GasUsage.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "%matplotlib inline\n",
    "from matplotlib import pyplot as plt \n",
    "import psycopg2\n",
    "import sqlite3\n",
    "import os\n",
    "from datetime import datetime\n",
    "import statsmodels.api as sm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {},
   "outputs": [],
   "source": [
    "gas_usage=pd.read_csv('/Users/rhl/Downloads/BillingData/gas.csv',skiprows=5)\n",
    "gas_usage['START_DATE']=pd.to_datetime(gas_usage['START DATE'])\n",
    "gas_usage['END_DATE']=pd.to_datetime(gas_usage['END DATE'])\n",
    "gas_usage['USAGE']=gas_usage['USAGE']\n",
    "gas_usage['COST']=gas_usage['COST'].str.replace('$','').str.replace(',','').astype(float)\n",
    "gas_usage = gas_usage[['USAGE', 'COST', 'START_DATE', 'END_DATE']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "metadata": {},
   "outputs": [],
   "source": [
    "hdd = pd.read_csv('/Users/rhl/Downloads/hdd.csv',skiprows=6)\n",
    "hdd.columns = ['DATE','HDD','%EST']\n",
    "hdd['DATE']=pd.to_datetime(hdd['DATE'])\n",
    "del hdd['%EST']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {},
   "outputs": [],
   "source": [
    "gas_usage['key']=1.0\n",
    "hdd['key']=1.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = gas_usage.merge(hdd, on='key', how='outer').query('(START_DATE <= DATE)&(DATE<=END_DATE)').groupby(['USAGE', 'COST', 'START_DATE', 'END_DATE']).agg({'HDD':'sum'}).reset_index()\n",
    "df['DAYS_IN_PERIOD']=df['END_DATE']-df['START_DATE']\n",
    "df['MIDDLE_DATE']=df['START_DATE']+df['DAYS_IN_PERIOD']/2.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>USAGE</th>\n",
       "      <th>COST</th>\n",
       "      <th>START_DATE</th>\n",
       "      <th>END_DATE</th>\n",
       "      <th>HDD</th>\n",
       "      <th>DAYS_IN_PERIOD</th>\n",
       "      <th>MIDDLE_DATE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>21.07</td>\n",
       "      <td>2018-09-15</td>\n",
       "      <td>2018-09-30</td>\n",
       "      <td>89.9</td>\n",
       "      <td>15 days</td>\n",
       "      <td>2018-09-22 12:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>8.0</td>\n",
       "      <td>34.55</td>\n",
       "      <td>2019-07-18</td>\n",
       "      <td>2019-08-15</td>\n",
       "      <td>22.0</td>\n",
       "      <td>28 days</td>\n",
       "      <td>2019-08-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>8.0</td>\n",
       "      <td>35.33</td>\n",
       "      <td>2019-06-18</td>\n",
       "      <td>2019-07-17</td>\n",
       "      <td>31.4</td>\n",
       "      <td>29 days</td>\n",
       "      <td>2019-07-02 12:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>8.0</td>\n",
       "      <td>36.36</td>\n",
       "      <td>2019-08-16</td>\n",
       "      <td>2019-09-16</td>\n",
       "      <td>104.6</td>\n",
       "      <td>31 days</td>\n",
       "      <td>2019-08-31 12:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>10.0</td>\n",
       "      <td>34.44</td>\n",
       "      <td>2017-08-16</td>\n",
       "      <td>2017-09-14</td>\n",
       "      <td>138.8</td>\n",
       "      <td>29 days</td>\n",
       "      <td>2017-08-30 12:00:00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   USAGE   COST START_DATE   END_DATE    HDD DAYS_IN_PERIOD  \\\n",
       "0    6.0  21.07 2018-09-15 2018-09-30   89.9        15 days   \n",
       "1    8.0  34.55 2019-07-18 2019-08-15   22.0        28 days   \n",
       "2    8.0  35.33 2019-06-18 2019-07-17   31.4        29 days   \n",
       "3    8.0  36.36 2019-08-16 2019-09-16  104.6        31 days   \n",
       "4   10.0  34.44 2017-08-16 2017-09-14  138.8        29 days   \n",
       "\n",
       "          MIDDLE_DATE  \n",
       "0 2018-09-22 12:00:00  \n",
       "1 2019-08-01 00:00:00  \n",
       "2 2019-07-02 12:00:00  \n",
       "3 2019-08-31 12:00:00  \n",
       "4 2017-08-30 12:00:00  "
      ]
     },
     "execution_count": 176,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:3: FutureWarning: M and Y units are deprecated and will be removed in a future version.\n",
      "  This is separate from the ipykernel package so we can avoid doing imports until\n",
      "/usr/local/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:4: FutureWarning: M and Y units are deprecated and will be removed in a future version.\n",
      "  after removing the cwd from sys.path.\n"
     ]
    }
   ],
   "source": [
    "def rng():\n",
    "    for yr in range(0,3):\n",
    "        start=(pd.to_datetime('10-15-2017')+pd.Timedelta(yr,'Y')).normalize()\n",
    "        end=(pd.to_datetime('05-15-2018')+pd.Timedelta(yr,'Y')).normalize()\n",
    "        yield (start,end)\n",
    "idx=pd.Series([False]*df.shape[0])\n",
    "df['HEATING_SEASON']=-1\n",
    "for x,y in list(rng()):\n",
    "    maxs = df['START_DATE'].apply(lambda z: max(x,z))\n",
    "    mins = df['END_DATE'].apply(lambda z: min(y,z))\n",
    "    df.loc[(maxs < mins),'HEATING_SEASON']=x.year\n",
    "    idx=np.logical_or(idx,(maxs < mins))\n",
    "df['IS_HEATING']=df['HEATING_SEASON'].replace(-1,np.nan).notnull()\n",
    "df['HEATING_SEASON']=df['HEATING_SEASON'].replace(-1,np.nan).fillna(method='bfill').astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax=df.set_index('MIDDLE_DATE')['USAGE'].plot()\n",
    "_=ax.set_ylabel('CCF')\n",
    "ax=df.set_index('MIDDLE_DATE')['COST'].plot(ax=ax.twinx(),color='r')\n",
    "_=ax.set_ylabel('DOLLARS')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    16.000000\n",
       "mean      0.528775\n",
       "std       0.359055\n",
       "min       0.258065\n",
       "25%       0.380850\n",
       "50%       0.467186\n",
       "75%       0.521025\n",
       "max       1.785714\n",
       "dtype: float64"
      ]
     },
     "execution_count": 179,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "non_htg=df.query(\"IS_HEATING==False\")\n",
    "(non_htg['USAGE']/non_htg['DAYS_IN_PERIOD'].dt.days).describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "metadata": {},
   "outputs": [],
   "source": [
    "def json_data1():\n",
    "    for root, dirs, files in os.walk('/Users/rhl/Downloads/NEST_DATA-3/Nest/thermostats'):\n",
    "        for x in files:\n",
    "            if 'summary' in x:\n",
    "                yield os.path.join(root,x)\n",
    "usage=pd.concat([pd.read_json(x).T for x in json_data1()]).sort_index()\n",
    "def parse_json(x):\n",
    "    startTs = pd.to_datetime(x['startTs'],utc=True).tz_convert('America/New_York')\n",
    "    duration = int(x['duration'][:-1])\n",
    "    return (startTs, startTs+pd.to_timedelta(duration,'s'),duration)\n",
    "flatten = lambda l: [parse_json(item) for sublist in l for item in sublist]  \n",
    "heating_cycle=pd.DataFrame.from_records(flatten(usage['cycles'].values.tolist()),columns=['startTs','endTs','duration'])\n",
    "heating_cycle['key']=1.0\n",
    "df['key']=1.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {},
   "outputs": [],
   "source": [
    "merged = df.merge(heating_cycle,on='key',how='outer')\n",
    "merged['START_DATE']=pd.to_datetime(merged['START_DATE']).dt.tz_localize('America/New_York')\n",
    "merged['END_DATE']=pd.to_datetime(merged['END_DATE']).dt.tz_localize('America/New_York')\n",
    "merged['START_DT']=merged[['START_DATE','startTs']].max(axis=1)\n",
    "merged['END_DT']=merged[['END_DATE','endTs']].min(axis=1)\n",
    "merged = merged.query('START_DT < END_DT')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {},
   "outputs": [],
   "source": [
    "heating_data = merged.groupby([x for x in df.columns if 'key' not in x])['duration'].sum().reset_index()\n",
    "heating_data['duration']=heating_data['duration']/3600"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 208,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>USAGE</th>\n",
       "      <th>duration</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HEATING_SEASON</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>2017</td>\n",
       "      <td>281.0</td>\n",
       "      <td>189.001667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2018</td>\n",
       "      <td>1492.0</td>\n",
       "      <td>2657.550833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2019</td>\n",
       "      <td>876.0</td>\n",
       "      <td>1283.322500</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 USAGE     duration\n",
       "HEATING_SEASON                     \n",
       "2017             281.0   189.001667\n",
       "2018            1492.0  2657.550833\n",
       "2019             876.0  1283.322500"
      ]
     },
     "execution_count": 208,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "heating_data.query('IS_HEATING==True').groupby('HEATING_SEASON')[['USAGE','duration']].sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1c2fb887d0>"
      ]
     },
     "execution_count": 216,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEFCAYAAAAPCDf9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOydd3zUVbr/32daJpl0AkmoCUgNJfRQVETABnbEjt7d9beuZVdXXd29q6xtbXvde6+re3V11RUL4lp37aJAKAkIUkMRAgRISC+TNuX8/vjOTIZkkkz5Tibl+369fCVzvud7vieYefLMcz7P8wgpJRoaGhoafQNdpDegoaGhodF1aEZfQ0NDow+hGX0NDQ2NPoRm9DU0NDT6EJrR19DQ0OhDaEZfQ0NDow9hiPQGOiIlJUVmZGREehsaGhoaPYqtW7eWSSn7+7rWrY1+RkYGW7ZsifQ2NDQ0NHoUQogj7V3TwjsaGhoafQjN6GtoaGj0ITSjr6GhodGH6NYxfV/YbDaKiopobGyM9FZ6FGazmcGDB2M0GiO9FQ0NjQjSqdEXQrwCLAZOSSnHu8aeBpYAzcCPwM1SyirXtQeAnwAO4E4p5eeu8fOB/wb0wN+klE8Es+GioiLi4uLIyMhACBHMEn0OKSXl5eUUFRWRmZkZ6e1oaGhEEH/CO68C57ca+xIYL6WcCOwHHgAQQowDrgayXPc8L4TQCyH0wF+AC4BxwDWuuQHT2NhIv379NIMfAEII+vXrp3060tDoIUgpaTp0OCxrd2r0pZRrgYpWY19IKe2ul5uAwa7vLwHellI2SSkPAweBGa7/DkopD0kpm4G3XXODQjP4gaP9m2lo9BysuRs4dOGFNOzcqfraahzk/gfwqev7QcAxr2tFrrH2xtsghLhFCLFFCLGltLRUhe2pT2FhIePHjz9tbMWKFTzzzDNs2rSJmTNnkp2dzdixY1mxYsVp8y655BJmzZrVZs033niDiRMnkpWVxaRJk/jpT39KVVUVAPPmzWP06NFkZ2eTnZ3NlVdeGbafTUNDI/I0FxYCULfmW9XXDukgVwjxO8AOrHQP+Zgm8f3HxWf3Finli8CLANOmTetxHV6WL1/OqlWrmDRpEg6Hg3379nmuVVVV8f333xMbG8vhw4c98fXPPvuMZ599lk8//ZRBgwbhcDh47bXXKCkpITExEYCVK1cybdq0iPxMGhoaXYu9+CQAdbnr6X/nHaquHbSnL4RYjnLAe51sab9VBAzxmjYYONHBeK/j1KlTpKenA6DX6xk3ruXo4r333mPJkiVcffXVvP32257xxx57jGeeeYZBgwZ57vuP//gPRo8e3bWb1/BQt249xQ8/HOltaPRRbMUlADTu3IXD9YlfLYLy9F1KnN8AZ0sp670ufQS8KYT4L2AgMBLIQ/kEMFIIkQkcRznsvTaUjQP84ePd7DlRE+oypzFuYDwPLckK+v677rqL0aNHM2/ePM4//3yWL1+O2WwG4K233uKhhx4iNTWVK6+8kgceeACA3bt3M2XKlA7Xve6664iOjgZg4cKFPP3000HvUaNzar/6iqp33iHlttsw9OsX6e1o9DHsxcXoLBacVivWTZuIP7+1liZ4OvX0hRBvARuB0UKIIiHET4DngDjgSyHEdiHEXwGklLuBVcAe4DPgNimlw3XoezvwObAXWOWa2yNp71BUCMGDDz7Ili1bWLRoEW+++Sbnu/5nlZSUcPDgQebOncuoUaMwGAzs2rWrzRo7d+4kOzubESNG8M4773jGV65cyfbt29m+fbtm8LsAe3kZAI27e+yvqUYPxlZSguWsM9HFxVG3fr2qa3fq6Uspr/Ex/HIH8x8DHvMx/m/g3wHtrhNC8chDoV+/flRWVp42VlFR4YnRjxgxgltvvZWf/exn9O/fn/Lyct555x0qKys9c2pqanj77bd59NFHycrK4vvvv+ecc85hwoQJbN++ndtvv52GhoYu/9k0FBzlimCtYdcuYs86K8K70ehLSCmxFxdjOm8R5ORgXZ+LlFI1BZ5WhiEIYmNjSU9P5+uvvwYUg//ZZ58xd+5c/vWvf+E+4jhw4AB6vZ7ExETeeustPvvsMwoLCyksLGTr1q2euP4DDzzAPffcQ1FRkecZmsGPLPaKcgAad++J8E40+hqOykqkzYYhNQ3L3LnYi4tpPnRItfV7XBmG7sLrr7/Obbfdxq9//WsAHnroIUaMGMHvfvc77rrrLmJiYjAYDKxcuZJjx45x9OhRcnJyPPdnZmYSHx/P5s2bufDCCyktLeWCCy7A4XCQmJjI+PHjOe+88zzzvWP6KSkpfPXVV137A/cxHGVuo6+FdzS6FntxMQCGtFTMYxUhiHX9eqJGjFBlfc3oB8m4ceNYs2ZNm3FvVY43x48fbzP2/fffe75fvnw5y5cv93nvt99+G9wmewHSbsdeUoJxkM+0jrDgbGzEabWiT07GXlyMvawMQ0pKlz1fo2/jVu4Y09IwDR6EKSODuvW5JLdjHwJFC+9odFuczc0U3X4HB887n2av0Fe4cVQo8fzYM+cCmrev0bXYXBp9Q2oqAJa5c6nPz8fZ1KTK+prR1+iWOJubOX7nL6n79luw27HmbuiyZ9vLldCOZa5i9Bs0o6/RhdiLS8Bg8Hy6tMyZjWxspGHrVlXW14y+RrfD2dzM8TvupO7bb0lb8RCG1FSsmzZ22fPdRt80dCimjAwad2lGX6PrsJcUYxwwAKFTzLNlxgwwGqlbn6vK+prR1+hWeAz+d9+RtmIFSVdfjSUnh/qNm5BOZ5fswS3X1PdLwTx+vBbe0ehSbMUlGNLSPK91FgsxU6ZgzdWMvkYvo63BXwZAzKwcHFVVNHnVMQonbrmmITkJc1YW9pIS7N20+J9G78NeXIwxLfW0McvcOTTt24ft1KmQ19eMvka3wNncTNEddygG/w9/8Bh8AIurKql146Yu2YujrBwRE4MuJgZzliKZ0+L6Gl2BlBJbSQmG1LTTxmPnzAFQ5WxLM/pBUlxczNVXX82IESMYN24cF154Ifv372f37t3Mnz+fUaNGMXLkSB555BFPslZJSQmLFy9m0qRJnnvcZReys7NJTk4mMzOT7OxsFixYEOGfsOtwG3zrd2sVg7/sqtOuG1NTMQ0f3mVxfXtFhafejnncOBBCC/FodAnO6mpkY2MbTz9qzBj0/fqpEuLRdPpBIKXksssuY/ny5R5d/vbt2ykpKeGmm27ihRdeYNGiRdTX13PFFVfw/PPPc9ttt/Hggw+ycOFCfvnLXwKwY8cOT9kFgJtuuonFixf3qXr5zqYmiu68s12D78aSk0PV++8jm5sRJlNY9+QoL8eQnAyAPjZWOczVMnM1ugBbiaLRb+3pC50Oy5zZWNetRzqdnkPeYNA8/SBYs2YNRqORn//8556x7Oxs9u/fz5w5c1i0aBEAMTExPPfcczzxhNIO+OTJkwwePNhzz8SJE7t2492M0wz+w+0bfADL7FnIhgYaduwI+77s5eXovSprmrOyaPRRHE9DQ23c2bitPX1QQjyOykoa9+wN6Rk929P/9H4oVrmdWNoEuKDjnu27du1i6tSpbcZ3797dZnzEiBHU1dVRU1PDbbfdxrJly3juuedYsGABN998MwMHDlR1+z0FZ1OTEtJZu04x+Fe1b/ABYmbMAJ0O68ZNxIS5mYy9opzoiRM8r83js6j55BPspaUY+vcP67M1+ja2k64SDK6eHN5Y3HH99euJHh98sUnN01eRjirhCSE477zzOHToED/72c8oKChg8uTJdNeWkOHkNIP/yMOdGnwAfXw85qwsrJvCe5grnU4cFZWnefrRWcobTDvM1Qg3tpJi0Ot9lv0wpKQQNXZsyHH9nu3pd+KRh4usrCxWr17tc3zt2rWnjR06dIjY2Fji4uIASE5O5tprr+Xaa69l8eLFrF27liuuuKJL9t0daGPwly71+15LTg7lf/87TqsVncUSlv05qqvB4cCQ3GL0o8a6DnN37SZu3rywPFdDA5RsXEP//gi93uf12LlzKP/7qzjqrOhjg3sPaJ5+EMyfP5+mpiZeeuklz1h+fj4jR45k/fr1ngqYDQ0N3Hnnndx3330AfPPNN9TXK43Gamtr+fHHHxk6dGjX/wARwtnURNHtisFPf/SRgAw+KHF97Hbqt2wJ0w6VQ1wAfb9kz5g+1oIpM1NT8GiEHXtJMcbUtvF8N5Y5c5T3QN7moJ+hGf0gEELw/vvv8+WXXzJixAiysrJYsWIFAwcO5MMPP+TRRx9l9OjRTJgwgenTp3P77bcDsHXrVqZNm8bEiROZNWsWP/3pT5k+fXqEf5quwWPw168n/dFHSAxCoRQ9eTLCZAqrXt/uysY19Dv947U5K0sz+hphp3U2bmuip0xBREdjDaEkQ88O70SQgQMHsmrVKp/X2iuFfO+993Lvvfe2u+arr76qws66H20MfpDhLJ3ZTPSUKVg3hk+v73Bn43p5+gDmrHHUfPwxtlOnMA4YELbna/RdpJTYios91V19oTOZsMyYQV1u8C0UNU9fI+xUvrES67p1IRl8N5acHJr27fMURVMbe5k7vHN6M/To8eMBrcyyRvhw1tUh6+vbaPRbY5kzB9uRozQfOxbUczSjrxF26tasIWrs2JANPrji+kD95uBjmh1hrygHnQ59QsJp4+axY12ZuVqSVl+i+uNPOPXf/90lz+pIo++Nu+R3sCqeTo2+EOIVIcQpIcQur7FkIcSXQogDrq9JrnEhhPgfIcRBIcQOIcQUr3uWu+YfEEKo0wJGo9vjqK2lfts2Ys88U5X1zFlZ6OLiwhbXd5RXoE9ObqOe0Flch7laklafovyVV6h49bUuqfBq87RJ7NjTN2VmYBw4kLr1wYV4/PH0XwXObzV2P/C1lHIk8LXrNcAFwEjXf7cAL4DyRwJ4CJgJzAAecv+h0OjdWDdsBIeD2LPUMfpCrydmxoywxfXtFS0lGFpjHq8d5vYl7GVlNO3di2xo8Hjh4cTm8fQ7NvpCCKWb1sZNSJst4Od0avSllGuBilbDlwCvub5/DbjUa/x1qbAJSBRCpAPnAV9KKSuklJXAl7T9Q6LRC6lbtxZdXBzR2dmqrWnJycFWVBSWFoqOsvLT5JreRGdlYS8tVaW8rUb3x/vTZNPhw2F/nr24BITwK+vbMmcOTquVhh9+CPg5wcb0U6WUJwFcX91yhkGA9+lCkWusvXGNXoyUEuu69Vhmz0YY1BOKueP64fD2lQqbvpugm12ZuZq33zewbtiAMJsBaD4UfqNvKynGkJKCMBo7nWuZlQN6PXVBxPXVPsj1VYNAdjDedgEhbhFCbBFCbOkpJQpWrFjBM888E/I6VVVVPP/8857XJ06c6NEVN5v2H8BeUqJaaMeNafhwDP37Ux+GuL6jvLyNXNON5zBXa5/Y65FSYs3NJfaceegsFpq7yNPvLJ7vRh8fT/TEiUHp9YM1+iWusA2ur+7Pu0XAEK95g4ETHYy3QUr5opRympRyWv9eWNzKbre3e6210R84cKDPcg89Bes6pSSFZa66Rl8IQcysHKyb1G2h6GxsxGm1ok/u5/O6zmLBNHy45un3AZp//BH7qVPEzpmDafhwmg4fCvsz7SVtO2Z1hGXOHBp37cJeWRnQc4I1+h8BbgXOcuBDr/EbXSqeHKDaFf75HFgkhEhyHeAuco31WB577DFGjx7NggUL2Odq4zdv3jy2uEoElJWVkZGRAShJV0uXLmXJkiUsWrSIuro6zj33XKZMmcKECRP48EPln+/+++/nxx9/JDs7m3vvvZfCwkLGu/XhjY3cfPPNTJgwgcmTJ7NmzRrP2pdffjnnn38+I0eO9JR86A7UrV1H1JgxGFPVT2ay5MzCUVFB04EDqq3pLsHQnqcPSpKWZvR7P245pGX2bEyZGTQfLgz7M23FbTtmdUTs3DkgJfUBhjk7DbQKId4C5gEpQogiFBXOE8AqIcRPgKOAu4jKv4ELgYNAPXAzgJSyQgjxCJDvmvewlLL14XDAPJn3JAUVBaEucxpjksfwmxm/6XDO1q1befvtt9m2bRt2u50pU6b4LLXszcaNG9mxYwfJycnY7Xbef/994uPjKSsrIycnh4svvpgnnniCXbt2eZqqFBYWeu7/y1/+AsDOnTspKChg0aJF7N+/H1AauGzbto2oqChGjx7NHXfcwZAhQ9rsoStx1NVR//339Lv55rCsb5mVAyhxffPo0aqsaa9wN0T37emDkqRV89HH2EpOheWPmUb3oG7DBkwZijQyavhwaj76OLyF/uqsOGtrA/L0zRMmoIuPp259LvEXXuj3fZ0afSnlNe1cOtfHXAnc1s46rwCv+L2zbsy6deu47LLLiImJAeDiiy/u9J6FCxeS7JICSin57W9/y9q1a9HpdBw/fpwSV8ec9li/fj133HEHAGPGjGHYsGEeo3/uueeS4EomGjduHEeOHIm40bdu3Ah2u+rxfDfG9HRMGRlYN26k3003qbKm3ePpt2/0vQ9zNaPfO3E2N1Ofl0/i5ZcDYMrIBKCpsNBTZltt7CUujX4Anr7Q67HMmoV1/foOy7q3pkfX3unMIw8nvv6BDQYDTleMubGx8bRrFi8PYeXKlZSWlrJ161aMRiMZGRlt5rfG3WfXF1FRUZ7v9Xp9h+cGXYV17Tp0sbGqSjVbEzMrh5oPP0LabH4pHjrDU2GznZg+gHnMGE/P3Lj554T8TI3uR8O27ciGBixzZgNgGq4Y/eZDh8Nm9D0a/XT/jT6AZe4caj//nOaDB4kaOdKve7QyDEFw1lln8f7779PQ0EBtbS0ff/wxABkZGWzduhWgwwPY6upqBgwYgNFoZM2aNRw5cgSAuLg4amtr233mypUrAdi/fz9Hjx5ltEphDbWRUlK3bh2WWbNUMcbtYcmZhbO+noad6nRPa6mw2X5MX2exYBoxXMvM7cVYN2wAVxIggGnYMNDpwqrgsRe7euP6qd5xE+sqyVAXgIpHM/pBMGXKFJYtW0Z2djZXXHEFZ7pKDNxzzz288MILzJ49m7Kysnbvv+6669iyZQvTpk1j5cqVjBkzBoB+/foxZ84cxo8f36Ya5y9+8QscDgcTJkxg2bJlvPrqq6d5+N2JpgMHsBcXYwlTaMeNZeYMEEI1vb6jvBxdTAy66OgO50VnZdGwRzvM7a1YN2wgOjsbfWwsALqoKIyDBoVVwWNzh3cCrOBqTE/HNGIE1kBKMkgpu+1/U6dOla3Zs2dPmzEN/+iqf7uyv/1N7hk9RjafPBn2Zx26/Ap5+LrrVFmr6Nf3yAMLFnY6r/y115Sfr7hEledqdB9sFRVyz5ix8tRzz502fuSWW+SPl1watuee+P2Dct/sOUHde/Kxx+TeiZOko6HBMwZske3YVc3T11CdurXriBo1qtMaImpgmZVDww87cLo6koWCo4O6O96YPWWWtRBPb6N+82aQklhXE3I3URmZNBcWhq3wmq2TjlkdETt3LrKpifotW/2arxl9DVVx1Fmp//77sKl2WhOTMwtsNuq3+vcL3xH28gr0PhpSt8Y8ZgzodFpmbi/EmpuLLi7O84fdjWn4cGRjI/aTJ8Py3ECycVsTM306wmj0O8SjGX0NVanftBFsNixnntUlz4uZOkX5hVehJIO93D9PXxcTQ9QILTO3tyGlxJq7AUvOzDa1okyZGQA0hakGj704sGxcb3TR0URPm4rVz25amtHXUJW6tevQWSzETJncJc/TRUcTPXlyyIe50unEUVHRboXN1pjHZdGwe3eHUlqNnoXtyBFsJ04ozcdbETV8OEBYFDzOhgYc1dUBafRbEzt3Lk0HDmLrJN8HNKOvoSLSLdWcHV6pZmsss3Jo2rs34Bok3jiqq8HpbLfCZmvMWVk4ysqwa2WWew11XqUXWqPv1w9dXFxYFDw2PztmdYSnm5Yf0k3N6GuoRvPBg9hPnsSiUpcsf7HMCr2Foj91d7wxaz1zex3WDRsxDh6MaejQNteEEJiGZ4alBo+9xK3RTw96jahRo9D3T/ErxKMZ/SCIdel33bz66qvcfvvtgFJmedCgQWRnZzNy5Eguv/xy9uxp6as6b948Ro8ezcSJExkzZgy33347VVVVXbr/cFG3TvmFiz2ra+L5bszjx6OLjQ0pru9piN5BNu5pzxzrPszVFDy9AWmzUb95s8/QjpuojEyaD3VPT18IQezsOVhzNyAdjg7nakY/DNx1111s376dAwcOsGzZMubPn493b4CVK1eyY8cOduzYQVRUFJdcckkEd6sedevWEjVyZJdINb0RBgMx06eHFNd3VATm6euio4kaMZwGzdPvFTTs3Imzrs5naMeNafhw7KdO4airU/XZnmzcICWbbixz5+KorqbRy8n0hWb0w8yyZctYtGgRb775ZptrJpOJp556iqNHj/JDEG3PuhNOq5WGLVvDnoXbHpZZOdiOHsV2/HhQ97tLMPgj2XRjzhpP4+492mFuL8CauwF0Oiw5M9ud41bwqB3isZUUo09MROfq0hUs7lpBnUk3e3TBteLHH6dpr7qllaPGjiHtt7/tcE5DQwPZXoXEKioqOqy0OWXKFAoKfO9Tr9czadIkCgoKmDRpUnCb7gZYN29G2mzEdpFUszXuuL510yYSr7gi4Pvt5WWg06F3VSv1B3NWFtUffIC9pKTLP91oqIs1NxfzhPEd/v9vUfAcInrC+HbnBUooGn1vDMnJmMeN67QOj+bpB0F0dDTbt2/3/Pfwww93OL8zT7A3eIp1a9eii4npMqlma0xnnKEcZG0ILsTjKK9An5yM0Pn/ltB65vYOHLW1NOzc2WFoB8A4dCjodKo3SQ8lG7c1lrlzaXD142iPHu3pd+aRdxe2bdvGtGnTfF5zOBzs3LmTsWPHdvGu1ENKiXXtOmJmzUKYTBHZgxACy0xXC8UAaou7URqi+3eI68ZzmLt7N3HntmkvodFDqN+8GRyONqUXWqMzmTAOGax6k3R7cQnREyaqspZl7hzKX3yxwzmapx9m3nvvPb744guuuaZtLxqbzcYDDzzAkCFDmDhRnf/pkaD50CFsJ04Q28VSzdZYZuXgKCsLqoViRw3R20M5zB1Bg6bg6dHU5eaii4kh2o/walTmcFUTtJxNTTgqKkJS7ngTk52NztXcqT26tdG3FRXhtFojvY2AefbZZz2SzTfeeINvvvkG7ybv1113HRMnTmT8+PFYrVZPj9yeSt3adQBdVm+nPTx6/U2BSzft5eV+yzW9MWdlaYe5PRzrhg3EzJjhV0KhKdNVeK0TWaS/qKHR90aYTMTMbP8wGrp5eMdRXU3Jk0+R/vAfIr2V06hrJdm66aabuMnVsm/FihWsWLGi3Xu//fbb8G0sQljXrcV0xgiMAwdGdB/GgQMxDhuKdcNGkm+8MaB7g/H0QckR0A5zey7NRUXYjhwl+fob/JpvGp6JbG7GdvIkpsGDQ36+Ghr91sSedSb89YV2r3drT9+QkkLVqlXUrlkT6a1otIPTaqU+f0vEVDutseTMoj4/HxlAy0hnQwPO+nr0fpZg8MacNQ5AS9LqoVhzNwAtcsfOiMp0t05UJ0nL4+mHUHenNYlXX93h9e5t9AcMIGr0aE7+5++xV1REejsaPrBuzlOkmhEO7bixzMpRcgYCaKHoT5vE9nCXWdaStHom1g0bMKSlYXIZ884wqVx4zePppwbWMasjOhMxhGT0hRB3CSF2CyF2CSHeEkKYhRCZQojNQogDQoh3hBAm19wo1+uDrusZfjyAgU89hbOmhpMPPqjFTbshdevWImJiiJ46NdJbAfDEM+s35/l9jzsbV+9HWeXW6KKjiTrjDE222QORDgfWTZuwzJntt9pLn5SELiFBtRLL9uISdPHx6CwWVdbzh6CNvhBiEHAnME1KOR7QA1cDTwLPSilHApXAT1y3/ASolFKeATzrmtcp5tGj6H/33dR99TXV//wn0Dt07V1NOP7N3FJNS04OughJNVtjSErC0L8/zceO+n2P3V1sLYBsXG/MWVk07tLKLPc0Gnfvxlld3ak+3xshBFGZmep5+ipq9P0l1PCOAYgWQhiAGOAkMB9Y7br+GnCp6/tLXK9xXT9X+PnnNXn5jcTMnEnJY49jkpLy8nLtDRYA0vVvZg4xzbs1zYcPYzt+vNuEdtwY0tM99Uz8weEKHfrTQMUX5qwsHBUV2F0f1TV6BtYNG0CIgIw+KAoetUosq5WNGwhBq3eklMeFEM8AR4EG4AtgK1AlpXSfohUBg1zfDwKOue61CyGqgX5Amfe6QohbgFsAhrpKnAqdjoF/fJxDF1+C+NN/Ufu7355WwEyjc8xmM4NVUBt4U7d2LUDE9fmtMaam0hTAQVtLhc3gjH70+JbMXGO6OtI7jfBjXZ+LeexYDElJAd1nGp6J4/33cdTWoo+LC2kPtuJiJcmvCwna6AshklC890ygCngXuMDHVLdL7surb+OuSylfBF4EmDZtmue6ceBA0h78PSfu+w39v/qazFt+FuzWNVTCunYdphEjMA4a1PnkLsSQnoY1N9fvzFxHRTm6mBh00dFBPS9qzBjQ62nYtYu4BQuCWkOja3FardT/8AP9bloe8L0eBc/hw0SHkFQpm5txlJWpqtzxh1DCOwuAw1LKUimlDfgnMBtIdIV7AAYDJ1zfFwFDAFzXE4CAJDnxS5YQd/75lP7v/3ZaPlQjvDjr66nPz+92Xj6AMTUNZ309Tj9L4PrbEL09dGYzUSNG0Lhb+53sKVjz85Vezp2UXvCFWgoe2yklWmFM7zlG/yiQI4SIccXmzwX2AGuAK11zlgPudNOPXK9xXf9GBhiYF0KQvuIhDImJHL/vPpyNjSFsXyMUrHndS6rpjftNZDt50q/59vKyoOP5bszjx9Oo9cztMVhzNyDMZqKnTAn4XtPgwaDXh6zgsZcoZ0A9xtOXUm5GOZD9HtjpWutF4DfA3UKIgygx+5ddt7wM9HON3w3cH8xz9YmJpD/+OM0Hf6T02WeD3b5GiFjXrlOkmu0Ukosk7pR2ux9NosFVYTPAYmutMWeNUw5z/fxDoxFZrBs2EDNtWlCqM2EyYRoyJHRPPwzZuP4QUhkGKeVDwEOthg8BM3zMbQsSY18AACAASURBVASWhvI8N7FnziXpuuuoeO11YufN89Rc0egaPA3QZ87sNlJNb9xvIr89/YoKor36IwRDtKvMcsPu3REvR6HRMbbiYpp//JHEK6/sfHI7mIYPpzlEBY+nY1YXq3e6dUZuRwy459eYMjM58cBvcVRXR3o7fYrmwkJsx45hOXNupLfiE0P//qDT+SWhlE4njooK9EFk43rjPszVkrS6P57SCwFKNb0xZWbQXHgkpMJrtpJidBYL+lY9t8NNjzX6uuhoBj71FPayMoofeTTS2+lTWNe5q2p2j3o7rRFGI4aUFGx+aPUdVVXgdGIIosKmNzqzWcnM3aUZ/e6OdcMG9P1TiBo1Mug1ooYPR9psQbfnhMho9KEHG32A6AnjSfnFrdR88gnV//pXpLfTZ6hbuw5TZqYqVQbDhSE9DXtx5+EdR3lgDdE7QimzrB3mdmek04l1wwZiZ/tfesEX7lo9geSDtMZW3PXZuNDDjT5Ayi23ED1pEsV/eNhzMKIRPpwNDdTn5XVL1Y43xrR0bCc7/33wNEQPosJma8zjs3BUVmqHud2YpoICHJWVIYV2oMXoh9Ik3V5crHn6wSAMBgY+9STSZuPEAw8gnc5Ib6lXU5+Xh2xuxtJNSim3hzEtFVtJSadet71cSQhXw9P3HOZqZZa7LdYNSjw/JkTxhyEpCX1iYtAllqXNhr20NCI9GHq80QcwDRtG6v33U79xE5VvrIz0dno1dWvXIaKjiZne/aSa3hjS0pH19Thrajqc5/B4+qHF9AGiRo92HeZqSVrdlbrcXKJGjcI4IPRSxoqCJzjZpr2sDKTE0MVyTeglRh8g8aqlxJ59Nqf+9CeaDh6M9HZ6LXXr1mGZMQNdVFSkt9IhHtlmJ4e59opy0OvRJySE/Eyd2UzUyJGagqeb4mxooGHr9yGHdtyYMjNoCtLot2j0NU8/aIQQpD/6CLqYGE7c9xtkc3Okt9TraC4sxHb0KJZuHs+HFu1zZ4e5jvJy9MlJCJ06bwVz1jgad+3SDnO7IfVbtiqhySBKL/giavhwHOXlQUnGw9Exy196jdEHRZ+d/sjDNO7ZQ+nzz0d6O72Olgbo3TueD3iqXXZ2mGsvrwhZrumNOSsLR1UV9hMnOp+s0aVYN2xAGI3ETFOn4Y/Jq/BaoEQqGxd6mdEHiFuwgIQrLqf8xZeo/35bpLfTq7Dm5mLKyMA0ZEikt9IphpQU0OmwlXRs9INtiN4e0ePHA2jtE7sh1g0biJ46Nehqqq3xyDaDUPDYi0sQ0dHo4uNV2Usg9DqjD5D6wG8xDhzIid/8BkedNdLb6TU0HT6Medy4SG/DL4TBgGHAAOydefoVFarINd1EjR4NRiP1efmqrakROvbSUpr27fO7Abo/mAYPBoMhKAWPu2NWKLkCwdIrjb4+1sLAJ5/AVlTEqSefiPR2egXS4cB28mS3q53fEca0tE49fXt5ecgVNr3RRUURv3Ah1R98gNOqORzdBevGjUBopRdaI4xGTEOH0lwYeHjHfjIyGn3opUYfIGbqVPr99KdUvbua2m++ifR2ejz20lKw2XqU0TekpXXo6Tvr65H19arINb1JuuF6nHV1VH3wgarragSPNXcD+qQkzGPHqrquKTMzqBLLtpKSiGTjQi82+gD977idqDFjOPmfv/c0v9YIDneNEWM3Lr3QGmNaGrbi4naVNPaKSgAMKhv96OxszBMmUPmPN7RkwW6AlBLrhg1YZs1STaXlJmp4Js1HjyLt9s4nu/fjcGA/dQpDFzdPcdOrjb4wmRj09FM46+o4+Z+/12R0IeAx+oN6TtlgQ1oqsrERZzuSOocrGzfUCputEUKQfOMNNBcWYl2/XtW1NQKn6cAB7KWlqsbz3Zgyh4PNhq2oyO977GXl4HBERKMPvdzoA0SNHMmAX99N3Zo1VK1eHent9Fg8Rr8H1Yo3upqptFeTyV13R21PHyD+vPMw9O9Pxev/UH1tjcBwl15QM57vxpSZARBQklZLxywtvBM2km64gZicHEr++ATNR49Gejs9kuaiIgz9+3f7TFxvPG0T2zH6jgpXhU0VD3LdCJOJxGuuxrp+fUiVGDVCpz4vH+OwoZ7cDTXxNEkPIK4fyWxc6CNGX+h0DPzj4wi9XsnWDSD+pqFgO36iRx3igndWbseevtoHuW6SrroKYTRS+cYbYVlfo3Okw0H9li1YZrRp5qcK+sRE9MnJASl4ItUxy02fMPqgZGim/efvaNi+ndqvNTVPoNiOH+95Rj8lBfT6drNy7eVl6CwWdGZz2J4ff9FFVH3wIY5OCr9phIem/ftx1tQQEyajD2AaHpiCx1ZSjDCZ0Ccmhm1PHdFnjD5A/OLFGNLTqXr33UhvpUfREzX6AEKvVxK02gvvqNAQvTOSb7wBWV9P1er3wvocDd/U5+UBEDN9etieEZWZGVApBrdGPxKJWRCi0RdCJAohVgshCoQQe4UQs4QQyUKIL4UQB1xfk1xzhRDif4QQB4UQO4QQU9T5EQLYr15P4hVXYM3Npbko+DZnfQ37qVNgt/c4ow8tsk1f2CvKw3KI64153Diip02lcuXKkPqpagSH1R3PD2MoxZQ5HEdFBfbKSr/mR1KjD6F7+v8NfCalHANMAvYC9wNfSylHAl+7XgNcAIx0/XcL8EKIzw6KxCsuByGoek9T8vhLi1yzBxr99DRs7VTadJSVqy7X9EXy9TdgO36cujVrwv4sjRak0xnWeL4bt4LH3y5a9uLiiGn0IQSjL4SIB84CXgaQUjZLKauAS4DXXNNeAy51fX8J8LpU2AQkCiHUP07vBGN6OpYz51L93j+1A10/6YkafTeG1DTsxb47aNkrOqiw2VgDpftV2UPcgnMxDEzX5JtdTNO+fTirq8Ma2gGlxDL4V21TOp3YTp3CGIGSym5C8fSHA6XA34UQ24QQfxNCWIBUKeVJANdXd4uaQcAxr/uLXGNdTtJVV2E/dYq6tWsj8fgeR3MP1Oi7MaanIZuacFRVnTYuHQ4clZUYUtox+mufhv87Cxr8+8jeEcJgIPnaa6nPy6Nx376Q19Pwj66I54PrE7DRSPPhzqW5jooKsNki0jHLTShG3wBMAV6QUk4GrLSEcnzh69SijfslhLhFCLFFCLGltLQ0hO21T+zZZ2Po35+qVdqBrj/Yjh/vcRp9N+4mFa0Pcx3V1eB0om/P0y/eAfYG2PVPVfaReOWVCLOZitdfV2U9jc6x5uVjHBoefb43wmDANGyoXyWW3Z3cIqXRh9CMfhFQJKXc7Hq9GuWPQIk7bOP6esprvnch9sFAm04TUsoXpZTTpJTT+vfvH8L22kcYDCRcfjl1a9e2e8in0UJP1Oi78SRotZJt2ss6aYh+qkD5+sNbquxDn5hIwiWXUPPxJ34f+GkEjzueHzMjvF6+m6jMTL9KLLdk4/ZAoy+lLAaOCSFGu4bOBfYAHwHLXWPLgQ9d338E3OhS8eQA1e4wUCRIXHolOJ1U/VMdT6430xM1+m7cCTCtD3MdFa7ELF+efkMl1BVDwhAoylcttp98/XXI5maq3lmlynoa7dO0fz/O6uqwH+K6MWUOp/nYMaTN1uG8SHbMchOqeucOYKUQYgeQDTwOPAEsFEIcABa6XgP8GzgEHAReAn4R4rNDwjR4MJbZs6lavVqT0nVAT9XouzH06wcGgycL0o276qrPmH6pK+5+9n0g9PDDm6rsJWrkSCyzZ1H51ludGgeN0OiqeL4bU2Ym2O00H+u48Jq9uBiMRvRhKP3hLyEZfSnldlcoZqKU8lIpZaWUslxKea6UcqTra4VrrpRS3ialHCGlnCCl3KLOjxA8iVctxX7ipKcgk0ZberJGH5TcDOOAAW09fZfR9/nmK3WFdjLPgjMWwA9vg1MdxyDphhuwl5RQ88UXqqyn4RtrXl6XxPPdRA131eDppByDrbgE44ABqpd4DoQ+lZHbmrj589EnJ2sHuh3QkzX6bgzp6T48/QrQ69EnJLS9oXQfGGMgYShkXwu1J+HQt6rsJfbsszEOG0qlJt8MG9LppD5/CzHTp3XZMz1N0juJ69uLI9cxy02fNvrCZCLhskupXbNG6Qyl0YaerNF3Y0xNbXNg76hQ2iT69LhO7YWUUaDTwegLwJwI29UJ8QidjuTrrqfhhx9o2LFDlTU1Tqer4/kA+vh49CkpnZZYtpWURFS5A33c6IMipcNup+p9rbWdL5p7haefhr1VBy17R3V3SvfBAFdbPUMUTLgSCj6BRt/NWAIl4fLL0FksVPxDq74ZDro6nu9GUfC0b/SllC5P349DXGs5fPGf0FSr4g4V+rzRj8rMJGb6dOVAV2tt1wbb8eMYBgxAZzJFeitBY0xNQzY34/CSStrLy3zX0W+ogtoT0H90y1j2tWBvhN3vq7IffWwsCZdfTs1nn2E7darzGzQCwpqXh3HIkC5PJjR1UnjNUVWFbG72Lxt357uw4X9hw3Mq7lChzxt9gMSrrsJ29KjHQ9BooSdr9N0YPFr9lsPcditslrnkmf29GmgPnAL9x6gW4gFFvondTtXbb6u2poYSz2/I7zp9vjdRZ5yBo6qK2m98l253Jwj65ekfcbXZ3Pic4vWriGb0gbhFC9EnJFC1StNPt8ZWVNTjjb67baK9pOUw115R4bvC5qm9AByLSeDLI18qY0LApGvg2GYoO6jKnkzDhhF79tlUvv0OzqYmVdbUUPrhOro4nu8m4bJLMY8fT9Evf+XT8PvdMUtKOLIBBs8AWz2s/y9V96kZfUAXFUX8JRdT++VXWrakF9Jux1Zc3AuMvuJZuT19Z309sr7ed4XN0n1giOblo5/x629/TWm964B/4jIQOtUydAGSbrgeR0UFNf/+VLU1+zr1myMTzwfQx8Ux9JWXMY8Z49Pwu52OTrNxS/dBfTlMuREmXg15L0G1eqXgNaPvImnpUqTNRvUHH3Y+uY/QotHvucodcLVDNBo9H6/trmxcnxU2S/dC/1HsqdiLRPL10a+V8fh0GHGuYvRV0uxbZs/GdMYIKv7xus8qoBqBU5+fh3Hw4IgVB9THxzP05b9hHjtWMfxff+25ZjtZDHp9+0X+3LhDOxlzYN79IJ2w9inV9qgZfRdRI0cSPXkyVatWaW9AF6pp9A9+BX9bAMU7VdhV4AidzpWgpXhajk6ycW0pozlYpYRxvjjilUSVfQ3UHIfD6lRnFUKQfP0NNO3ZS8PWraqs2ZeRTif1eflhbY3oD6cZ/l/d5TH89uJiDAMGIPT6jhcozIW4gZCUCUnDYNrN8P0/oPxHVfanGX0vEpcupfnwYe0N6MIt1zSFYvSlhG8eVWrYvHI+7I9MJqohPQ27K7xj92TjtjL6jdVQc5xDCanYnDYyEzLZWrKVsgalOBujL4KoBFUPdBMuuRhdQoJWa18F3PH8SBzitkYfF9fG4/dLoy8lHMlVvHx3O8Uz7wG9CdY8rsreNKPvRfz556GLjdV66Lpwe/qGUD4qH9sMJ7bBWfdC8nB4axlsflGlHfqPMS0dmyum6qm70zqm7yqsttdVQvq27NtwSiffHHXFZo1mmHAF7P1YabKiArroaJKWXkntV19hO9Gm6KxGALjj+ZYIxPN94TH8WeMo+uWvaNy5s3PlTsUhqCuBYbNbxuJSIefnsGu1Kp+WNaPvhS4mhoSLl1Dz2edKvfU+ju34idA1+pueB3MCzL0Lbv4URp0Pn94L/75Ptdi4PxjTUpUELacTR7mrwmZr9U6potwpkPVEG6JZMHQBGfEZLSoegOzrlDr7e9RL5ku65hoAKt9U7xNEX6Q+P1+J53cj4YE+Lo6hf/sb0VlZOK3WzjX6ha54/rC5p4/P+aXyKfObR0Pek2b0W5G4dCmyqYnqjz6O9FYiTsgllauOwt6PeWv0mcx7/yJWFf4b51Wvw6zbIe//4K1rwpJx6AtDWjrSZlMaWFeUo4uNbdsUxqXc2Ws9waikUeh1ehYOW0h+cT6VjS5V16Cp0G+kqiEe46BBxC1YQOW7q3HW16u2bl9CqbcT+Xi+L/RxcQx5+W8kXHopcQsXdDz5SC5YBkDKyNPHo5Ngzp2w/zM4utn3vX6iGf1WmMeOxTx+PFXvvtvnD3RDNvp5L7HDZOKp2t3YpZ1HNj3CjZ/fzP4Zy+GiPykHvK9coKocrT08ss3ikvYbop/aizPlDPZV7mNM8hgAFg5biEM6WkI8QigZukc3qnawBpB84w04q6up+VSTbwZD04GDOKqqukU83xf62FgGPvFHYqZObX+SlMoh7rDZLfF8b3JuBUt/+PphZW6QaEbfB4lLl9K0fz+NfbggVsga/WYr1dte495Bg0m1pPGvy/7FY3Mf40jNEZZ9vIw/6+pouPoNqCyEl+Yrcf8wYnAnaBWfbL8heuk+ivplYrVZGZusZOSOSR7DkLghp4d4Jl3t0uyrl00bPXUqxmFDtU+YQeLOpu8u8fygqDoKNUUwbI7v6yaLcjZ2ZD0cWhP0YzSj74P4iy5CxMRQ2YczdEPV6Mvtb/L7OCOncPD0WU+TEJXAxSMu5qNLP2LxiMW8vOtlLtv1P6y/+EnQG+HvF0LBv1T+KVrwbpvoKC9vK9dsrIGaIvbGJgIwpp/i6QshWDhsIZtPbqa6yXXOEz8Qhs9zafbVqdckhCBh8RLq8/I8B84a/lOfl4dx0KBuFc8PmCO5yteMdow+wNSblJLfIXj7mtH3gT7WQsJFF1Lz709x1NVFejsRISSNvtPJym1/YY0lhrun3s2E/hM8l5LMSTwy5xFeOe8VTHoTt259gvsmzqes/yh4+zrY+JeQPrq2hz4pCWE0Yi8pxl5e3lau6aq5U2AAgzAwMrElprpo2CLs0t4S4gHlQLf6WEsijQrEL74IpKTmX/9Wbc2+QHeO5wdEYa4Su/eu+9QaQ5SSsHVim6IiCwLN6LdD4tKlyIYGaj4Jn/fZnQlFo79r+9/5U5SdcxJGcf24G3zOmZ42ndVLVnNb9m18fSKXi2ObWTVyFs7Pfwv/uhsc9pD23xqh02FIS6P5+HEclZU+5JpKt6y9thqGJw7HpG9RLI3rN45BsYNOD/GMuQii4lU90I3KzMQ8fjw1n3yi2pp9gZZ4fg83+kfWK6GdzrpqTVym9Hv45tGgFHCa0W8H84QJRI0e3Wc1+8Fq9Guaa7hn53MMcMIjC/+K8HUg5cKkN/HzST/nvYvfY1y/LB6xF3Hj6Gz2//APeHOpavXr3RhTU2kq2AdStvX0T+0Fg5mC2iOeQ1w37hDPxpMbqWl26fON0ZB1Gez5UFUFUvzii2jcs4emTjowabQQqfr5qlJ9XDnf8tbnt4feAOf8Dsr2wY53An5U9zb69shVHxRCkHjVUhp376Zh9+6I7SNSBKPRl1Ly0Dd3UyJtPD34QhIs/f26LyMhg5cWvcTjcx/nqJAsGzyIZyu30fDyIuVwSyUM6ek0FxYq37eO6ZfuozRlBOWN5Z5DXG8WDluI3Wnnu2PftQxmX6dUQdyjXr2m+AsvBCE0bz8A6vPzMQ4ahGlwT47nu/p0t3eI25pxl0B6Nqz5Y8B2snsb/VN7oShy/dMTlixBmM190tu3HT+OcfDggO55s+BNvirZzK+q6pg4576A7hVCsGTEEj669COWnHEJryTEcZm5jvWvLYQidcpiGNPSPOcFbRqilxawN0n5VNPa0weYkDKBNEsaXxR6lZEYMgOSR8B29SpvGgcMICZnJtUff9LnJcP+0Gvi+UfWK8lXaRM6nwuKpPPc30P1Ufj+9YAeFbLRF0LohRDbhBCfuF5nCiE2CyEOCCHeEUKYXONRrtcHXdczOl1cb4BP71NNIREo+vh44s87j5qPP+lzSTOBavR3l+/mT1v+xLyGJm7MuAgsKUE9N9GcyMNzHlYOehOGcGuiifs+uprGQ991fnMneKfAn1ZLv6kWqo9REB0D+Db6QggWDF1A7olc6prr3IOKZv/IeqjouDdqICQsXoLt2LE+LRn2l6aDB3FUVobP6G9/C758MDxre1OYC0NzQNdJMTZvRpyrfDL47ilotvp9mxqe/i+BvV6vnwSelVKOBCqBn7jGfwJUSinPAJ51zeuYuIFwfCvsiFx3ocRlV+G0WvtU0kyLRt+/eH5tcy33fHsP/XQmHj1Vipj1i5D3MD1tOqsv+4hfjLuJz2LM3P/l/8NRvCukNY3p6Z7vTzP6rpo7BdgYEjeEWFOsz/vPyzgPm9PGd0Vef4AmXQ0IVTX7cYsWIkwmqj/WQjydUZ+XD4Qpnm9rVPrU5v6353ckLNSWQPkB/+L53ggB5z4I1lOw+f/8vi0koy+EGAxcBPzN9VoA84HVrimvAZe6vr/E9RrX9XNFR6d8ADHJMHg6fLVCtQJXgRI9eTKmESOoWtV3Qjz2khKXRr9zT19KyUMbHqLYWszT5TUkDJsLqVmq7MOkN3Hr9F/zmwn/j6/NRh77YBmyMvgYv6fCocGALj6+5YJbudNU6tPLdzOx/0QGRA84XcWTMBiGn62qZl8fF0fsvHnUfPop0q6uiqm3UZ+Xh3HgwPDE83e+C/WuCqtbXlF/fTdHXfH8jLkdz/PF0BwYeR7k/lnp7+wHoXr6fwbuA9y/7f2AKiml+ze1CHD/3xgEHANwXa92zT8NIcQtQogtQogtpaWlcMGTStW5tU+HuNXgEEKQuPRKGn74gcb9Yfxr340IRK75zr53+PLIl9w58ByyK45DTuhefmuum3o7P8m8mHejdfzfu5dAfUVQ6xhcRt+QlITwlsWV7qXWYKaovsTnIa4bndCxYNgC1h9fT73NK9yXfR1UHWl586pA/JLFOMrLsW7cpNqavY2wxvOlhE0vwIAsGH+lIs0NIIQSEIW5YLRA+qTg7j/394rSbcP/+DU9aKMvhFgMnJJSep+y+fLcpR/XWgakfFFKOU1KOa1///5Kgavs65X/ASr1Jw2UhEsuQRiNVL27uvPJvQDbcaXEb2ee/p7yPTyV/xRnDjqT5YU7lKYPI88Ly55+eeajXJyaw1+i7Kx+a3FQb0B9UhLCZPJRXXMfBf0zAd/xfG8WDltIk6OJtUVejVTGLAZTnKqa/dizzkIXF6epeDogrPH8w9/Bqd1KvZsZP4OmatgZpvf/kVwYOlPJTA+GtAkw/grFRtad6nR6KJ7+HOBiIUQh8DZKWOfPQKIQwuCaMxhwFwkvAoYAuK4nAP65bOc+CAYzfP7bELYbPIakJOIWLaL6ww9xNjZGZA9die34cRACg1cMvDV1zXXc8909JJuTeWz4UnRF+TDz550nlgSJEIIVi55nbsIoHtHXsOadK8FhC3gNQ1pa24bopwooiFfGxvbrIBsSmDxgMinRKad31DLFQNalsPsDaFIng1sXFUXceYuo/fJLnA0NqqzZ26jPd8Xzw2H0N70AMSkwYSkMmQmp4yH/JfWzxa3lcGpP4PH81pzzO0W6ue5PnU4N+h0qpXxASjlYSpkBXA18I6W8DlgDXOmathxwi5g/cr3Gdf0b6a8mLS4Vzr4PDnwOB77sfH4YSFy6FGdNDbVfRKbzU1diO368Q42+lJI/bPwDJ+pO8PTZT5P0/RtKdurk68K6L6POyJ8u+gfjYtK4136U7e/fHPCbMPW+e+l3yy0tA011UH2UApOB/tH9SYnuWHWk1+k5d+i5vkM8NmvQqfG+SFi8GGd9PXVrgi+u1Zupz8sPTzy//EelhPH0nyiNc4RQvi/eqb6E/OhG5Wvr+vmB0m8ETL5eOXvoJLclHG7Zb4C7hRAHUWL2L7vGXwb6ucbvBu4PaNWZP4d+Z8Bn94O9Wc39+kXMzBkYhw3tEwe6nck1393/Lp8VfsYdk+9gsjlVaSgy+QaIigv73mKMMfzl4lWkmRK4reZ7fvzs1wHdH7dgAZaZXp5h2T4A9jrqOw3tuFk0bBEN9gbWH/equzM0RwlvbV8Z0H46Imb6dAwDBlDdR0uBdISUkvq8vPB5+XoTTPtJy9iEq5QQXv7f1H3WkVwlijFoSuhrnX0fIODbjoWRqhh9KeW3UsrFru8PSSlnSCnPkFIulVI2ucYbXa/PcF0PLM/cYILzn4Dyg7D5r2psOyCEECQtXUr9li29PkW+I6NfUFHAk3lPMnfQXG4ef7PyJpBOmHmLz/nhINmczF+XvIPJEM3PT3xK8boQDvlL99Ek4FDjKb+N/pTUKSSbk09X8bg1+4XroPJI8PvxQuj1xF94IXXr1uGo8k+Z0Vdodsfz1ZZqNlQqf7jHX6lEGNxExUL2NbD7n0pIRi0K1ysKRUNU53M7I2EwTP8p/NDx2VL3zshtzciFykHhd08p2tYuJuHSS8Fg6NUHuh1p9K02K/d8dw+JUYk8NvcxdLZG2PJ3GH0hJGV06T4Hxw/hhQtfp9Zg4taCv1G9LcjG4qUFHDTH4JDOTuP5bgw6A+cOPZfvir6j0e51xjPpauVrEPVQ2iN+yWKw2aj5vPeHFQPB6q63M1NlT//715XSGjm3tr027SfgaIZgf9da01ClhIyCkWq2x5l3gzGmwyk9y+gDnP9HsDcq9aS7GENKCnHz51P9wQc4m7s+xNQV2EtKwOHw6ek/nf80x2qP8dTZT5FsToadq6ChwvcbpAsYk5LFf89/jkKjiTvzHqFx/+eBL3KqgL1Jys/qr6cPioqnwd5A7onclsHEoZB5luIpqnTgZx43DlNmJjUfa81VvKnPy8cwMF3d+vkOO2x+ETLOhPSJba8PGKNc2/KKOv2dj24CZOiHuN5YUmDWbR1O6XlGv98Ixchsf0O1miyBkLh0KY7KSuq+/rrLn90VtKfRd0onXx75ksXDFzM1dapLx/xXRS7mb5GoMDBzyJn8cdYKtkVFcf83d+I4HuDvRGkBBZZ44oxxDI71v9bQtLRpJEYlnl6LB2DStUq1RPcBXYgIIYhfspj6LVuwnTypypo9HXc83zJ9RodVXANm70dK56qOck2msrRSLAAAIABJREFU/0TJyTiowvv/SK5ydjDYd4jKKYNM9pt9R4eXe57RB6VlWGxqROryWObMxjhwYK/tqtWeRn9/5X5qmmuYNXCWMnDoWyjdCzNv9d3Psws5f/QV/GbSrXwdbeLxj65H+pvP0WyFqiPs1UtGJ48OyIAYdUbmD53Pd0Xf0eTwqnI47mJFybT2GdW8/YSLLgKg5l/agS54xfPVPsTd9IJyGD+qg1yTMYsV26PGge6RXCUPyRjddisnNzFz5UwOVwdR06kTQUXPNPrmeFiwAo5vUTV+6g9CpyNx6ZXUb9xE81H1yv52F9rT6OcXK5roaanTlIFNLyhNmsdf0dVb9Ml1k3/Bf4y4jFUxBv5v9eX+nfmU7ccBHLBVBxTacbNo2CKsNisbT3h59SYLnPNb+PFr5dBPBUzDhmGeNFFT8biwuvX5asbzi7ZAUZ4SRfAqema1WTlV75XwpDcqLQsPfKF8oguWplo4sb3dT8n/2PMPGh2NfHpY/ZpfPdPoA0y8Wvkr+dVDqjax8IeEyy8HnY6q1e916XO7gvY0+vnF+QyNG0qaJU3RMR/4XDnYMpojtNO2/GrOH7g4bQ5/iZZ8/sWvOr/hVAFHjAYanDa/D3G9mZE+g3hT/OkqHoAZtyi1zj+93+96KJ2RsHgJTQUFNB04oMp6PZmwxPM3Pa98Qsu+1jNU1lDG1Z9czaUfXMqxmmMtc6csB6ELrR7Psc0gHT7j+SfqTrCuaB0C0fZ3SwV6rtHX6eCCp1x1eZ7p0kcbU1OJPftsqv75T6QtsKzQ7o4vuabD6WBLyRamp7lij5v/6tIx/0cEdtg+QghWLPxfhooo3ivf3nnGbmkBe6OUj9bBePpGnZFzhpzDmqNraHZ4Hezr9LDkz0qxLpUEB/EXnA96fZ/39lvi+dPVi+dXFynZ1FNu9IRGKhsr+dkXP6OkvgQE3P3d3S1KrYRBMOZC+P4fSiXOYDiyAYReyfZtxer9qxFCcFPWTRysOsihKnUl4j3X6AMMnqYcnG38i+J9diGJVy3FUVZG7bffdulzw40vo7+/cj+1zbVMS5umeK7bViphHW8dczfBqDMyP20m+SY9dZ2peUoLKEgYgElnIjMhM6jnLcpYRK2tlk0nWxVGGzgZZvw/xRs8lh/U2t4YUlKwzJpFzSd9u7lK848/4qioUDeen/cSIJVPaEB1UzW3fHkLx2qP8dz853jizCcoqCjgibwnWu6Z/lNFubbng+CeWZir/I5EnV7G2+aw8c8D/+SsQWdx/bjrAVT39nu20QdY8FBE6vLEnnkmhtTUXtVVy6PRb5XWflo8f9s/lHIDM38eiS36xdlZ12MXgg27Xut4YmkBe81mRiaNxKg7vdjVd/tL+e37O3E4OzawOek5xBnjfL8x5/8O4tLhk18FXCfIF/GLL8J2/DgN27aHvFZPxaPPV8voN1th66swdgkkDaOuuY5bv7qVH6t+5M/n/JkZ6TM4a/BZ/GzCz3jvwHt8cNBl5DPPhn4jgzvQba5X+oRktI3nf3PsG8oby7lq9FUMiBnA5AGTT6/zpAI93+jHpcHZ9yq1Mg581WWPFQYDiVdcjnXdemwnTnR+Qw/ArdFvLdfML3HF880pio552BwYmB2hXXZOdvp04oWB78p2tN8/tLkeWXmEApp8hnaeX3OQNzcf5cW1HX+0NulNzBsyj2+OfoPN2cqwR8UppcFLdikH3yESt2AhIiqKmk/6rma/Pi8fQ3p6wK082+WHt6CxCnJ+Qb2tntu+vo295Xv509l/Yu6glqSpX2T/ghlpM3h006Psq9jXUo+nKF85kA2Eonxw2nwe4q7at4pBsYOYPVCJ9S8ctpD9lfsprC4M5ac8jZ5v9EGRDSaP6PK6PIlXKMqV3nKg69boe4d3HE4HW0u2KvH8ff9WenJ2Yy8flIzZuf0msi5Kj+NAO15S2X6K9Tqqnc1tauiX1TWRX1iBxaTn2S/3s6+4Y6HAwmELqWmuIe9kXtuLY5fAqAvg2z+G3ORdH2shdv451Hz6Wa87S/IHKSX1+flYZqgUz3c6lVyTgVNoTJ/End/cyfbS7Txx1hOcM/Sc06YadAaePOtJ4k3x3P3t3dQ218Kka8AQDVtebucB7XBkg3IQPDTntOFD1YfIK87jylFXoncpiBYMXQDAV0fVc2h7h9E3mJRM3fIDkPdilz3WOGgQlrlzlQNdhwoZehHGVtTW6J8Wz89/CRKGwpiLIrVFv5k3ZimVej07d7STMl9awN4oRaE0pt/pnv6Xe0pwSvi/G6YRZzZw96rt2Bzt54PMHjQbi9HiO8QjBFz4lPL9v+8NWbufsGQJjspKrBvUa9jSU2g+dAhHebl6oZ2DX0H5AZpn3MJd391NXnEej855lPMyfOv0U6JTeObsZzhed5wHcx9EmhNg4lLY8W5gKq0juUpSoznhtOHV+1dj0Bm49IxLPWPpselMTJnYNgkwBHqH0QcloeKMhfDdk341ElCLxKVXYi8upm7dui57Zrhwa/Q9bQXxiufHjVCKQ01aFljzZhddffg4e/CZ6BF8d2qr74YrpQUURJnRCR2jkkaddumzXcUM6xfDnDP68fjlE9h9oobnvmk/4StKH8XZg/9/e+cdFsXV/fHP3QWWDgICCggqiohdLLEbscQSNRGjSUyPb3peTfWXxPQ3iTHN9GYSY5qaGGvsJXbFDoLYQEEpIr1uub8/ZsEV6VJ1Ps+zz+7O3Jn5MuycuXPuuecMZuPZjRhMZZQ3dG2lxO7Hrrnm1MuOAwagcXG5YernSpOJwtNnyFyxkpSPPgJq0Z+/+wv0Ti14Lm0n2xO3M/um2YxrO67CTXp49WBGzxlsOLuBBccWQK+HwZCvuImqgqFQce+USqVcYChg2cllhLUKuyq993D/4URfiuZc9jlqg+vH6IPS29fnw8bX6+2QTkOHovXwuC6SsOkTE7Hy8kJYxOiX+PPPRSjZNIMrvijK4uutpwh6ZQ3Tvt/Dd9tOczIlp85vAi46F3q4BLLF1grKiuJJiSHa0ZUA5wDsrC7PiMzM17Pz1EVGdfJGCMHIEG8mdvfhs80nOZqQWe7xRviPIKMwo+QmeRV9HgWvzsos8muo9yxsbHAeOZLsjRsx5eVVvkETQkpJUXw8matWkfzeHOLvuZfYXr05PXo05597jtxt23EaMaJ2/PnJxzCe3sz/+bdjU8IWXuz9IpPaT6p8O+CejvcQ1iqMj/Z/xAGNAXx7m7PNVuE3nbhfyR1WKj5/bdxasoqymBw0+apNwvzNLp742nHxXF9G36Md9H1ECSlMPFAvhxTW1rhOnEjOli3ok+vvCaMuKB2ueYU/P3qF4trxLiMRVQX8G5vKu2ti6NjCmfMZ+by1KpqwD7cycM5mXv77KBuOJZNXVDfFvwcH3spJGxsSj5bRC0uNIdpac9Ug7sboZPRGyaiQy087r40LwcPRhpmLDlGgL9uN19+nP3ZWduWH12mtlNj97CTY/HaN/yZQonhkfj7ZGzdd034aEiklRefOkbVmDSlz5xJ/3/3E9u7DqZGjOP/Ms6QvXIipoADnW8fR4u23aL3sb4Ii9uE775Na8eebdn/BbE9P1uTGMbPnTO4KrnoBICEEb/R/Ax9HH57d+iwXu01RUr6f2Vr5xnHmBH2ljP6i2EW0dml9eca7Bb5OvoS4h9Ra6Ob1ZfQBBj2vpAf454V6y8vjGj4JjEYyl9bOtPuGQjH6l1MqF/vze7l3htOblV5+NS64c5fyeOr3gwR5OfHrw33Y+MwQtj0/lDcndKKDtxN/HUjkoQURdHt9PXd/V/tPAYP9hgCwNWnvlb1rfT7pmWdJlvqrBnHXRCbRwsWWrr6uJctc7K157/YunEjJ4aMNsWUey9bKlkG+g9h4diPG8jIw+oYqER97vr6mTol9aChW3t5Npn6ulBJ9YiJZa9eR8sGHnH3gQWL73sSp4SNI/O8M0n5agCk7G+fRo/F+8w1a//UnQfsjaL3oD1q8+iqut9+ObVAQwrqGNWRL68lJ5a1z/7DcwZbHuj2m1IWoJk42Tnw45EOyirJ44eJ2DPbuVQvfjN+hFFu3dytZFHMphiOpR5jcfnK5N7Th/sM5evEo53OuPVLQqvImTQxbZyV2f9njSurf4hzndYhNq1bY9+1LxuIluE+fjqijOrF1iTQY0Ccn42zR09+bpESjhOZmKXnEg8dWeX8FeiOPLNyP0ST56u6e2NsoPzU/N3um9fVnWl9/Cg1G9p1JZ8vxFLbEpvLWqmjeWhWNbzM7hgQ1Z0h7T/oFupdsW10CXAIIsPdia14+dx5fffm3cDGWaBvFgFgO4uYWGtgam8rU3q3QaK68+IYEeTK1dyu++fc0Izp60dPfjdKM8B/B2ri1HEg5cHn2cmmGzVaemlb+Fx7apDwBVBOh0eA8ZjSXflqAIT0dq2bNqr2PukJKiSEpiYKoKPIjIymIjKIgKgpjerrSwMoKXft2OI8YgW1ICLadOqFr367c0px1oW/OusdZ7GTPg20m8kiXmkeiBbkF8XLfl3llxyt80bYHT0WuhsxEZcZuWRj1cG7vVWVFFx1fhE6rq3A8Ybj/cD4+8DEb4jdwT8g9NdYM16PRB2WW7r7vYf2rSqRJPZTxazY5nMSZz5C7cxeOAxou1XBN0SddHaMfkRSBv7M/Xqf+VZ6eypgyXhZSSl5aGknU+Sy+vzeUAA+HMtvprLQMaOfBgHYevAwkpOexNTaVLcdTWXogkYW7z2Kj1dCnjRvjurZkVCdvnG2r19sbHDCSX3N/IvfoIhyKjX7qcWJ0yn4se/pbjqdSaDAxqpN3WbvipTHBbDuRyjOLDrP66YFX3YwG+AzAVmvL2ri15Rt9WxelAtyS+5VIs5sqSONbAS7jxnHp+/lkr1lDs6lTa7SP2kCfnEJBVCQFkZHkR0VREBmFMc1cWUqrRdeuHY43D8WuUyezgW+PRlcLVaJqgJSSjyM+YGF2NHcLV54e8Po1u4omBE7gUMohvj3xJ11tbRh84Cdl0L4szh9SJjZauHZy9bmsOr2KUQGjcNG5lL0d0Mq5FR3cOrA+fr1q9MtEo4HR78N3w5Tq8GGv1fkhHcPC0Lq6krF4cdM0+qVi9Iv9+SNaDYN/5ytpF6oYtbNwz1n+PJDA08PaMSy46qkafJvZc1cff+7q40+RwURE3CW2xKayLiqJ55cc4ZW/IwkL9mJ8t5YMCfLExqryJ6rBfkP46dgCdl3YQ1jeJeWxOiWaGJ2OFvbeV1xoa6KScHewoVfA1b14AEedFXPDuzLlm928908Mr4/vdMV6e2t7BvoOZOPZjczqPask1voqQibCoV8V337H8eX3DCtAFxSErl0gmStW1pvRN6Smlhh2xchHYky9qKzUaNC1bYvjoEHYdgrBLiQEXYcOaGwbT0K+rw5/xfxjPxGelc3zoz6stdw9s/rM4ljaMWahYdH+7/G1dQX/m5SBe8snuXhzTWWLSVkrT60kz5DHHUF3XLHPzHw9iyPOcUcvP5zMHZ3h/sP59OCnJOcm4+VQ8xQo16fRB3NenqlKXp7u05TiK3WIxsYGlwkTuPTLLxjS0rByd6/T49U2pY3+8fTjZOuz6SXsoSgHgm+t0n72x6fzxooohgY15+lh7Wqsx8ZKQ79AD/oFejDrlg4cTsjk74OJrDh8nlVHL+Bqb83ozi2Y2N2Hnq2aXeWOKaabZzecrOzZYpdDWPQK6HkvpB4n2s6BDhaZNQv0RjZFJ3Nrt5Zoy9kXQN827jzQvzXzd5xhRIg3/QOvDq9bH7++YhePEDBmLnzeV4nmmVL9YupCCJzHjCX1448pSkjExrf6N46KMKSlXXbRRB2jIDJSmbFt1m/Ttg2O/fphG6L04G07BKGxr7hMX0MyP3I+Xxz+gvEGG14WzRGBYbW2b51WxwdDPuCO5ZN4ppmJhWtnYQ1g4wh+vaFVP+UmcHoreLQHR09AefL4I/YPgt2C6eRxZQdi7trj/Lw7nqUHE/npgd54OOpKjP6GsxuqNfBcmqbnfK4OYa8p2SDXvlQvh3MNnwR6PZl/1zAJUwNSOka/JD4/5ZSScrb1oEr3kZJdwGO/7KeFix0f39G9XENcXYQQdPNz5bVbQ9j9f8P44f5eDG7fnKUHEgn/ahcD52zm/bUxnEi+euastcaaAb6D2ebgiDFSCavNSz1GvEZe4drZfuIiuUVGRnVqcdU+SvP8qCDaeDjw/JIjZBdcOTN2sO9gHK0dmb5+OjO3zGTLuS1Xp2cApabwkBcgZiXErK7eCTHjPLZ2iqsY0tPJ2badi199TcKTT3Ji6M2c6D+Ac9P/w8VPP6Po9Gnse/fGa9aL+C/8maCIfbRduZKW772H2z3TsO/RvVEb/DVn1vDR/o+4pXkor587iabvY7Ve+MfPyY83B/yPY1awcOzrcPv3yhhSdrLyRPfjGCUYwqKXfzj1MCfSTxAeFH7FU8fJlBx+3XuWfm3dOZWaQ/hXuzh3KY/WLq0JdA285iieGvf0hRB+wALAGzAB30gpPxFCuAF/AAFAHDBZSpkulL/qE2A0kAfcJ6WsMIQhr+gaZ7k6ecOgZ2HDa8rsu1q8u5eFrm1b7EJ7krFoMW4PPFC7pdzqmNIx+hFJEfg7tcLrxCZl4ptVxQNteqOJJ345SGa+nqWP9cbFvnYiLUpjrdUwNMiToUGe5BYaWHcsib8PnufLLaf4fPMpQlo6M6GbD3f1bVXicx/iN4R/4v4h8sJeumacJTbvAtLF84pwzTVRSTjbWnFTm8qf0Gyttcyd3JVJX+7krZXRvDfpchirvbU9v4z5hcXHF7P6zGrWx6/HzdaNMW3GML7teILcgi7v6KYn4MgiZaZu60FXZVysDBtfX+y6dydr5Uo8/jO9StsYMzPNPXhlgLUgMrLkKQ+Ugi32PXqYB1lDsO3YEa1j9XQ1JoqMRXy0/yOC3YJ5O7MQrZ1btYI75m08QUxSFp9M6Y61tuI+8jD/YQzxG8KXx3/llgnL8e5sjvvPT4eze+D8QehyOQ5/0fFFOFg7MKb1lTPc31sTg521lk+ndicuLZf7f9jHpK92suCBPozwH8GXh78kNS+V5vbNq34iLLiWnr4BeEZKGQz0BR4XQnQEXgQ2SinbARvN3wFuAdqZX9OBSjNQnUrNYf72GpQLs6TvY+DWBtbMqpVMh5XRLDycovh48vZeezrd+sQyRr/Ynx/q4Ad5aUqJuEp4Z3UMe+Mu8e5tXQhu4VzXcgFw0FkxsbsvPz3Qmz3/F8ar4zpipRG8vTqat1dFl7Tr79MfrdCw1U4H/75PtLVyMygunKI3mlh/LJmwYK8qjRMA9GjVjEcGt+WPiHNsirmySlcblza80PsFNoRvYN7QefTw7MFvMb8xacUkJi2fxM/HfiYtP02pwjT2Y6Uu65Z3anQOnMeNpfDECQqOH79qnTE7m9zde0j7/nsSZszg5PARxPbpy9kHHiT1ww8pOHYM2y6d8XzuWVr9+CPt9+6h7do1+HwwF/cH7sehd+8mbfABFscu5nzuef7bbgrWx1dD6P1llicsi+WHz/Ph+lhWH03i3X9iqrTNC71ewCRNzI2wqPFh1wyCRsHQWSVu5oyCDNbGrWVsm7HYW19+Stp9Oo31x5J5dEhb3B119PR3Y/Ej/ZASJn+9Cx+bPkgkG8/WvEZvjY2+lPJCcU9dSpkNRAM+wHigOKftT0BxIonxwAKpsBtwFUJU+CztYmfNGyuP8cWWKtY8LQsrHYx8By7G1kteHqeRI9E4Oze5lMuWMfol/vy8HCVtdSVPSMsOJTJ/xxnu6xfAhO6161uuKs2ddNzfvzXLnhjApJ6+/H0wkdxCZdKXi86F7p492OLkCgd/IUZng6u1E172ymDY7tNpZObry43aKY+nw9rRwduJF/48Snru1Yn+rDXWDG01lI+GfsTm8M3M6j0LK40Vc/bNIWxxGE9ufJINMpuiHvcoWTgvHK723+08SimukrF4Cbl795I2/wcSn3mWUyNHEdurN2fvu4+U9+dScOQoth070nzmTFrN/572u3cRuH4dvh99hPuDD+LQtw9a5/q5WdcXefo8vjnyDb28e3HTmb1KIEKvh6u0bWxyNi8sOUKofzPu7tuK77ef4Z+jlRem93Xy5aHOD7E2bu2VZTRLsezUMopMRVfMwDWZJP9bHU0LF1seHHC5vkOQtxN/PtqPZvbWvPBbMl62ra7JxVMrPn0hRADQHdgDeEkpL4ByYwA8zc18AMvkEQnmZaX3NV0IESGEiLAz5jK+W0vmrDnOh+tjaz5pp/1IxXBteRdyUmu2jyqisbXFZdw4steuxVAcm9zIKY7RL+7pl/jz4w9C25srdDtEX8jihT+P0DvAjZfGVL/kYF0wtbcfuUVGVllcpEP8hnBCKzmvhWgbGzq4dyxxv62JTMLeRsug9tV7XNZZaflgclfSc4t4dXlUhW1dbV25M/hOfh/7O0tvXcq0jtOISotixpYZ3Jx7gLc9vYha+TiUN7GrHKzc3HAY0J/0hQs5e8+9pMyZQ97BA+jat6f5f/+L33ff0W7XTgI3bsD3k4/xmP4wDv36oXV1rXznTZyfj/3MpYJLPB3yEOLgQgi5DZwrH7PJKtDzyM/7cbS14ou7ejB7bAjd/Fx5bskRzlwsI49TKe7vdD9+Tn78b8//rqyoZkZKyeLYxXT37H5F3qcVR85zJCGTZ0cEYWt9ZeSXn5s9ix/pR5vmjiQkBLIvKUJ5WqwB12z0hRCOwJ/Af6WUFSUVKcvBfZUVl1J+I6UMlVKGNm/enA8nd2NyqC/zNp7g3X9iamb4hVB6+/q8esnL4zp5MlKvJ2v58jo/Vm1QOkY/IikCf3svvDISKsy1k5mv55GF+3G2teazuyr3edYXPVo1o21zB/7Yd7mPMchXGYjeaG/PSRsbgt07AmA0SdZGJTM0yPOqC60qhLR04elh7Vh++Dyrq9ATBAhsFsjM0Jmsm7SOL8O+5KaW/fjLQccUm0z+WvtktTV4Pf88zWfOxO+br2m3YzvtNm3C99N5eDzyHxwH9G9Uk7fqi4yCDH6M+pGhfkPpGrMeirKVoueVIKXk2UWHib+Ux+d39sDT2RYbKw2f39UDK63g0YX7ya9krFGn1TGr9yzisuKUpGyl2JO0h/iseMLbh5csK9AbmbPmOB1bODOxnKfl5k46fpvelyCn/khMvL2lZt6Ea7pKhRDWKAb/FyllcQ6C5GK3jfm9OCFNAuBnsbkvUOmcYq1G8O5tXZjW15+v/z3Na8ujMFVSzahMmrdX8sAfXKhM2rqGpFeVYRvUHtuuXUhfvLhJlLYrCdf09b3sz8deqeHZflSZ25hMkhl/HOJ8Rj5f3t0DT6fGE48thGBKr1bsj08viehp7dIaf2d/fnbzQC8u18TdH5/OxZzCart2LHl0SFu6+rrw8t+RpGaXU7SlDKw0VgzwGcD7g99n8x1b6SEc+PjCFrLi/q3W8XVt2+Ix/WEcBw1qcqHCdcX8yPnk6nN50i0Utn8I3e4Gnx6Vbvfl1lOsO5bM/40Opnfry/M1fFzt+PiObhxPzuaVZZGVXtcDfQcyrNUwvjnyDRdyruwMLDq+CBedCyMCRpQsW7ArjsSMfF4aE1xh1JuzrTW/33cbtnjyz+l1fFQDD0iNjb45Gud7IFpK+aHFquXAvebP9wLLLJbfIxT6ApnFbqBKRWoEb4wP4eGBrflpV3yVytiVyZBZSuz+jo/h055wYEG1H6erSrPwcIpOnmoSpe0sY/Rj0mPI1mfTOy0BAgZckSPEknmbTrApJoVXxnYsMyVBQzOxhw/WWnFFb3+w72AuCOX/XZx+YU1kEjZWGoZ28CxzP1XBSqvhg8ldySk08NLSozW60TvrXHgx7BMyNBq+Wfs4FJSf0VOlYpJzk/k15lfG+g6l3drZ4NVJmaxZCdtPXGTu2uOM7dKCB/oHXLV+SJAnTw4NZMn+BBZFVJ7m+PlezyOl5P2Iy8dOzUtl89nNTGg7AZ1WmZmcnlvEp5tOMjSo+VXzPsrCzsaKqSFjsXY8xSdbDjF7WfU6wtfS0+8PTANuFkIcMr9GA+8Cw4UQJ4Dh5u8Aq4HTwEngW6Ba88+FEPzf6GCeujmQ3/ed49nFhzFUUNiiTHSOMPFLeHgTuLWG5U/CN4OVPPG1jPMtt6BxcCBj0aJa33dto09MBI0Gay8vIpIiAAhNjSvXtbMpJpmPN5zgth4+TOvrX49Kq46Ho47hHb3462AihQbF0A/2HQyAnZUd/k7+SClZG5XEoHYeOOqubZ5ioKcTz40IYt2xZJYeTKx8gzIIbtmHiS0H8otOEr/sP9dccOVG5asjX2GURh47fRiMBpi8AGwqnkeQmJHPU78fJNDTkfdu71JuuPXTYe0ZEOjBK8uiiDpf8Y25pWNLpneZzvr49exIVLJr/nXiLwzSQHjQZdfOvE0nyC00MGt01cfERrUeicREWI9Uft4dz9N/HKLIUDV7eC3RO9ullEJK2UVK2c38Wi2lTJNSDpNStjO/XzK3l1LKx6WUbaWUnaWUEdU9phCCmSOCeG5kEEsPJvLU7wer/IdegU9PeGAtTJqvVLz5cQz8cTdcusbwUAs0Dg44jx1L1po1GLPqzpVUG+gTEkpi9Pcl7SPA2hlPo7HMCllxF3P57++H6NjCmf9N7Nyo5yLc0asVl3KL2HBM8TB29+qOk7UT7Zu1R6vRciQhk8SM/CpNyKoKDwxoTa+AZry6PIrzGfk12seTA9/ERmPN3LR9ypOoSrWIz4pn6YmlhOt88E08CBM+r3Q2fqHByGML91NkMPHV3T1xqKADoNUIPpnSDTd7Gx775QCZ+RWHgd8bci/+zv68s/cd8g35LDmxhL4t+uLvrHSW4i7msnB3PHf08qO9V9VzhAW7BePj6IO1SySzbunAisPneWhBRJXSlDeOkbdq8vjQQF7qbqfZAAAgAElEQVQeE8zqo0k89sv+cnOcV4gQSj6ZJ/bB0Jfh5Eb4vLcykauW/P2u4eHIggIyG3kK3OJwzRJ/fkEh+ISCc8sr2uUVGXhk4X6EEHw9rWeNBj7rkwGBHvi42vH7PqU2rbXGmtf6vcaT3ZXB0jVRSVhpBGHBNXftWKLVCOaGd8VokoR/tYtD56pRQs+Mh50HD3d7jC0O9uza9DKkRFe+kUoJnx/8HBuhYXrMDmWOTsfxlW7z+opjHE7IZG54V9o0r3xegrujjs/u7E5iej7PLT5coTvPRmvDrN6ziM+K58mNT5KUm3RFmOactTFYazXMCGtf7j7KQgjBCP8R7L6wmyl9PZhzexe2n0jlzm/3lBk+bEmTNPoADw1sw5sTOrEhOoWHF0RUOqJeLtZ2MPg5ePIAdJoE2z+qNX+/XacQdB2DyVjUuAd0i84nYmP25+foc+iVllima2f2siiOJ2czb2p3/Nwa77T7YrQaQXioL9tPXuTcJaXK1IiAEfRp0QcpJWsik7iprTuu9rWX1tff3YHfpysFr8O/2snPu+Kq/b+fFnIPPvbezHFzxrD4Pii6vipk1RUxl2L4J+4f7s7MwqNlKIRVHqm3KOIcv+45y6ND2lZrMD80wI0Xb+nAumPJfLvtdIVt+/v0Z7j/cPYk7aG5XXOGmOs87I9PZ/XRJKYPaoOnc/UDIYb7D8dgMrA1YSuTe/nx5d09OXYhi8lflz8/AJqw0QeY1tef9yd1YcfJi9z3w15yCq+hApNzizL8/UOu2d/fbPJkCmNiKIiMvKb91BVSr8eQpMTol/jzCwqvMvq7T6exZH8Cjw5uy+BqxrM3JOGhSsDY4v0JVyw/npzNmYu51xS1Ux5dfF1Z9dSAEt/vjD8OVas6mE6r45nez3PSSsNfBQmw5sXKN1Jh3v6PcJZwX54Rwn+oNHVIZGImL/8dSf9Ad54ZXr2eNsCDA1ozKsSb99YcZ++ZSxW2fb7X8zhZOzGlwxSsNdZIKXl71TE8nXRMH9Sm2scG6OTRCW8Hb9bHKRO1RoZ489P9vbmQWVDhdk3a6INyUX88pTsR8enc8/2eSn1slXKFvz/d7O+fVmN/v/PYsQg7OzIWNc4ZuvrkZDCZsPbxUfz50gpP96Ar/KB6o4lXl0Xh42rHkzfXPHNmQ+Djasegds1ZEnHuioivNZFJCAHDO9Y8RW1FuNrb8P29vXh2RHuWHz7P+M92cDIlp8rbh7UKo6dXTz7z9Cb74AKI/LNOdF4v7E+KYNv5nTyYnoHzbd+CS8V1dNNzi3hk4X7cHWyYN6U7VjWYYyKEYE54F/ya2fHErwcqDNf1dvBmffh6Hur8EAD/RCZx4GwGz4xoX+MiQUIIhvsPZ8f5HeQUKb+tm9q6lzxplkeTN/oAt3Ztyed39uBoYiZ3fbe7Up9WpVzl799QY3+/1tER51tuIWvVKow5lc/mq2/0CUqkibZFC/YnRRCak3FVrp2fdsZxPDmbV8d1xM6mcfvxy2JKLz/OZxaw7cTl2dhrIpPo5e9Wp/MLNBrBEze34+cH+3Apt4jxn21n5ZGqlbsTQvB8r+fJMOn52i8IVvy3VgMNrieklHyy7WWaGwxM7TK90rQhRpPk6T8OkZJVyJd398TdseZFXZxtrfny7p5k5ut56reDFYaSO1g7oBEaigwm3lsTQ5CXE5N6+pXbviqM8B+B3qRna8Ll+rydfMovxgLXidEHGNXJm2+mhRKbnMPUb3dXa5JMudSSv7/Z5HBMeXlkrb62FLh1QXGM/lnHAnIMufTKL7jCtZOSVcDHG04wJKh5nfWK65phwV64O9iUxOyfuZhLTFJ2nbh2yqJ/oAernhpIhxbOPPHrQV5bHlWlqLOO7h2ZEDiBX6yKiLfSwpIHwHCNHZrrkG1HF3AwL5FHrLyxK69qlQWfbDzBv7GpvHprR7r5XXs6iuAWzrw1oRO7TqfxwbqrE9+VZuHueOLT8pg1ukOFtRuqQpfmXfC082Rd3Loqb9OojX6uPpfsoqtzpJfH0A6e/HBfL+LT8pjyzS6SKvFtVZlr9Pfbdu2Krl07MhYvqR09tYDU68n48y8ufvklQqcjwqj0IkNtPMC7c0m7/62Opshg4rVxIY06PLMibKw03N7Tl/XHkrmYU8iayCQARtaT0QfwdrHl9+l9eXBAa37cGccd3+yqUljnUz2ewkZrwwfte8P5A7DpjXpQ23Qw5afzScRc/IySibf9Vml1t43RyczbeIJJPX25s3erWtMRHurHHaF+fLHlFFO/2c3++LLzbmXm65m36QQDAj1qZWxMIzSE+YexPXE7ufqqeRIatdGPy4qj32/9GP3XaGZumcm3R75le+J2LuZfLHeb/oEe/PRAb5KzCpn89S4S0msx8qGG/n4hBK7h4RQcPUpBdMOG4JmKikj//XdOjRzFhZdeQuvsjN+XX7A3dS8BegOeHW4tKTCx+3Qafx86zyOD25Rb57apMDnUD4NJ8teBBNZEXqCrrws+rlVLsVtbWGs1vDK2I5/f2YPYpGzGfrr9CpdTWXjYefBwl4fZnBHN7q4TYOencOLaimhcN0jJP3/dRawWnuj8MNZOFd/E49NymfHHIUJaKj3z2u7EvDWxE6+O68iJlGxu/3InD/y476oJXF9sPklmvp5ZozvU2vGH+w+nyFTEtoRtVWovGnMoYYcuHeQzC58h+lI00WnRJORcjsBobtecYPdgOrh1INgtmGD3YFo6tCw5kYfOZXDP93twsrXml4f61L7R0ufDzs+UvB4mA9z0OAyYCbZlp6c1ZmRwYtBgXCfdjvfs2bWrpQqY8vPJWLyYtO++x5CSgl3Xrng89igOgwZhlEYG/tKXURkXefXW38D/JvRGE2PmbSO30MiGmYObpC+/NOFf7eTspTySswp5YVQHHh1StyU0K+JUag6PLtzPiZQc/jusPU/eHFhuzpVCYyHj/x6PvZUdi88no81Ogkd2VClj5PWMfuc8xh/7AnsHbxZN2YRGlN+HzS8yctuXOzmfkc/KJwfUachxXpGBH3fG8fXW02Tm6xnTuQUzhrdHZ6Vh2AdbubVbS+aGd6214xlNRoYtHkYPrx58OETJiCOE2C+lDC2rfaM2+qGhoTIi4vLE3eyibGIuxRCdFq28X4rmTOYZjFLxrzvbOJfcBDq4d8BK78uLfyRho7Xi14f7EOhZ9RlvVSbrPGx8Aw7/Bg6eMOwV6HZXmY+Zic8/T87mLbT7dysau/rpZerPnydz+Qou/fwzxrQ07Hv1wuOxR7Hv27fkBhl1MYopq6YwJ7OIW544Bhot3207zVurovlmWk9GhNSfG6QuWbI/gWcXKznrNz87hNYN/PSSV2Tg5aWR/HUwkcHtm/PxHd1o5lB2mOG6uHU8s/UZXuk0nclr/qc8dd6zrMrF6q87zu5m0ZJw3nR35fObP2eQX/nlPKWUzFx0mL8PJTL/vl4MDaqdyXiVkZmv57ttp5m//Qz5eiM+zexIzS5k87NDaOFSu9f/W7vfYvmp5WyZvAV7a/vrx+iXRYGhgBPpJ5SngUvRxKTFEJseS5FJGfDSaW0pyvNCFPnwcJ+BhLUNJahZUO37pxP3K9W5zu0B7y4w6l0I6H9Fk7x9+4ifdg8t3nkH14kTytnRtWPMyCBrzVqyVq4kz3z+HPr1U4x96NW/gx+PfMsHB+exybU/zcd/RXJWAcM+2EqvgGbMv69Xk/XllyavyECftzfi08yONf+tvOZvfSCl5Ne9Z3l9+TGaO+n4ZEo3QgOuTmAnpeT+tfdzOuM0qwLvxWnlTBj6Egx+vgFUNzC5F8n/aiBj3Kzw8+zCj6MXlvsbzSsy8M7qGH7eHc+MsPY8HVb/IcdpOYV8ueUUC3bH8/iQwDrRsOfCHh5a9xAfDvmQ4f7Dr2+jXxZ6k54zmWdKngoOJkcSdTEGNMrArreDN2GtwhgZMJIuzbtU+FhYLaRU4qnXv6qUwOs4Hoa/oRTBRrlwT48eg7ZZMwJ+/aV2jmnGlJ9PzpYtZK5YSc62baDXY9OmDS7jxuI8Zgw2rcoftHp82WTOphxhxdDPIDCMp38/yD9Hk1g3Y1CT9+WXZvPxFJxtrenp37hyzB9NyOSxX/eTmJ7Pfwa3ZUZY+6tKNx5LO8aUlVO4t+M9PHP2OEQugftWgX+/BlLdAJiMsPB2vr90iI+bOfHTqJ/o4VV2yuR9cZd4bvFh4tLyuL9/AK+M6Vhh2uK6ptBgxEarqZNOlMFk4OZFN9O3RV/mDJ5z4xn9sohPy2HKD6vJMsXS3DuWDCIxSj2edp6E+YcxImAE3Zp3Q1sbj8tl+fsHPgM6J9Lm/0DKnDm0WbEcXbtru+NLg4Hc3XvIWrGC7PXrMeXlYeXpifOYMTiPHYNtx46V/sAMJgMDfw7lltw8Zk+PZPfZbKZ8s5unbg5k5oigCrdVqV1yCg28tfIYv+87R3ALZz66oysdvK8cI5q9YzYrTq9g2S2/0urXO8FQCI9sLzcF9nXH5nfI2vYet7RpR1fvXnwR9sVVTQr0RuauPc73O87g42rH+5O6clPb67/OwOu7Xmf16dVsvWMrdtZ2qtEHuJCZz5w1x9l56iLJOZlYOcbg4BYFdjGY0NNM586IAOUJoIdnj2u/AVzl75+NodUoTgwdhtudU/GaNatGuzWkppL2w49kLl+O8eJFNE5OOI0cgcvYsdj36oXQVl13VMphpvxzN3N0gYSF/8mYedvIKzKyfsb1MXjbFNlwLJkX/zpCVr6BZ0a056GBbUriuVPzUhm7dCx9W/Tlkw73w3dh0G44TPm1JOqqXjGZwFhkfumv/mzSl738qs+Gytvo8yHyT+YF3cS3RQksGbeEILcrOyYHzqbz7KLDnL6Yy919WzHrluAKs2ZeT+w8v5P/rP8Pnwz9hGH+w1Sjb4mUkri0PPacTmP36TR2nblAmukQVs6RWDvGgEaPvdaVAS2HMKnDaHp598JKcw0/nFL+/oQDAeQdPkHgv1vR6Ko+G1CfnELa99+R8ccipMGA08034zxuLI6DB1drP5bM2fA0PyduYlPXF1iW3e+6G7xtqqTlFPJ/S4+yNiqZ3gFufDC5a0nEyXdHv+OTA5/w3Yjv6BN/QMnN0/9paNG1HINZnuEttcxUzvKK9ifrpggRGivQWIPWBrSX3y96tme0KZ4hrYYyZ9CckuYFeiMfbYjl239P08LFjvdu78KAdpUXJLme0Jv0DF00lIE+A3l30Luq0a8IKSVnL+Wx+3QaO06dZ9eFHWRrD2DlFIPQFGGFIx2d+3Fru1FMDB6MjbYGWRkt/P25sSmc3eJBy9eex2XK/ZVuqk9OJu3b78hYtAhpNOIyfjwe0x/Cxj/gmnp3i7bO5s24pUzIK+SJybsYNi/iuhu8bcpIKfnrQKJSIlRKXh0XQnioL0WmIiWE09qexWMWoV10Dxyvwmzvq4xoeZ+tS32uoK3Gugr7K2/fFexTU/Y429u732ZJ7BKWTVhGK2dlnOrwuQyeWXyYkyk5TO3tx/+NDsbJ1ro2/xVNhtk7ZrM+fj2779qtGv3qIKXk3KV8tp08zz+ntxCVsR29LhKhLQSjHZ5WoQxocTOTQoYS4u1WvcEhfT5y+6eceu47rB2M+L8yDXx7Q2E2FGaZ35WXPuUiaZtOkXEgDSklLkFaPDoXYaPLVoq8u7WBkInQcYIyi7aqhjo/nTXLH+L5/OMMNFnz8aj5PPuvkgRq/YxB+LtfX4O3TZ2E9DyeXXyY3acvERbsxTu3deZg2lae2foMs2+aTXjg7ZAarfSOyzOoGquGcf9cI3qjnqyiLLKLsjmfe57HNz7OxMCJzL5pNoUGI59uPMmXW0/R3FHHe5O6NKkMsHXBtoRtPLbxMSLvi1SN/rUgpeT0xQx+i9zIv4kbuaCPAE0B0miLJr8THZwGMLzNQAa09SbIy6lKN4GL8+aS+sX3tB2TjI3TlY/I+nwb0o67khFrhQRcOzniPqglNt7uoHNSXtb2cG4vnPlXecR2a6vcAEImgldI+Rd47Fp2rvkvjztr6WLryVcTlnIo0cDUb9XB28aMySSZv+MMc9Yex0lnxdsTO/HbuReJy4pj5cSVONnUwRyUWkBv0pNTlEN2UTbZRdlkFWWVGHHLZWV9zi7KpsB4ZSoVOys7VkxYQWqGLc8uPkxMUjbhPX15eWxHXOxuzN69JXqjnsF/DGbXXbtUo1+bFBmLWBm7laWx/xCZsQMDeUijDkNOMLrC7oR69qV/2xb0beNOB++ybwL6lBRODr0Z9ykT8HxwEuic0Kfnc3HBH2Qs/RsQuN52Gx7TH8bax6d8MbkXIXoFRC2FuG0gTeDeznwDmACeHZUbQH4GrH2Jo8cW8WBLb/yc/Phh3O/YaR3VwdsmRGxyNjP+OETU+SxGdNezu2A294Xcx8zQmXVyPKPJSI4+p1xDXXp5acOeb6g4v5BWaHG2ccbJxumKV/Gy0usCnAL5c28On28+iZuDDe/e3pmbOzTNRIB1xUvbX+J/A/+nGv26Qm/UszdpL0tjV/Nv4mbyjdlg0qHP7oAhuzP2xhD6BHjTt407fVq7EdzCuSQS49wTT5B/4CABv/1K2vwfyPjrLwBcb78Nj+nTsW7ZsqJDX01OKsQU3wC2KzcAj/bQfiRE/sXpgovc4+eHk31zvhvxE0fiTCzcE8+Ok2l8e09ok82ieaNRZDAxb+MJvthyEtdWSzE57Ge5hY/bEpM0kaPPKdtgF2aRrS+7x138vbIkXhqhwcHKEUdrRxysnXGwdsTe6vLLTqu8bLWO2God0Gkc0GkcsdE4oNPYo5E6TBKMUmI0SQxG87tJYjSZzO+y5H3VkQscu5DFbd19eHVcCC72au++NPuT9xPqHdp4jL4QYhTwCaAFvpNSvlte26Zg9C3Rm/TsS9rH+vj1rI/bQGZRBlp0aAuCybzYEUNOB5x19vRu7U7fNm70TT2O9v+UHpqwtsY1fBLuDz+MxsubIoOJQoORQoOJQr2JfL2B5NxUzmSeJC77FAm5cThY29HBvR1dvdoT5B6Im61FrHZOCkQvh6i/IW47FzzbMc3NngIJfW1ns/GokfQ8PV7OOu7r15pHBrdRB2+bGPvj05mxZCuXmr2Jg9YTR60XepmHXuail3kYyMNIPlDxNS5MdghphzDZgckOTLZIkx3SaIc02GI02mIy2mI06DAZzcuNdkiTLZhsqM+8jZ5OOt6a0EmNLquERjM5SwihBWKB4UACsA+YKqU8Vlb7pmb0LTGYDOxP3s/6+PVsiN9AWkEa1kKHu6YLOZc6ciGpDRqjNbMjFpLj6MqyjsO4YONCocGIXuah1SWj0SWhMb9rdckIq8sZQ00GB4TQI7SX86trTI44iJa42fjhYx9AG9c2dPQIxEsneW7fM6QXXSLnzMNYGXwZHuJFeE9fBrZrfs05vVUajrwiA9P/nseR7L8R0hYh7dFKOzQo71ocsMIeK2GPNQ5YC3tsNA7YaOyxEY7YCDusrbRoNRqsNAKtRpR612ClFWhEqeVai/WltytprymjfanlVx1TU6rt1ds05KzapkJjMvo3Aa9JKUeav88CkFK+U1b7pmz0LTGajBxIOVByA0jNT8VaY0OgY09s9V0xGE3kk0iOTCTLeI48U1rJtjqNPZ62AbSwC8DHoTWtnNrQ2ikQd3s3svKLOJ6WQOylk5zLjiOlIJ5MYyJF4jxoL/tSpdSA1OCd/xTTug1lfLeWtVoMXEVFpXFRkdGv76lqPsA5i+8JQJ961lDvaDVaenn3opd3L17s/SKHUg6xPn496+LXkZKvVK630djQ1rUtvVz7EdgskEDXQNo3a4+XvVeFbpdBeAJX5h6RUnI+O5UDF2KISj3Bmax4RviHcXtI40gypqKi0nDUd08/HBgppXzI/H0a0FtK+aRFm+nAdIBWrVr1jI+Przd99Y1Jmjh+6Ti2Vrb4Ofld26xfFRUVFTONqaefAFhWAvYFrqgULaX8BvgGFPdO/UmrfzRCQ7B7cEPLUFFRuYGo73KJ+4B2QojWQggbYAqwvJ41qKioqNyw1GtPX0ppEEI8AaxFCdmcL6WMqk8NKioqKjcy9e5EllKuBlbX93FVVFRUVOrfvaOioqKi0oCoRl9FRUXlBkI1+ioqKio3EI064ZoQIhVoKoH6HsDFhhZRDVS9dYeqte5Q9VYNfyllmcUFGrXRb0oIISLKmwzRGFH11h2q1rpD1XvtqO4dFRUVlRsI1eirqKio3ECoRr/2+KahBVQTVW/doWqtO1S914jq01dRUVG5gVB7+ioqKio3EKrRV1FRUbmBUI3+dYxQi97WGeq5rTvUc1u3qEa/GgghbhVCtG1oHdVAW/xBvZBqHVcAIUSjr3wjhLhTCNHV/Lkp/A5siz80Eb0IIZqMLW0yQhsSIUSYEGIX8D3QoqH1VIYQYowQYj3woRBiEIBsxCP2QogJQog3G1pHVRBCuAgh1gFrQEkX3sCSysX8u90GfAx0h0b/OxghhNgJfCaEuAsavd5bhRAzG1pHdVGNfjkIBUchxArgZfNrN+BvXt8oz50QIgB4G/gUiAamCyGKy1M2Gs3m86s1a5sLvCiEGNjQuqpAAZAOdDKX/0QIoa14k/rDfF7thBCLUH6zbwFLAHvz+kaj1RIhRHPgDWAO8CtwhxBilnldo/ndgvJ0J4R4AZgHzBVCdJNSmhrruS1NozqZjQmpkAMslFIOkVJuROndjTevNzWowPJpC2yXUi4HfgC+A54UQjQz/zAbxeOy+fwagZMovdDHgEbd2zdf1K4oN/87UG6sSCmNjey85gO/mH+3a4GdwDTzemODCiwD87nzAg5LKf82X2svAs8KITwa0+8WSp7ujgMdgJnA1+blje7cloVq9EshhHhKCPFucS9OSvmHebkWyADOCSF0DanREiHEJCFEH4tFCcDtQgidlLJASrkF5aJ/pUEElsJ8fr8tfvoAtkops6WU3wIOQogHze0a/LdpofUBIYQwX9RZwBgp5UrgiBBithCik5RSNqRhstD6MICUcpl5uRY4A0QJIfwq2kd9IoS4VwgxHEpcODlAPyGEm3nZMWAx5htrQ2NhFyabF60yX18fA55CiDvN7awbTmUVkVKqL8VtKIAZwA5gEopr5D6guUWbfkBMQ2s1a/EEtqIUlv8b0FisWwB8bPF3dUV5xPdqYM33ofSSR5m1zwLaWqy/BYgCmjWC81ta6/+hPEV5Am+Z2zwAGIAI83frRqS1jcX6zij1qZ0awXltZv4tXgCOAFqLdQuAn0u13QO0bkC95dkFT4s2E4HEhj63VX01eG+qsSCV/95Q4GUp5RKUf3RXlAupuM1OIEEIcWvDqLyMlDIFWIai7wLwH4vVbwBjhRAh5r+rAMhG6U01JMOA96SUa4BnUKI07ipeKaX8h8vjEE7FT1sNRFlaw4F84BbzYO5TwCYup/9uqEHd0lptgLuLV0opj6LontIw8i4jpUwH1gHBwH5gtsXqJ4BRQohe5u+5wGGgqF5FWlCBXRhp0WYpECuEeBaUAfSG0FpVbkijX/ox3MKVEAEMBDBfQLFAiBCig7mdMxAD6OtPbYV6PwWOoVxEY4QQLQCklCdRIo2+EEIMQDEAnkCDjENY6D0IjDVrjAB2AS2FEP0tmr8AvAOcALzrUydUqHUn0BoYAKwH9kopu0kpRwBDhBCtzQaiMWjdjcV5Nf9+1gG2DeyCKj72AillBvAFcJsQwh9ASpkFvA68IoS4F2UguhMN1Fmpgl0Ismj+KDBHCJEE+NSr0GpyQxp9wM7yi7w8KHsScBJCdDZ/3wq4AI7mdlmAL8qgU31Spl4ppV4qg0o7UW5GT1u0eQfF8D8IBAEPSmWAr84pjmIovsgtzu8OQCPMYaRAJMpTSktz+0AUQ/A30ENKWef+3GpojQISASdgtpTyZYvdtJJSnmlEWq84r+abkSeQW583pjL0SvN7gfl9H/APSrQZ5mWfoYSY9kSJlJskpcxsIL2V2QUnc/tuwLfAnyi/25/qQ29NuaGMvhCirxDiT+BzocQEF/+TiyfY7AWMwHAhhJVUBpN8AMsiCFOklD82sF5Rqsd2EVgOtBdC+AohPM3ROguA/0gpJ0spk+pB701CiG+BGUII5+KL3OL8nkAxnncIIbRSygSU3nyAeX0m8ISU8jYp5flGpvUcihH1l1IWCSXcVAMgpcxtZFpLn1eAZ6WU8+tSZxX0lpwzCz4DAoUQIUIILyFEoJRyEzBDSnlvXf8OzLr6CyF+Al4WQrhZ6C0elK3MLqQBj0kpw+tD77Vywxh9IcQQlF7kXyjhVncDzYQQGnNvudgtsg8IRAkZAygE4or3U9xLaWC9UkophRA6oUTpGKWU/6Jc+JEoPREPs9568Yeae5mfofi4WwKzhBAjzBqKfd3ZwDYUn/Nc80XVDOWiQUqZKqU80Yi1ulpoNcp6CNutjfNqbtsYfgdGqYRf2gkhip+ezwJLgaMov1vn4rb1pLcNynW2GeXJ4k0hxGizBr35vTy7EG9ef848btIkuGGMPtAF2Cel/AVYCFgDOcUXrhDiLSHE9yiDS/OA3kKI/cAlFH9oY9P7BkoMfgvz90dQBnO/BrrUh/EsRSiwQ0r5G8qEIC9gqhDCy6zvLZRJN5kog3fNUAxVJlDfj8Oq1obT+wbwC9DG/H0qyhyNuUBnKeWBetbbG4g2P70/CxwCxhWPj1ViF9bWs9ZaodHnDakpQoi+wCUpZax50b/Aa0KI8yg/smiUgc61wDmUH+FsKWWcefs7ASvzgFNj1BtoqRfF79jP3CtpCL3Hga5CiJZSyvNCiByUp43xQogtKOf3RSnlKfP2DwAOUspsVWvT1FpDvYHAc8V6UeYQDKmPMRGz3nEoPfoIKeVuFNfNk0KIVlLKs0KIHSjn9A4hRAQNbBfqguuupy+EcBVCrEKJsJhs8Rh5CCW80R/F/y6/PZYAAAa6SURBVDYEZQAsDGWA604p5UkLP21Offxja0Gv1tx+Q30Y/PL0oviVs4AfzeMQfihRJc5Syliz3lMW59dU14ZJ1dqo9Rb/bnfXh8EXQrQQSkqV51Cehn4QQoyUUp5GiSIrDg8+jhIR5wIcbSi7UJdcd0YfcEB57HrS/Lkkn4uUci/QnMtx1ZtQ/LTpoIRo1YefthTXqre+p36X1luc0C0WJUb8HWCxlHIiyvjC0OING+D8qlobr976/t2GoqQnGSSlfBP4BJhuXrcd6CyE6GPWlQgMkuaooQY6v3XGdWH0hRD3CCEGmyMFElHqUi5CmZTURwhRHBKoQwlvfMy86TDAzdyO+vrHXmd6exfrlVIWSSk3Syl/N2/aEyUkj/rSq2pV9ZbSO8R8HW1EmfFbTBpKvD0o8xoOAh+Zn1hCgHghhH196q0vmqzRFwothBCbgXtRZnZ+KZQETQVSyjxgA8qj3DAAKWUhSmijoxDiX2AqSohgiqr3mvTeXGrbAebBroHASlVr09R6nei9E5gP2EspL4jLYZgtzJqRUiZJKT9BMf7zUSLl3jP/bdcfshHkgqjuC3O+DqA9ShZMUAalPwX+KtV2BkoUgStgZ15mh0VuElVvreh1QRlABCVUb7SqtelqvV71WrRZAYSZP3tatG3w/ER1/RLmP7ZJIJTJKG+gVIRajRLTO0lKea95vUBJQDZFSrnVvMwR5cfYH2iFMmMuUdVb63r7oQw695TK5CBVaxPUeiPoFULYAF8Br6GEOY9FiSBKrw+9DU2Tce8IIQajxMo2QwlPfBMlB85QIURvKJnm/QbKP7OYMSg+8UMoccD1ZUBvNL2HzXrrw4iqWlW9NdH7unkzW5RMmRtRUimE3SgGH2g67h0Uv+A0i+9foCQ5ug/Yb16mQZl+vggIMC8bjzISr+q9TvSqWlW916DXF2VC1gKgW33rbQyvBhdQjX+uPaDjsk/uLuAd8+dDwJPmz6HAb6re61evqlXVW0O9vze03sbwajLuHSllnpSyUF6O7x0OpJo/3w8ECyFWAr8BB+DqlMT1iaq37lC11h3Xud790LB6GwNNLg2DUGbySZScHsvNi7NRqgV1As5Isx9cmm/xDYmqt+5QtdYdqt7rlybT07fAhJJ87CLQxXwXfwUwSSm3y3oa+KwGqt66Q9Vad6h6r1ca2r9UkxfQF+WfvB2lOEiDa1L1qlqvJ62q3uv31aTi9IsRQvgC04APpTJrtVGj6q07VK11h6r3+qRJGn0VFRUVlZrRFH36KioqKio1RDX6KioqKjcQqtFXUVFRuYFQjb6KiorKDYRq9FVUVFRuIFSjr9LkEEJIIcTPFt+thBCp5gk5CCHuE0J8Zv78mhAiUQhxSAhxQgjxlxCio8W2W4QQx4UQR4QQMUKIz4QQrhbrc8o4vuU+i1+upduZ2w4RQmQKIQ6aj/OvEGJsGe0OCyF+s/j+uXm/x4QQ+RbHmSSE+FEIccZi2c6ankuVG48ml4ZBRQXIBToJIeyklPko+VYqmnH5kZRyLoAQ4g5gkxCis5SyOEfLXVLKCHOe9XeAZcDgSjSU7LMKbJNSjjUfvxvwtxAiX0q50bwsGKUDNkgI4SClzJVSPm5eFwCslFJ2K96Z+abxnJRySRWPr6JSgtrTV2mq/IOSwx2UMpK/VdC2BCnlH8A6lDJ6pdcVAc8DrYQQXWtJZ+ljHELJ7f6ExeI7gZ/Num6ti+OqqBSjGn2VpsrvwBQhhC3QBdhTjW0PAB3KWiGVbI2Hy1tvwQwL98rmahy7rOPfAfyBcuOaWsV9vG9x/F+qeXyVGxjVvaPSJJFSHjG7PqailMirDpWl1q1K6t3quHfK3b8QoheQKqWMF0IkAPOFEM1k5ZWcVPeOSo1Qe/oqTZnlwFyq6NqxoDsQXdYKc4rezuWtryUsjz8V6CCEiANOodR3vb0Oj61yg6MafZWmzHzgDSnl0apuIIS4HRhBGTcKIYQ1ykDuOSnlkVpTeeUxuqCk/P1cCKEBwoEuUsoAKWUAStnBqrp4VFSqjereUWmySKX49idVaDpDCHE34ABEAjdbRO4A/CKEKEQpu7cBxfAWY292uxTzYal9FjNBShlXzvEHCiEOopT2SwGeklJuFEIMARLllbne/wU6CiFaSCkvVPA3vS+EeNnie2/zQLSKSoWoWTZVVFRUbiBU946KiorKDYTq3lFRqQWEECOB90otPiOlnNgQelRUykN176ioqKjcQKjuHRUVFZUbCNXoq6ioqNxAqEZfRUVF5QZCNfoqKioqNxCq0VdRUVG5gfh//ltmFkPlgmsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "heating_data.sort_values(by=['START_DATE']).set_index('MIDDLE_DATE')[['USAGE','COST','duration','HDD']].plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 220,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>duration</th>\n",
       "      <th>HDD</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>4.361111</td>\n",
       "      <td>89.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0.000556</td>\n",
       "      <td>16.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>16.439444</td>\n",
       "      <td>63.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>15.995278</td>\n",
       "      <td>194.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>40.724444</td>\n",
       "      <td>273.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>58.763611</td>\n",
       "      <td>114.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>80.812778</td>\n",
       "      <td>636.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>188.401111</td>\n",
       "      <td>472.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>159.905833</td>\n",
       "      <td>382.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>220.985833</td>\n",
       "      <td>655.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>10</td>\n",
       "      <td>142.797222</td>\n",
       "      <td>483.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>11</td>\n",
       "      <td>225.894444</td>\n",
       "      <td>533.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>12</td>\n",
       "      <td>32.160833</td>\n",
       "      <td>841.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>13</td>\n",
       "      <td>348.833056</td>\n",
       "      <td>694.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>14</td>\n",
       "      <td>223.241389</td>\n",
       "      <td>688.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>15</td>\n",
       "      <td>108.188889</td>\n",
       "      <td>840.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>16</td>\n",
       "      <td>321.580833</td>\n",
       "      <td>816.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>17</td>\n",
       "      <td>324.622778</td>\n",
       "      <td>1024.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>18</td>\n",
       "      <td>336.861389</td>\n",
       "      <td>1203.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>19</td>\n",
       "      <td>414.843611</td>\n",
       "      <td>1135.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>20</td>\n",
       "      <td>478.366944</td>\n",
       "      <td>1034.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>21</td>\n",
       "      <td>522.378056</td>\n",
       "      <td>1179.9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      duration     HDD\n",
       "0     4.361111    89.9\n",
       "1     0.000556    16.3\n",
       "2    16.439444    63.6\n",
       "3    15.995278   194.2\n",
       "4    40.724444   273.4\n",
       "5    58.763611   114.3\n",
       "6    80.812778   636.4\n",
       "7   188.401111   472.6\n",
       "8   159.905833   382.9\n",
       "9   220.985833   655.1\n",
       "10  142.797222   483.3\n",
       "11  225.894444   533.2\n",
       "12   32.160833   841.0\n",
       "13  348.833056   694.3\n",
       "14  223.241389   688.9\n",
       "15  108.188889   840.5\n",
       "16  321.580833   816.3\n",
       "17  324.622778  1024.4\n",
       "18  336.861389  1203.3\n",
       "19  414.843611  1135.7\n",
       "20  478.366944  1034.2\n",
       "21  522.378056  1179.9"
      ]
     },
     "execution_count": 220,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "heating_data[['duration','HDD']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}