXXnoJe/bsQVVVFVatWmX3rc/Bgwcxa9YsbN68GWazGT169MAHH3yArl27SmrDF7dmry2rQvGaclQZG6zL9DotCguyMSpHr0gbgbC2rAozP/gW55pabJYLAjB1cGvCsmRLBRztMAJgs1wjAPcPNmD29dl2ZUs+KcfSrRUwi/blr+iahKc/3I8z5y7I6oMgAI726MSYCACwqVfKnLmba0fr5bZFJJXcc1BHPXcFC46vsjz5/JaczMycOVNyAK+88oqkcp9++im++uor5Obm4uabb7ZLZg4dOoS8vDzce++9mDBhAnQ6HQ4ePIh+/fohNTVVUhtKJzNry6rwwPK9dh/olrx70aRcVe60a8uq8LvlexWvd9oQ24Sm5JNyvL6lQvF25HA3Z+7meuoQg9PkztO2iKSSew7qqOeuYMHxVZ5Pkpnhw4fbvN+zZw9aWlrQvXt3AMAPP/yAsLAwXHnllfjiiy88DloQBLtk5vbbb0dERATeffddj+uzUDKZaTGLGDTvC5usuy0BQLpOiy9njVDV14otZhEDSzbgRF2j4nVrBOC7P45GZLgGTc1mXPb7T22+kQk0Z3Pmbq6B1r550he17h8UPOSegzrquStYcHx9wycPzdu4caP1VVBQgGHDhuHYsWPYu3cv9u7di6NHj2L48OG44YYbvO4A0Hpdzscff4xLL70U1113HVJTU9G/f3989NFHLrdrbGyEyWSyeSllZ0Wtyw83EUCVsQE7K4LnCchS7Kyo9UkiA7R+2L+7vRJA63+DKZEBnM+Zu7kGPEtkXLVFJJXcc1BHPXcFC45v4Ml6zszLL7+MkpISmz9rkJSUhDlz5ih2N1NNTQ3q6+sxd+5cjBo1CuvWrcNNN92E8ePHY/PmzU63KykpgU6ns766dOmiSDwAUFPn+sPN03LBwtfxHq49Z/PfYNR+DHw5JmrbPyh4yD0HddRzV7Dg+AaerGTGZDLhxIkTdstrampQV1fndVDAr3dMjR07Fo899hj69OmDp59+GjfeeCMWL17sdLvZs2fDaDRaX0ePHlUkHgBIjdcqWi5Y+DrezOQYm/8Go/Zj4MsxUdv+QcFD7jmoo567ggXHN/BkJTM33XQT7r77bvzzn//EsWPHcOzYMfzzn//Evffei/HjxysSWKdOnRAeHo7sbNu7YS6//HIcOXLE6XZRUVFISEiweSklz5AMvU4LZ794Cmi9ct1ya65a5BmSkRYf5ZO6NQIwOT8LQOt/g+3nYmdz5m6ugda+edIdte4fFDzknoM66rkrWHB8A09WMrN48WLccMMNmDRpEjIzM5GZmYk77rgDo0ePxsKFCxUJLDIyEv369cP3339vs/yHH35AZmamIm14KkwjoLCgNblqv9Na3hcWZKvuAq8wjYDisT18Uvf9gw3W58dEhmtw///f4h0MXM2Zu7kWAGtfpMy2mvcPCh5yz0Ed9dwVLDi+gScrmYmJicHChQtx6tQpfPPNN9i7dy9qa2uxcOFCxMbGSq6nvr4epaWlKC0tBQBUVFSgtLTU+s3Lk08+iffffx9Lly7Fjz/+iAULFmDNmjV48MEH5YStiFE5eiyalIt0ne3Xhek6rapvvRuVo8fiSbmIiQyzWycIrbdYTxticPkvj7Y0gv1t2QAw+/psTBtisPuGxlJ+8aRc63Na5BCcBJgYE2FXr7s5czfXs6/Pdrg+SUZbRFLJPQd11HNXsOD4BpbXD807evQoBEFA586dPd5206ZNdrd8A8CUKVOwbNkyAMCbb76JkpISHDt2DN27d0dxcTHGjh0ruQ1fPDQP6LhPeeQTgB2PCZ8ATMGGTwAOThxf5fjkOTNtNTc3o7i4GH/5y19QX18PAIiLi8PDDz+MwsJCRETI/5e10nyVzBAREZHvePL5LetvM02fPh2rVq3C/PnzkZ+fDwDYvn07ioqKcPLkSZd3GxEREREpSdY3MzqdDitXrsTo0aNtln/66ae4/fbbYTQaFQvQW/xmhoiISH188gTgtrRaLbKysuyWZ2VlITIyUk6VRERERLLISmYeeugh/PGPf0Rj46+PwG9sbMQLL7yA6dOnKxYcERERkTuyrpn55ptvsGHDBnTu3Bm9e/cGAHz77bdoamrC1VdfbfPgvA8//FCZSImIiIgckJXMJCYm4uabb7ZZpuTfQCIiIiKSSlYy89ZbbykdBxEREZEssq6ZISIiIgoWsr6ZOXXqFJ5//nls3LgRNTU11r9wbVFbW6tIcERERETuyEpmJk2ahEOHDuHee+9FWloaBGd/EIeIiIjIx2QlM19++SW+/PJL651MRERERIEi65qZyy67DOfPn1c6FiIiIiKPyUpmFi5ciGeffRabN2/GqVOnYDKZbF5ERERE/iL7OTNGoxEjRoywWS6KIgRBQEtLiyLBEREREbkjK5m54447EBkZiRUrVvACYCIiIgooWclMWVkZvvnmG3Tv3l3peIiIiIg8IiuZ6du3L44ePcpkpgNrMYvY8dMpbD90CoCI/lkp0IQJOFnfiNR4LfIMyQCAnRW1qKlrsC4L0wh29bgrIzc+OfV6sp27su3XX5mZhD2HT6Pa1IDa+kYkx0YiXRdt3U5K223LdIqLAkTg5NlGRcdO7ph4Oh5S4vXV/uENb2MKxj4pwVm/Omp/SV1kJTMPP/wwHn30UTz55JPo2bMnIiIibNb36tVLkeAoMNaWVeHpD/fjzLkL1mULcMimTGJM65y3LaPXaVFYkI1ROXprPcVrylFlbHBaRm58cur1ZDt3ZR2t1wiAWbRvV6/TYkxvPVZ/W+WybUd1tq/H27FrT+qYyBkPd/H6av/whrcxBWOflOCsX1L2ayJ/EERRdHD6dU2jsb8JShCEoLwA2GQyQafTwWg0IiEhIdDhBL21ZVX43fK9sra1/Fts0aRcAMADy/ei/c7Vtoyck93asipZ9XqynbuyU4cYsGRLhd16T0kZL3exekPqmMgdD1fxyp1HX/I2pmDskxKc9csZtfeXgocnn9+ykpnDhw+7XJ+ZmelplT7DZEa6FrOIq+Z+gWqT428GpBAApOu0EEUR1aZGl2W+nDXC46/vB837wuk3F87q9WQ7AG7LCk6+gZFDAJCWEAVAkDTucseuPaljsvnJ4Rj60kan5QDn30g5i1fuPPqStzEFY5+U4K5fzqi1vxRcPPn8lvWcmczMTJcvUqedFbVeJTIAIAKoMjY4TWTaltlZ4dnf8NpZUevypOqsXk+2k1JWqUTGUl+1qVHyuMsdu/akjsm72yvdfpC5Gg9H8cqdR1/yNqZg7JMS3PXLGbX2l9RL1jUzFuXl5Thy5Aiamppslo8ZM8aroCgwauq8S2R83Z7U8u3Lyd0umHkbq9TtD9ee86odR+0F43x4G1Mw9kkJ/trPiLwlK5n56aefcNNNN2H//v3Wa2UAWJ83E0zXzJB0qfHaoG5Pavn25eRuF8y8jVXq9pnJMV6146i9YJwPb2MKxj4pwV/7GZG3ZP3M9Oijj8JgMODEiROIiYnBgQMHsGXLFvTt2xebNm1SOETylzxDMtITvDv5CGi9myE9IQrOfim3lLHc3u1JfHqd1uN6PdlOSlklLwEQAKQnRCE9wXmb7cvLGbv2pI7J5Pwsl+WA1vHwZE7kzqMveRtTMPZJCe765Yxa+0vqJSuZ2b59O/7whz/goosugkajgUajwaBBg1BSUoJHHnlE6RjJT8I0AorGZMve3nLCKyzIRtGYHjbLHJXx9MLAMI2AwoJsj+v1ZDspZe8fbGi9ENij6O1Zti8a08M67q7q9Gbs2pM6JpHhGpflBLSOh7t62sYrdx59yduYgrFPSnDVL2fU3F9SL1nJTEtLC+Li4gAAnTp1wvHjxwG0Xhj8/fffKxcd+d2oHD0WT8q1PkfGmcSYCLsy6Tqt9XbMUTl6LJqUi3Sd1mkZufHJqdeT7dyVnX19tsP1zs7bep0W04YYoHfRtrM2Pemjp6SOidzxcBWvr/YPb3gbUzD2SQnO+iVlvybyF1m3Zg8ePBiPP/44xo0bh4kTJ+L06dN47rnnsGTJEuzZswdlZWW+iFUW3potD58AzCcAezsefAJw8PRJCXwCMPmbz58z89lnn+Hs2bMYP348fvrpJ9x444347rvvkJKSgvfff9/ur2kHEpMZIiIi9fF5MuNIbW0tkpKSbP6C9rFjx5CRkeHwicH+wmSGiIhIfXz+0DxHkpOTbRIZAMjOzkZlZaVSTRARERHZ8elXJgp96UNERETkVOB+/yEiIiJSAJMZIiIiUjUmM0RERKRqPk1m2l8QTERERKQ0XgBMREREqibrr2ZLVV5ejoyMDF82QURERCFOVjLT0NCAv/71r9i4cSNqampgNptt1u/duxcA0KVLF+8jJCIiInJBVjJzzz33YP369bjllluQl5fHa2OIiIgoYGQlMx9//DE++eQTXHXVVUrHQ0REROQRWRcA/+Y3v0F8fLzSsRARERF5TFYy8/LLL2PWrFk4fPiw0vEQEREReUTWz0x9+/ZFQ0MDLr74YsTExCAiIsJmfW1trSLBEREREbkjK5mZMGECfv75Z7z44otIS0vjBcBEREQUMLKSmW3btmH79u3o3bu30vGQDC1mETsralFtakBtfSOSYyORrotGniEZYRrBur6mrgGp8VrkGZIBwG6Zs7JhGs+TVaXq8XWdSvJlfMHY91CcY3cs8VedOY9vjp6GCMCQEovJ+VmIDHf8q771+DWeR+3ZJiTFROL0uSa749jifFMLitfsx6bvTwIAsvUJuK5HGoznmlFebYLxfBOO1Z6DIAjomhyDKzKTcMLUiMzkGJdxKNFvX86b2vcN8i1BlPGY3tzcXCxcuBADBgzwqvEtW7bgpZdewp49e1BVVYVVq1Zh3LhxDstOmzYNS5YswZ///GfMmDFDchsmkwk6nQ5GoxEJCQlexRuM1pZVoXhNOaqMDXbr9DotxvTWY/W3VTbrE2NafxY8c+6C27J6nRaFBdkYlaP3KiY59fi6TiX5Mr5g7HsozrE7ro5FjQDcP9iA2ddnS97Gou0Y3P/OLqwvr5Edo7M4vOGPeVP7vkHyePL5LSuZWbduHYqLi/HCCy+gZ8+edtfMSE0aPv30U3z11VfIzc3FzTff7DSZ+eijj1BUVIRffvkFTz75JJOZ/7e2rAoPLN8LX/7RCMu/exZNypV00nAWk6f1+LpOJfkyvmDseyjOsTtSj8VpQ35NJDw5fgUAPTsnYN8xk7eh2sXhDX/Mm9r3DZLPk89vWd83jho1Ctu3b8fVV1+N1NRUJCUlISkpCYmJiUhKSpJcz+jRozFnzhyMHz/eaZmff/4Z06dPxz/+8Q+7pCmUtZhFFK8p92kiA8Baf/GacrSYXbfmKiZP6vF1nUryZXzB2PdQnGN3PDkWl26tQFOz2ePjVwQUS2TaxuENf8yb2vcN8h9Z18xs3LhR6TgcMpvNmDx5Mp588kn06NFD0jaNjY1obGy0vjeZlDsBBJOdFbUuv5pWkgigytiAnRW1yO+WIjsmqfX4uk4l+TK+YOx7KM6xO54ci2YReHd7JbIzdH47fl3Fce/gi2XX4Y95U/u+Qf4jK5kZOnSo0nE4NG/ePISHh+ORRx6RvE1JSQmKi4t9GFVwqKnz/4nQXZtSY/Ikdl/UqSRfxheMfQ/FOXbH07gO155Dp/goH0XjWRze8Me8qX3fIP+Rlcxs2bLF5fohQ4bICqatPXv24LXXXsPevXs9uvV79uzZmDlzpvW9yWTqkH/wMjVeG3RtSo3Jk9h9UaeSfBlfMPY9FOfYHU/jykyOCYq+ZCbHeLW9P+ZN7fsG+Y+sZGbYsGF2y9omHC0tLbIDsti6dStqamrQtWtXm3off/xxvPrqq6isrHS4XVRUFKKiAv+vHl/LMyRDr9Oi2tjg8+tmBADpul9v6ZYbk9R6fF2nknwZXzD2PRTn2B1PjkWNAEzOz0KYRoBepw3YT02WOLzhj3lT+75B/iPrAuDTp0/bvGpqarB27Vr069cP69atUySwyZMnY9++fSgtLbW+MjIy8OSTT+Kzzz5TpA01C9MIKCxQ7vZKZywpamFBtttnOrSNqX1JT+rxdZ1K8mV8wdj3UJxjdzw5Fu8fbEBkuMa6jdQeCQB6dVbubkxLHN7wx7ypfd8g/5G1N+t0OptXp06dMHLkSMyfPx9PPfWU5Hrq6+utiQoAVFRUoLS0FEeOHEFKSgpycnJsXhEREUhPT0f37t3lhN3hjMrRY9GkXOh1jr9i1eu0mDbEYLc+KSbC+qwZd2XTdVqPbn20xJTuZT2+rlNJvowvGPseinPsjrtjUSPY3w7tbhsL/f+PwerpgzEyO9WrOB3F4Q1/zJva9w3yD1nPmXHm4MGD6NevH+rr6yWV37RpE4YPH263fMqUKVi2bJnd8qysLMyYMYPPmWmHTwAODnwCcMefY3f4BGA+AZiU4/OH5u3bt8/mvSiKqKqqwty5c3HhwgV89dVXnlbpM6GQzBAREXU0nnx+y7oAuE+fPhAEAe3zoAEDBuDNN9+UUyURERGRLLKSmYqKCpv3Go0GF110EbRa3h5HRERE/iUrmcnMzMSGDRuwYcMG1NTUwGy2fSw2v50hIiIif5GVzBQXF+MPf/gD+vbtC71e79FD7YiIiIiUJCuZWbx4MZYtW4bJkycrHQ8RERGRR2Tdo9fU1ISBAwcqHQsRERGRx2QlM/fddx9WrFihdCxEREREHpP1M1NDQwOWLFmCzz//HL169UJEhO3TZF955RVFgiMiIiJyR1Yys2/fPvTp0wcAUFZWZrOOFwMTERGRP8lKZjZu3Kh0HERERESyKP9HOoiIiIj8iMkMERERqRqTGSIiIlI1JjNERESkakxmiIiISNWYzBAREZGqybo1m4JXi1nEzopa1NQ1IDVeizxDMsI0gtN1AJyWl6Kp2Yx3t1ficO05ZCbHYHJ+FiLDlcmRvY3X1VgoUZ6Cn7s59fech9o+Fmr9pcBhMtOBrC2rQvGaclQZG6zL9DotCguyAcBuXWJM65Obz5y7YFd+VI7ebXsln5Rj6dYKmMVfl73wyUHcP9iA2ddnK94XT+J1NRaO+uZpeQp+7ubU33MeavtYqPWXAksQRVF0X0y9TCYTdDodjEYjEhISAh2Oz6wtq8IDy/ei/WQKgN0yVyz/Zlo0KdflCafkk3K8vqXC6fppQ+QnNM764oijeF2NRfuycspT8HM3p1OHGLBkS4Xf5jzU9rFQ6y/5hief37xmpgNoMYsoXlPu8MPf00zVUr54TTlazI63bmo2Y+lW54kMACzdWoGmZrOHrbvuiyPt45UyFm375ml5Cn7u5lRE6/7przkPtX0s1PpLwYHJTAews6LW5qtcb4kAqowN2FlR63D9u9sr4e48ZBZby3lKTl/axutu+/Z987Q8BT8p+5Cr/VfpOQ+1fSzU+kvBgdfMdAA1dcolMlLqPVx7TtL2UstJaVPpbS1lpW7jqzEm5Sk1V/6up6PsY6HWXwoOTGY6gNR4rV/rzUyOkbS91HJS2lR6W0tZqdv4aoxJeUrNlb/r6Sj7WKj1l4IDf2bqAPIMydDrtFDqhkcBrXcdWG6Fbm9yfhbc3V2pEVrLeUpOX9rG62779n3ztDwFPyn7kEaA3+Y81PaxUOsvBQcmMx1AmEaw3n7d/gQiOPl/ZyxlCguynT4PIjJcg/sHG1zWc/9gg6znzbjqiyPt45UyFm375ml5Cn7u5lQArPuvP+Y81PaxUOsvBQcmMx3EqBw9Fk3KRbrO9qvbdJ0WiyflYrGDdUkxEdZnt7QtL+W2ydnXZ2PaEIPdNzQawbvbsl31RWq8rsbCUd88LU/Bz92czr4+269zHmr7WKj1lwKPz5npYPgEYNfb8wnAoYVPAA6sUOsvKcuTz28mM0RERBR0+NA8IiIiChlMZoiIiEjVmMwQERGRqjGZISIiIlVjMkNERESqxmSGiIiIVI3JDBEREakakxkiIiJSNSYzREREpGpMZoiIiEjVmMwQERGRqjGZISIiIlVjMkNERESqxmSGiIiIVI3JDBEREakakxkiIiJStfBAB0ChrcUsYmdFLWrqGpAar0WeIRlhGiHQYRERkYoE9JuZLVu2oKCgABkZGRAEAR999JF13YULFzBr1iz07NkTsbGxyMjIwJ133onjx48HLmBS1NqyKgya9wUmLN2BR1eWYsLSHRg07wusLasKdGhERKQiAU1mzp49i969e2PBggV2686dO4e9e/fi97//Pfbu3YsPP/wQP/zwA8aMGROASElpa8uq8MDyvagyNtgsrzY24IHle5nQEBGRZIIoimKggwAAQRCwatUqjBs3zmmZXbt2IS8vD4cPH0bXrl0l1WsymaDT6WA0GpGQkKBQtOSNFrOIQfO+sEtkLAQA6Totvpw1gj85ERGFKE8+v1V1AbDRaIQgCEhMTHRaprGxESaTyeZFwWVnRa3TRAYARABVxgbsrKj1X1BERKRaqklmGhoa8PTTT2PixIkuM7SSkhLodDrrq0uXLn6MkqSoqXOeyMgpR0REoU0VycyFCxdw++23w2w2Y+HChS7Lzp49G0aj0fo6evSon6IkqVLjtYqWIyKi0Bb0t2ZfuHABt956KyoqKvDFF1+4/d0sKioKUVFRfoqO5MgzJEOv06La2ABHF2xZrpnJMyT7OzQiIlKhoP5mxpLI/Pe//8Xnn3+OlJSUQIdECgjTCCgsyAbQmri0ZXlfWJDNi3+JiEiSgCYz9fX1KC0tRWlpKQCgoqICpaWlOHLkCJqbm3HLLbdg9+7d+Mc//oGWlhZUV1ejuroaTU1NgQybFDAqR49Fk3KRrrP9KSldp8WiSbkYlaMPUGRERKQ2Ab01e9OmTRg+fLjd8ilTpqCoqAgGg8Hhdhs3bsSwYcMktcFbs4MbnwBMRESOePL5HdBrZoYNGwZXuVSQPAKHfChMIyC/G38+JCIi+YL6mhkiIiIid5jMEBERkaoxmSEiIiJVYzJDREREqsZkhoiIiFSNyQwRERGpGpMZIiIiUjUmM0RERKRqTGaIiIhI1ZjMEBERkaoxmSEiIiJVYzJDREREqsZkhoiIiFSNyQwRERGpGpMZIiIiUjUmM0RERKRq4YEOgIgoVLSYReysqEVNXQNS47XIMyQjTCMEOiwi1WMyQ0TkB2vLqlC8phxVxgbrMr1Oi8KCbIzK0QcwMiL1489MREQ+trasCg8s32uTyABAtbEBDyzfi7VlVQGKjKhjYDJDRORDLWYRxWvKITpYZ1lWvKYcLWZHJYhICiYzREQ+tLOi1u4bmbZEAFXGBuysqPVfUEQdDJMZIiIfqqlznsjIKUdE9pjMEBH5UGq8VtFyRGSPyQwRkQ/lGZKh12nh7AZsAa13NeUZkv0ZFlGHwmSGiMiHwjQCCguyAcAuobG8LyzI5vNmiLzAZIaIyMdG5eixaFIu0nW2PyWl67RYNCmXz5kh8hIfmkdE5AejcvQYmZ3OJwAT+QCTGSIiPwnTCMjvlhLoMIg6HP7MRERERKrGZIaIiIhUjckMERERqRqTGSIiIlI1JjNERESkakxmiIiISNWYzBAREZGqMZkhIiIiVWMyQ0RERKrGZIaIiIhUjckMERERqRqTGSIiIlI1JjNERESkakxmiIiISNWYzBAREZGqMZkhIiIiVWMyQ0RERKoW0GRmy5YtKCgoQEZGBgRBwEcffWSzXhRFFBUVISMjA9HR0Rg2bBgOHDgQmGCJiIgoKAU0mTl79ix69+6NBQsWOFw/f/58vPLKK1iwYAF27dqF9PR0jBw5EnV1dX6OlIiIiIJVeCAbHz16NEaPHu1wnSiKePXVV/Hss89i/PjxAIC3334baWlpWLFiBaZNm+bPUImIiChIBe01MxUVFaiursa1115rXRYVFYWhQ4di27ZtTrdrbGyEyWSyeREREVHHFbTJTHV1NQAgLS3NZnlaWpp1nSMlJSXQ6XTWV5cuXXwaJxEREQVW0CYzFoIg2LwXRdFuWVuzZ8+G0Wi0vo4ePerrEImIiCiAAnrNjCvp6ekAWr+h0ev11uU1NTV239a0FRUVhaioKJ/HR0RERMEhaL+ZMRgMSE9Px/r1663LmpqasHnzZgwcODCAkREREVEwCeg3M/X19fjxxx+t7ysqKlBaWork5GR07doVM2bMwIsvvohLLrkEl1xyCV588UXExMRg4sSJAYyaiIiIgklAk5ndu3dj+PDh1vczZ84EAEyZMgXLli3DU089hfPnz+PBBx/E6dOn0b9/f6xbtw7x8fGBCpmIiIiCjCCKohjoIHzJZDJBp9PBaDQiISEh0OEQERGRBJ58fgftNTNEREREUjCZISIiIlVjMkNERESqxmSGiIiIVI3JDBEREakakxkiIiJSNSYzREREpGpMZoiIiEjVmMwQERGRqjGZISIiIlVjMkNERESqxmSGiIiIVI3JDBEREakakxkiIiJSNSYzREREpGpMZoiIiEjVwgMdQEfTYhaxs6IWNXUNSI3XIs+QjDCNEOiwiIiIOiwmMwpaW1aF4jXlqDI2WJfpdVoUFmRjVI4+gJERERF1XPyZSSFry6rwwPK9NokMAFQbG/DA8r1YW1YVoMiIiIg6NiYzCmgxiyheUw7RwTrLsuI15WgxOypBRERE3mAyo4CdFbV238i0JQKoMjZgZ0Wt/4IiIiIKEUxmFFBT5zyRkVOOiIiIpGMyo4DUeK2i5YiIiEg6JjMKyDMkQ6/TwtkN2AJa72rKMyT7MywiIqKQwGRGAWEaAYUF2QBgl9BY3hcWZPN5M0RERD7AZEYho3L0WDQpF+k625+S0nVaLJqUy+fMEBER+QgfmqegUTl6jMxO5xOAiYiI/IjJjMLCNALyu6UEOgwiIqKQwZ+ZiIiISNWYzBAREZGqMZkhIiIiVWMyQ0RERKrGZIaIiIhUjckMERERqRqTGSIiIlI1JjNERESkakxmiIiISNU6/BOARVEEAJhMpgBHQkRERFJZPrctn+OudPhkpq6uDgDQpUuXAEdCREREnqqrq4NOp3NZRhClpDwqZjabcfz4ccTHx0MQ1P8HH00mE7p06YKjR48iISEh0OGENM5F8OBcBAfOQ/DoCHMhiiLq6uqQkZEBjcb1VTEd/psZjUaDzp07BzoMxSUkJKh2B+1oOBfBg3MRHDgPwUPtc+HuGxkLXgBMREREqsZkhoiIiFSNyYzKREVFobCwEFFRUYEOJeRxLoIH5yI4cB6CR6jNRYe/AJiIiIg6Nn4zQ0RERKrGZIaIiIhUjckMERERqRqTGSIiIlI1JjMqsmXLFhQUFCAjIwOCIOCjjz4KdEghqaSkBP369UN8fDxSU1Mxbtw4fP/994EOK+QsWrQIvXr1sj4ULD8/H59++mmgwwp5JSUlEAQBM2bMCHQoIaeoqAiCINi80tPTAx2WXzCZUZGzZ8+id+/eWLBgQaBDCWmbN2/GQw89hB07dmD9+vVobm7Gtddei7NnzwY6tJDSuXNnzJ07F7t378bu3bsxYsQIjB07FgcOHAh0aCFr165dWLJkCXr16hXoUEJWjx49UFVVZX3t378/0CH5RYf/cwYdyejRozF69OhAhxHy1q5da/P+rbfeQmpqKvbs2YMhQ4YEKKrQU1BQYPP+hRdewKJFi7Bjxw706NEjQFGFrvr6etxxxx1YunQp5syZE+hwQlZ4eHjIfBvTFr+ZIfKS0WgEACQnJwc4ktDV0tKClStX4uzZs8jPzw90OCHpoYcewg033IBrrrkm0KGEtP/+97/IyMiAwWDA7bffjp9++inQIfkFv5kh8oIoipg5cyYGDRqEnJycQIcTcvbv34/8/Hw0NDQgLi4Oq1atQnZ2dqDDCjkrV67E3r17sWvXrkCHEtL69++Pd955B5deeilOnDiBOXPmYODAgThw4ABSUlICHZ5PMZkh8sL06dOxb98+fPnll4EOJSR1794dpaWlOHPmDP71r39hypQp2Lx5MxMaPzp69CgeffRRrFu3DlqtNtDhhLS2lyH07NkT+fn56NatG95++23MnDkzgJH5HpMZIpkefvhhrF69Glu2bEHnzp0DHU5IioyMxG9/+1sAQN++fbFr1y689tpreP311wMcWejYs2cPampqcOWVV1qXtbS0YMuWLViwYAEaGxsRFhYWwAhDV2xsLHr27In//ve/gQ7F55jMEHlIFEU8/PDDWLVqFTZt2gSDwRDokOj/iaKIxsbGQIcRUq6++mq7O2buvvtuXHbZZZg1axYTmQBqbGzEwYMHMXjw4ECH4nNMZlSkvr4eP/74o/V9RUUFSktLkZycjK5duwYwstDy0EMPYcWKFfj3v/+N+Ph4VFdXAwB0Oh2io6MDHF3oeOaZZzB69Gh06dIFdXV1WLlyJTZt2mR3txn5Vnx8vN31YrGxsUhJSeF1ZH72xBNPoKCgAF27dkVNTQ3mzJkDk8mEKVOmBDo0n2MyoyK7d+/G8OHDre8tv4FOmTIFy5YtC1BUoWfRokUAgGHDhtksf+utt3DXXXf5P6AQdeLECUyePBlVVVXQ6XTo1asX1q5di5EjRwY6NKKAOHbsGCZMmICTJ0/ioosuwoABA7Bjxw5kZmYGOjSfE0RRFAMdBBEREZFcfM4MERERqRqTGSIiIlI1JjNERESkakxmiIiISNWYzBAREZGqMZkhIiIiVWMyQ0RERKrGZIaISIaioiL06dMn0GEQEZjMEBERkcoxmSEiIiJVYzJDRE4NGzYMjzzyCJ566ikkJycjPT0dRUVFAIDKykoIgoDS0lJr+TNnzkAQBGzatAkAsGnTJgiCgM8++wxXXHEFoqOjMWLECNTU1ODTTz/F5ZdfjoSEBEyYMAHnzp2TFNPatWsxaNAgJCYmIiUlBTfeeCMOHTpkXW+J64MPPsDgwYMRHR2Nfv364YcffsCuXbvQt29fxMXFYdSoUfjll1+s223atAl5eXmIjY1FYmIirrrqKhw+fNi6fu7cuUhLS0N8fDzuvfdeNDQ02I3VjBkzbJaNGzeOf6+LyA+YzBCRS2+//TZiY2Px9ddfY/78+fjDH/6A9evXe1RHUVERFixYgG3btuHo0aO49dZb8eqrr2LFihX4+OOPsX79evz1r3+VVNfZs2cxc+ZM7Nq1Cxs2bIBGo8FNN90Es9lsU66wsBDPPfcc9u7di/DwcEyYMAFPPfUUXnvtNWzduhWHDh3C888/DwBobm7GuHHjMHToUOzbtw/bt2/H1KlTIQgCAOCDDz5AYWEhXnjhBezevRt6vR4LFy70aAyIyIdEIiInhg4dKg4aNMhmWb9+/cRZs2aJFRUVIgDxm2++sa47ffq0CEDcuHGjKIqiuHHjRhGA+Pnnn1vLlJSUiADEQ4cOWZdNmzZNvO6662TFWFNTIwIQ9+/fL4qiaI3rjTfesJZ57733RADihg0bbOLo3r27KIqieOrUKRGAuGnTJodt5Ofni7/73e9slvXv31/s3bu39f3QoUPFRx991KbM2LFjxSlTpsjqFxFJx29miMilXr162bzX6/WoqamRXUdaWhpiYmJw8cUX2yyTWuehQ4cwceJEXHzxxUhISIDBYAAAHDlyxGWbANCzZ0+HbSYnJ+Ouu+7Cddddh4KCArz22muoqqqylj148CDy8/Nt6m//nogCh8kMEbkUERFh814QBJjNZmg0racPURSt6y5cuOC2DkEQnNYpRUFBAU6dOoWlS5fi66+/xtdffw0AaGpqctmmo2Vt23zrrbewfft2DBw4EO+//z4uvfRS7NixQ1JMAKDRaGzGAnA+HkSkLCYzRCTLRRddBAA232C0vRjYF06dOoWDBw/iueeew9VXX43LL78cp0+fVqz+K664ArNnz8a2bduQk5ODFStWAAAuv/xyu8Sm/fuLLrrIZixaWlpQVlamWGxE5Fx4oAMgInWKjo7GgAEDMHfuXGRlZeHkyZN47rnnfNpmUlISUlJSsGTJEuj1ehw5cgRPP/201/VWVFRgyZIlGDNmDDIyMvD999/jhx9+wJ133gkAePTRRzFlyhT07dsXgwYNwj/+8Q8cOHDA5qeyESNGYObMmfj444/RrVs3/PnPf8aZM2e8jo2I3OM3M0Qk25tvvokLFy6gb9++ePTRRzFnzhyftqfRaLBy5Urs2bMHOTk5eOyxx/DSSy95XW9MTAy+++473Hzzzbj00ksxdepUTJ8+HdOmTQMA3HbbbXj++ecxa9YsXHnllTh8+DAeeOABmzruueceTJkyBXfeeSeGDh0Kg8GA4cOHex0bEbkniO1/5CUiIiJSEX4zQ0RERKrGZIaIgsaRI0cQFxfn9NX+9msiIoA/MxFREGlubkZlZaXT9VlZWQgP530LRGSLyQwRERGpGn9mIiIiIlVjMkNERESqxmSGiIiIVI3JDBEREakakxkiIiJSNSYzREREpGpMZoiIiEjVmMwQERGRqv0fqPNM8es/tH8AAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "\n",
    "data = resource\n",
    "std = data['num_amsdu'].std()\n",
    "mean = data['num_amsdu'].mean()\n",
    "correlation = data['num_amsdu'].corr(data['num_ampdu'])\n",
    "print(f'num_amsdu std is {std}')\n",
    "print(f'num_amsdu mean is {mean}')\n",
    "print(f'pearson between num_amsdu and num_ampdu is {correlation}')\n",
    "#['num_amsdu'] = data['num_amsdu'].round()\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# 绘制散点图\n",
    "plt.scatter(data['num_amsdu'], data['num_ampdu'])\n",
    "plt.title('Scatter plot of num_amsdu vs num_ampdu')\n",
    "plt.xlabel('num_amsdu')\n",
    "plt.ylabel('num_ampdu')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "52733b79",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:52:49.132282Z",
     "start_time": "2024-09-24T12:52:49.127698Z"
    }
   },
   "outputs": [],
   "source": [
    "# 删除无用的值\n",
    "\n",
    "drop_columns=['pd','ed','ap_from_ap_0_mean_ant_rssi','ap_from_ap_0_max_ant_rssi','num_amsdu']\n",
    "\n",
    "# drop_columns=['pd','ed','ap_from_ap_0_mean_ant_rssi','ap_from_ap_0_max_ant_rssi'\n",
    "#               ,'sta_to_ap_0_max_ant_rssi','sta_to_ap_0_mean_ant_rssi'\n",
    "#               ,'sta_to_ap_1_max_ant_rssi','sta_to_ap_1_mean_ant_rssi'\n",
    "#              ,'sta_from_ap_0_max_ant_rssi','sta_from_ap_0_mean_ant_rssi'\n",
    "#              ,'sta_from_ap_1_max_ant_rssi','sta_from_ap_1_mean_ant_rssi']\n",
    "\n",
    "resource=resource.drop(drop_columns,axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "5bf54a7b",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:52:50.866961Z",
     "start_time": "2024-09-24T12:52:50.858428Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 389 entries, 0 to 388\n",
      "Data columns (total 18 columns):\n",
      " #   Column                      Non-Null Count  Dtype  \n",
      "---  ------                      --------------  -----  \n",
      " 0   protocol                    389 non-null    object \n",
      " 1   nav                         389 non-null    int64  \n",
      " 2   ap_from_ap_0_sum_ant_rssi   389 non-null    object \n",
      " 3   sta_to_ap_0_sum_ant_rssi    389 non-null    object \n",
      " 4   sta_to_ap_1_sum_ant_rssi    389 non-null    object \n",
      " 5   sta_from_ap_0_sum_ant_rssi  389 non-null    object \n",
      " 6   sta_from_ap_1_sum_ant_rssi  389 non-null    object \n",
      " 7   sta_from_sta_0_rssi         389 non-null    int64  \n",
      " 8   nss                         389 non-null    int64  \n",
      " 9   mcs                         389 non-null    int64  \n",
      " 10  per                         389 non-null    float64\n",
      " 11  num_ampdu                   389 non-null    int64  \n",
      " 12  ppdu_dur                    389 non-null    float64\n",
      " 13  seq_time                    389 non-null    float64\n",
      " 14  throughput                  389 non-null    float64\n",
      " 15  pro_pd                      389 non-null    float64\n",
      " 16  pro_ed                      389 non-null    float64\n",
      " 17  pro_nav                     389 non-null    float64\n",
      "dtypes: float64(7), int64(5), object(6)\n",
      "memory usage: 54.8+ KB\n"
     ]
    }
   ],
   "source": [
    "resource.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "48128984",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:52:52.121603Z",
     "start_time": "2024-09-24T12:52:51.756104Z"
    }
   },
   "outputs": [],
   "source": [
    "# 对RSSI求平均值处理\n",
    "for i in range(2,7):\n",
    "    resource.iloc[:,i]=resource.iloc[:,i].apply(lambda x: sum(eval(x))/len(eval(x)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "4ada103a",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:52:52.884239Z",
     "start_time": "2024-09-24T12:52:52.879743Z"
    }
   },
   "outputs": [],
   "source": [
    "nss_mcs_dict=dict()\n",
    "phy=[8.6,17.2,25.8,34.4,51.6,68.8,77.4,86.0,103.2,114.7,129.0,143.4,17.2,34.4,51.6,68.8,103.2,137.6,154.9,172.1,206.5,\n",
    "    229.4,258.1,286.8]\n",
    "num=0\n",
    "\n",
    "for i in range(1,3):\n",
    "    for j in range(0,12):\n",
    "        nss_mcs_dict[(i,j)]=phy[num]\n",
    "        num+=1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "c3d5706d",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:52:53.931343Z",
     "start_time": "2024-09-24T12:52:53.928213Z"
    }
   },
   "outputs": [],
   "source": [
    "def get_phy(nss,mcs):\n",
    "    return nss_mcs_dict[nss,mcs]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "5a54ce0c",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:52:54.809040Z",
     "start_time": "2024-09-24T12:52:54.799765Z"
    }
   },
   "outputs": [],
   "source": [
    "resource['phy_rate']=resource.apply(lambda row:get_phy(row['nss'],row['mcs']),axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "c12b10ce",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:52:55.659043Z",
     "start_time": "2024-09-24T12:52:55.654655Z"
    }
   },
   "outputs": [],
   "source": [
    "resource=resource.drop(['nss','mcs'],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "8aa6c7aa",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:52:57.761407Z",
     "start_time": "2024-09-24T12:52:57.756956Z"
    }
   },
   "outputs": [],
   "source": [
    "resource.iloc[:,0]=resource.iloc[:,0].astype(str).map({'tcp':1,'udp':0})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "07d4bd08",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:53:01.466009Z",
     "start_time": "2024-09-24T12:53:01.458083Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 389 entries, 0 to 388\n",
      "Data columns (total 17 columns):\n",
      " #   Column                      Non-Null Count  Dtype  \n",
      "---  ------                      --------------  -----  \n",
      " 0   protocol                    389 non-null    object \n",
      " 1   nav                         389 non-null    int64  \n",
      " 2   ap_from_ap_0_sum_ant_rssi   389 non-null    object \n",
      " 3   sta_to_ap_0_sum_ant_rssi    389 non-null    object \n",
      " 4   sta_to_ap_1_sum_ant_rssi    389 non-null    object \n",
      " 5   sta_from_ap_0_sum_ant_rssi  389 non-null    object \n",
      " 6   sta_from_ap_1_sum_ant_rssi  389 non-null    object \n",
      " 7   sta_from_sta_0_rssi         389 non-null    int64  \n",
      " 8   per                         389 non-null    float64\n",
      " 9   num_ampdu                   389 non-null    int64  \n",
      " 10  ppdu_dur                    389 non-null    float64\n",
      " 11  seq_time                    389 non-null    float64\n",
      " 12  throughput                  389 non-null    float64\n",
      " 13  pro_pd                      389 non-null    float64\n",
      " 14  pro_ed                      389 non-null    float64\n",
      " 15  pro_nav                     389 non-null    float64\n",
      " 16  phy_rate                    389 non-null    float64\n",
      "dtypes: float64(8), int64(3), object(6)\n",
      "memory usage: 51.8+ KB\n"
     ]
    }
   ],
   "source": [
    "resource.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "f60210c4",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:53:03.830557Z",
     "start_time": "2024-09-24T12:53:02.504290Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "D:\\Anaconda\\Lib\\site-packages\\sklearn\\cluster\\_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
      "  super()._check_params_vs_input(X, default_n_init=10)\n",
      "D:\\Anaconda\\Lib\\site-packages\\sklearn\\cluster\\_kmeans.py:1436: UserWarning: KMeans is known to have a memory leak on Windows with MKL, when there are less chunks than available threads. You can avoid it by setting the environment variable OMP_NUM_THREADS=2.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cluster centers:\n",
      "          0         1         2\n",
      "0  0.897046  0.003141  0.924525\n",
      "1  0.063595  0.001679  0.034990\n",
      "2  0.770176  0.001579  0.150017\n",
      "cluster\n",
      "0    178\n",
      "1    170\n",
      "2     41\n",
      "Name: count, dtype: int64\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ShenFei\\AppData\\Local\\Temp\\ipykernel_17752\\904818509.py:10: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df['cluster'] = kmeans.fit_predict(df)\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.cluster import KMeans\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.metrics import silhouette_score\n",
    "\n",
    "df=resource.iloc[:,[13,14,15]]\n",
    "k = 3\n",
    "kmeans = KMeans(n_clusters=k, random_state=42)\n",
    "df['cluster'] = kmeans.fit_predict(df)\n",
    "\n",
    "# 查看每个簇的平均特征值\n",
    "cluster_centers = pd.DataFrame(kmeans.cluster_centers_)\n",
    "# columns=['pro_pd', 'pro_ed', 'pro_nav']\n",
    "\n",
    "print(f\"Cluster centers:\\n{cluster_centers}\")\n",
    "\n",
    "# 查看每个簇的样本数量\n",
    "print(df['cluster'].value_counts())\n",
    "resource['state']=df['cluster']\n",
    "\n",
    "# # 可视化聚类结果\n",
    "# plt.scatter(df['pro_nav'], df['pro_pd'], c=df['cluster'], cmap='viridis', label='Cluster')\n",
    "# plt.colorbar()\n",
    "# plt.xlabel('pro_nav')\n",
    "# plt.ylabel('pro_pd')\n",
    "# plt.title('KMeans Clustering of WLAN Communication States')\n",
    "# plt.show()\n",
    "# 3D plot of the clustering results\n",
    "# fig = plt.figure(figsize=(10, 7))\n",
    "# ax = fig.add_subplot(111, projection='3d')\n",
    "\n",
    "# # Plot the clusters in 3D\n",
    "# scatter = ax.scatter(df['pro_nav'], df['pro_pd'], df['pro_ed'], c=df['cluster'], cmap='viridis')\n",
    "\n",
    "# # Set labels and title\n",
    "# ax.set_xlabel('pro_nav')\n",
    "# ax.set_ylabel('pro_pd')\n",
    "# ax.set_zlabel('pro_ed')\n",
    "# ax.set_title('3D KMeans Clustering of WLAN Communication States')\n",
    "\n",
    "# # Add a color bar\n",
    "# plt.colorbar(scatter)\n",
    "\n",
    "# plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "af497117",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T13:32:50.888701Z",
     "start_time": "2024-09-24T13:32:50.854162Z"
    }
   },
   "outputs": [],
   "source": [
    "# 信干比采取dbm直接相减得到db\n",
    "def func(x, y, state):\n",
    "    if state == 0:\n",
    "        return x + 100\n",
    "    else:\n",
    "        return x - y\n",
    "\n",
    "    \n",
    "# 创建一个新列来存储 SINR 计算的结果\n",
    "sinr = []\n",
    "df = resource\n",
    "# 遍历 DataFrame 的每一行\n",
    "for i in range(len(df)):\n",
    "    if df.iloc[i, list(df.columns).index('bss_id')] == 1:  # 偶数行\n",
    "        result = func(df.iloc[i, list(df.columns).index('sta_from_ap_1_sum_ant_rssi')], df.iloc[i, list(df.columns).index('sta_from_ap_0_sum_ant_rssi')],df.iloc[i, list(df.columns).index('state')])\n",
    "    else:  # 奇数行\n",
    "        result = func(df.iloc[i, list(df.columns).index('sta_from_ap_0_sum_ant_rssi')], df.iloc[i, list(df.columns).index('sta_from_ap_1_sum_ant_rssi')],df.iloc[i, list(df.columns).index('state')])\n",
    "    sinr.append(result)\n",
    "\n",
    "# 将计算结果添加为新列\n",
    "df['sinr'] = sinr\n",
    "resource = resource.drop('bass_id',axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "37b9350f",
   "metadata": {},
   "source": [
    "# read from here"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "0902e4a8",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:44:28.324171Z",
     "start_time": "2024-09-24T12:44:28.078410Z"
    }
   },
   "outputs": [],
   "source": [
    "# resource.to_excel('programe3_2ap_processed.xlsx')\n",
    "resource = pd.read_excel('programe3_2ap_processed.xlsx')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3a738aca",
   "metadata": {},
   "source": [
    "# 处理"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "19b9bf94",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:54:44.821078Z",
     "start_time": "2024-09-24T12:54:44.817062Z"
    }
   },
   "outputs": [],
   "source": [
    "target=resource.iloc[:,12]\n",
    "resource=resource.drop('throughput',axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "c2c38934",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:54:46.401031Z",
     "start_time": "2024-09-24T12:54:46.396501Z"
    }
   },
   "outputs": [],
   "source": [
    "resource=resource.drop('sta_from_sta_0_rssi',axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "5a5318ca",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T12:54:48.121775Z",
     "start_time": "2024-09-24T12:54:48.111930Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 389 entries, 0 to 388\n",
      "Data columns (total 17 columns):\n",
      " #   Column                      Non-Null Count  Dtype  \n",
      "---  ------                      --------------  -----  \n",
      " 0   protocol                    389 non-null    object \n",
      " 1   nav                         389 non-null    int64  \n",
      " 2   ap_from_ap_0_sum_ant_rssi   389 non-null    object \n",
      " 3   sta_to_ap_0_sum_ant_rssi    389 non-null    object \n",
      " 4   sta_to_ap_1_sum_ant_rssi    389 non-null    object \n",
      " 5   sta_from_ap_0_sum_ant_rssi  389 non-null    object \n",
      " 6   sta_from_ap_1_sum_ant_rssi  389 non-null    object \n",
      " 7   per                         389 non-null    float64\n",
      " 8   num_ampdu                   389 non-null    int64  \n",
      " 9   ppdu_dur                    389 non-null    float64\n",
      " 10  seq_time                    389 non-null    float64\n",
      " 11  pro_pd                      389 non-null    float64\n",
      " 12  pro_ed                      389 non-null    float64\n",
      " 13  pro_nav                     389 non-null    float64\n",
      " 14  phy_rate                    389 non-null    float64\n",
      " 15  state                       389 non-null    int32  \n",
      " 16  sinr                        389 non-null    float64\n",
      "dtypes: float64(8), int32(1), int64(2), object(6)\n",
      "memory usage: 50.3+ KB\n"
     ]
    }
   ],
   "source": [
    "resource.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "4e497ad6",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T13:03:37.963210Z",
     "start_time": "2024-09-24T13:03:29.156456Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "\n",
    "rfg=RandomForestRegressor(n_estimators=100,criterion='friedman_mse',random_state=0,oob_score=False)\n",
    "\n",
    "\n",
    "# target_0_phy=pd.DataFrame(target_0_phy)\n",
    "\n",
    "rfg=rfg.fit(resource,target)\n",
    "\n",
    "# rfg0=RandomForestRegressor(n_estimators=100,criterion='friedman_mse',random_state=0,oob_score=False)\n",
    "# rfg0=rfg0.fit(resource,target_0.to_numpy())\n",
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "res=cross_val_score(rfg,resource,target,cv=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "2a1bfe56",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T13:03:40.552144Z",
     "start_time": "2024-09-24T13:03:40.547769Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.2900571945404747"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "c92f7758",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T13:03:41.919892Z",
     "start_time": "2024-09-24T13:03:41.914721Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.86379903,  0.96674704,  0.92845924,  0.83177476,  0.89834726,\n",
       "        0.52420142,  0.22583225,  0.52067251, -3.63591665,  0.7766551 ])"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "f38f60bf",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T13:03:46.477126Z",
     "start_time": "2024-09-24T13:03:46.461241Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1.19124459e-03, 4.45031323e-04, 2.65996446e-03, 3.41066465e-03,\n",
       "       3.23494945e-03, 2.58660543e-03, 1.26456161e-03, 6.35404717e-01,\n",
       "       6.20061103e-03, 1.62703754e-02, 2.38008579e-01, 1.53716861e-03,\n",
       "       1.89227020e-04, 2.79631704e-03, 7.72017944e-02, 2.33485904e-03,\n",
       "       5.26333021e-03])"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfg.feature_importances_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "db14b4ff",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-09-24T13:03:52.046385Z",
     "start_time": "2024-09-24T13:03:52.029027Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['pro_ed', 'nav', 'protocol', 'sta_from_ap_1_sum_ant_rssi', 'pro_pd',\n",
       "       'state', 'sta_from_ap_0_sum_ant_rssi', 'ap_from_ap_0_sum_ant_rssi',\n",
       "       'pro_nav', 'sta_to_ap_1_sum_ant_rssi', 'sta_to_ap_0_sum_ant_rssi',\n",
       "       'sinr', 'num_ampdu', 'ppdu_dur', 'phy_rate', 'seq_time', 'per'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "resource_columns = resource.columns\n",
    "resource_columns[np.argsort(rfg.feature_importances_)]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7f14b64e",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.4"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": false,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}