STLM_onevar_multiSteps.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>The code below used STLM by using only Capacity field to predict the RUL(STLM using one variable with multisteps)</b> \n",
    "<p>We built the model only on Battery B0005</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "import sys \n",
    "import numpy as np # linear algebra\n",
    "from scipy.stats import randint\n",
    "import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv), data manipulation as in SQL\n",
    "import matplotlib.pyplot as plt # this is used for the plot the graph \n",
    "import seaborn as sns # used for plot interactive graph. \n",
    "from sklearn.model_selection import train_test_split # to split the data into two parts\n",
    "#from sklearn.cross_validation import KFold # use for cross validation\n",
    "from sklearn.preprocessing import StandardScaler # for normalization\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.feature_selection import SelectFromModel\n",
    "from sklearn import metrics # for the check the error and accuracy of the model\n",
    "from sklearn.metrics import mean_squared_error,r2_score\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "df=pd.read_csv(\"discharge_onerow.csv\")\n",
    "featurs=['Batt_name','cycle','amb_temp','voltage_battery','current_battery','temp_battery','current_load','voltage_load','time','Capacity']\n",
    "f1=['cycle','Capacity']\n",
    "df=df[featurs]\n",
    "dataset=df[(df.Batt_name =='B0005')| (df.Batt_name =='B0006') | (df.Batt_name =='B0007')|(df.Batt_name =='B0018')]\n",
    "# Feature Selection\n",
    "dataset=dataset[(dataset['Batt_name']=='B0005')]\n",
    "dataset=dataset[f1]\n",
    "data_train=dataset[(dataset['cycle']<90)]\n",
    "data_set_train=data_train.iloc[:,1:2].values\n",
    "data_test=dataset[(dataset['cycle']>=90)]\n",
    "data_set_test=data_test.iloc[:,1:2].values\n",
    "\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "sc=MinMaxScaler(feature_range=(0,1))\n",
    "data_set_train=sc.fit_transform(data_set_train)\n",
    "data_set_test=sc.transform(data_set_test)\n",
    "\n",
    "X_train=[]\n",
    "y_train=[]\n",
    "#take the last 10t to predict 10t+1\n",
    "for i in range(10,89):\n",
    "    X_train.append(data_set_train[i-10:i,0])\n",
    "    y_train.append(data_set_train[i,0])\n",
    "X_train,y_train=np.array(X_train),np.array(y_train)\n",
    " \n",
    "X_train=np.reshape(X_train,(X_train.shape[0],X_train.shape[1],1))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Applied STLM </b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "regress=Sequential()\n",
    "regress.add(LSTM(units=50, return_sequences=True, input_shape=(X_train.shape[1],1)))\n",
    "regress.add(Dropout(0.3))\n",
    "\n",
    "regress.add(LSTM(units=50, return_sequences=True))\n",
    "regress.add(Dropout(0.3))\n",
    "\n",
    "\n",
    "regress.add(LSTM(units=50, return_sequences=True))\n",
    "regress.add(Dropout(0.3))\n",
    "\n",
    "regress.add(LSTM(units=50))\n",
    "regress.add(Dropout(0.3))\n",
    "\n",
    "\n",
    "regress.add(Dense(units=1))\n",
    "\n",
    "regress.compile(optimizer='adam',loss='mean_squared_error')\n",
    "\n",
    "regress.fit(X_train,y_train,epochs=200,batch_size=50)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Epoch 1/100\n",
    "79/79 [==============================] - 3s 43ms/step - loss: 0.4145\n",
    "\n",
    "Epoch 2/100\n",
    "79/79 [==============================] - 0s 924us/step - loss: 0.3334\n",
    "\n",
    "Epoch 3/100\n",
    "79/79 [==============================] - 0s 904us/step - loss: 0.2314\n",
    "\n",
    "Epoch 4/100\n",
    "79/79 [==============================] - 0s 915us/step - loss: 0.1041\n",
    "\n",
    "Epoch 5/100\n",
    "79/79 [==============================] - 0s 928us/step - loss: 0.0579\n",
    "\n",
    "Epoch 6/100\n",
    "79/79 [==============================] - 0s 934us/step - loss: 0.0703\n",
    "\n",
    "Epoch 7/100\n",
    "79/79 [==============================] - 0s 917us/step - loss: 0.0401\n",
    "\n",
    "Epoch 8/100\n",
    "79/79 [==============================] - 0s 904us/step - loss: 0.0533\n",
    "\n",
    "Epoch 9/100\n",
    "79/79 [==============================] - 0s 925us/step - loss: 0.0419\n",
    "\n",
    "Epoch 10/100\n",
    "79/79 [==============================] - 0s 905us/step - loss: 0.0424\n",
    "\n",
    "Epoch 11/100\n",
    "79/79 [==============================] - 0s 909us/step - loss: 0.0375\n",
    "\n",
    "Epoch 12/100\n",
    "79/79 [==============================] - 0s 893us/step - loss: 0.0414\n",
    "\n",
    "Epoch 13/100\n",
    "79/79 [==============================] - 0s 904us/step - loss: 0.0325\n",
    "\n",
    "Epoch 14/100\n",
    "79/79 [==============================] - 0s 933us/step - loss: 0.0275\n",
    "\n",
    "Epoch 15/100\n",
    "79/79 [==============================] - 0s 873us/step - loss: 0.0318\n",
    "\n",
    "Epoch 16/100\n",
    "79/79 [==============================] - 0s 918us/step - loss: 0.0192\n",
    "\n",
    "Epoch 17/100\n",
    "79/79 [==============================] - 0s 918us/step - loss: 0.0207\n",
    "\n",
    "Epoch 18/100\n",
    "79/79 [==============================] - 0s 909us/step - loss: 0.0172\n",
    "\n",
    "Epoch 19/100\n",
    "79/79 [==============================] - 0s 901us/step - loss: 0.0186\n",
    "\n",
    "Epoch 20/100\n",
    "79/79 [==============================] - 0s 922us/step - loss: 0.0141\n",
    "\n",
    "Epoch 21/100\n",
    "79/79 [==============================] - 0s 920us/step - loss: 0.0172\n",
    "\n",
    "Epoch 22/100\n",
    "79/79 [==============================] - 0s 903us/step - loss: 0.0123\n",
    "\n",
    "Epoch 23/100\n",
    "79/79 [==============================] - 0s 904us/step - loss: 0.0131\n",
    "\n",
    "Epoch 24/100\n",
    "79/79 [==============================] - 0s 885us/step - loss: 0.0113\n",
    "\n",
    "Epoch 25/100\n",
    "79/79 [==============================] - 0s 914us/step - loss: 0.0123\n",
    "\n",
    "Epoch 26/100\n",
    "79/79 [==============================] - 0s 919us/step - loss: 0.0127\n",
    "\n",
    "Epoch 27/100\n",
    "79/79 [==============================] - 0s 930us/step - loss: 0.0107\n",
    "\n",
    "Epoch 28/100\n",
    "79/79 [==============================] - 0s 928us/step - loss: 0.0131\n",
    "\n",
    "Epoch 29/100\n",
    "79/79 [==============================] - 0s 924us/step - loss: 0.0168\n",
    "\n",
    "Epoch 30/100\n",
    "79/79 [==============================] - 0s 896us/step - loss: 0.0104\n",
    "\n",
    ".\n",
    ".\n",
    ".\n",
    ".\n",
    "\n",
    "Epoch 100/100\n",
    "79/79 [==============================] - 0s 889us/step - loss: 0.0091\n",
    "\n",
    "Out[213]: <keras.callbacks.History at 0x7f9ad516aef0>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Test the Model that was built by STLM</b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#### predictions\n",
    "data_total=pd.concat((data_train['Capacity'],data_test['Capacity']),axis=0)\n",
    "inputs=data_total[len(data_total)-len(data_test)-10:].values\n",
    "inputs=inputs.reshape(-1,1)\n",
    "inputs=sc.transform(inputs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test=[]\n",
    "for i in range(10,89):\n",
    "    X_test.append(inputs[i-10:i,0])\n",
    "X_test=np.array(X_test)\n",
    "X_test=np.reshape(X_test,(X_test.shape[0],X_test.shape[1],1))\n",
    "\n",
    "pred=regress.predict(X_test)\n",
    "pred=sc.inverse_transform(pred)\n",
    "pred=pred[:,0]\n",
    "tests=data_test.iloc[:,1:2]\n",
    "rmse = np.sqrt(mean_squared_error(tests, pred))\n",
    "print('Test RMSE: %.3f' % rmse)\n",
    "\n",
    "metrics.r2_score(tests,pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Test RMSE: 0.030\n",
    "\n",
    "\n",
    "R^2: 0.8916473658827782"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "RUL(Error)=|126-125|\n",
    "=1 cycle "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred4=[1.5109049, 1.5067059, 1.5034488, 1.5005403, 1.498761 , 1.4981421,\n",
    "       1.4992642, 1.5018138, 1.503656 , 1.5056738, 1.5075662, 1.492796 ,\n",
    "       1.4839845, 1.4767292, 1.4711597, 1.465779 , 1.4615036, 1.4575909,\n",
    "       1.454027 , 1.4506783, 1.4484158, 1.4462972, 1.4443446, 1.4424661,\n",
    "       1.4380679, 1.4311802, 1.425846 , 1.4214693, 1.4188418, 1.4155321,\n",
    "       1.4114283, 1.4077026, 1.405254 , 1.4034641, 1.4015648, 1.4002702,\n",
    "       1.3994751, 1.3991467, 1.3991414, 1.3986106, 1.3982402, 1.3933313,\n",
    "       1.3868399, 1.3825758, 1.3791816, 1.3761727, 1.3734736, 1.3712626,\n",
    "       1.3694792, 1.3682021, 1.3673415, 1.3668255, 1.3659699, 1.3655413,\n",
    "       1.3632015, 1.3588042, 1.3560386, 1.3534404, 1.3517599, 1.3495598,\n",
    "       1.3468499, 1.3443987, 1.3423759, 1.3409755, 1.3396008, 1.3388804,\n",
    "       1.3388563, 1.3395094, 1.340758 , 1.3418714, 1.3428435, 1.3428377,\n",
    "       1.3375453, 1.3343163, 1.3319844, 1.3299532, 1.3282046, 1.3267752,\n",
    "       1.3251127]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred2=[1.5095047, 1.505182 , 1.5017349, 1.4985278, 1.4964548, 1.4956222,\n",
    "       1.4967338, 1.4995174, 1.5016944, 1.5042136, 1.5067403, 1.4912944,\n",
    "       1.482186 , 1.4747189, 1.4690324, 1.4634969, 1.4590889, 1.4550277,\n",
    "       1.4513285, 1.4478728, 1.4456096, 1.4435389, 1.4416875, 1.4399569,\n",
    "       1.435547 , 1.4284523, 1.4229819, 1.4185165, 1.4159185, 1.412598 ,\n",
    "       1.4084224, 1.4046086, 1.4020915, 1.4002208, 1.3982018, 1.3968226,\n",
    "       1.3960024, 1.3957292, 1.3958644, 1.395504 , 1.3953686, 1.3904271,\n",
    "       1.3837754, 1.379454 , 1.3760293, 1.3729705, 1.370195 , 1.367903 ,\n",
    "       1.3660536, 1.3647503, 1.3639095, 1.3634626, 1.3626862, 1.3623906,\n",
    "       1.3600874, 1.3555952, 1.3528095, 1.3501998, 1.348567 , 1.346395 ,\n",
    "       1.3436819, 1.3412246, 1.3391671, 1.3377022, 1.3362108, 1.3353803,\n",
    "       1.3352875, 1.3359427, 1.3372804, 1.3385375, 1.3397032, 1.3398769,\n",
    "       1.3344688, 1.331196 , 1.3288473, 1.326803 , 1.3250507, 1.3236343,\n",
    "       1.3219578]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     cycle  Capacity       pre\n",
      "469     90  1.605819  1.510905\n",
      "470     91  1.563849  1.506706\n",
      "471     92  1.548092  1.503449\n",
      "472     93  1.532376  1.500540\n",
      "473     94  1.526953  1.498761\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/khaled/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:2: 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: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  \n"
     ]
    }
   ],
   "source": [
    "len(pred4)\n",
    "data_test['pre']=pred4\n",
    "print(data_test.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Discharge B0005')"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmYAAAFNCAYAAACqr6PiAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAIABJREFUeJzs3XmcjeX/x/HXfc6Zc86cmWHsa6slJZVCfO1GSZbJiEipSKSxJIr4qvDNVrQh8gupKGVXZIvqW7aUNmVrQZGss5z9/v0xXyPZzmHOnJnxfj4eHnHOue/7Op+OM2/Xdd3XZZimaSIiIiIiUWeJdgNEREREJIuCmYiIiEgeoWAmIiIikkcomImIiIjkEQpmIiIiInmEgpmIiIhIHqFgJiIRN3ToUCZMmHDex8+dO5eOHTvmYItERPImBTMRuSBNmjThuuuuo3r16tSoUYMOHTowa9YsgsFg9muGDRvGI488EsVWRs7cuXO5+uqrqV69OtWrVycpKYm33377pNf88MMPpKSkcP3115OSksIPP/yQ/ZxpmowdO5abb76Zm2++mTFjxvD35SXPduzLL79M1apVs69dvXp1fvvtt8i/aRGJGAUzEblgr776Kps3b2b16tV069aN1157jcGDB0e7Wafl9/tz/Jw33HADmzdvZvPmzbz00kuMHTuW77//HgCv10vPnj1p3bo1GzZs4I477qBnz554vV4A3nnnHVasWMGCBQtYuHAhH3/8MbNnzw7pWIDmzZtnX3vz5s1ccsklOf7+RCT3KJiJSI5JSEggKSmJF154gXnz5vHTTz8BMHDgQMaPHw/AwYMH6d69OzVq1KBWrVrcfffd2b1rv//+O6mpqdSuXZubb76ZYcOGnXT+0aNHU7NmTZo0acKaNWuyH3///fdp3rx5do/V8WADsG7dOho0aMCUKVOoW7cugwYNAuC1116jXr161KtXjzlz5nDVVVfxyy+/AFmBaPTo0TRq1Ih//etfDB06FLfbHVINqlatSoUKFdixYwcA69evx+/3c99992G32+ncuTOmafLFF18AMH/+fLp06ULp0qUpVaoUDzzwAPPmzQvpWBEpeBTMRCTHXXfddZQuXZqNGzee8ty0adMoVaoUn3/+OZ999hn9+vXDMAwCgQDdu3enbNmyrFq1irVr13L77bdnH7dlyxauuOIKvvjiCx588EEGDx6cPeRXrFgxJk+ezJdffsnIkSMZOXIk3333XfaxBw4c4MiRI6xevZrhw4ezdu1apk+fzrRp01i+fDnr168/qY1jx45l165dzJ8/n48++oj9+/eHPEduy5Yt/Pzzz1x77bUAbN++nauuugrDMLJfc9VVV7F9+3YAtm3bRpUqVbKfq1KlCtu2bQvpWIDVq1dTq1YtWrRoccoQqojkPwpmIhIRJUuW5MiRI6c8brPZ+PPPP9m7dy8xMTHUqFEDwzDYsmUL+/fv5/HHH8flcuFwOKhRo0b2cWXLlqV9+/ZYrVbatGnDn3/+yYEDBwBo1KgRl156KYZhUKtWLerWrXtSKLRYLPTu3Ru73Y7T6eTDDz8kJSWFSpUqERsbS2pqavZrTdNkzpw5PPnkkyQmJhIfH0/37t1ZsmTJGd/r119/TY0aNahevTrt2rUjOTmZyy+/HID09HQSEhJOen18fDzp6ekAZGRkEB8fn/1cQkICGRkZmKZ5zmObN2/OBx98wOeff87w4cOZOHEiixcvPuv/FxHJ2xTMRCQi9u3bR+HChU95vGvXrlx22WV06dKFpKQkpkyZAmQNY5YtWxabzXba8xUvXjz797GxsUBWqAFYs2YN7du3p1atWtSoUYO1a9dy6NCh7NcXKVIEh8OR/ef9+/dTunTp7D+XKVMm+/cHDx4kMzOTlJQUatSoQY0aNXjwwQdPOt8/XX/99WzcuJHNmzfz2WefsW3bNsaNGwdAXFwcaWlpJ70+PT2duLg4AFwuV3bQAkhLS8PlcmEYxjmPrVixIqVKlcJqtXLjjTfSuXNnli1bdsZ2ikjep2AmIjluy5Yt7Nu3j5tuuumU5+Lj4xk4cCArV67k1VdfZdq0aXz++eeUKVOG33//PezJ+V6vl969e9OlSxc+++wzNm7cSIMGDU66s/HvQ4GQ1Zu3b9++7D///vvv2b8vUqQITqeTJUuWsHHjRjZu3MimTZvYvHlzSO0pXrw4zZo1Y/Xq1UBWePrxxx9Pas+PP/5IxYoVAahUqRJbt27Nfm7r1q1UqlQppGNP5++vFZH8R8FMRHJMWloaq1evpl+/frRu3ZqrrrrqlNesXr2aX375BdM0iY+Px2q1YrFYuO666yhRogTPP/88GRkZeDweNm3adM5rer1evF4vRYsWxWazsWbNGj777LOzHnPbbbcxd+5cduzYQWZm5knzxywWC+3atePZZ5/lr7/+ArJ6/z755JOQanDo0CGWL1+eHZ5q1aqF1WrljTfewOv18uabbwJQu3ZtAJKTk5k2bRr79u1j3759TJs2jTZt2oR07IoVKzhy5AimabJlyxZmzpxJUlJSSO0Ukbzp9GMGIiJh6NGjR3bAqlixIg888AAdOnQ47Wt/+eUXhg8fzsGDBylUqBAdO3bk5ptvBrKW3RgxYgSNGzcGoFWrVqftdfu7+Ph4hgwZQt++ffF6vTRu3JgmTZqc9ZiGDRty77330rlzZwzDoGfPnixYsAC73Q7AgAEDmDBhAu3bt+fQoUOUKlWKjh07Ur9+/dOe76uvvqJ69eoAOJ1O6tSpk71ciN1uZ8KECQwZMoTnn3+eChUqMGHChOxrdejQgd9++41WrVoBcOedd2bX7lzHfvDBBwwePBiv10upUqXo1q1bdqgTkfzJMNXvLSIXuR07dtCyZUu++eabM85xExHJDRrKFJGL0vLly/F6vRw5coSxY8fSuHFjhTIRiToFMxG5KM2ePZs6depwyy23YLVaefrpp6PdJBERDWWKiIiI5BXqMRMRERHJIxTMRERERPKIfDfTNRgMEghEfvTVajVy5ToFgWoVHtUrdKpVeFSv0KlWoVOtwvPPesXEWMM6Pt8Fs0DA5PDhjIhfJzHRlSvXKQhUq/CoXqFTrcKjeoVOtQqdahWef9arRImEs7z6VBrKFBEREckjFMxERERE8ggFMxEREZE8QsFMREREJI9QMBMRERHJIxTMRERERPIIBTMRERGRPELBTERERCSPUDATERERySMUzM7CNOGzz6x4PNFuiYiIiFwMFMzOIi0NUlJiGTzYEe2miIiIyEVAwewsEhLg4Yd9vPGGnVWrwtuEVMJz4IDB99/r4ygiIhc3/SQ8h4EDPVSpEqBvXyeHD0e7NQXXmDF2WrRwkaF9ckVE5CKmYHYOTie88oqbAwcMBg50Rrs5EbF7t8HixTZmzbIxZUoMP/6Y+x+LH3+0kJ5usGqVLdevLSIiklfop2AIrrsuSL9+XsaMcdCypZ+WLf3RblKO6tIllq++OjFUW7x4kM8+S6dIkdxrw44dWWFw0SJbgauviIhIqNRjFqI+fbxce22AQYMcHD0a7dbknEAAtm610LGjjw0b0li8OJ1DhwxGjMi9Gx7S0mD/fgt2u8lHH9nIzMy1S4uIiOQpCmYhiomBcePc/Pln7oaWSNu928DtNqhZM8Bll5nUqhXkoYd8zJxpZ/363Pl47NqVdZ0OHXykpxt8/LE6ckVE5OKkYBaGG24I0q2bj+nT7axbVzDu0ty+PesjULFiMPuxAQM8lCsXZMAAJz5f5Nuwc2dWGzp39lGkiMnChQpmIiJycVIwC9MTT3goXz7IY485+PJLC19/beG77yzs3Wvg9Wa9Ji0NvvnGwqJFNj76yMqGDRZ+/tmIbsPP4Kefsj4ClSqdCGbx8fDssx5++MHK5MkxEW/D8fllFSsGuf12H8uW2bSor4iIXJTUNRGm+HgYM8bN3Xe7uO22uNM8b5KWdvoQ1ru3hyFDvJFuYli2b7dQtGiQYsXMkx5v3txP8+Y+xoxx0Ly5nwoVzDOc4cLt3GmhbNkgLhe0auXnrbfsrFlj5dZbAxG7poiISF6kYHYemjYNsGpVOnv3GgQCBn4/HDpk8OefBgcPGpQsaXLllUEuvzyIz5f13HvvxfDSSw5q1w7QtGneCRzbtllOGsb8uzFjPNSvb6N371gWLszAGqHR2507LVx5ZVYb6tcPkJhosnBhjIKZiIhcdBTMztO11wa59trQX1+3boAffrCQmupk1aoMypaNXA9UOLZvt3DbbadfnqJUKZNnn3XTs2cskyfH0LNnZCac7dxp0KJFVgiLicnqrVu40MaTTxp5pk4iIiK5QXPMconTCVOnZuJ2G/To4cSfB5bqOngQDhw4c48ZQNu2fm67zceoUQ62bcv5j8vhw3DwoIUKFU60oW9fD8EgPP64E1O5TERELiLqMctFFSuajBnjJjU1lkqV4rniiqzhzvh4sFpNLBawWsFiyeo5uu8+LxUrRi6ZbNuWNTb594n//2QYMHZs1pBm3bpxFCliUqpUEJsNjhwxOHzYoH59mDYt67XhOn5H5vGhTIArrjAZONDDU085mTfPRkpKHkixIiIiuUDBLJe1b+/H4chk/Xoru3ZZ2LrVQmamQSCQtdhrMAjBoEFaGixfbmPVqnRcrsi05XRLZZxOqVIm8+ZlsGyZjX37DPbtM/D7Da6+OojbDYsWxfDeezbatQs/QJ0IZicH0Ice8rFgQQyDBzto0CBA8eLqOhMRkYJPwSwKkpP9JCefPcR88omVtm1djBzpYPjwyKwdsW2bBYfD5NJLzx16qlYNUrXqqXeUBoPQqpWNp5920KyZn0KFwmvDjh0WLBaTyy47ORxarfDCC26SklwMGeLg1Vfd4Z1YREQkH9Icszyqfv0A99/vZcqUGL74IjK3Q27fnnU35IXcbWmxwMsvBzlwwGD06PB3RNi1y0L58iaO0xxapUqQRx/1MnduDMuWFYwFfUVERM5GwSwPGzrUwyWXmPTp4yQjI+fP/9NPlrPOLwvVjTfCfff5+L//i2HzZgu//Wawfr2FL76w8uuvxll3D/j7Uhmn07u3l6uvDvD4484CtUepiIjI6WgoMw+Lj4fx4920bevi8ssTiI83SUg4/gsSEkwefthL48bhr/fldsOvvxqkpFx4MAMYNMjDokU2mjU7ddFdwzC59dYA06dnntQ7Z5pZQ5nt2p05udntWUOazZu7eOYZB88/ry0BRESk4FIwy+Pq1w/w5psZbN5sJS3N4NgxOHbM4Ngxg59+stC9eyyffppOyZLhTY7ftctCMGhQuXLOBLMiRWDaNDeffGKlTBmTMmWyhkj37jX46isr06fbmTEjhi5dToSwAwey3sfZeswAqlcP0qOHj4kT7bRp46dePS08KyIiBZOCWT5w662B066Cv327QaNGcQwe7OC118KbHH/8jsycGMo8rnbtALVrn9rOjh397Nxp4T//cdCihZ9SpbJC5PE7Mv++htmZPP64hw8/tJGa6mTq1Exq1Mi5douIiOQVEZtjNmjQIOrUqUPLli1P+/yxY8fo0aMHrVu3pkWLFrz//vuRakqBVbGiSb9+XhYsiGHp0vAmxx9fLPZcvVU5wTCy9hf1euHf/z4xy3/XrqyFz6644txtcLlgypRMLBZo2dLF8OF2bXQuIiIFTsSCWUpKClOnTj3j82+99RYVKlRg4cKFzJw5k9GjR+P15q0NvvOD1NSsyfFPPOHk2LFTnzdNyMjIGlL8/nsL331nYdcug2++sVC+fJC4U6eERcSVV5r06eNl/vwYZs+2MX++jTlzYrDZQluuA+D664OsWZNOx44+Xn7Zwa23uvjpJ92/IiIiBUfEhjJr1qzJ7t27z/i8YRikp6djmibp6ekULlwYm00jq+Gy22HcODe33+7i2mvjKVrUpHBhk0AADh/OWpnf4zn9kvxNmuTuivq9enmZOzdrU3SAmBiT5GQ/4fxvT0iA8eM9tGjhp3dvJ7fe6mLMGDft22t3ABERyf+iloQ6derEww8/TP369UlPT2f8+PFYLOr9OB833RRkxoxM/vtf2//CWNYCrUWKmCQmmiQm8r//mhhGVg9aRoZB/fq5G2YcDnjzzUw2brRyzTVBKlcOYref37maNg2walUG3bs7SU2NZf16L2PHes5rWygREZG8wjDNyG0TvXv3bnr06MHixYtPeW7p0qV8+eWXDBo0iF9//ZUHHniAhQsXEh8ff9ZzBoNBAoHIb89jtVoIBDTBPBTRrJXfD4MHG4wfb2Hq1CCdO+f9rZv02QqdahUe1St0qlXoVKvw/LNeMTHhzQGPWo/Z3LlzeeihhzAMg8suu4zy5cuzc+dOrrvuurMeFwiYHD4cgdVW/yEx0ZUr1ykIol2rJ56A//43lgEDrNSpk0GJEtEPZ14vZ+wNjHa98hPVKjyqV+hUq9CpVuH5Z71KlEgI6/iojR2WKVOGzz//HIADBw6wa9cuypcvH63mSD5mscDzz3tISzv5rs9o+f57C1deGc/y5dpGSkREwhOxHrN+/fqxfv16Dh06RIMGDejVqxd+f9acpo4dO9KzZ08GDRpEq1atME2T/v37U7Ro0Ug1Rwq4ypWD9Onj5bnnHLRr5yMpKXqL0H77rQWv16BfPyeffJJOYmLUmiIiIvlMROeYRYLPF9BQZh6TV2rl8UBSkov0dIM5czKoWDE6H+1x4+yMGuXAajVp29bPK6+cvPhvXqlXfqBahUf1Cp1qFTrVKjz5dihTJKc5HDBhghu3G5o1iwt70d2csnu3QfHiWT14774bw7JlGtIUEZHQKJhJgXL99UGWL8+gQoUgnTu7GDvWTm73Cf/2m4VLLsnaleGaawL07+/kyJHcbYOIiORPCmZS4JQvb7JwYQZ33eVj7FgH772Xuzcf796dtauC3Q4vvuhm3z4L06ef54JtIiJyUVEwkwLJ6YQXXnBTu7afgQOd/Ppr7qw8a5qwZ49B+fJZ3XTXXx+kUSM///d/MWjHMRERORcFMymwrFZ45RU3pgmpqU4CuXCj5p9/GrjdBpdccmJxwe7dvfzxh4VFi7TlmIiInJ2CmRRol15qMmqUmy++sDF2rJ2ffrLw008W9u2LTA/a7t1Z5y1f/kQwa9w4QKVKASZPzv35biIikr8omEmB166dn+RkH+PGOahXL4569eK4/vo4xo2z53gv2u7dWX+ljg9lQtYCuN26+fjqKyvr1+sOTREROTMFMynwDANeftnNtGmZTJmS9euOO/yMGuWgffvYHO09++23rHP9fSgToH17H4mJJpMnx+TYtUREpOBRMJOLgtMJLVr4ueOOrF+TJrkZP97Nxo1WmjRx8fnnJ/dkLVli4667Ytm0Kby/Irt3W0hIMClc+OTHXS7o3NnLBx/Y2LHjQt+NiIgUVApmclEyDOjUycfSpRkUKgRt28byf/8Xg88HQ4c6eOCBWD75xErLlq6whjyz7sgMnva5Bx/0ERsLPXtaCJ7+JSIicpFTMJOL2tVXB1m2LJ0mTQIMGuSkRo04Xn3VTteuXr7+Op3k5KwhzzZtYklLO/f5ji8uezqlS5sMH+5h9WqDadM0pCkiIqdSMJOLXqFC8MYbmfTv7yEQgClTMhk50kOJEiaTJrl56aVM1q2zMnKk45znOr647Jl06uTjtttMhg1zsHNn7qytJiIi+YeCmQhZd04+/riXb79N5447/NmPGwZ06ODngQd8TJ0ac9Y5Z8eOwZEjZx7KPH6+V18N4nBAampsrqytJiIi+YeCmUgIBg/2ULq0Sb9+Tny+07/mt99OXSrjdMqWhWefzbrxYNIkDWmKiMgJCmYiIUhIgNGj3fzwg5WJE0+/7+XpFpc9k7Zt/bRo4WPUKAdbt+qvoYiIZNFPBJEQ3XZbgJYtfTz3nJ1t2079q3O8x+xMk///zjBgzBgPhQqZpKaeuRdOREQuLgpmImEYOdJDXJzJww87T9mUfPduC3a7SYkSoe27VKKEyZgxHrZssfLCC6fvhRMRkYuLgplIGEqVMnn++aww9dxzJ4ep3bsNypUzsYTxt6plSz9t2/oYP97O0qXarklE5GKnYCYSphYt/HTq5OXFF+188cWJMHWupTLOZORIN1dfHaRzZxdPP+3QsKaIyEVMwUzkPAwf7uGyy0weecSZvdfmb78Zp+yRGYrERFiyJIP77/cycaKd5GQXe/ZojTMRkYuRgpnIeYiPh0mTMvnrL4NmzVxs2GBh/37LOZfKOBOnM+tmgMmTM/nhBwtNmsSxfLmGNkVELjYKZiLn6aabgixalIFhQHKyCwhtqYyzadPGz8qV6ZQtG6RTJxfDhtnx+899nIiIFAwKZiIXoFq1IMuWZXDDDVmB7MorL3x38iuvNPnggww6d/byyisORo3SHZsiIhcLW7QbIJLflSxpMm9eBhs3WqlZ88KDGUBsLDz3nAe/H155xc5tt/mpUSNnzi0iInmXesxEcoDDAXXrBjByeM7+sGEeypQx6d3bSWZmzp5bRETyHgUzkTysUCEYP97N9u1WRo50RLs5IiISYQpmInlco0YB7rvPy+TJMSetmyYiIgWPgplIPvDUUx4uuSRrSDM9PdqtERGRSFEwE8kH4uPhxRfd/PyzhREjNKQpIlJQKZiJ5BN16wbo1s3L//2fnU8+0ZCmiEhBpGAmko8MHuzhiiuC9O3r5NixaLdGRERymoKZSD7icsHLL2eyZ49By5Yutm/XnpoiIgWJgplIPlOrVpBZszLZv9/gllvimDdP60SLiBQUCmYi+VDjxgFWrszgmmuCdO8eS2qqk0OHot0qERG5UApmIvlU2bIm8+dn8OijHubOtVG3bhyLFqn3TEQkP4tYMBs0aBB16tShZcuWp31+6tSpJCcnk5ycTMuWLbn66qs5fPhwpJojUiDFxMCgQV6WLcugbFmTrl1j6dvXgccT7ZaJiMj5iFgwS0lJYerUqWd8/sEHH2TBggUsWLCAfv36UbNmTRITEyPVHJECrVq1IEuXZtCvn4e337aTkuJi/37dGCAikt9ELJjVrFmTwoULh/TaJUuWnLFnTURCY7PBwIFeXnstk2+/tdCsmYtvvtFsBRGR/MQwTdOM1Ml3795Njx49WLx48Rlfk5mZScOGDfnoo49C6jELBoMEAhFrcjar1UIgEIz4dQoC1So8uVGvzZuhbVsLBw/C1KlB7rwzopeLGH22wqN6hU61Cp1qFZ5/1ismJrwFwaM+U3j16tXceOONIQ9jBgImhw9nRLhVkJjoypXrFASqVXhyo15XXAFLlxrcf38sd99tZdMmDwMGeLHksw40fbbCo3qFTrUKnWoVnn/Wq0SJhLCOj/rX9JIlS2jRokW0myFS4JQsaTJvXgYdOvh4/nkHI0bYo90kERE5h6gGs2PHjrFhwwaSkpKi2QyRAsvhyNr8/N57vbzyioM1a7THpohIXhaxocx+/fqxfv16Dh06RIMGDejVqxd+vx+Ajh07ArB8+XLq1q2Ly+WKVDNELnqGAcOHe1i3zkpqqpPVqzMoXjzy8zRFRCR8EZ38Hwk+X0BzzPIY1So80arXt99auO02F40bB3jjjUyMfLCahj5b4VG9QqdahU61Ck++n2MmIrnj2muDDB3qYdkyGzNmxES7OSIichoKZiIXkW7dfDRq5Ofppx388ks+6DITEbnIKJiJXEQMA8aNc2OxwKOPOglqaSIRkTxFwUzkIlO+vMmwYR4+/VRDmiIieU3UF5gVkdzXqZOPhQttPPOMA5uN7IVnmzf3UbRodNsmInIxUzATuQgdH9Js2tTFY485sx9/++0YFi3KyHc7BIiIFBT6+hW5SJUvb7JxYzobNqTx5ZdpjBnjZsMGK9OmaXhTRCRaFMxELmLx8XDZZSbly5vcd1/WHZsjRjjYs0d3bIqIRIOCmYgAWcObY8e6MU144gkn+WvpaRGRgkHBTESyXXaZycCBHj76yMaUKTEKZyIiuUzBTERO0q2bj8aN/fz7307uvTeWffs0rCkiklsUzETkJFYrzJqVyYgRbtautVK/fhwLF+oGbhGR3KBgJiKnsFjgoYd8rF6dToUKQR58MJYBAxxkZka7ZSIiBZuCmYicUYUKJgsXZpCa6mHGDDvNm7vYuVNDmyIikaJgJiJnFRMDQ4d6mT07g99/t9ClSyx+f7RbJSJSMCmYiUhImjQJMG6cm++/tzJlihahFRGJBAUzEQnZ7bf7adbMz5gxDnbv1pCmiEhOUzATkZAZBjz7rBuAJ590RLk1IiIFj4KZiITlkktM+vf3sHRpDO+9p2U0RERykoKZiISte3cfN94YoGfPWAYOdJCeHu0WiYgUDApmIhK2mBiYNy+D7t29vP66naSkODZt0teJiMiF0jepiJyX2FgYPtzD3LkZeL3QqpWLl16yEwxGu2UiIvmXgpmIXJB69QKsXp3O7bf7GTHCQbt22l9TROR8KZiJyAUrXBhee83NuHFuNm600qiRixUrrNFulohIvqNgJiI5wjDgnnt8LF+eQalSJnff7WLIEAceT7RbJiKSfyiYiUiOqlw5yNKlGTz4oJcpU+y0auVi714NbYqIhELBTERynNMJzz7rYfr0TLZvt3DLLS7WrdPQpojIuSiYiUjE3H67nw8/zCA+HlJSYnn/fS1IKyJyNgpmIhJRV10VZNmydG66KcBjjzn55RcNa4qInImCmYhEXGIiTJrkxmKB/v2dmGa0WyQikjeFFMwOHz4c6XaISAFXrpzJkCEe1qyx8c47GtIUETmdkIJZ+/bt6d27N2vWrMHUP3VF5Dzdf7+PWrX8DB3qZP9+DWmKiPxTSMFs2bJl3HXXXSxYsIBbbrmFcePGsWvXrki3TUQKGIsFxo/3kJEBDz7o5I8/FM5ERP4upGBmGAZ169Zl3LhxjBgxgnnz5tGuXTvuueceNm/eHOk2ikgBUqlSkBdfdLNli5WGDeNYtEjDmiIix4UUzA4dOsSMGTNISUnh9ddf59///jdffPEFTzzxBI899thpjxk0aBB16tShZcuWZzzvunXrSE5OpkWLFtxzzz3n9w5EJN9p29bPypXpXHZZkK5dY+nTx0laWrRbdeH27zd4+20FTRE5fyEFsw4dOpCWlsbEiRNOgXJIAAAgAElEQVSZMmUKt956KzabjWrVqtGhQ4fTHpOSksLUqVPPeM6jR4/yzDPPMGnSJJYsWcKLL754fu9ARPKlChVMlizJ4NFHPbzzjo3GjePYsCF/3yj+7rs2+vaN1ZIgInLeQvoW7NOnD4888gilS5fOfuzDDz8E4KGHHjrtMTVr1qRw4cJnPOeiRYu45ZZbKFu2LADFihULudEiUjDExMCgQV7mz88kGITWrV1MnRoT7Wadt7/+yvpK/emn/B0wRSR6Qvr2eO211055bMqUKRd04Z9//pmjR49y7733kpKSwvz58y/ofCKSf9WuHWD16nSSkgIMHerg++/zZ7A5eDCrp+zHH/Nn+0Uk+s46GWLNmjWsXbuWffv2MWLEiOzH09LSsFovbN+7QCDAd999x/Tp03G73XTo0IHrr7+eK6644qzHWa0GiYmuC7p2KKxWS65cpyBQrcKjep1eYiLMmAHVqsHAgS4+/jh4zloFg1n/teSRHHT0aFZDdu2yk5iY+z1/+myFTrUKnWoVngut11mDWalSpbj22mtZtWoVVatWzX48Li6OQYMGnfdFAUqXLk2RIkVwuVy4XC5q1KjB1q1bzxnMAgGTw4czLujaoUhMdOXKdQoC1So8qteZWa3w1FM2evWK5eWXffTpE3PGWpkmPPCAkz//tLBkSd6o5759LsDKt9/mzvfUP+mzFTrVKnSqVXj+Wa8SJRLCOv6swaxKlSpUqVKFVq1aYbPl7J1GSUlJDBs2DL/fj8/nY8uWLdx///05eg0RyX/at/fzzjt+hg93cNddQZzO079uzhwbH3yQ1Su1fbtBxYrRX/z6r79ODGUGg3mnJ09E8o+zpq0+ffrw4osv0qZNm9M+v2jRojMe269fP9avX8+hQ4do0KABvXr1wu/3A9CxY0cqVKhA/fr1ad26NRaLhTvvvJPKlStfwFsRkYLAMGDMGDeNGsVx330WXn7ZoGTJk0PXvn0GQ4Y4ueaaAN9/b2Xhwhj69fNGqcUnHDxo4HKZZGQY7NljcMkl0Q+LIpK/GOZZ9ljav38/JUuWZM+ePad9vly5chFr2Jn4fAENZeYxqlV4VK/QvP22jSeecOJywYgRbu68049hnBjCXLnSxqpVGfTr5+DYMYOPP45uTX0+KFcugQYN/Kxda2PWrAySkgK52gZ9tkKnWoVOtQrPhQ5lnrWjvWTJkgAEg0GKFy9OuXLlKFeuHMWKFdOemSISUXff7WfDhiAVKgR55JFYGjZ0cd99Tnr2dPLBBzEMGOClUqUgrVv7+f57K9u3R3ftsEOHsq5fp05WGNu6VeOYIhK+kNcxM4wTX3oWi4U+ffpErFEiIgBVqsCiRRmMHOmmXDmTnTstfPihjZtv9tOzZ9bQZcuWfgzDZOHC6K5/dnypjIoVg5QoEeSnny7sznURuTiFNKM/EAhgt9uz/2y32/H5fBFrlIjIcVYrdO3qo2vXrO+c4531x/+tWKaMSa1aARYssEV1ntnxYFa0qEmVKkGtZSYi5yWkb46iRYuycuXK7D+vWLGCIkWKRKxRIiJnYhgnQtlxycl+fvjByrZt0QtDx+/ILFrUpHLlrGCmGR8iEq6QvsWeeeYZJk+eTKNGjWjYsCFTp05l2LBhkW6biEhITgxnRm8D8ePBrFgxk6uuCpKennVnpohIOEL6Frv00kt59913SU9PxzRN4uPjI90uEZGQlS5tcvPNAWbNiuHhh724orBI+d+HMq+6KmtLgp9+slC+fO7emSki+VvI/7z8+OOP2bZtGx6PJ/ux1NTUiDRKRCRcjz/upW3bWJ5+2sGYMZ5zH5DDDh40SEgwsduhcuWsYPbjjxaaNFEwE5HQhTSUOXToUD744APefPNNAJYtW8bevXsj2jARkXDUqxfg4Yd9TJ9u56OPcv+OyL/+MihaNGtSWbFiJsWL6wYAEQlfSN8amzdvZsyYMRQqVIjU1FRmz57NH3/8Eem2iYiEZdAgD9deG6BvXyf79+fu/K6DBw2KFTsx2/+qq4L8+KOWzBCR8IQUzJz/26wuNjaWffv2ERMTw+7duyPaMBGRcDkcMGmSm7Q0g+7dnRw6lHvXPnjwRI8ZZAWzn37SnZkiEp6QglmjRo04evQoXbt2JSUlhSZNmtCiRYtIt01EJGxXXRXkuefcrF9vJSkpjvXrc2c48e9DmQBVqwY5dsxgyxYNZ4pI6M66V+bpeL1ePB4PCQnh7f2UU7RXZt6jWoVH9QrdhdRq82YL3brFsmePQadOPkqVMomNzVr8NSkpcMpaaBfq8svj6dzZx7BhWTceHDkC118fT4sWfiZMcOfsxc5An63QqVahU63Cc6F7ZYZ0V6bH4+Htt99m06ZNGIbBTTfdRMeOHXE4HOG1VkQkl1SvHmTVqnSeeMLJu+/G4HafSGKtW/sYM8ZN0aI5c62MDMjIMChe/MS/cwsXho4dfbzxRgxDhxqUKqUxTRE5t5D62B9//HG2bdvGPffcQ6dOndixYwcDBgyIdNtERC5IoUJZc85+/TWNP/44xs6dxxgyxMOHH9po0CCOlStzZnL+8Q3M/z6UCdCtmxe/H15/Pbr7eIpI/hFSMNu1axfPPvsstWvXpnbt2gwfPpyff/45wk0TEck5FgvEx0Pv3l6WLcugWDGTjh1dvPSS/YIn6P99cdm/u/JKk2bN/MyYEUNm5oVdQ0QuDiEFs2uuuYavvvoq+89ff/01N954Y8QaJSISSddeG2TZsgxSUnyMGOEgNdWJ5wLWpP37Ppn/1KOHj4MHLcyZo14zETm3kOaYff3118yfP5+yZcsCsHfvXipUqECrVq0AWLRoUeRaKCISAU5n1jDnVVcFGTnSwZ49BnPmZBJzHvnp7/tk/lOdOgGqVQswZUoM997ry/GbDkSkYAkpmE2dOjXS7RARyXWGAY8+6qVMmSC9e8cyYYKdvn29YZ/nTEOZx6/RvbuX1NRYVq+2aosmETmrkIYyy5UrR7ly5XA6nRiGkf3r+OMiIvlZhw5+kpN9jB1r54cfwl937K+/DAzDJDHx9JPV7rjDT6lSQV591X6hTRWRAi6kb6CVK1dy6623kpSUxD333EOTJk3o1q1bpNsmIpJrRo70UKiQSZ8+Tvz+8I49vuq/9Qw3edrt0KWLj48/trF1qxacFZEzC+kb4sUXX+Sdd97h8ssvZ9WqVUyfPl2T/0WkQCle3GT0aA9ffWVl4sTwerb+uR3T6XTu7MPpNJkyRTcBiMiZhRTMbDYbRYoUIRgMEgwGqV27Nj/88EOk2yYikqtat/bTqpWPMWPs/Phj6D1boQSzYsVM2rf3MWdODH/+qTsAROT0QvrmKVSoEOnp6dSsWZP+/fszYsQIbLaQ7hsQEclXRo3yEB8f3pDmP/fJPJPu3X14PAYzZqjXTERO76zB7JdffmHTpk1MnDiR2NhYBg0aRP369UlMTGTIkCG51UYRkVxTooTJyJEevvzSyquvhhag/vrLOO1SGf9UqVKQpCQ/06bFsH+/es1E5FRnDWbPPvsscXFxuFwuLBYLNpuNNm3a0LBhQ1555ZXcaqOISK664w4/t9/uY/RoB9u2nX1gwTRDG8o87rHHPKSlGTRt6mL9et0IICInO+u3wp49e6hSpcopj1erVo09e/ZErFEiItFkGDB6tAeXC9q2jaVePRdVqsRRs2Ycmzef/LV57Bj4/aEHsxo1gixZkoHTCXfc4eL//k/DmiJywlmDmecse5S43e4cb4yISF5RqpTJpEmZVKoUpHLlIMnJfkwTkpNdLFp0Yo7t2Vb9P5Nrrw2yfHk6SUkBBg1yMneu5uyKSJazBrNq1arx7rvvnvL4nDlzqFq1asQaJSKSFzRpEuD99zN5/XU3o0d7+PDDDKpVC9K1ayzjx9sJBk+s+h9OMAMoXBimTcvkppsCPPGEk717NedMRM6xJdOTTz5JamoqixYtyg5i3377LT6fT3PMROSiU6KEyfvvZ/Doo05GjnSwYYOV1q19wOm3YzoXmw0mTMikSZM4evVyMmdOJhZNOxO5qJ01mBUvXpzZs2fzxRdfsG3bNgAaNmxInTp1cqVxIiJ5jdMJEye6qVkzwNChDlavzlru/3yCGcCVV5oMG+ahf38nU6fG8NBDvpxsrojkMyFNbKhduza1a9eOdFtERPIFw8jaYqlmzQDdusWyd2/WzgHn6957fSxbZmP4cAfFi5ukpIS5J5SIFBjqNBcROU/VqgVZuTKd1avTiY8///MYBrz4opvrrgvSo0cs/fo5yMjIuXaKSP6hYCYicgHi4qBChf/1lrnd2DZvylrcLEzFi5vMn59Bnz4e3norhttuc2nDc5GLkP7Wi4jkEPunayjSrDGFU1pi/XFr2MfHxMDgwV5mz87kwAGDZs1cvPVWzPnkPBHJpyIWzAYNGkSdOnVo2bLlaZ9ft24dN910E8nJySQnJ+suTxHJ97xNbuHY2BewffcNRRr/i7inBmMcOBD2eRo3DrB6dQY1agR49FEnDz/s5NixU19n2fcHjnfeJuHhBynSuC5sDT8MikjeErFglpKSwtSpU8/6mho1arBgwQIWLFhAampqpJoiIpI7LBbc93Xh4H+/xN2+I7GvvkKxm6oSP/AxLL/+EtapSpUyeffdTAYN8jB/vo2kpDi2fBnEtu4LXM8OI7FJPYpVq0yhXj2wr1mFdcc2rKNGRuiNiUhuiVgwq1mzJoULF47U6UVE8iyzeHHSXpjAoU834G5zJ86Z0yla63oK3Xc3MWtWhzwHzWqFxzr+wlcPjufVP+6gxm2XUqTVrbheHo+ZkEDakKc5uPJT/vp2O5n3P4jxzmwsv/wc2TcnIhEV1TlmX331Fa1bt+bBBx/MXidNRKSgCFSqTNoLEzi4YQuZqX2JWf85ie2SKVK7OnFPDsC+ZBHGX3+dHNQCASw7d+B4dxaF299B0RuuptqUx2hc8ls+LX8X7XiXySN+48iCD8ns3Y9AtevAYiHz4VSwWHBNejl6b1hELphhmpGbVrp792569OjB4sWLT3kuLS0NwzCIi4tjzZo1/Oc//+Gjjz465zmDwSCBQORnwlqtFgKBYMSvUxCoVuFRvUJX4GrldmO89x6WWW9hfPYZxv/WxDCdTihZElxxsGsnxv/2KTYvu4xgp3sI3t0JKlcmGISmTS18+y18/XWQMmVOPr2tx0Pw9tv4t+2AUqVy+93lKwXusxVBqlV4/lmvmBhrWMdHLZj9U5MmTXjvvfcoWrToWV/n8wU4fDjyC/wkJrpy5ToFgWoVHtUrdAW6Vl4vti83EbN5E5Z9f2D5cz9GejqBy6/Af1UVAlWuxn99df65R9OOHQaNG8fRpImfadPcGH/bYjNx/25s1aqS0ecxMp4cesZL//yzQe/eToYN83DDDRfnD9wC/dnKYapVeP5ZrxIlEsI6PqSV/yPhzz//pHjx4hiGwZYtWwgGgxQpUiRazRERyV12O/7adfDXDm+LuwoVTAYM8DJ8uIPFi/20avW3XQIqV8bbMpnY118jM7UPZqFT5/kGAtCrl5N162yMGQNvv515oe9ERHJQxIJZv379WL9+PYcOHaJBgwb06tULvz/rC6Rjx44sW7aMWbNmYbVacTqdjBs3DuPv//QTEZHTevhhLwsW2HjiCQfVqgW4/PITAx8ZfR8jcclC4vv34djkafCP79XJk2NYt87GTTcFWLHCxtatFqpUuTh7zUTyoogOZUaChjLzHtUqPKpX6FSrM/vuOwvJyS6CQRg92k27dv7sesW+NJ74EU+R9syzWTcF/M/WrRZuucVF48Z+XnjBTfXq8dxxh58XX3RH8Z1Ehz5boVOtwnOhQ5la+V9EJB+qWjXI6tXpVK0a4JFHYunRw8nRo1nPZfbqi6dlMnHD/k3Mp2sB8PmyhjDj402ee85D0aLQsaOP996zsW+fRitE8goFMxGRfOqSS0zmzctk4EAPCxbYqFPHwvffW8AwOPbSRAJXVqBQt/uw/PIzEybY+fprK2PGeChZMmugpHt3L4EAvPZaTJTfiYgcp2AmIpKP2WzQr5+XuXMzOXYMmjd38c47Nsz4BI7OmAWBAHHJybw99k9at/addLPAFVeYtGjhZ8YMO2lpUXwTIpJNwUxEpACoUyfA+vVBbrwxQK9esTz2mIP08pU49PZc/H8cYFngFkY/tvuU4x55xMuRIwYTJtij0GoR+ScFMxGRAqJ0aZgzJ5M+fTzMnGmnRQsXz3z4L5oHl3CF9RcqPtz6lE3Vb7wxSEqKj5dftrNtm34kiESb/haKiBQgNhsMHuzlzTcz+PVXCy+/7MDepA5pb83GunM7RW5tiG3LVycdM3y4B5cL+vd3ENTKGSJRpWAmIlIA3XprgBUr0nnoIS/jx7vxN2rM4cUfgWmS2PJWHHNmZ7+2RAmTp5928/nnNmbN0o0AItGkYCYiUkBddpnJiBEeypTJugvTf90NHPpoDb4ba1DokYeIe3IAeL0AdOzo51//8vP00w7279fyGSLRomAmInIRMUuU4MicBWR074lr6mQS77gdy949GAY895ybtDSYNEk3AohEi4KZiMjFJiaG9OGjOPradKw/fE+RpvWxL1pAxQpBWrb0M3NmjJbPEIkSBTMRkYuUJzmFwx99TKB0WQp3vZdCndrRN/knjh41eOcdzTUTiQYFMxGRi1igUmUOf/QxacNHEvP5f2nYswaTy/ybtyZ7dYemSBQomImIXOxsNjK7P8Khzzbgue12Hvp9BMt+rsKuoW9BIBDt1olcVBTMREQEgGDZchybMp0/Fyznj5hLqT3lYYo0qoN90Xz+3n32vxs5RSQCFMxERORkdW5m3uMf0453+WMvFO7amaOVGjCy5jKqVHZRvnwCjz3mwOeLdkNFCh4FMxEROcW9nX18WyWF2nFbeLTEDGyeDMb90o4tVGNCnenMngkdO8Zy9Gi0WypSsCiYiYjIKYoUgbVrM9i8xc2T37XB9fN6jr76f5QqY9Dz8wc4FluSbp904fkGK/h9pzvazRUpMBTMRETk3Gw2PCntOLT6vxyZ9R5mckvaxy1m4t4UKv/rMhLu7YDzrTew7Psj2i0Vydds0W6AiIjkIxYL3qRb8SbdCs/7+HTEf9k9aRmd1y2g8LIPAPBXrYa3cRLexkn4atUGhyPKjRbJP9RjJiIi5ycmhrpPN2Rh0/GUcf/MNzP/S9qQZwgmJhL76isktm1F8cqXUviuNthXLIt2a0XyBQUzERE5b4YBzz/vJsZu0O2VmqSnPsqReUv466dfODLzHTI7dca6YzuF725HwkP3Y+zbF+0mi+RpCmYiInJBypQxGTHCzbp1NgYMcODxgBmfgLdZc9KfHcvB/24ifeAQHB8spmi9mtg/WJxj1z50CAYMcDB7tmbmSMGgYCYiIhfsrrv8PPKIl5kz7bRs6WLXLuPEk3Y7Gf0e59CazwlUqEChrvdiXzjvgq+5Zo2Vhg3jmDHDzqhRDm0hJQWCgpmIiFwww4CnnvIwY0YmP/9soWnTOO69N5bOnZ106eJk6VIrgQqVOPLeQvw1alGoexccC+ae9/XGjbPTrp2LQoVMevXysHevhU2b9CNN8j99ikVEJMc0b+5n5cp0/vWvAHv3Gvz2m4Uvv7TSubOLp55y4HUkcHjW+/hq3kxCj64433oj7Gu43fDCC3aaNfOzfHkGffp4sdtNFi6MicA7EsldCmYiIpKjLr3UZObMTFauzGD16gzWrUunSxcvkybZadMmlt+PJXBk1vv46jUg4dFU4h9NhczMkM+/bp0Vt9ugc2cvsbFQqBA0bhxg8WKbhjMl31MwExGRiHI4YNQoD5MmZfLtt1aSklys3VSII7Pnkt5vALFvvUGR25ti3bEtpPOtWWMlJsakTp1A9mOtWvnYs0fDmZL/6RMsIiK5om1bPx99lEHRoibt2sUy7sVY0h7/N0fenoNlz28UafQvXOPGgNd71vOsWWOjRo0A8fEnHrvtNr+GM6VAUDATEZFcU7lykKVLM2jTxs+oUQ46dYrljxubceiT9Xhua0HcqBEUSaqH/cMl4POdcvyBAwbffGOlYcPASY9rOFMKCgUzERHJVfHxMGmSm9Gj3XzyiZWkpDg27C7Lsdemc+StdzEyMih8X0eK3XA1cU8Nxrpze/axn3xiBaBhQ/8p5z0+nPnll/rRJvmXPr0iIpLrDAMeeMDH4sUZWCzQurWLKVNi8DS9jYNfbObIzHfw1byZ2NcmUeRfNUjodj/Wb7awZo2VwoVNbrjh1G6x48OZc+dqOFPyLwUzERGJmhtuCLJiRTpNmgQYMsRJp06x/HnYjrdZc45Of4u/vtpKZq9Hsa9aQdGkenSa24GUG7djtZ56rkKFIDnZz4wZMXz3nX68Sf5kmKZpRrsR4fD5Ahw+nBHx6yQmunLlOgWBahUe1St0qlV48nO9TBNefz2Gp592UKiQSYsWfvx+8PkMmjb1c0ejA2SOnkzRqeOx24J4+/Ylo9ejEBt70nn++sugfn0XZcqYLF2aQcwZOs/yc61ym2oVnn/Wq0SJhLCO1z8pREQk6gwDunb1sWxZBuXLmyxaZGPlShsrV1rp1i2WhweV5tWS/6YKWznaqAVxz42iaJ0bcb49E/wn5psVK2YydqyHb76x8uKL9ii+I5Hzo2AmIiJ5xjXXBFm2LIMffkjn66/T2bIlnccf9zBvno1nn3VgubQcwbencXjBhwTLlCGh7yMUaVQHxztvY6QdA6BFCz8pKT7GjbPzzTf6MSf5S8Q+sYMGDaJOnTq0bNnyrK/bsmULV199NUuXLo1UU0REJJ+y2aB/fy8LF2ZQuXKAdu2yltDw1anL4Q9WcuT1N8E0KdSrB8WqViSh2/045r3H6Md+o2hRk3vuiWXLFoUzyT8i9mlNSUlh6tSpZ31NIBDgueeeo169epFqhoiIFAA1awb59NMMnnjib4vPGgbelq059OkGDi1ejrtDJ+yfrqFQ9y5UrFuBHa5rGXb0UV65/WM+fM8TvcaLhMEWqRPXrFmT3bt3n/U1M2fOpFmzZnzzzTeRaoaIiBR0hoG/1s2k1bqZtP+MwbblK2I+XYt97Rq6/D6FB70v4elpZ++wGylxS1XM66ph3HgdlkLFCZYpC3bNRZO8I2LB7Fz27dvHihUrmDFjRp4LZku+28cHW/fj92v56FDYbBbVKgyqV+hUq/CoXsfFQKkkaJcEbYNw5ChHfzmMNeMYCf40rF9mwpfrTrzcZgOrFdNqzboL4TiLBWwxmDG2/73Ghmm1YsbHYxYqnPtvK0oK8ueq9bWlaVG1VLSbcZKoBbP//Oc/9O/fH+vpFqM5C6vVIDHRFaFWZXG57BiGgc2meQmhUK3Co3qFTrUKj+p1OhYoUZSiJYpy8CBs2GZgeNxcWiKT0oluDK8na+unQAAjEMhat+O4QAB8XoyMdPD7IPC3cFKmDGaFClnhrYAryJ8rl8ue45nCarVc0DmjFsy+/fZb+vXrB8ChQ4dYs2YNNpuNpk2bnvW4QMCM+Hoqja8oQpvq5bRuS4i0xk14VK/QqVbhUb3OLS0NnnnGwYyX7Nx8s5+pU92UKhXicp7BIMbRI7heeJ7Y8U8SqFCRY5Om4r++emQbHWUF/XOV0+8t365jtmrVquxfzZo146mnnjpnKBMREbkQ8fEwdqyHGTOCfPONlaQkF59/HuLIjcWCmViE9KdHcOS9hRjp6SS2aoZ9xbLINlouKhELZv369aNDhw7s2rWLBg0aMGfOHGbNmsWsWbMidUkREZGQdOxo8uGHGcTHQ5s2sTz3nJ1AIPTjffUbcmjlp/grV6FQ54445r8fucbKRUVbMp1BQe+6zUmqVXhUr9CpVuFRvUJ3vFZpafD4407eey+GOnX8TJzoply50H8sGkePUOjeDsR88V/S/jMad9fuJ91AYJrgdp+yc1S+os9VePLtUKaIiEi0xcfDxIluXnklky1brDRt6uKLL0K/Kc0sVJgjs+fivaUZCU8+TuE7bse69Yfs53v1clK/fhxudyRaLwWRgpmIiFz02rf3s3x5OoULQ9u2scyeHca9cbGxHH1jNseefwnbD99RpEld4p4azOp3DvLuuzH8+quF+fOjdq+d5DMKZiIiIkDFiiZLl6ZTp06A3r1jGTrUgc8X4sEWC+577+fgf7/E3a4Dsa++QsteVzO9aF/qXrmbqVPt5K+JQxItCmYiIiL/k5gIs2Zl0rWrl1dftZOc7GL3buPcB/6PWbw4aS9O5ImWX/G+cSf3Hn6Fj3+pwGNbHuDHWV9HsOVSUCiYiYiI/E1MDIwc6eG11zLZutVCkyZxLF0a+ryztWutjF10HRtSX+PQ55tI73Q/bZhH/b71Sby9KY5Zb0J6egTfgeRnCmYiIiKnkZzsZ8WKdC65JEjnzi769nVw9OjZj0lPh379nFx5ZZD+/b0Er7gS73PPMfSBnTxmGUfgr8MU6tOTYtUqE9/3EewrPwKPNliXEzQbUURE5AyuvNLkgw8yeP55Oy+9ZGfNGhu9e3uJjc2aMFalSpAbbjixVdOoUQ5+/dXCggUZJy2RcffDsdw8vS/c8TBDmnxC7FszcCycT+zbMwnGJ+BNugVf/Yb46tUncEWFk/fslIuK1jE7A63bEjrVKjyqV+hUq/CoXqE7n1pt2mQhNTWWHTtODDZZLCZPPeWhRw8fGzdaaNnSxf33+xg9+tResHvvjeWzz6y8/LKbFi384PFg/+Rj7B8sxr7iI6x//A5AoFRp/NdWI1C1Gv6rryFw6WUEL7mUYMlSUdmbU5+r8FzoOmYKZmegD2LoVKvwqF6hU63Co3qF7nxr5fPB3r1ZvVmBAIwY4WDx4hjuvtvLxo1W0tMN1q5NJ+E0P4t/+82gS5dYvv7ayv33e3nmGc+JXjXTxLpzOzGfrCVm/RfYvvsW67YfMfz+7ONNmw0zsQjBokUJFi9BsFx5ApdcSrBcecnzxLoAABcnSURBVIKJiZgJhQiWKk2gytU52uOmz1V4FMwiRB/E0KlW4VG9QqdahUf1Cl1O1SoYhDFj7Iwb5wBg1qwMkpLOvLeT1wv/+Y+DSZPsXH99gEWLMnA6z/Bijwfrrp1Yd/+K5ddfsf6+F+PgQSyHDmL5cz+WPbux7N2DEQyedFjg0svxJLfB3eZOAtdWu+D3qM9VeBTMIkQfxNCpVuFRvUKnWoVH9QpdTtfqww9t7N1r0LVraAufLVpko2vXWHr18vDvf3vP/8I+H5Y/92McOYJx5Ai2Hdv4//buPDrKKs3j+LeqUllJgISBSAybLbSy9iA9gkBLAohAIB2NEwTb6XEZooKQVlpoRhsYtghKRBoj4mHadqIcloROWiJhFdAJjCJggzAskqgEyYJJKkmlqt75I90Z0ASTmKRe4Pc5x3OsN8VbTz3nvlVP3Xvfe/3+nI59zy4sLheVCZMp+/1/YISGNfkl1K4aR1syiYiIeNm997oaXJQBxMS4eOghJ6tW+fLxxz/iq9hux9M5Avdtt+O6czCVk3/FpXc2UXj0JI7pSfhteJfQu+7Ab30aXDYs2hTnz1uIjQ0gL083JrQkFWYiIiJe8PvfV3HTTQZPP+3f7HtpGqFhlM/9PcXb9uDu1p2Qp/6N0IF9CFy6EGt+XpPOuXOnjf37fUhLszdvsHIFLZchIiLiBcHBsHx5JQkJgTz7rD933OGmvBwqKy243TXz18LDDSZPrsanid/W7t59KMnchu/Wv+D/p3UEvpRM4EvJVA+7m8qEB6kaGwOBgQ0615EjNYvsbtniw7PPOrWiRwtRYSYiIuIlUVFuHn7YyX/+py/vvlt3T9SmTT6kplYSHt7EKeE2G85xMTjHxWDNO4d/2p/wX59GyBOPYQQGUf3zf6J68F01/93xc+qrAo8cqRlkO3HCxrFjVm6/3VPn8+TH0VCmiIiIFyUnV3HgQBlHjpRx6lQpX35ZyvnzpRQUlLJqVQWffmojKiqQ3bsbvi1UfTyRXXDMmkNR7qeUpP+FyoQHsRYUELR4Ae0mjCGsz09o8/QT+L7/Hpfv4O7xwNGjNiZOrMZqNdiyRf06LUWFmYiIiBdZLNC1q0GnTgbBwTV7dVqtNcfj411kZzsICzN44IEAkpN9cde/GkfDWa1UDxlK2ZLlFO/+kIvHz3Bp7R9xjhiJX9afaTvlnwnr/1OC5v07fP45Z89aKC+3cPfdboYOdZOebufaWtPh2qHCTERExMR69fKwdauD+HgXy5b58cADAVy40LwTvIzQMJwxsZSufoPCv57i0p/epXrQPxGQugp73950nzSC6aQwMOIrJk50cfq0laNHVUK0BK1jVg+t29JwylXjKF8Np1w1jvLVcNdirgwD0tJ8eO45f0JCDF57rZKhQ5uj+6x+loIC2mdt4qvlaUR+cwjDYqGy3x0sPhxD8D+P5F9e+mm9c9Ka/JqXSrCdOY313BdYnH9b481mwxk1EqNtu2Z9rZagBWZbyLV40XqLctU4ylfDKVeNo3w13LWcq88+s/LoowGcOWPhmWeczJzpxPbjp5/Vq127QEaPNgjOP07aL/8L35xs7J98DIAREICrd19c/QdQ3f9nuPoNwN2zV4OKNcvFi/j89Sg+xz7DduJzfE58ju1/T2AtLKzz+e4uXfl2zTpcPxvYrO+vuakwayHX8kXb2pSrxlG+Gk65ahzlq+Gu9VyVlcGsWf5s2GCnf383kZEerFbo0MFg5kwnnTo131d727aBRERYGDXKTUpKzYJr6amF7Pz3/SyY8CGRFz7GfvhTLI5yAAw/Pzw3dcZ9U+eajdft9poJc4aBtagQy4UL2M5/hfXixdrX8ISG4r61F66evXB3vwV39x64u3aDwJrNRK35+QTPeBLrhQLK5i2i8l8fa9b9QJuTCrMWcq1ftK1JuWoc5avhlKvGUb4a7nrI1d+HNt980xens2ZT9XPnrAQHG7z0UiVjxjTPMGd5eSDdu9tYtKiSRx+tuVOzrAyGDQsiKMggJ8eBv92N7fQpfA4fwufoEaxf5WM9fx7rhQIs1S6gptTwhIbi6dgJT8dONYXY7b1x3dYbo2PHH4zDUlxE8FP/ht+2bKrv+DmOaTNx3nNvzZ0SJqLCrIVcDxdta1GuGkf5ajjlqnGUr4a7XnP1+edWEhP9OXrUxkMPOVmwoKqh68fWa9++QH75Sxtbtji4887/L/Z27LCRkBDI9OlVzJ37I/b7bAyPB/+31hG48mVs577A1bMXlVMepiomFk/Eza0Tww/QXpkiIiIC1NzB+d57Dp580slbb/ly772BnDr144b8Dh2yYLEY9OlzZQ9cVJSbBx908uqrvnzySSuVE1YrlQ//K0UffcK3r63FCAikzfNzCPvZ7bQbN4qAV1OwHT/GtbyWh3rM6nG9/ppqCcpV4yhfDadcNY7y1XA3Qq527LDxxBP+VFVZWLGikokTm7aJ+WOPteHIEYOPPir/3t8uXYLhw4No29bg/fcd+Pv/2Kgbz3b6f/Hbko7vlnTsRw8D4I7sgvMXI6i+axjVdw3DE35Tq8WjHjMRERH5nqgoN9u3O7jtNg+PPRbA7Nl+VFU1/jyffgp9+9Y9X61tW3jppUqOH7cxblwgp0+3/oR8d4+f4JjxDCU79lJ46Bily1Jw9e6L35Z0QhIfJaxfL0IH9iHkkV8R8MrL2D/cR7PvGt+MVJiJiIhcpyIiDDIyHCQmOlm71pcJEwI5d67hxVNxMZw9a6Fv3/r3xYyOdvPWWw7y8qyMHBlERob3tmvydI6g8le/5ts/plH4+VmKc/ZQNm8R1f94Bz6HD9HmP16g3cR76fCTm6n6+RjOvPM/Xou1PirMRERErmN2O8ybV8W6dRWcOmUlOjqI7OwfXvjM5YJly/wA6N//6nd43nOPmx07yunVq6Z3btYsP+93StlsuPoNoCLxKUrXrKPowGEuHjvDpT++w1f3JXLqrA9n/rvuNdO8SYWZiIjIDWDsWBc5OeV07erhoYcCmT/f9/J9yq9QVAQJCQGsWePL1Kkehg374aU3br7ZYMuWmhsP1q3zZexY7wxtXo0RFoZzzFjm+iVzj98u+v3uHm+H9D3aHl5EROQG0a2bQWamg+ef9+PVV/3YtcuHm2/2UF5uobLSgt1u4OsLJ05YuXjRwooVFTzxhC8lJQ07v90OL7xQxZAhLqZN8yc6OojlyyuJi2vajQctoaQE1q+3ExfnokMH893/qB4zERGRG4i/PyQnV/HaaxV4PJCXZ6Wy0oK/v4FhQGmphW7dPGRkOHjwwaYVVKNGudmxw0Hv3m6mTg0gKckPh0lugn37bTsOh4XHHmultdcaST1mIiIiN6C4OFeL9mR17myQnl5BcrIvKSm+5ObaGD7cTUCAQUgIJCRUN+vWUQ3hcsGbb/oyZIiLPn3qv6HBm9RjJiIiIi3CxwfmzHHyzjsVAGzYYOf1131ZuNCPqKhA9u9vwd3X67B1qw95edbaraXMSIWZiIiItKgRI9zs3evgxIky8vLK2LOnnJAQuO++AFatsrfaQv1vvGEnMtLDvfeaZ87bd7VYYTZ79mwGDx7M+PHj6/x7Tk4OMTExTJw4kbi4OA4ePNhSoYiIiIiJ/PSnHt5/v5wxY1zMm+fPgw8GcP58y97BmZ9vYf9+Hx56qBpb63bUNUqLzTGLi4tjypQp/Pa3v63z74MHDyY6OhqLxcLx48eZMWMGW7dubalwRERExESCg+HNNytZu9bNggV+DB8exOLFldx5p5vycgsOR82cMI8HrFbo18+Dr2/TX+8vf6kpeWJizDuMCS1YmA0aNIj8/Px6/x4UFFT7/xUVFVgs5lrrRERERFqWxQKPPlrNiBEunnoqgMTEgHqfO3Cgm7VrK+jcuWnjnllZPtx2m5tbbjHfEhmX8+pdmdu2bWP58uUUFRWRmprqzVBERETES265xeDPf3awZYsPDoeFoCCDwEADu72meMvPt/L8836MHBnI2rWVDB78wwveXu7CBQsffWQjKcmcS2RczmIYLTflLj8/n6lTp5KZmXnV5x04cIBVq1axbt26Hzynx+PB7W75atdms+J2m/NWWrNRrhpH+Wo45apxlK+GU64aziy5+utfIT7eypkz0KULVFWB2w3x8Qbz5xtcNhD3PW+8YeGJJ6wcPOimX7+WjfO7+bLbGzehzRTrmA0aNIhz585RVFREaGjoVZ/rdhuUlLT8KnXt2gW2yutcD5SrxlG+Gk65ahzlq+GUq4YzS646d4b33oMXX/Tjm28s+PlBaSmsXGlnyxYPKSmVDBlSd0/a+vUBdO8OkZGOBu9i0FTfzdc//ENwo/691wqzL774gi5dumCxWPjss8+orq6mffv23gpHRERETC4kBBYsqLri2P791Tz9tD+xsYE89VQVc+Y48bmsuikpgb17bUyd6uRamM7eYoVZUlISubm5FBcXM3z4cKZNm4bLVbNuyKRJk8jOziYjIwMfHx/8/f15+eWXdQOAiIiINMqQIW527SrnhRdq9v/89FMbqamVtftgZmf74HJZGD/evGuXXa5F55i1hOpqt4YyTUa5ahzlq+GUq8ZRvhpOuWq4aylX77zjw6xZ/oSGGkyZUk1YmMHGjT7k51v5+ONyrK2wrP41O5QpIiIi0pwSElz07u0gMdGf5GS/2uOJic5WKcqagwozERERuW707eth714H1dVQVGShpMRC9+7ev6u0oVSYiYiIyHXHbodOnQw6dbqmZmxpE3MRERERs1BhJiIiImISKsxERERETEKFmYiIiIhJqDATERERMQkVZiIiIiImocJMRERExCRUmImIiIiYhAozEREREZNQYSYiIiJiEhbDMK6tvQpERERErlPqMRMRERExCRVmIiIiIiahwkxERETEJFSYiYiIiJiECjMRERERk1BhJiIiImISPt4OwGz27NnDwoUL8Xg8xMfH8/jjj3s7JNP4+uuvmTVrFhcvXsRqtfLAAw/w8MMPs3LlStavX09oaCgASUlJ/OIXv/BytOYQFRVFUFAQVqsVm83Gpk2bKCkpYebMmXz55ZdERESwYsUK2rZt6+1Qver06dPMnDmz9nFeXh7Tp0+ntLRUbetvZs+eza5duwgLCyMzMxOg3rZkGAYLFy5k9+7d+Pv7s2TJEnr37u3ld9C66srX0qVL2blzJ3a7nS5durB48WJCQkLIz89n7NixdO/eHYD+/fszf/58b4bfqurK1dU+11NTU9mwYQNWq5W5c+cybNgwr8XuDXXla8aMGZw5cwaA0tJSgoODycjIaFrbMqSWy+UyoqOjjXPnzhlVVVVGTEyMcfLkSW+HZRoFBQXG0aNHDcMwjNLSUmP06NHGyZMnjVdeecV44403vBydOY0YMcIoLCy84tjSpUuN1NRUwzAMIzU11UhOTvZGaKblcrmMIUOGGPn5+Wpbl8nNzTWOHj1qjBs3rvZYfW1p165dxiOPPGJ4PB7jk08+Me6//36vxOxNdeXrgw8+MKqrqw3DMIzk5OTafOXl5V3xvBtNXbmq79o7efKkERMTY1RVVRnnzp0zoqOjDZfL1Zrhel1d+brc4sWLjZUrVxqG0bS2paHMyxw+fJiuXbsSGRmJr68v48aNY/v27d4OyzQ6duxY+6u7TZs29OjRg4KCAi9Hde3Zvn07sbGxAMTGxpKTk+PliMzlww8/JDIykoiICG+HYiqDBg36Xs9qfW3p78ctFgsDBgzg22+/5cKFC60eszfVla+hQ4fi41MzUDRgwADOnz/vjdBMp65c1Wf79u2MGzcOX19fIiMj6dq1K4cPH27hCM3lavkyDIP33nuP8ePHN/n8KswuU1BQQHh4eO3jTp06qfCoR35+PseOHaN///4AvP3228TExDB79mwuXbrk5ejM5ZFHHiEuLo53330XgMLCQjp27AjUFLtFRUXeDM90srKyrvhQU9uqX31t6bufZeHh4fos+46NGzcyfPjw2sf5+fnExsYyZcoUDh486MXIzKOua0/fk1d38OBBwsLC6NatW+2xxrYtFWaXMerYncpisXghEnMrLy9n+vTpzJkzhzZt2jBp0iS2bdtGRkYGHTt2ZMmSJd4O0TTS0tLYvHkza9as4e233+bAgQPeDsnUnE4nO3bsYMyYMQBqW02kz7KrW716NTabjQkTJgA1Re3OnTtJT0/nueee4ze/+Q1lZWVejtK76rv21LauLjMz84oflk1pWyrMLhMeHn5F13ZBQUHtr1GpUV1dzfTp04mJiWH06NEAdOjQAZvNhtVqJT4+niNHjng5SvPo1KkTAGFhYYwaNYrDhw8TFhZWO6x04cKF2sm1UnPzTe/evenQoQOgtvVD6mtL3/0sO3/+vD7L/mbz5s3s2rWLZcuW1RYUvr6+tG/fHoA+ffrQpUuX2oncN6r6rj19T9bP5XKxbds2xo4dW3usKW1Lhdll+vbty9mzZ8nLy8PpdJKVlUVUVJS3wzINwzD43e9+R48ePfj1r39de/zyuSs5OTnceuut3gjPdBwOR+0vI4fDwb59+7j11luJiooiPT0dgPT0dKKjo70ZpqlkZWUxbty42sdqW1dXX1v6+3HDMDh06BDBwcH68qSm8F+zZg2rV68mICCg9nhRURFutxuouSP47NmzREZGeitMU6jv2ouKiiIrKwun01mbq379+nkrTFPZv38/PXr0uGKotylty2LU1S95A9u9ezeLFi3C7XZz3333kZiY6O2QTOPgwYNMnjyZnj17YrXW1PRJSUlkZmZy/PhxACIiIpg/f76+BKi5CJ988kkA3G4348ePJzExkeLiYmbMmMHXX3/NTTfdREpKCu3atfNytN5XUVHB3XffTU5ODsHBwQA8++yzalt/k5SURG5uLsXFxYSFhTFt2jRGjhxZZ1syDIP58+fzwQcfEBAQwKJFi+jbt6+330Krqitfr7/+Ok6ns/Z6+/vSBdnZ2bzyyivYbDZsNhvTpk27oX6U15Wr3Nzceq+91atXs3HjRmw2G3PmzLnhlrCpK1/x8fE899xz9O/fn0mTJtU+tyltS4WZiIiIiEloKFNERETEJFSYiYiIiJiECjMRERERk1BhJiIiImISKsxERERETEKFmYhIHfLz83/UfnciIk2hwkxERETEJHy8HYCISEtJT09n7dq1WCwWIiMjOXbsGNnZ2djtdsrKypgwYQLZ2dl89dVXvPDCCxQVFWGz2UhJSaldRBlqFghetmwZubm5OJ1OJk+eTEJCghffmYhcr1SYich16eTJk6xevZq0tDRCQ0MpKSlh6dKl7N69m5EjR5KVlcXo0aOx2+0888wzPP7444waNYqqqio8Hg+FhYW159qwYQPBwcFs3LgRp9NJQkICd9111w2/bY+IND8NZYrIdemjjz5izJgxtRt7t2vXjvvvv5+NGzcCsGnTJuLi4igrK6OgoIBRo0YB4Ofnd8U+igD79u0jIyODiRMnEh8fT0lJCV988UXrviERuSGox0xErkt17TY3cOBA5s2bR25uLm63m549e9ZuNP9D55o7dy7Dhg1riVBFRGqpx0xErkuDBw9m69atFBcXA1BSUgJAbGwsSUlJxMXFAdCmTRvCw8PJyckBwOl0UlFRccW5hg4dSlpaGtXV1QCcOXMGh8PRWm9FRG4g2sRcRK5bmzdvZu3atVitVm6//XaWLFnCN998Q3R0NHv37iUkJASAs2fP8vzzz1NcXIzdbiclJQWLxcLUqVPJzMzE4/GwYsUKdu7ciWEYtG/fnj/84Q8EBwd7+R2KyPVGhZmI3FC2bt3K9u3befHFF70diojI92iOmYjcMBYsWMCePXt4/fXXvR2KiEid1GMmIiIiYhKa/C8iIiJiEirMRERERExChZmIiIiISagwExERETEJFWYiIiIiJqHCTERERMQk/g+ASaN4nnj6nwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "plot_df = dataset.loc[(dataset['cycle']>=1),['cycle','Capacity']]\n",
    "plot_per=data_test.loc[(data_test['cycle']>=90),['cycle','pre']]\n",
    "sns.set_style(\"darkgrid\")\n",
    "plt.figure(figsize=(10, 5))\n",
    "plt.plot(plot_df['cycle'], plot_df['Capacity'], label=\"Actual data\", color='blue')\n",
    "plt.plot(plot_per['cycle'],plot_per['pre'],label=\"Prediction data\", color='red')\n",
    "#plt.plot(pred)\n",
    "#Draw threshold\n",
    "plt.plot([0.,168], [1.4, 1.4])\n",
    "plt.ylabel('Capacity')\n",
    "# make x-axis ticks legible\n",
    "adf = plt.gca().get_xaxis().get_major_formatter()\n",
    "plt.xlabel('cycle')\n",
    "plt.title('Discharge B0005')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.6 (tensorflow)",
   "language": "python",
   "name": "tensorflow"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}