{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "GHCN V2 Temperatures ANOM (C)  CR 1200KM 1880-present\n",
    "\n",
    "GLOBAL Temperature Anomalies in .01 C     base period: 1951-1980\n",
    "\n",
    "http://climatecode.org/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import git\n",
    "\n",
    "\n",
    "if not os.path.exists('ccc-gistemp'):\n",
    "    git.Git().clone('https://github.com/ClimateCodeFoundation/ccc-gistemp.git')\n",
    "\n",
    "if not os.path.exists('madqc'):\n",
    "    git.Git().clone('https://github.com/ClimateCodeFoundation/madqc.git')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It seems that\n",
    "\n",
    "http://data.giss.nasa.gov/gistemp/sources_v3/GISTEMPv3_sources.tar.gz\n",
    "\n",
    "and \n",
    "\n",
    "http://data.giss.nasa.gov/pub/gistemp/SBBX.ERSST.gz\n",
    "\n",
    "are down, so let's use a local copy instead."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "!mkdir -p ccc-gistemp/input\n",
    "\n",
    "!cp GISTEMPv3_sources.tar.gz SBBX.ERSST.gz ccc-gistemp/input"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/home/filipe/Dropbox/Meetings/2018-PythonSul/keynote/notebooks/ccc-gistemp\n"
     ]
    }
   ],
   "source": [
    "%cd ccc-gistemp/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We don't really need `pypy` for the fetch phase, but the code is Python 2 and the notebook is Python 3, so this is just a lazy way to call py2k code from a py3k notebook ;-p\n",
    "\n",
    "PS: we are also using the International Surface Temperature Initiative data (ISTI)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fetching ftp://ftp.ncdc.noaa.gov/pub/data/globaldatabank/monthly/stage3/recommended/results/recommended-ghcn_format.monthly.stage3.v1.1.1.20180307.tar.gz to input/isti.v1.tar.gz\n",
      "  116679933/116679933 [100%]00/116679933 [2%][4%]6679933 [6%]79933 [9%]0/116679933 [11%] [13%]840000/116679933 [15%]79933 [17%]/116679933 [21%][23%]76000/116679933 [25%]9933 [27%]116679933 [31%]34%]2000/116679933 [36%]933 [38%] 46968000/116679933 [40%]16679933 [42%]4%]000/116679933 [46%]33 [48%]59104000/116679933 [50%]6679933 [52%]%]00/116679933 [56%]3 [58%]1240000/116679933 [61%]679933 [63%]]0/116679933 [67%] [69%]376000/116679933 [71%]79933 [73%]/116679933 [77%][79%]12000/116679933 [81%]9933 [83%]0/116679933 [86%]]6679933 [90%]107376000/116679933 [92%]933 [94%]56000/116679933 [96%][98%]\n",
      "  ... input/isti.merged.inv from merged.monthly.stage3.v1.1.1.inv.\n",
      "  ... input/isti.merged.dat from merged.monthly.stage3.v1.1.1.dat.\n"
     ]
    }
   ],
   "source": [
    "!pypy tool/fetch.py isti"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "QC the ISTI data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 100% ZIXLT831324 TAVG 1960 180       \n"
     ]
    }
   ],
   "source": [
    "!../madqc/mad.py --progress input/isti.merged.dat"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We need to copy the ISTI data into the `input` directory."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "!cp isti.merged.qc.dat input/isti.merged.qc.dat\n",
    "\n",
    "!cp input/isti.merged.inv input/isti.merged.qc.inv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here is where `pypy` is really needed, this step takes ~35 minutes on valina `python` but only ~100 seconds on `pypy`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fetching ftp://ftp.ncdc.noaa.gov/pub/data/ghcn/v3/ghcnm.tavg.latest.qca.tar.gz to input/ghcnm.tavg.latest.qca.tar.gz\n",
      "  12780302/12780302 [100%]%]780302 [21%]44000/12780302 [41%]2%]2 [82%]\n",
      "  ... input/ghcnm.tavg.qca.dat from ./ghcnm.v3.3.0.20180405/ghcnm.tavg.v3.3.0.20180405.qca.dat.\n",
      "input/GISTEMPv3_sources.tar.gz already exists.\n",
      "  ... input/oisstv2_mod4.clim.gz from GISTEMP_sources/STEP4_5/input_files/oisstv2_mod4.clim.gz.\n",
      "  ... input/sumofday.tbl from GISTEMP_sources/STEP1/input_files/sumofday.tbl.\n",
      "  ... input/v3.inv from GISTEMP_sources/STEP1/input_files/v3.inv.\n",
      "  ... input/ushcn3.tbl from GISTEMP_sources/STEP1/input_files/ushcn3.tbl.\n",
      "  ... input/mcdw.tbl from GISTEMP_sources/STEP1/input_files/mcdw.tbl.\n",
      "  ... input/Ts.strange.v3.list.IN_full from GISTEMP_sources/STEP0/input_files/Ts.strange.v3.list.IN_full.\n",
      "  ... input/antarc2.list from GISTEMP_sources/STEP0/input_files/antarc2.list.\n",
      "  ... input/antarc3.list from GISTEMP_sources/STEP0/input_files/antarc3.list.\n",
      "  ... input/antarc1.list from GISTEMP_sources/STEP0/input_files/antarc1.list.\n",
      "  ... input/antarc1.txt from GISTEMP_sources/STEP0/input_files/antarc1.txt.\n",
      "  ... input/antarc2.txt from GISTEMP_sources/STEP0/input_files/antarc2.txt.\n",
      "  ... input/t_hohenpeissenberg_200306.txt_as_received_July17_2003 from GISTEMP_sources/STEP0/input_files/t_hohenpeissenberg_200306.txt_as_received_July17_2003.\n",
      "  ... input/antarc3.txt from GISTEMP_sources/STEP0/input_files/antarc3.txt.\n",
      "input/SBBX.ERSST.gz already exists.\n",
      "... creating directory log\n",
      "... creating directory result\n",
      "... creating directory work\n",
      "====> STEPS 0, 1, 3, 4, 5  ====\n",
      "No more recent sea-surface data files.\n",
      "\n",
      "Load ISTI.MERGED.QC.DAT records\n",
      "(Reading average temperature)\n",
      "Step 0: closing output file.\n",
      "Step 1: closing output file.\n",
      "Region (+64/+90 S/N -180/-090 W/E): 0 empty cells.\n",
      "Region (+64/+90 S/N -090/+000 W/E): 0 empty cells.\n",
      "Region (+64/+90 S/N +000/+090 W/E): 0 empty cells.\n",
      "Region (+64/+90 S/N +090/+180 W/E): 0 empty cells.\n",
      "Region (+44/+64 S/N -180/-135 W/E): 0 empty cells.\n",
      "Region (+44/+64 S/N -135/-090 W/E): 0 empty cells.\n",
      "Region (+44/+64 S/N -090/-045 W/E): 0 empty cells.\n",
      "Region (+44/+64 S/N -045/+000 W/E): 0 empty cells.\n",
      "Region (+44/+64 S/N +000/+045 W/E): 0 empty cells.\n",
      "Region (+44/+64 S/N +045/+090 W/E): 0 empty cells.\n",
      "Region (+44/+64 S/N +090/+135 W/E): 0 empty cells.\n",
      "Region (+44/+64 S/N +135/+180 W/E): 0 empty cells.\n",
      "Region (+24/+44 S/N -180/-150 W/E): 31 empty cells.\n",
      "Region (+24/+44 S/N -150/-120 W/E): 2 empty cells.\n",
      "Region (+24/+44 S/N -120/-090 W/E): 0 empty cells.\n",
      "Region (+24/+44 S/N -090/-060 W/E): 0 empty cells.\n",
      "Region (+24/+44 S/N -060/-030 W/E): 10 empty cells.\n",
      "Region (+24/+44 S/N -030/+000 W/E): 0 empty cells.\n",
      "Region (+24/+44 S/N +000/+030 W/E): 0 empty cells.\n",
      "Region (+24/+44 S/N +030/+060 W/E): 0 empty cells.\n",
      "Region (+24/+44 S/N +060/+090 W/E): 0 empty cells.\n",
      "Region (+24/+44 S/N +090/+120 W/E): 0 empty cells.\n",
      "Region (+24/+44 S/N +120/+150 W/E): 0 empty cells.\n",
      "Region (+24/+44 S/N +150/+180 W/E): 26 empty cells.\n",
      "Region (+00/+24 S/N -180/-158 W/E): 1 empty cell.\n",
      "Region (+00/+24 S/N -158/-135 W/E): 40 empty cells.\n",
      "Region (+00/+24 S/N -135/-112 W/E): 80 empty cells.\n",
      "Region (+00/+24 S/N -112/-090 W/E): 19 empty cells.\n",
      "Region (+00/+24 S/N -090/-068 W/E): 0 empty cells.\n",
      "Region (+00/+24 S/N -068/-045 W/E): 9 empty cells.\n",
      "Region (+00/+24 S/N -045/-022 W/E): 29 empty cells.\n",
      "Region (+00/+24 S/N -022/+000 W/E): 1 empty cell.\n",
      "Region (+00/+24 S/N +000/+022 W/E): 0 empty cells.\n",
      "Region (+00/+24 S/N +022/+045 W/E): 0 empty cells.\n",
      "Region (+00/+24 S/N +045/+068 W/E): 3 empty cells.\n",
      "Region (+00/+24 S/N +068/+090 W/E): 0 empty cells.\n",
      "Region (+00/+24 S/N +090/+112 W/E): 0 empty cells.\n",
      "Region (+00/+24 S/N +112/+135 W/E): 0 empty cells.\n",
      "Region (+00/+24 S/N +135/+158 W/E): 0 empty cells.\n",
      "Region (+00/+24 S/N +158/+180 W/E): 2 empty cells.\n",
      "Region (-24/-00 S/N -180/-158 W/E): 0 empty cells.\n",
      "Region (-24/-00 S/N -158/-135 W/E): 0 empty cells.\n",
      "Region (-24/-00 S/N -135/-112 W/E): 55 empty cells.\n",
      "Region (-24/-00 S/N -112/-090 W/E): 67 empty cells.\n",
      "Region (-24/-00 S/N -090/-068 W/E): 7 empty cells.\n",
      "Region (-24/-00 S/N -068/-045 W/E): 0 empty cells.\n",
      "Region (-24/-00 S/N -045/-022 W/E): 0 empty cells.\n",
      "Region (-24/-00 S/N -022/+000 W/E): 2 empty cells.\n",
      "Region (-24/-00 S/N +000/+022 W/E): 0 empty cells.\n",
      "Region (-24/-00 S/N +022/+045 W/E): 0 empty cells.\n",
      "Region (-24/-00 S/N +045/+068 W/E): 0 empty cells.\n",
      "Region (-24/-00 S/N +068/+090 W/E): 29 empty cells.\n",
      "Region (-24/-00 S/N +090/+112 W/E): 1 empty cell.\n",
      "Region (-24/-00 S/N +112/+135 W/E): 0 empty cells.\n",
      "Region (-24/-00 S/N +135/+158 W/E): 0 empty cells.\n",
      "Region (-24/-00 S/N +158/+180 W/E): 0 empty cells.\n",
      "Region (-44/-24 S/N -180/-150 W/E): 25 empty cells.\n",
      "Region (-44/-24 S/N -150/-120 W/E): 37 empty cells.\n",
      "Region (-44/-24 S/N -120/-090 W/E): 48 empty cells.\n",
      "Region (-44/-24 S/N -090/-060 W/E): 2 empty cells.\n",
      "Region (-44/-24 S/N -060/-030 W/E): 21 empty cells.\n",
      "Region (-44/-24 S/N -030/+000 W/E): 18 empty cells.\n",
      "Region (-44/-24 S/N +000/+030 W/E): 15 empty cells.\n",
      "Region (-44/-24 S/N +030/+060 W/E): 5 empty cells.\n",
      "Region (-44/-24 S/N +060/+090 W/E): 21 empty cells.\n",
      "Region (-44/-24 S/N +090/+120 W/E): 43 empty cells.\n",
      "Region (-44/-24 S/N +120/+150 W/E): 0 empty cells.\n",
      "Region (-44/-24 S/N +150/+180 W/E): 0 empty cells.\n",
      "Region (-64/-44 S/N -180/-135 W/E): 74 empty cells.\n",
      "Region (-64/-44 S/N -135/-090 W/E): 72 empty cells.\n",
      "Region (-64/-44 S/N -090/-045 W/E): 0 empty cells.\n",
      "Region (-64/-44 S/N -045/+000 W/E): 20 empty cells.\n",
      "Region (-64/-44 S/N +000/+045 W/E): 44 empty cells.\n",
      "Region (-64/-44 S/N +045/+090 W/E): 5 empty cells.\n",
      "Region (-64/-44 S/N +090/+135 W/E): 60 empty cells.\n",
      "Region (-64/-44 S/N +135/+180 W/E): 1 empty cell.\n",
      "Region (-90/-64 S/N -180/-090 W/E): 4 empty cells.\n",
      "Region (-90/-64 S/N -090/+000 W/E): 0 empty cells.\n",
      "Region (-90/-64 S/N +000/+090 W/E): 0 empty cells.\n",
      "Region (-90/-64 S/N +090/+180 W/E): 0 empty cells.\n",
      "\n",
      "Step3: closing output file\n",
      "Step4: closing output file\n",
      "WARNING: Bad mix of land and ocean data.\n",
      "  Land range from 1880-01 to 2018-02; Ocean range from 1880-01 to 2015-09.\n",
      "Step 5: Closing box file: result/landBX.Ts.GHCN.CL.PA.1200\n",
      "Step 5: Closing box file: result/oceanBX.Ts.ERSST.CL.PA\n",
      "Step 5: Closing box file: result/mixedBX.Ts.ERSST.GHCN.CL.PA.1200\n",
      "... running vischeck\n",
      "See result/google-chart.url\n",
      "====> Timing Summary ====\n",
      "Run took 223.5 seconds\n"
     ]
    }
   ],
   "source": [
    "!pypy tool/run.py -p 'data_sources=isti.merged.qc.dat;element=TAVG' -s 0-1,3-5"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Python `gistemp` saves the results in the same format as the Fortran program but it ships with `gistemp2csv.py` to make it easier to read the data with `pandas`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "!pypy tool/gistemp2csv.py result/*.txt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "\n",
    "df = pd.read_csv(\n",
    "    'result/landGLB.Ts.GHCN.CL.PA.csv',\n",
    "    skiprows=3,\n",
    "    index_col=0,\n",
    "    na_values=('*****', '****'),\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's use `sklearn` to compute the full trend..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import linear_model\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "\n",
    "\n",
    "reg0 = linear_model.LinearRegression()\n",
    "\n",
    "series0 = df['J-D'].dropna()\n",
    "y = series0.values\n",
    "X = series0.index.values[:, None]\n",
    "\n",
    "reg0.fit(X, y)\n",
    "y_pred0 = reg0.predict(X)\n",
    "R2_0 = mean_squared_error(y, y_pred0)\n",
    "var0 = r2_score(y, y_pred0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " and the past 30 years trend."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "reg1 = linear_model.LinearRegression()\n",
    "\n",
    "series1 = df['J-D'].dropna().iloc[-30:]\n",
    "y = series1.values\n",
    "X = series1.index.values[:, None]\n",
    "reg1.fit(X, y)\n",
    "y_pred1 = reg1.predict(X)\n",
    "R2_1 = mean_squared_error(y[-30:], y_pred1)\n",
    "var1 = r2_score(y[-30:], y_pred1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAi0AAAIaCAYAAADsln1nAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzs3Xd4lFXax/Hvk94rJAKhNwGBEBBXsaEiYN11VXQtqLgqK4tidy0ouCoWYMHey2JbfbEiShUbImpAQUkQBEICpPcySc77xzMzSUghgRmSwO9zXbmcecp5zkR0bu5zzn0sYwwiIiIibZ1Pa3dAREREpDkUtIiIiEi7oKBFRERE2gUFLSIiItIuKGgRERGRdkFBi4iIiLQLClpERESkXVDQIiIiIu2CghYRERFpFxS0iIiISLvg19od2B8dOnQwPXr0aO1uiIiIiAf88MMPWcaYjvu6rl0GLT169GDt2rWt3Q0RERHxAMuytjXnOg0PiYiISLugoEVERETaBQUtIiIi0i4oaBEREZF2QUGLiIiItAsKWkRERKRd8EjQYlnWS5Zl7bEs65dGzl9iWdZ65883lmUNrXXuD8uyfrYsK9myLK1jFhERkQZ5KtPyCjCuifNbgZOMMUOAmcBze50fbYxJNMaM8FB/RERE5BDjkeJyxphVlmX1aOL8N7XergYSPPFcEREROXy0xpyWScCntd4b4HPLsn6wLOuaVuiPiIiItAMHtYy/ZVmjsYOW42sdHmWMSbcsKw5YYlnWb8aYVQ3cew1wDUC3bt0OSn9FRESk7ThomRbLsoYALwDnGmOyXceNMenOf+4BFgIjG7rfGPOcMWaEMWZEx4773FNJREREDjEHJWixLKsb8H/AZcaYlFrHQy3LCne9Bk4HGlyBJCIiIoc3jwwPWZb1JnAy0MGyrDRgOuAPYIx5BrgXiAWesiwLoNK5UigeWOg85ge8YYxZ7Ik+iYiIyKHFU6uHLt7H+auBqxs4vgUYWv8OERERkbpUEVdERETaBQUtIiIi0i4oaBEREZF2QUGLiIiItAsKWkRERMSrrnx5Df/3Y9oBt6OgRURERLymqtqwYlMmKzZlHnBbClpERETEawpKHQBszy4+4LYUtIiIiIjX5DuDlj+ySw64LQUtIiIi4jV5zqAlv9RBfonjgNpS0CIiIiJe48q0AGzLObAhIgUtIiIi4jV5JRXu19sOcIhIQYuIiIh4TUGtTMv2HAUtIiIi0kblOeexRIX4s+0AVxApaBERERGvySt1EBrgS5+OYQe8gkhBi4iIiHhNfqmDyGB/usWGsF1Bi4iIiLRVeSUOIkMC6B4Tyq6CMsocVfvdloIWERER8ZqCUgeRwX50jw0BYMcBTMZV0CIiIiJek1daQVRwAN2cQcuBLHtW0CIiIiJe45rT0j3GGbQo0yIiIiJtUV6Jg6gQf2JCAwgP9DugjRMVtIiIiIhXlDmqKK+sJjLEH8uy6BYbckDLnhW0iIiIiFe49h2KDPYHoHtsyAFVxVXQIiIiIl7hroYbHABAt5hQ0nJLqKo2+9WeghYRERHxioYyLY4qQ3pe6X61p6BFREREvMK1w3NUiDNoca4g2t8hIgUtIiIi4hV7Z1oOtFaLghYRERHxCnfQ4sy0dIoMJsDXZ793e1bQIiIiIl6RV+LA18ciPNAPAF8fi4SYYGVaREREpG3JL3UQEeSHZVnuY91jQva7Kq6CFhEREfGKvFIHUSEBdY51jw1le3YxxrR82bOCFhEREfGK/FIHEc5JuC7dYkIorqgiu7iixe0paBERERGvyC+pIGqvoKV3XBi9Ooa6l0O3hJ+nOiYiIiJSW36pg+6xoXWOndSvI8tvPnm/2lOmRURERLzCntPiv+8Lm0lBi4iIiHhcdbUhv9RRb3joQChoEREREY8rLK/EGOpNxD0QClpERETE4/JdOzzvteT5QChoEREREY/be98hT1DQIiIiIh6XV1p3h2dPUNAiIiIiHqdMi4iIiLQLea45LQpaREREpC1zZVq0ekhERETatPxSB0H+PgT5+3qsTQUtIiIi4nF5JRVEBTew3DkzE26/HfLyWtymghYRERHxuPxSR8OTcO+6C2bPhoyMFrepoEVEREQ8Lq/EQeTey53XroUXXoCpU2HAgBa3qaBFREREPK5epqW6GqZMgbg4mD59v9r081DfRERERNzySx0Mrh20vPYafPcdvPoqRETsV5vKtIiIiIjH5ZXUyrTk59uTb489Fi69dL/bVKZFREREPKq8sopSR1VNCf/77rNXDX36Kfjsf75EmRYRERHxKHcJ/5AA2LAB5s+Hv/8dkpIOqF2PBC2WZb1kWdYey7J+aeS8ZVnWPMuyNluWtd6yrKRa5yZalpXq/Jnoif6IiIhI6ylwBS1BfvZKoYgI+Pe/D7hdT2VaXgHGNXF+PNDX+XMN8DSAZVkxwHTgGGAkMN2yrGgP9UlERERagWvfoT6rPoPly2HmTOjQ4YDb9UjQYoxZBeQ0ccm5wGvGthqIsiyrEzAWWGKMyTHG5AJLaDr4ERERkTYuv9RBkKOMPo9MhyFD4NprPdLuwZqI2wXYUet9mvNYY8dFRESkncorcTB59bsE7EyDN98AP8+EGwdrIq7VwDHTxPH6DVjWNZZlrbUsa21mZqZHOyciIiKeU/X771z33XtUXHgRnHCCx9o9WEFLGtC11vsEIL2J4/UYY54zxowwxozo2LGj1zoqIiIiByZp7kwqfXzxfexRj7Z7sIKWD4HLnauI/gTkG2MygM+A0y3LinZOwD3deUxERETao8WL6fPtMl446W/4dk3waNMeGWSyLOtN4GSgg2VZadgrgvwBjDHPAIuAM4DNQAlwpfNcjmVZM4HvnU3NMMY0NaFXRERE2qqKCrjhBjLiu/LFGZdwg4eb90jQYoy5eB/nDXB9I+deAl7yRD9ERESkFc2dCykpzLnqITrHR3m8eVXEFRERkQOXng4zZ1J99tm8Fz+E7rEhHn+EghYRERE5cLfdBg4HGdMfpKra0D0m1OOPUNDSirKzs0lMTCQxMZEjjjiCLl26uN9XVFR49FkJCQnk5eXVOz527FgKCwub3U5xcTHXXnstkyZNIjs7G4Dff/+dq666ihEjRnDGGWfw/PPPU11d7b4nKSkJh8Nx4B8CqK6u5uGHH/ZIW4157LHHeOONNwC49NJL6dmzJ4mJiQwdOpQVK1YAsGnTJm644QbGjx/Phx9+CMAFF1zAli1bvNo3EZE26csvYcECuPVWfg+PB/BKpgVjTLv7GT58uDnUTJ8+3Tz66KP1jldXV5uqqqoDbr9Lly4mNzf3gNuZP3+++eqrr0xKSoq59957TVZWlunWrZt5+eWXTV5entm2bZuZNGmSyczMNMYYk5KSYv785z8f8HNdHA6HiYyM3K/7mqOiosIMGTLEVFZWGmOMueSSS8zChQuNMcZ8/vnn5sgjj6xz/XvvvWfmzp1rjDFm6dKl5rrrrmtx30RE2rXKSmOGDjWma1djiovNa9/+Ybrf/rHJyCttdhPAWtOM739lWtqgzZs3c9RRR3HdddeRlJRERkYGn376KcceeyxJSUlMmDCB4uJiwM6g3HfffQwbNowhQ4aQkpICQGZmJmPGjCEpKYnJkydj/5moz5WBcT1z0qRJDBo0iPHjx1NWVlbvej8/P3bt2sWuXbvw8/Nj9uzZTJgwgSuuuILIyEi6devGCy+8QGxsLACLFy9m/PjxAHzyySckJSUxdOhQTj/9dACKioq44oorGDlyJMOGDeOjjz4C4IUXXuD8889n7Nix9O3blzvvvBOAO+64g8LCQhITE7n88svZvHkziYmJ7v49/PDDPPDAAwAcf/zx3HXXXZx44ok88cQT9OrVi8rKSgDy8vLo2bMnVVVVdT7fkiVLOProo/H19a332Y899lh27tzpfr906VJWr17N1KlTATj55JNZvHhxvTZFRA5pzz4L69bB7NkQEsL27GIC/XyICw/0+KMUtLRRGzduZNKkSfz000/4+/vz8MMPs2zZMn788UeGDBnCf/7zH/e18fHx/PTTT1x99dXMnj0bgOnTpzN69Gh+/PFHxo0bR3p6gzX76ti0aRM33ngjGzZsIDg4mPfff7/eNRMnTmTlypUsWLCAG2+8kR9//JHjjz++3nWWZRc7Xrx4MePGjWPXrl1MnjyZhQsXsm7dOt566y0AZsyYwbhx41izZg3Lly/n5ptvdgdL69at491332X9+vX897//JT09nYcffpjw8HCSk5N57bXX9vmZCgoKWLVqFTfeeCOjRo1i8eLFALzxxhtceOGF9YKTr7/+muHDhzfY1uLFi/nzn/8MwJo1a7jsssvIycnhpZfsxW++vr706NGDX35pcLNzEZFDT1YW3H03nHIK/PWvAPyRXUK3mBB8fBoqen9gDtbeQ9JCvXv35uijjwbgm2++YePGjRx33HEAVFRU1AkUzjvvPACGDx/OokWLAFi1apX79bnnnkt4ePg+n9mnTx8GDx7sbuuPP/6od01wcDDz5893v3c4HPj7+zfYXllZGXv27KFbt24sXLiQ0aNH0717dwBiYmIA+Pzzz/n000/d81TKysrYvn07AKeddpq730ceeSTbt28nLi5un5+jtosuusj9+uqrr2bevHmcddZZvPzyy7z++uv1rs/IyGDYsGF1jk2bNo1p06aRlZXFmjVrABg5ciQZGRn17o+LiyM9PZ2hQ4e2qJ8iIu3SXXdBQQHMmwfOv6xuzy7xznwWlGlps0JDa2ZdG2MYN24cycnJJCcns3HjRp577jn3+cBAOwXn6+vrHv6AmmxHc7naaaitxgwcOJCffvqpwXNffPEFJ554ovszNNQfYwzvv/+++7Nt376dfv36Nbs/fn5+dSb97j2kVfv3eNJJJ5GSksKKFSvw9/fnyCOPrNdecHBwvTbmzJnD5s2bmT59OldccUWDn7X284ODg5u8RkTkkPDDD/D88/DPf8KgQYD9//TtOSV088LKIVDQ0i4cd9xxfPHFF+6VKcXFxaSmpjZ5z4knnsiCBQsA+Oijj1q0Qqglpk6dylNPPcXatWsB+w/sW2+9RWFhYZ35LKNGjWL58uVs27YNgJwcu/Dx2LFjmTdvnru9xgIgFz/nTqGuAOaII44gPT2d3NxcysrK+OSTT5q8/9JLL+WSSy7hyiuvbPD8gAED2Lx5c73jvr6+3HzzzZSUlLBs2bJG209NTWWQ8z9eEZFDVnW1Hax07Aj33ec+nFlYTqmjih4dlGk5bMXHx/Piiy8yYcIEhg4dynHHHeeecNuY+++/n6VLl5KUlMTKlSvp0qWLV/rWp08fXnnlFSZPnkzv3r0ZMGAAX3/9NYGBgaxatYoTnLt7xsfH8/TTT3PuuecydOhQLrnkEsCee1NSUsLgwYMZNGgQ99X6w9+YSZMmMWTIEC6//HKCgoL417/+xdFHH80555zDwIEDm7z3kksuIT8/nwkTJjR4/owzzuCLL75o8JxlWdx999088sgjDZ5PT08nMjISbegpIoe8//4Xvv0WZs2CyEj34W05JQB0i/FO0GI1tqqkLRsxYoRx/c1e2qZt27YxZcoU92qgtuKtt97is88+4+WXX270mnPOOYe5c+fSq1evFrX96KOPEhcXx8SJEw+0myIibVd+PvTvDz17wtdfg09N/uPdH9K45X/rWHHLyfTs0PwhIsuyfjDGjNjXdZqIK17RvXv3NhewTJ48maVLl7pXEDVm1qxZpKentzhoiY2N5dJLLz2QLoqItH0zZsCePfDxx3UCFoBt2cX4WNAlyjtz+xS0yGHj6aefbtZ1AwYM2K/2r7rqqv26T0Sk3fj1V3ul0KRJMKJ+YmRbdgmdo4IJ8PPO7BPNaREREZF9MwamToWwMHjwwQYv2ZZTQo9Y76wcAgUtrc7X19e931BiYmKDtVFq69GjB1lZWQCEhYXts/2tW7dyzDHH0LdvXyZMmNDgnkYOh4OJEycyePBgBgwYwEMPPQTYxeZq9y0iIoK5c+c2+Jy5c+e6i73l5OQwZswY+vbty5gxY8jNza13/YoVK+q0HRQU5C5mt2zZMpKSkkhMTOT44493r+Z54oknmpyLIiIiXrRwISxdCjNn2quGGrA9u5huXqrRAmjvodYWGhraouu7d+/u3tenOfdecMEF5s033zTGGHPttdeap556qt41CxYsMBMmTDDGGFNcXGy6d+9utm7dWueayspKEx8fb/7444969zscDjN48GD3/j633nqreeihh4wxxjz00EPmtttua7KP2dnZJjo62hQXFxtjjOnbt6/ZuHGjMcaYJ5980kycONHdt8TExH1+ZhER2T8l5ZUmt7i8/oniYmO6dTNm8GBjGtnLLa+kwnS//WPzzMrNLX4u2nuo/XrllVeYMmWK+/1ZZ53FypUrW9yOMYbly5dz/vnnA3YJ/oZK81uWRXFxMZWVlZSWlhIQEEBERESda5YtW0bv3r3dFW1rW758OUlJSe4aKh988IF7BU1jz6zt3XffZfz48YSEhLj7U1BQAEB+fj6dO3cGICQkhB49erir0oqIiGf9e9FGLnpudf0Ts2bB9u0wfz74NTwddnu2vdzZW9VwQRNxW11paal7w7+ePXuycOHC/WonMTGR5OTkOseys7OJiopyBxMJCQl1NvxzOf/88/nggw/o1KkTJSUlzJkzx11m3+Wtt97i4osvbvDZe+/Xs3v3bjp16gRAp06d2LNnT5N9f+utt7jpppvc71944QXOOOMMgoODiYiIYPXqmv+ARowYwZdffsnIkSObbFNERFru57R8UvcUUVlVjZ+vM6+xdasdtFx0EZx0UqP3bsuxN/L1VjVc0JyWVhccHOwuYb+/AQtQL2ABGtzZuaFS+mvWrMHX15f09HS2bt3K448/7q6+C/ZeRx9++CEXXHBBg8/OyMjY74JqGRkZ/Pzzz4wdO9Z9bM6cOSxatIi0tDSuvPLKOgGNa28fERHxLGMMW7KKqao2ZOTX2s7kppvs7MqjjzZ5/zZnpsWbc1oUtLRB+9pPp7k6dOhAXl6eu+R9Wlqae6iltjfeeINx48bh7+9PXFwco0aNonbxvk8//ZSkpCTi4+MbfM7e+/XEx8e7NxPMyMhocpPDd955h7/85S/uTRczMzNZt24dxxxzDAATJkzgm2++cV+vvX1ERLwjq6iCwjLn90VuqX3w88/h/fftnZwTEpq8f3t2CR3CAgkL9N4gjoKWNqhHjx4kJydTXV3Njh079nsOh2VZjB49mnfffReAV199lXPPPbfedd26dWP58uUYYyguLmb16tV1NhN88803Gx0agvr79Zxzzjm8+uqrTT6zsbajo6PJz893b1OwZMmSOnVTUlJSOOqoo/b10UVEpIW2ZhW7X6fllkBFhb3EuW9fmDZtn/dvyyn26nwWUNDSJo0aNYqePXsyePBgbrnlFpKSkvZ5j2tezN5mzZrF7Nmz6dOnD9nZ2UyaNAmADz/8kHvvvReA66+/nqKiIo466iiOPvporrzySoYMGQJASUkJS5Ys4bzzzmv02ePHj2fVqlXu93fccQdLliyhb9++LFmyhDvuuAOAtWvXcvXVV7uv++OPP9ixYwcn1Roj9fPz4/nnn+evf/0rQ4cO5fXXX+fRWinJr7/+mtNOO22fvw8REWmZLZlF7tc7ckvtInKbNsHcuRAYuM/7t2WX0N1Lew65aO8h8Yi//OUvPPLII/Tt29drz/jpp5+YPXs2r7/+uteeISJyuHpw0a+88s0fRIf4Mz6mmvtu/SucfDI0Y0uWMkcVA+5dzA2n9uXG0/q1+NnN3XtImRbxiIcfftg9j8VbsrKymDlzplefISJyuNqSWUyP2BC6x4Ryyiuz7eGhOXP2eZ+jqpqnVmzGGFq0SeL+0JJn8Yj+/fvTv39/rz5jzJgxXm1fRORwtiWriH5x4Qzc+jMnfvcZ3HUX9OnT5D0b0wu49d11bEgv4KwhnRg76Aiv9lGZllZUVlbGyJEjGTp0KIMGDWL69Onuc80pv98alixZwvDhwxk8eDDDhw9n+fLlDV5333330aVLF3eZ/kWLFrnPrV+/nmOPPZZBgwYxePDgRldHnX/++e6l1z/88AODBw+mT58+TJ06tcHl3Pn5+Zx99tnu32ftkv/jxo0jKiqKs846q849F110EampqS3+PYiItBe/Zxbx5IrNDf5/08VRVc327BJ6xQRx0WuPkB7REcdttzfZ7hvfbeecJ75id0E5z1w6nCf+lkSQv6+nu1+HgpZWFBgYyPLly1m3bh3JycksXrzYXUjt9ttvZ9q0aaSmphIdHc2LL7540PpVVVXV6LkOHTrw0Ucf8fPPP/Pqq69y2WWXNXrttGnT3DVozjjjDAAqKyu59NJLeeaZZ9iwYQMrV650L3eubcOGDVRVVdGrVy8AJk+ezHPPPUdqaiqpqaksXry43j1PPvkkAwcOZN26daxcuZKbb77ZHezdeuutDc6FmTx5Mo888kjTvxARkXbstW/+4NHPNlFQWtnoNWm5pVRWG0796gPifv+Vf598FRmOpgOQp1ZuZnBCJEumnci4o7ybYXFR0NKKLMtyb3rocDhwOBxYltXs8vsvvvgi02otQ3v++efdhdj++9//MnLkSBITE7n22mvdgcjkyZMZMWJEvcxOjx49mDFjBscffzz/+9//Gu3zsGHD3LVeBg0aRFlZGeXl5c3+zJ9//jlDhgxh6NChAMTGxuLrW/8/jAULFriXSmdkZFBQUMCxxx6LZVlcfvnljW5HUFhYiDGGoqIiYmJi3NWATz31VMLDw+vdc8IJJ7B06VJ3LRsRkfYov9RBVlHD/y9O3pEHQE5J4xn7LZlFRJUWMPSZx8g/ZhSfHHm8vey5EcXllaTllnJK/ziiQwMOrPMtoKCllVVVVZGYmEhcXBxjxozhmGOOaXb5/YsuuogPP/wQh8MBwMsvv8yVV17Jr7/+yttvv83XX39NcnIyvr6+LFiwAIB///vfrF27lvXr1/PFF1+wfv16d3tBQUF89dVXXHTRRTzzzDM888wzTfb9vffeY9iwYQQ2shTuiSeeYMiQIVx11VXunZ5TUlKwLIuxY8eSlJTUaJaj9tYAO3fuJKFWUaPGfh9Tpkzh119/pXPnzgwePJj//Oc/+Pg0/Ufcx8eHPn36sG7duiavExFpy+5+/xcmvlS/pleZo4qNGfZebjnFjf8Fc0tmMbeseh3fwgJKH58DllVTYK4Bm/fYy6P7xtf/y6A3KWhpZb6+viQnJ5OWlsaaNWv45Zdfml1+PzQ0lFNOOYWPP/6Y3377DYfDweDBg1m2bBk//PADRx99NImJiSxbtsw9N+Sdd94hKSmJYcOGsWHDBjZu3Ohub8KECe7X1113Hdddd12j/d6wYQO33347zz77bIPnJ0+ezO+//05ycjKdOnXi5ptvBuzhoa+++ooFCxbw1VdfsXDhQpYtW1bv/tpbAzT39/HZZ5+RmJhIeno6ycnJTJkyxb3xYlO0NYCItHdpuSVsSC8gs7BuYLIhvQBHlf3/0JxiR6P3l65Zy9/WLca6/npi/zQcH4smMy0puwsB6Bcf5oHeN5+CljYiKiqKk08+mcWLFzdaft+VlUlMTHQXhrv66qt55ZVX3FkWsL/kJ06c6J5PsmnTJu677z62bt3KY489xrJly1i/fj1nnnlmnUmwoaHNW6qWlpbGX/7yF1577TV69+7d4DXx8fH4+vri4+PD3//+d3dV34SEBE466SQ6dOhASEgIZ5xxBj/++GO9+2tvDZCQkEBaWlqd5ze0HcHLL7/Meeedh2VZ9OnTh549e/Lbb7/t8/NoawARae9yiu2hn++2Ztc57hoaAsgtbmR4yBhOf+bfFIZGwv334+/rQ6fI4CYzLal7igjw86F7rHeXOO9NQUsryszMJC/P/gNVWlrK0qVLOfLIIxstv+/KyiQnJzNjxgwAjjnmGHbs2MEbb7zhLod/6qmn8u6777p3V87JyWHbtm0UFBQQGhpKZGQku3fv5tNPP21xn/Py8jjzzDN56KGHGDVqVKPX1a7ZsnDhQnfp/bFjx7J+/XpKSkqorKzkiy++YODAgfXur701QKdOnQgPD2f16tUYY3jttdca3Y7AlbXZvXs3mzZtck/kbUpKSgqDBg3a53UiIm2VK2hZvaV+0BLjnHOS3VjQsmABR/6+nsWX3ghRUQB0iQ5mxz4yLb07huHrUz/r7U0KWlpRRkYGo0ePZsiQIRx99NGMGTPGvSS3sfL7DbnwwgsZNWoU0dHRAAwcOJAHHniA008/nSFDhjBmzBgyMjIYOnQow4YNY9CgQVx11VVNBh2NzWl54okn2Lx5MzNnznRnfVzB0dVXX+3eaPG2225j8ODBDBkyhBUrVjDHWaAoOjqam266yT10lZSUxJlnnlnvOWeeeSYrV650v3/66ae5+uqr6dOnD71792b8+PH1+nnPPffwzTffMHjwYE499VRmzZpFhw4dAHvC7QUXXMCyZctISEjgs88+A+zgJjg4mE6dOjX6uxARacsqKqvdGx1++/veQUsuf+oVQ6CfD7kNTcQtLKT61ttI7tSP7Atq9oFLiN5HpmV30UEfGgKV8T8knHXWWUybNo1TTz21tbviMaWlpYwePZqvv/66wdVFnjJnzhwiIiKaDApFRNqyPQVljHxwGV1jgtmRU8qau04lLjyIrKJyRjywlLvOGMBLX29lVJ8OPHbB0Lo333YbPPoo5172OJNv+5t76fKcJSnMW57KppnjCfCrm98oKq/kqOmfcevY/lw/uunic82lMv6Hgby8PPr160dwcPAhFbCAPafl/vvvb3CVkCdFRUUxceJErz5DRMSbXEuZzxhsZ4xXb8kBIHm7Pf0gsVsUMaEB7iEkt99+gzlz2HbuBNZ17k+vjjXzUxKigzEGMvLrZ1tSnZNw+8Yd/EyLgpZ2LCoqipSUlCbrqrRnY8eOpVu3bl59xpVXXuleWi4i0h7lFNnByIl9OxIW6Oee15K8Iw9fH4ujOkfWD1qMgRtugNBQFv3tBnws6B5bs0NzQrT9uqEhotTd9nLnfgd5uTNo7yEREZF2zZVp6RgeyMgWZrGGAAAgAElEQVSeMax2zmv5aUcuRx4RTnCAL9EhAWzLrjWx9oMP4PPPYe5cNlQGkhAdQqBfzVB8QrS9onJHTv3JuCm7Cwn086FrTEi9c96mTIuIiEg75lrKHB0SwJ96xbAlq5iM/FLW78hnWDd7NVBMaEDNkufSUpg2DQYNgn/8gy2ZxfV2Z+4UGYSvT8MF5lL2FNEn7uCvHAJlWkRERNo111LmqBB/ju1lr5h847vtFJZXktjVXlUaExpAYXklFZXVBDz6KPzxByxfjvHzY2tWMcf0iqnTpp+vD50igxosMJe6u5A/9Yr17odqhDItIiIi7VhucQWRwf74+/owsHME4UF+vPrNHwAkdq3JtADk/5oKDz0EF14Io0ezq6CMUkcVvTrWn1Tb0LLngjIHGfll9G2F5c6goEVERKRdyylxuIMSXx+LY3rGUFBWSXiQH72cwz6u84F33Ao+PvDYYwBszSoGoGcDlW0TokPqBS3uSbhxB38SLihoERERaddyisuJDvF3v3cN3SR2jcLHOe8kOiSAUX8kE7HoI/jXv6BrVwDS85zbpUTX38okITqY3YVllFdWuY+luvccUtAiIiIiTh+vT+eyF7/DUVXd5HU5xQ5iQgPd72sHLS6xgRb3LX2W4oTu4NzAFiA9z86kHBEZVK/dhOgQjKkJbABSdhcR7O/bYJBzMChoERERaYO+2JTJl6lZvPdDWpPX5RSXExNak2kZ1DmCu88cwCXHdHcf6/z6C/TN3sHqG+6BoJoAJSO/lA5hAQT516883tUZmNSejJu6p5A+cWHuDM7BpqBFRESkDUp3VqOdv3wzFZUNZ1uMMeQWO4h2zlkBsCyLq0/oVZM9ycgg9KEHWN5rBD8nnlDn/p15ZXSOajhrkuCsw/L7niL3sZTdha02CRcUtIiIiLRJGXllxEcEsjOvlHcbybYUV1RRUVVNbK2gpZ477sAqK2POmf+oV8o/Pa+UzpENBy1HRATRPTaEBz75lTlLUsgqKmd3QXmrzWcBBS0iIiJtjjGGnXmlnD2kM4ldo3hyRcPZFlcJ/+iQRoKWb76B116Dm2+mqFvPOkGLMcYOWhrJtPj6WLz/j1GcPbQz/1mWypnzvgRold2dXRS0iIiItDE5xRWUV1bTJTqYaWP6sTOvlHfW7qh/nbOEf0xDmZaqKvjnP6FLF7jrLrsqbklN0JJf6qCkoorOUfUn4bpEhwYwZ0IiL1w+AmPAsuDIIyIO/APuJ1XEFRERaWNcK3Y6RQZzYt8OJHWzsy0XjEios0eQqzR/g0HLCy/Ajz/Cm29CWBjRIQF1JtW6ntFYpqW20wbGc3SPGH7PKmrW9d6iTIuIiIgXPfrZb/y0PbdF97gm4XaJCsayLKaN6UdGfhn/W1t3bkt2Y0FLTg7cdRecdBJMmOC8xr9OpsW13Lm5QUhkiD9J3aJb9Dk8TUGLiIiIlziqqnlyxe8s+jmjRffVBBT20M3xfTrQNSaYb507OLu4N0vcO2i55x7IzYV58+wxHSAmNJDcYgfGGPsZ+XWf0R54JGixLGucZVmbLMvabFnWHQ2cn2NZVrLzJ8WyrLxa56pqnfvQE/0RERFpCwrLKgEoKK1s0X3peaUE+vm4MyiWZdGnYxi/ZxbVuS6npAJ/X4vwwFqzPZKT4Zln4B//gCFD3IdjQv2pqKqmqNzuy868Uvx9LTrUKkzX1h3wnBbLsnyBJ4ExQBrwvWVZHxpjNrquMcZMq3X9P4FhtZooNcYkHmg/RERE2prCMgdgbzTYEunO+imWVVPErVfHML7dkk11tXEXd8spqiA6JKDmOmPsybcxMTBjRp02XSuMcosdhAf5k5FXRqfI4FYrFLc/PJFpGQlsNsZsMcZUAG8B5zZx/cXAmx54roiISJvmzrS0NGjJL603bNOzQyhljmoyCmrK6ueUVNSdz/LGG/DVV/DggxBdd/5JbFiA+x5w1mhpR0ND4JmgpQtQex1WmvNYPZZldQd6AstrHQ6yLGutZVmrLcv6swf6IyIi0ia4gpb80pZmWuoXfevV0d6JeUutIaKc4lpBS2Eh3HorjBgBV11Vr01XpiWnuLzmGa24Emh/eCJoaSivZBq59iLgXWNMVa1j3YwxI4C/AXMty+rd4EMs6xpncLM2MzPzwHosIiJyELiHh1owp6Wispo9heX1AoreHe2ibluzit3HcosraibhPvAAZGTA/PngW38voVjn3JWcYgeVVdXsKiijy2EYtKQBXWu9TwDSG7n2IvYaGjLGpDv/uQVYSd35LrWve84YM8IYM6Jjx44H2mcRERGv25/hod0FZRhTf1VPXHggoQG+bMmsCVpySirsEv6bNsGcOXDFFfCnPzXYbrRzU8Xc4gr2FJZTbew6MO2JJ4KW74G+lmX1tCwrADswqbcKyLKs/kA08G2tY9GWZQU6X3cARgEb975XRESkParJtNQsNd6XxuqnWJZFz46hbHFmWiqrqskrcRAd7A833gjBwfDww422Gxboh7+vRXZxRb0l1e3FAa8eMsZUWpY1BfgM8AVeMsZssCxrBrDWGOMKYC4G3jJ1/60NAJ61LKsaO4B6uPaqIxERkfbMlWmpNvbmhmGB+/7aramfUj8L0qtDGD86C9XlOefJDP1pFSxeDLNnQ3x8o+1almWX8i+uYGdeTfG69sQjZfyNMYuARXsdu3ev9/c1cN83wGBP9EFERKStKSyvmctSUOpoXtDiKq/fwNBNr46hfLQ+nTJHFbnFFQQ6yjlm/gMwcCBMmbLPtqNDApyZFuc2AYdj0CIiIiL1Fdaay1JQ5qAz+w4SduaVEh3iT3BA/cm0PTuEYgxsyy4ht6SCa9b8H6E7t8OrS8Hff59tuzZNzMgvJSLIr1lBVFuiMv4iIiJe4hoeguavIMpoYimyawXRlswiSlO38I/V71Jw5rlw6qnNats1PNQelzuDghYRERGvKSyrxM9ZcbagmbVaXNVwG9Kzg7NWS1YxvR65D4DyWY80uz8xofbw0M689rfcGRS0iIiIeE1hmYMjIu0VOs0tMJeeV9poQBEa6Ed8RCA+K1bQffkinvrT+UT0b7C8WYOiQwLIL3WQllOiTIuIiIjUKCyrdAcgzanVUlDmoLC8kk6RjS9F7hsdxJnPP0hufAILTpxAoF/9uS+NcZXyLyyvpFM7W+4MClpERES8pk7Q0ow5LRmulUNNZEEu/v5Duu36g3cvvZmwyLAW9cdVyh/a33Jn0OohERERrykscxAVEkBogG+zMi2NFZZz272bMe88zcqew/mgWxLRDZTrb0psrc0VNTwkIiIiAFRVG4orqggP8iMi2L9ZE3H3WfTtjjvwqyjn/tOu4dddRcSE7HuZc23R7TxoUaZFRETEC4qcy53Dg/yICPJvdqbFz8eiY3hg/ZOrV8Mrr1A49Sa2BneBakNMaAPXNcG1I7SPBfENPaONU6ZFRETECwrL7SDFzrT4NWtOS3peKfERQfg6l0m7VVfDP/8JnTsTNvM+/H3t8zGhLcy0OOe0xEcE4efb/kKA9tdjERGRdqDQnWnxJzK4mZmW/Ebqp7z0EqxdC48+im9EON1j7XottYd7miPAz4fwQL92OTQEClpERES8onA/h4fq7bycmwt33gknnAAXXwxAL2eRudgWBi0AXaKD6d0xtMX3tQWa0yIiIuIFrn2HwoP8iQj2J7+k6aClqtqwK7+Barj33gs5OTB/Plj2sFCvjmHA7jpLmJvr1atGEuTfslVHbYWCFhERES+om2nxo7C8kupqg8/e81WcMgvLqaw2dXdeXr8ennoKJk+GoUPdh12Zlpj9yLTER7S/onIuClpERES8oCbTYi95NgaKKiqJCGp48mxGvr3cuZMrqDDGnnwbHQ0zZtS59vRB8aTs7snghEjvfYA2SEGLiIiIFxQ4My0RQf7uQKWg1NFo0LIr366G6y6v/9ZbsGoVPPssxMTUuTYqJIC7zxropZ63XZqIKyIi4gWFZZX4+1oE+vkQEWznCJpa9ryrwA5ajogIgqIiuOUWSEqCSZMOSn/bA2VaREREvKCo3EFYoB+WZdVkWppYQbQrv4wAXx97nsq/7oP0dHj3XWhhqf5DmTItIiIiXlBYVkm4M1iJCK4ZHmrMroIy4iMDsTZvhscfh8svh2OPPSh9PWh27rSXcO8nBS0iIiJeYAct9oBGTaal8eGhjPwyOoUHwQ03QFAQzJp1UPp5UJSWwr//Df36wf3373czGh4SERHxgsIyhztoiWxOpiW/jEv3rINPP7UzLUcccVD66VXG2ENct94K27bBeefB1Kn73ZwyLSIiIl5Qe3gozBm8NDanxRhDTk4BFyyYDQMG2EudDwX33gsXXgiRkbB8Obz3HvTqtd/NKdMiIiLiBbWHh3x9LMID/chvJNOSW+Jg4jfvEb1rB7y+BPxbthFim7J7N1RUQNeucMUV0KUL/P3vHplQrEyLiIiIFxSU1a3JEhHs3+iS56yNm7l+9TvsOvUMOO20g9VFz6qogMceg759azJFvXvDddd5bAWUghYREREPq642FJXXZFrArozb2PBQ5L134mMMWTMeOlhd9Bxj4KOPYNAge+7KiSd6bRKxghYREREPK66oxBgIC6wJWuxMSwNBy4oVxC/+kKf+dAGxg/oexF56yDPPwDnngJ+fPYn444+hf3+vPEpzWkRERDysqNy1WWKt4aEgf3bmlda9sLISpk4lP74Lzx9zHlPDAg9mN/dfTg7s2gUDB8LFF0N1NVxzjdfn4ijTIiIi4mG1d3h2iQj2q59peeop+OUXFl52MxExEfj5tvGv5cpKeOIJ6NMHLr3UHhqKioLrrz8ok4fb+G9HRESk/am9w7NLRJB/3Tkte/bYS4JPP51l/Y/liMjgg93Nlvn8cxg61J5km5QEr74KlnVQu6CgRURExMMKyhoYHgr2p6i8kupqYx+4804oKYF589hVUM4REW14aOjDD2HsWCgvh/ffhyVLYPDgg94NBS0iIiIe5hoeiqiVaYkM9scYKCyvhDVr4KWX4MYboX9/duWX0amtZVry8+1+AowfD08/DRs2wLnnHvQMi4uCFhEREQ+rGR6qPRHXWRW3uBymTIFOneCeeygqr6SwvJIjIoNapa/1VFXB88/b9Vb+/Gc7u+Lvb9dbCWzdbJCCFhEREQ9reCKuHcD4vPIyfP89PPIIhIezK78MgCMi2kDQsmoVjBhhrwTq18+uv9LKgUptWvIsIiLiYYVlDnwsCAmoqQQbEeRPRFkRcQ/NgFGj4JJLAGqCltbOtKxdCyedZJfff+ste8+gVhoGaowyLSIiIh5WVFZJWKAfVq0v/YhgP6Z9tQC/vByYP98dEOwqsIOWTq0RtBQVwdKl9uvhw+0VQb/9BhMmtLmABRS0iIiIeFztHZ5dYrds4rIfP+H3v1wCw4a5j+/KtwvOxR/M4aHqanjtNXsI6OyzISvLDlIuvxxCQg5eP1pIQYuIiIiHFZTV3XcIY+jwr1spDAzl2ytuqHPtroIyokP8CfL3zKaC+7R6NRx7LEycCAkJsHw5dOhwcJ59gDSnRURExMMK99rhmXfewe/LVTw29h/EBobXuXZXftnBKyyXlgbHHw9xcfDKK3DZZeDTfvIX7aenIiIi7URh7UxLURHccgsMG8bHI8+sV8o/I7/Mu4XlSkth4UL7dUICvPcebNpkZ1raUcACClpEREQ8rrDcURO0PPigneGYP5/w0KC6pfyB3QVeyrQYA++8AwMGwHnn2YEK2MXhwsObvreNUtAiIiLiYe6JuJs3w+OP25sLjhpl7z9UK9NSXllFVlGF52u0/PijvXx5wgR7Q8OVK6F/f88+oxVoTouIiIgHGWMoLKskLMjPLtMfEGAXksO103Ol+9o9BeWAh5c7FxXBKafYz332WZg0CXwP0iRfL1OmRURExIPKHNVUVRuO+mkVfPIJTJ9ul+yn/k7PrhotB1xYrqLCXsJsDISF2XNYUlLsyraHSMACClpEREQ8qrDMQWBlBSc8+aA9JDN1qvtcRHDd4aGMA62Ga4y9A/OgQfbE2hUr7OOjR9vDQocYBS0iIiIeVFBWyaTv3ydi5zaYN88epnGyMy01w0O7DyRo+eUXOP10e2Ktvz8sXmwPCx3CNKdFRETEg8q2bGXKt2+Teep4Op5+ep1zEcF+FJVXUllVjZ+vDxn5ZYQG+BIe2MKv46oqewfm7Gz4z39g8mQ7cDnEKWgRERHxoLiZ9+JjDOnT/03Hvc65Cs7974c0QgJ8Sd6RS3xkUJ09ihrlcMDLL9ul9oOC7E0Ne/RoN9VsPUFBi4iIiKesXEncoveZO+pixvXtXe901xh7X587/+9n97Gxg+L33e5nn8G0afDrrxAcbFeyHTHCY91uLxS0iIiIeEJlJUydSnGnBJ4+5nwuCKo/XDNmYDxf3T6aispq97HOUU0UlktJgZtuslch9ekDH3xgb3B4mFLQIiIi0kJ3vLeezlHBTD21b83Bp5+Gn3/mm5lPUV4USFgj81QSoluwi/I119iF4h55xF6FFOjFcv/tgFYPiYiItNDKTZl8sj6j5kBmJtx7L5x2Gj+PtFfwNBa0NKmqCl54AXbvtt8/9xykpsKttx72AQt4KGixLGucZVmbLMvabFnWHQ2cv8KyrEzLspKdP1fXOjfRsqxU589ET/RHRETEW4wxZBeXszmziDJHlX3wX/+yK9HOm0dheSVhgX74+jRjcm1tX3wBw4fD3/9uT7gF6NcP4psx5+UwccBBi2VZvsCTwHhgIHCxZVkDG7j0bWNMovPnBee9McB04BhgJDDdsqzoA+2TiIiIt+SXOnBUGaqqDam7i+D77+HFF+GGG2DAANJyS4kLb0FWZOtWOP98OPlkyM2Ft9+G22/3Wv/bM09kWkYCm40xW4wxFcBbwLnNvHcssMQYk2OMyQWWAOM80CcREZF6isor60yC3R9ZReXu1xvScmHKFIiLg3vvxRhD8o48hnZtQTXa6dPh009h5kz47Te48EJozhLow5AngpYuwI5a79Ocx/b2V8uy1luW9a5lWV1beK+IiMgBO//pb5jx8YYDaiOzsML9OvCN12HNGnuibEQE6fllZBaWM6xbE0FLdbW9T9Avv9jvZ82yVwndfbe9nFka5YmgpaFw0Oz1/iOghzFmCLAUeLUF99oXWtY1lmWttSxrbWZm5n53VkREDk/GGLZkFbPo511UVTf4VdMs2cV2pqWrTwWjX5kDxx0Hl14KwE/bcwFIbCzT8u23cOyx9j5Bzz9vH+vUCbro7+vN4YmgJQ3oWut9ApBe+wJjTLYxxpVPex4Y3tx7a7XxnDFmhDFmRMeOe9cYFBERaVpJRRUVldXkFFe4g4v9kVVof53NSH6XiMI8qv8zD3zsr9Pk7XkE+Plw5BERdW9KS7MDm+OOgx074NVXYc6c/e7D4coTQcv3QF/LsnpalhUAXAR8WPsCy7I61Xp7DvCr8/VnwOmWZUU7J+Ce7jwmIiLiUTnFNcM6S3/ds9/tZBVV0D9rGyd9/jZvJo5lW48j3eeSd+RxVOcIAvz2+np99ll491246y57KOjyy92BjjTfAf/GjDGVwBTsYONX4B1jzAbLsmZYlnWO87KplmVtsCxrHTAVuMJ5bw4wEzvw+R6Y4TwmIiLiUa6gJcDPh2W/7t7vdrIKy5i54nmqw8J59MTL2ZCeD4Cjqpqfd+YzrFs0GGOvAlq2zL7p9tvtEvwPPABhYQf8WQ5XHqmIa4xZBCza69i9tV7fCdzZyL0vAS95oh8iIiKNySmxg5bTB8bz8foMtmUX0z02tMXtdF+5mJFbknHMm09RRiQb0ws4a0hnNu0qpLyympMKt8OJ18JXX9lLmU891Q5UFKwcMOWmRETksJDrzLRcMMKeSrlfQ0TFxZz/xmz+6NoP/39Mpk9cGBszCgD49acUZi36DydccgZs2mRXs33rLY/1XxS0iIjIYcI1PJTYNYq+cWH7N0T08MN0zN3DB1fdDr6+DOocyYZ0O2ip/vhjztu4wt6NOTXVrmzr6+vJj3DYU9AiIiKHhZziCvx8LCKC/Dh1QDxrtuaQX+pofgO//4555BE+PGo0hSP+BMYw7vfvOPa7z9lTWMbzvU/k3of+h/X44xAZ6b0PchhT0CIiIoeF3JIKokMDsCyL0wbEUVlt+CKlBXW/pk2DgAAeOPEK+mZugzFjGPOva7nkp0Ws/j2bzdllJIw4ynsfQDwzEVdERKStyy6qICYkAIBh3aKJCQ1g2a+7OWdo533f/Omn8NFH5N9yB/9c9jYXPrMYIiMpfXwuf9vVk5Fr7OLujRaVE49QpkVERA4LuSUVxITaQYuvj8Xo/nGs3JRJZdU+9iIqL4epU6FfP/YMSeLi5MXsvPgKSE0l+KYb6BQbxrdbsrEsGJKgYSFvUtAiIiKHhZzimqAFYPSRHckvdbhX/zRq8mTYvBnmzWPLsBM48boXyJs1G2JjARjU2a5+2zcujPAgf6/1XxS0iIjIYWLvoKVTZLD7eIM2bYLTToOXX4bwcBg9muzictIj4ugQFui+bGAnO7uioSHvU9AiIiKHvKpqQ16pg+haQUtksD2ts6Cssu7FeXlw881w1FHwxRfg5wfffQcBAWQ5d3iODatpZ6Az05LYNdrLn0IUtIiIyCEvv9SBMRATUjN84xrKKSzba9nz7t3w5JMwdixUVsKdd8KAAQBkFZUTFeKPv2/N1+fxfTow6fienDH4CO9/kMOcghYRETmkLPt1N6UVVXWO5RTbOzPH1BrWiXAGLQWllbByJdxxh32if397DsuOHdCtW81x7KAltla2BiA4wJd7zhpIVEjd4+J5ClpEROSQsXlPIZNeXcv//ZRW53hOsZ1NiakVWAT5+9Azfxen3D0ZRo+2S+7nOPfsff99WL8eHn8cQkLc92QVldeZzyIHl4IWERE5ZPyy014JtD27pM5xV6YlOtQ5PFRUhHXXXXz2/GR6rv0SZs60d2GOiYHMTLjnHjjlFPjrX+u0k11UQYdwBS2tRUGLiIgcMlzLl9NyS+scd2VaYkOdAUd5OTz7LCuGnswDs9+Hu++GYHs1EXfdBUVFMH8+WFaddjKLyumoTEurUdAiIiKHjI3OzQt35NbNtOSWVJCU9isdbvwHVFfbNVZSUnjiinvZEVxr1c/atfDCC/DPf8LAgXXaKHNUUVhWWW9Oixw8KuMvIiKHBGMMG9Lzgb0yLTt2cOL9N3L9F59A586wdSv07g2xsYQH+VHoWvJcXW0HK3FxMH16vfaznfVcNDzUepRpERGRQ8KugjJySxx0igwip7iC4oJimDED+ven/zdLeeXUSyElxQ5YnMKD/ChwLXl+/XVYvRpmzWpwl+bsIntejCbith4FLSIickhwDQ2NHWTXS9mZXwYLFsDZZ3PHA2+z8M/XQmhonXsigvztTEt+Ptx+O/zpT3DZZQ22n+UOWjQ81FoUtIiIyCFhQ3oBg3dvZspTdxBSUcqOokr4/nt4+21Sg2PrVMN1CQ/yp6DUYWdk9uyBJ54An4a/Gl3VcJVpaT2a0yIiIu3frl0Mv/9mpnz1McR2oE/8Kfa8lgHxgL2/UN/4sHq3RQT70Sl9K+bVeVhXXw3Dhzf6iEwND7U6ZVpEROSgWLcjj8te/I7yyqp9X9xcVVXwyCPQrx8jv/mU5eMvxUpNYVPXI0mrtYIot6SiTmE5l/BAP+5b+iwmPBwefLDJR2UXVRAa4EtwgK/n+i8tokyLiIgcFJ9t2MWXqVnsKSina0zIvm9oDh8fWLwYx0knMabT2Vxw8SmcFhVFQnSwewVRmaOKkoqqBoeHBny7hOO2rSN31myiO3Ro8lFZReVaOdTKlGkREZGDImV3EUDNEuP99fPPcNZZ9t5AlgUff8yP/3mFP2K6uHdcTogOcddqyXEuVa5XX6WkhGFzZ/Jrxx7snHD5Ph+rEv6tT0GLiIgcFKl7CgEortjPoCUrC/7xD0hMhG+/tcvuA4SEuCvhDupkBy1dY2oyLa6gpV6mZdYsgjN2Mn3MdRQ2o0sNbZYoB5eCFhER8brSiiq259iZj6L9ybTMnw99+8Jzz8H110NqKpx+uvv0hvQCOoQFEhcRBNiZlrwSB4VljoYzLVu2wKxZ5J17Pmu6HlVTq6UJ2neo9SloERERr/s9swhj7NeF5fsRtCQnw8iR9s7L8+bZGxvWsjG9wD00BJAQbe8jtDOvlNySBjItN90Efn4UP/CQ3ad9BFKVVdXklFRoeKiVKWgRERGvS9ld6H7drEzLpk32vJXvv7ffP/UULF5cbz8ggIrKalL3FDKwU+2gxZ7ouyOn1J1pca8e+uwz+OADuOceQnp2A7BrtTQhp6QCY6CjCsu1KgUtIiLidSm7i9wbJheVNxEg5ObCtGlw1FHw5Zf2PkEAgYH1dlx2Sd1TiKPKMKiBTEtabgk5xRX4WBAZ7A8VFTB1qj3UdOONhAfZi2j3lWlxFZaLVaalVWnJs4iIeF3q7kL6xoWRuqeIovJG6rS8/DLceivk5MDf/w4zZ9qbF+6Dq3x/7eGh2NAAgv19ScstpcxRRXRIAD4+Fsyda+8/tGgRBAbiB4QE+O5zTkt2sQrLtQUKWkRExOtS9hQyNCGK9LyyxoeH0tPtDMvcufYKoWbamFFASIAvPWJr9hWyLIuE6GB25JTg52vZ81nS0+1A6OyzYfx497X2/kNNBy3ad6ht0PCQiIh4VUlFJTtySukXH05YoF/N8NCWLXDeefDee/b722+HFStaFLCAPV+mb3w4vj51h49cBeayiyqICQ2A224DhwPmzKlzXXiQHwWlTQ8PZRY6gxatHmpVClpERMSrNu+xi8r1iw8jLMgPR14+3HknDBgAn39u118B8PNrdN5KU3bll9ElKqje8YToENJyS8gtqeDoHRvtHZ9vuQV6965zXUSwP4VNzbMB0vPKCA/0IyLIvwXCgcYAACAASURBVMX9E89R0CIiIl7lqoTbJy6cMRu/Yvot58HDD8NFF9nzS669lvwSB+c+8ZU7wGmJPYXlxIXXD1q6xgRTUFbJzqwiLlnwKHTtagdLewkP8tvnRNz0vFI6NRAYycGlOS0iIuJVqbsLCfCx6BEbQrhPFbtjjiDq80V23RWnzZmFrEvL58ftufSJq78bc2NKKiopLKskLqL+sI1r2fNfvv+Ezn9sgrffhtDQetdFBPnzR1Zxk89Jzy+lc1Rws/sl3qGgRUREvGf7dkbffwMdjuiDn+8Z/HzCGbx/5Il8XitgAch31knJLqpoUfN7Cuy5JvENZFoSooOJLsnnli9fJz3pWDpfcEGDbTQv01LGkISoFvVNPE/DQyIi4nklJXDffXDkkQz78Qs6htpzQUKD/Btc8uwKWnKcS4uba3dBGQDxEQ3Pabnly9cJKy9h4x0zGp0vExHsT0GZA+Mq2buXMkcVOcUVdFGmpdUp0yIiIp71+ecwaRKkpVF5/gWc0mE8F51/PABhgX4NlvF3rd5paaZlt3NVT3wDw0PRv/3Mxcmf8crws+k9dGijbYQH+eGoMpRXVhPk71vvfHqevfFip0jNaWltyrSIiIhnVFfb/wwLs4vCrVrFL7OfY2dkHH3jwwE7QCgqr6yX1XAPDxW3dHjIzrTE7Z1pMQZr6lQKwiKZe/zfakr4N8C1IqixUv7pefYzNKel9SloERGRA5ORAVdeaZfHBzjuOFi7Fk44wb3nUL94e3JtWKAfxkBJRd0hopqgpWXDQ3sKywn08yEiaK+Bg//+F775hoUXTKEgKIzo0MaXKrtK+Rc0Mq/FlWnR8FDrU9AiIiL7p6wMHnoI+vWza6BERODeytk5fyR1dyEBfj50d1arDXMGCMV7DRG557S0dHiooIz4iCCs2vNVCgrs7QBGjmTb2fbk29jQxovCRQQ7My2NVMVNzy/FshqeNyMHl+a0iIhIy337LVxyib2h4bnnwmOPQZ8+9S5L2V1E745h7mq1YYHODQrLK6m9q5AraMkqrsAYUzcIaYIdtOwVkMyYAXv2wEcfcXZcV/z8/QgOqD9X5f/ZO/P4uOpy/7/P7HuWydIlSRealJatdIEiylpQcMENr6Je3DfwCld/V9TrclFQXBEFBdSriIgbXhHFpexICxRaWtrSpk3btE2apElmyezL+f1x5pyZySyZSSaZSfN9v168SGbOzJyEIeczz/N5Po+Ka4KliX2eEM0OMyaD+JxfbYRoEQgEAkHpJBKg18O8eeB2w113wYYNBQ/vHvCzbkmj9r0qWsbvH1JFSzSeJBBNaMdNxKAvwoqMRYns2QPf/z584AOwbh1rgDWLGoo+h7MET4vws9QGQrQIBAKBYGKGhuBLX4IjR+DPf4YlS+C554rG7gejcfq8YZY1p8PiNNEyrj2UKRiGxyKlixZ/hPMz9wEtWaK0rP7930t6PKSNuAUrLd4QK+a58t4nmFlErUsgEAgEhYnFlK3LnZ1w992KKIilBMYELRx16qatMV2lcBRoxfhCMW2DcqkTRGOROGOReLbXxGRS9gu1tBR+4DjSRtzcSossy/R5QiwQEf41gRAtAoFAIMjPrl1w+ulw/fWwfj1s3w633QbG0pYG9nuVqZsFdWnR4jQrj81nxF3SpJh1S81qGdSC5aa2edlm0qPXSfjziJbRYIxwLMn8OtEeqgWEaBEIBAJBNmolpa0NmpvhoYfg4Ydh5cqynkYdFc70g6iVlsz2UCyh+FiWNiltpFJTcQeKRPiXgyRJOC0GLeAuk3w/g6B6CE+LQCAQCBRGR+F//geeeAKef14ZYX7yyUk/3VFPGEmCeRlJsnazMsWTKVpUP8uSZqXScrzUSou/QLDcJHBZjHkrLSKjpbYQlRaBQCCY68Tj8KMfKb6V225Tti+HQlN+2n5PiBanGaM+fakxG/SY9LosT4s6OdTqMmMz6Rkp0dOiLkvMt+G5XJwWQ95wuXSlRXhaagFRaREIBIK5zOHD8PrXw44dcMEFium2yJ6ecujzhvK2VRwWA2ORdFVDFQt1ViNuh4nhsVLbQ2GsRj3OEieNilGw0uINYzboaLQXXgMgmDlEpUUgEAjmImGltcL8+bB4Mfz+9/DooxUTLFA438RhNmTltKiVFpfFSKPdXPL00IA/QqvLXHIQXTEKeVqOehThVYnXEEydiogWSZJeJ0nSHkmS9kmSdEOe+/9TkqRdkiRtlyTpEUmSFmXcl5AkaVvqnwcrcT4CgUAgKIDPB5/9rNIK8njAYIAHH4S3vW3CEeZy0EaF82xGdpgNjEXSu4dU0VJnNdJkN5U8PTTgC1fEzwJKlH++Sku/GHeuKaYsWiRJ0gO3A5cBK4F3SZI03mK+FVgry/LpwO+Bb2bcF5JleVXqnzdN9XwEAoFAkIdkEn72M2VP0De/CRdfrHhZpomRQJRIPFlSeyhTtDTaTWV4WsIV2wdU2NMSzhrZFlSXSlRazgL2ybLcI8tyFLgfuCLzAFmWH5NlOZj6djPQVoHXFQgEAkEp+HyKufaDH4SlS5Uk25//HJqapu0l1WC5gu2hPNNDLqsRt8PMcCCCrC5eLIAsywz6I7Q4p27CBaU1NRaJk0imXzeWSDLgDzNfTA7VDJUQLQuBwxnfH0ndVogPAg9nfG+RJGmLJEmbJUl6cwXORyAQCAQAgYDyb5cLVq9WNjH/61+wbt20v/TRIqPC+TwtZoMOi1GP224ilpDxR4pXgcYicYLRxJSD5VTUVNzM8zrmDSPLsFC0h2qGSoiWfE3QvBJZkqT3AGuBb2Xc3CHL8lrgKuBWSZJOKvDYj6TEzZahoaGpnrNAIBCcuAQC8OUvK+FwPT3KbXfdBVddVVHfSjGKhbIp7aHsSovLqiTlutUo/wl8LVqwXAU9LZAd5d/vLVwtElSHSoiWI0B7xvdtQN/4gyRJ2gB8AXiTLMvaPJssy32pf/cAjwNn5nsRWZbvkmV5rSzLa5ubmytw2gKBQHCCIctw331w8slw443w2teCuTKViHLp94YwG3Q02HIj/53j2kPeUIy6lGhQR4snSsVVI/xbppiGq+LKsxNJpOHWHpUQLc8DnZIkLZEkyQS8E8iaApIk6UzgThTBMphxe4MkSebU103AucCuCpyTQCAQzC3icSVn5d3vVpYFPvUU3H8/LCzWrZ8++jxhFhYYFXaYDYRjSWKJJJAtWpocisiaKBV3QEvDrZynBbIrLWqLSxhxa4cpJ/LIshyXJOla4O+AHviZLMs7JUm6Edgiy/KDKO0gB/C71Bu4NzUptAK4U5KkJIqA+oYsy0K0CAQCQal4PFBfr4wuX3ghXH01vO99oKtuDJeab5IPdf9QIBKn3mbCG4ppbZ50paW4aBmscHvImRItmZWWfm+IBpsRq0lfkdcQTJ2KJOLKsvxX4K/jbvtSxtcbCjzuGeC0SpyDQCAQzCnCYfje9+Dmm+Gvf4XXvAa+8pVqn5VGnyfEBcvzt/Lt5nQrRhUtXa1OIC1aJkrFHfBFsJv0OCqQhgvgsirPo04yQeFwPEH1EIm4AoFAMJuQZfjDH5SNy5//PGzYULUWUCGi8SRDYxHmF2irqLH7qq/Fl9EeshgVITJRKu6Av3IZLZBZackULYWrRYLqIESLQCAQzCbe+lZ4+9vBboeNG+GPf1SyV2qIAZ86Kly8PTQWiZNMKuPNqhEWSO0fmqg9FK6YnwXSI8+ZAXNHCyT6CqqHWJgoEAgEtc7x49DYqPhULrsMLrkEPvIRxcdSgxydYOrGkVFp8YfjyHJ65BgoKRV3wBdhVXt9hc4YjHodVqNeq7Q8d2AEfzhOh9tesdcQTB1RaREIBIJaJRqF734Xli1TguFAESuf+ETNChbIHBXOX6XIDHLLjPBXcdvNHC/iaVHScMMVC5ZTcVmVpYnHxyJ88tcvstht4x1rRYB7LSFEi0AgENQasgwPPQSnnQaf/jS86lUzkmJbKSYKZbNnVFryi5bilRZfOE44lqyopwUUX4s3FOP632xjNBjj9nev1rwugtqgdqW6QCAQzFU+9jElwXb5cvjLX+Dyy6t9RmVx1BOi0W7CYsw/Kqy1hwpVWhyKaJFlOW/OixYsV2HR4rIY2Lh7gHhS5utvPY1TFtRV9PkFU0dUWgQCgaAWGB2FkNJW4Y1vVNpC27dPm2CJJZL8feexCRcT5uOlwx56h4MF71embgoLCrspNfIciWthbnW2bE9LPCnjC+XfP6RF+FdoWaKK02IknpR586oFvHNd+8QPEMw4QrQIBAJBNYnH4Y47oLMTvv1t5bY3vAGuvx5Mpml72Sf2DPHRX77Arn5f2Y+9/jfb+OpfCueA9nvCRVNkdToJh9lAIKM95Mpow2ipuAWi/A+NKIsgKz2OvHyekxXzXdz0ltPyVngE1Ue0hwQCgaBaPPIIXHcdvPyykmb7pjfN2EsPpwTBaCA2wZG5+MJxdh71Fry/zxPinJPcRZ9D3fScrz2UmYp7Up58um29HhrtJtoaKitaPn/5CpJJGZ1OCJZaRVRaBAKBoBp85StKMFwgoITFPfIInHHGjL282nrJDFMrlXAsQZ83zGges6wvHMMfiRdtD0F607M3FMOgk7BlROVPlIq77bCHM9rqpqUaIgRLbSMqLQKBQDBT+HwQi4HbrVRVLBal0mKZ+QAz1UviK1O0yLJMKJYAYFe/j3OXNWXdX+pmZLvZgD8lWuqsxiwBoraH8qXi+sIx9g2N8cYzFpR13oITA1FpEQgEgukmkYCf/hS6uuAzn1FuW70abrihKoIF0jt2MhcElkI0kSSRVMy7u/py/TD9nuLjzipOs4GxcCxrw7NKg135Pl8q7vbDXmSZigbLCWYPQrQIBALBdPL003DWWfChD8FJJynBcDWAGlefuSCwFMLRpPZ1PhOvmoZbKMJfRTHiJvCFYllpuABmgx6nxZA3q2Xb4VEAzhCiZU4iRItAIBBMFz/6kbJ9eXAQ7rtPETA1EhKnihVfmZUWtTUEsLMv14zb5wlh1Es0O4qPI6uelnyiBZSAuXypuNsOe1jabM+pzgjmBsLTIhAIBJUkEICREWhvV3wrg4NKS8heWztsJutpUUVLW4OV/UMBwrFEVojc/qExFtRbJzS0OswG/OEYRr2Ud7+P22HOqbTIsszWXg8XLG8p65wFJw6i0iIQCASVQJaV/UDLl8N736t8v3AhfPnLNSdYIHN6qMxKS1QRLWsXNZBIyuwd8Gv3ReIJnu4+zqtOair0cA1nqtLiCcWos+Z+fm605256PjIaYjgQZVWHaA3NVYRoEQgEgqny3HPKfqD3vAfmzYObboIaDydT81HK9bSolZY1ixsB2Jlhxt3cM0IgmmDDiokrIXazgaQMnmCuERegyWFiwB8mmUwn9m497AHgTOFnmbMI0SIQCART4YEH4Oyz4eBB+NnPFAFz7rnVPqsJUdtCk620dLY4cJoNWRNEj+wewGLU5YxB50PdPwTkFS3nLmvCE4zxlx392m3bej2YDTqWz3OWdc6CEwchWgQCgaBcwmHYs0f5+rWvhRtvhL174f3vB13t/1mNJZIEU+Jjsp4Wu8nAivkubYJIlmUe2T3Iq5c1F1yUmInTkhYtrjyblC87dT6dLQ6+/0i3NmK99fAop7fVYdTX/u9YMD2I//ICgUBQKrKspNeuWKHsB4rHFb/KF78Iztnz6V+truh1UvmVlpRosZp0rFzgYne/j0RSZne/n6OeEJesLM0kO1GlRa+T+NSGTvYNjvHQ9j6i8SQ7+3win2WOI0SLQCAQlMK2bcp+oLe/XREod94Jhtk5gKn6WOa5LPjDsbI2PYdTFRqLUc/KBS6C0QSHhgM8snsAgAtProxoAbj81Pksb3Vy2yPd7OzzEo0nWdXeUPK5Ck48hGgRCASCiXjmGVizRllseMcd8OKLcNFF1T6rSaO2hBY2WEnKEIgmJnhEGq3SYtSzcr4LUELmNr4yyBnt9bQ4S0v4dWS2hwqIFl2q2rJ/KMDNf90NwJlicmhOI0SLQCAQ5CMaVcQJwPr1cMst0N0NH//4rK2wqKjjzuqW5HKWJqqixWYy0NnqwKCTeHzPEC8d9nBJCVNDKqVUWgBed8o8Tp7n5PmDo7Q4zcyvq87aA0FtIESLQDDHeeDFI/z5pb5qn0btIMvw0ENw6qlw8cXg9Srm2s98BhpOjNaEWmlpS0XtqyKmFFQDr9mgw2zQ09nq5P+2HgXg4hWtJT9PlmixFRYtOp3EdRs6AWXf0HRsdhbMHoRoEQjmMLIs8/WHX+GnTx+o9qnUBrt2wWWXwRvfqAiVX/0K6uqqfVYVR81oaWuwAeVVWpQEXJ2WeLtyvot4UmZhvZWTyxhFVttDkgQOU/HK1aUr5/HOde2866yOkp9fcGIyu2ucAoFgSvQcDzDkj2ASI6TQ0wNnnKFMA33ve3DNNWA8MffbqEbchan2UDljz6FoAmvGSPPKBS7+8CJsWNFSVhXEbNBj0uuwmvQTRv7rdBLfeNvpJT+34MRF/KUSCOYwm/YPAzDkj5Q1QXLCEI/DU08pXy9dCrffrvhWrrvuhBUsoIgUvU6i1aX4Q8oZew7FskXLusUNSBJcftr8ss/DYTGIxYeCshCiRSCYw2zuUURLNJEsy9dwQvDPf8KqVcoY84FUe+wjH4Hm5uqe1wzgC8VxWQy4Ujt/yonyD8USWE1p0XJ6Wz3Pf2EDZy91l30edrNeiBZBWQjRIhDMUWRZZnPPCPbUBWhoLJxzzIMv9fFvd24qy/NQKzzw4hGuue/F3Du6u+GKK+DSSyEUgt//HhYvnvHzqya+sLLvR02i9ZVRaQlHs0ULQJPDPKnzcJqN1Bcx4QoE4xGiRSCYo+wfGuP4WIRLT5kHwKA/knPM091DPHtghBse2DHr2kcbdw/w95ePZZ+3xwOrV8Ojj8I3vqEYb9/85ppfblhpfKEYLqsRi1HxlZTjaQmO87RMhRsuO5lPXdxZkecSzA2EaBEI5iiqn+WNZyhehKE8oqXfG0avk/jL9n7u3XxoWs4jlkjyy00HiSWSFX3eQ8NB4kmZQCgKf/+7cmN9vbLUcO9e+OxnwTy5CsFsxxeOa1UWl9VQtqellN1CpXBeVzNrU9uiBYJSEKJFIJijbO4ZYUGdhTUdykUjn2g55g1z0cktXLi8ma8+tJsdR7wVP4+nu4/zxT/t5Mm9QxV7TlmW6R0Osu7wy5jOORte9zrYtEm588orYX75ptETCaXSovhZnBZjWZ6WcKxylRaBoFyEaBEI5iCKn2WY9UvduKwGTAZdftHiC7OgzsJ337GKJoeJT9z3gpbxUSkODgdS/w5W7Dk9u7v5+m9v4nf33QDHh+HXv1ZSbaeZeCJJJF56JH618IZiWqXFaSm/0jLe0yIQzBRCtAgEc5DuwTGGA1HWn+RGkiSaHeYc0RKIxPGH48yrs9JgN/GDq1bT7wnznX/sqei5HEqJlUMp8TJlYjHsGy7k4n3P8d1Xv5vnH34a3vnOGfGtfOXPO7nq7men/XWmii8c0/b9uCzGsnNabEK0CKqEEC0CwRxEHXU+JzWm2uw0MzSWLVqO+ZRpInXXy5pFDbz+9Pn8cetRwrHKVRNUsXJoKpWWZBL++EdIJMBo5PkvfYeLPvxjbjv3XYzIMzed8tJhLy/2jtb0tFUkniAcS+KyqO2h6nlaBIJyEaJFIJiDbNo/zMJ6q7Ywr8WZW2k55lVEixpABnDlmnb84Tj/2DVQ9PljiSSfe2AH3QP+Cc/l0IgiVnpHJilann0WXvUqeOtblfFlYMuyNfS7lLwVTzA6ueedBIeGA8gybJ8G70+lUAVKVqWlnJyWCk4PCQTlIkSLQDDHSCZlnj0wwvqlbi12vdlpzhl57vdmV1oAXnWSm4X1Vn635XDR19ja6+HXz/Xy1x3Hih6XSMocGQkhSXBkNEgiWcZY9dGj8O//rnhVDh2C//1fxWQLHBoJaNkhnuDMVD08waiWd7LtsGdGXnMyqAJlMp6WWCJJPCkL0SKoGkK0CARzjL2DfkYCUdYvTY+aNjvNjASiWWPHA6n20LwM0aLTSbxt9UKe3necfm+o4Guo7adDI8V9Ksd8YaKJJKcvrCOWkOnzFH7OLGRZyVf5zW/gc59TRpjf9z5lySHQOxzkpGY7dpMeT4WNw4XIbG9t7R2dkdecDKqwUpNoXVYjoViipJHzUKotKIy4gmohRItAMMd44ZByQT17STp2vdmpVCWGx9KtlH5viHqbMce/8LY1bcgyPPDi0YKvoYqW3gl8Kqqf5TWdSiunaItIluEPfwC/XzHV3nEH7N4NN98MzuztwgeHgyx226m3mRidofaQ2uY6daGLbYc9NRvGp1VarGlPC5S2fygcFaJFUF2EaBEI5hh7j/lxmA20N1q125pTrZRMX8sxb5h5GX4WlUVuO2ctaeR3Ww7nvTBH4glNGB2awKeiVide09kEpMefc9i6FS64AN7+drjrLuW2deuUJYfjCETiHB+L0OG2UW8z4p2h9tCh48q5v+mMBRwfi3JktMSq0QyjTgpp4XKpf5diHtYqLaI9JKgSQrQIBHOMvQNjLGtxaH4WgJaUOMncP3TMF87ys2Ry5Zo2Dg4HNXGSybZeD5F4klXt9Qz5IwQihT/BHxoOYtRLnNnRgEmvy63MDA7Chz8Ma9Yokfs//rGygbkIarVmUUq0TLXS8pOnerjoO4/z2CuDRY87NBKkxWnmnKWKAKtVX4tXq7SkPS1ASQszg1EhWgTVRYgWgWCO0T3op6vVkXWb2h4a9I2rtBQQLZefNh+bSc/vthzJuW9TzzCSBG9f0wYUb/n0jgRob7BhMuhoa7Tmjj1/7GPw858rQqW7Gz76UdAXv2Cqz7GoUWkPTcXTsmn/MDf/dTf9njDv//nzfPq3LxWs3PQOB1nktnHyfCdmg46tvbUpWlRxko7xL7/SYhHtIUGVEKJFIJhDjASiHB+L0tWa7QFpcpiAdHsoGk9yfCzKPJc15zkA7GYDl582n4e29xGMZn9C39wzzMr5Lk5vqwOK568cGg7S4bYBsKjRpnhc/vxn6O1VDrjlFnj5Zfjud5W9QSXQmzL/drht1FuNk54eGvJH+I/7t7LYbeeZGy7ikxct4/+2HeWS7z3Bi3mMtodGAixy2zHqdZy2sI5th2vTjOsLxzDqJSxG5c+/VmkpQbSERaVFUGWEaBEI5hB7U7kpneNEi9mgp85q1ALm0pNDhRcKvuusdgLRBPc926vdFo4leLHXwzlL3SxqtANpETEedT/QokZFtKwN9PPfP7gO3vQmuPVW5aDOTli+vKyf8dBwkHqbkTqrkQabCU8wSrKcUWqUUexP3b8VXyjGHe9ZTYPdxKcvXc6frjmXpAy3P7ov6/hwLMGAL6L9LKva63m5z0c0XtklkJXAl4rwV9uDasXFV4IRV3haBNVGiBaBYA6hhr2Nbw9BKhU3VWk5pomW/JUWgDWLGnnVSW5+/EQPodQn8K29HqLxJOuXuqlLCYdClZbRYAx/JM4yYww++Uk+ft3bOPXoXsZu+bZSYZkkh4aDLHIrgqneZiQpw1i09MRXgNse6eaZ/cN89YpTOXmeS7v91IV1nN/VnDMdpLbA1KrRqo56ovEkrxzzTfrnmC584bjWEoIM0VJCG00VLSLGX1AthGgRCOYQewfGcJoNeaeCMvcPHcsTLJeP6y/p4vhYhHs3HwIUP4tOgnVLlAyYxW5bQdGijjuf/+s74I476HvHe7ngI3ex598+AMbJR+8fGgloFY96m9L28gRKbxENj0W47dFu3nLmQq5c25Zz/6qOeoYD2dNBB1OTQ6pYWtWutLJq0deiVFoM2veOMkaeVXEqYvwF1UKIFoFgDrF3wE9na/bkkEqLK71/KF+Efz7WLW7k1cua+PET+wlG42zuGeaUBXVacFmH254/YO4f/8DztLJYMH7D52HbNsK3/oBRW92UFifGEkn6PGEWpSoe9anz8IRKnyAaGosgy7BhRWve39OZqiDJmA7SJpZSYmlhvZVmp7kmJ4gylyUC6HUSDrOhJE+LCJcTVBshWgSCOUT34BidLc689zU7zAz6IsiyTL83jM2kz/pEXojrL+lkOBDl7icPsK3XwzknpUPrFjXa6POE02mr3d2KZ+W1r2XhXbcBsKCrA047jfZGK5I0tcWJR0dDJJIyHVqlRbk4j5Zhxk3v5sn/s588z4nFqGNbRhXl0HAQp8WgvZ4kSaxqr9dEiyzL3PdsLzf8YXt5qwrKZNAf5qO/3ML+obGCx/hC2aIFwFVilH9IGHEFVWbiv0gCgeCE4PhYhJFAlM48fhZQPC2hWIJANMGATxl3zldpGM+aRY28prOJ2x7tJpGUs9YDdLhtJJIyfYeOseiO78Jtt4HZDLfcwt0LL2DeIb/WajAb9Mx3WSa/OJF0mF3a05JqD5WR1aJ6O5yW/C0qQ2o6aGvGdNChEWXcOfP3taq9nn/uGmDHES9ff3g3z+xXUoL/89IuWpzFK1iTIZGU+dSvt7GpZ5hzlro5qTn/f2dvKK75WFScJS5N1EaehWgRVAlRaREI5gh7NRNugUqLM52K2+8N5fW9FOL6S7pIJGXFz7I4LVrUdkn8rruVseX3vleptvzXf3HAF9faOCodbtuU2kO9w6q3JLvS4i0jq0WrtBSpMp3Z0cDOjOmg3uGAJpS0Y1JtpCtuf5rtR7xcurIVgLESlxOWy/cf6WZTan1CsTRepT2U/bOVujQxFEtgNujQ6yYWswLBdCBEi0AwR+geUFoGpYiWAV+kYLBcPlZ3NPDaU1o55yR3ukLx5JN0bnsGgM2veyds2QI//SnMmwekqxOZLHbbp1RpOTgcxGLU0ZL6WVRPy2gZRlzV21Go0gJKFSUaT7K730c8keTIaEgTaCqnt9fTaDfx6s5m/n79ebxjbTtQmuG1XJ7cO8QPHu3m7Wva7CCiEgAAIABJREFUWNbiKChawrEE0Xgyp9LishpLzmkRfhZBNamIaJEk6XWSJO2RJGmfJEk35LnfLEnSb1L3PytJ0uKM+z6Xun2PJEmvrcT5CASCXPYO+HFaDLS68mevqKLlmC+stIfKqLQA3H7Vau75wNlw8CBceSWcfz4N3/smFqOOnrEErF6tHRuMxhnyR3KqEx1uG8fHoowVif4vxqHhIIsa7VqbxqDX4TQbyjLiqqLCWaTSok4HbTvsoc8TJp6UcwSYw2zguc9fzD0fOIuF9Vbt+Sb7sxXimDfM9b/ZRmeLg69ecSrtDVYOj+YXftreIev49lDplRbhZxFUkymLFkmS9MDtwGXASuBdkiStHHfYB4FRWZaXAd8Dbkk9diXwTuAU4HXAHannEwhmLWOROOd98zGeTZXqa4XugTG6Wp0FfSqqz+KVfh/xpDzhuPN4DKEg+i99EU4+Gf7yF7jxRqR//IOOxtyxZ/X7jnHVCTWQbrItot6RgJaVolJnKy8V1xeKYTLoivo25tdZaHGa2do7qk1HdTTac44z6NN/YssZLS6HGx/aSSiW4I53r8Zq0tPWYCtYaUlH+GcLMpeltEpLMCpEi6C6VKLSchawT5blHlmWo8D9wBXjjrkC+EXq698DF0vKX84rgPtlWY7IsnwA2Jd6PoFg1nJkNEjvSJAteZYJVgtZltmbZ+dQJvVWIwadxI6jXqB4sFxe/vY3uOkmZRPz3r3wxS+C1UpHoz0nFVfbDzROYKjf5yxOLAFZlukdCea0adRU3FLxhXONquORJIkzO5TpoEI/y3ic5tJ3/JTD1l4PrztlHstSU2FtDVa8oVheETJRpSXf1u5MwrGEMOEKqkolRMtC4HDG90dSt+U9RpblOOAF3CU+ViCYVaif6vs8hc2QlSYaT3Lrxr0FNyofH4viCcYKjjsD6HQSTQ4z24+kREsp7aFnn4Vf/Ur5+m1vg61b4d57oS0dyrbIbaN3JDguQTZlmG3MbQ9BegqoHAb9EcKxZI54UDY9l+dpKWXUe1V7AweHg2w/4sFk0E34+5qO9lAwGqffG2Zpc/r32Nag/PxH81Rb1AmhfJ6WRFLWtjgnkzI/enx/zns4FBOeFkF1qYRoyVdrHi/XCx1TymOVJ5Ckj0iStEWSpC1DQ0NlnqJAMHOon+pnUrRsOTTCrRu7efSVwbz3d08wOaTS7DRrkzZFjbhHjyqTQOvXw403QiIBkgSrVuUcushtIxxLMuhPb5A+NBykzmqkzjY+L8RIo900qayWF1KVreUZsfugjD2XOz3ktE6cyKv6Wh5++RjtDVZ0E0zU2M0p0VLB9tCBVBLvkqZ0Ba2tQamQHc4j/NT9QnV5Ki2Qbl292DvKLX97hb9s7886LhRNiAh/QVWphGg5ArRnfN8G9BU6RpIkA1AHjJT4WABkWb5LluW1siyvbW5ursBpCwTTQ7rSEp7ycx31hEoKI1Pj9wt5GfYW2TmUiWrGNeol3HZT7gGhEHzta9DVBb/7HXz+8/DCC6AvfCFTfSuZQqR3JMjiAu2UjkZbwSWLxdi4e4A6q5HVHdnboJVNz+XltJRSaTm9rQ6dpFzoxxuK82Ey6DAbdPgnUWkZ9IfzLl/sGVJ+T5mVlvbU7zvfe0GrtFhzPS2Qbl1t3K2I3+OBSNZxoVhStIcEVaUSouV5oFOSpCWSJJlQjLUPjjvmQeDq1NdvBx6VlVrxg8A7U9NFS4BO4LkKnJNAUDU8ocq0h0YDUS789uPcunHvhMemRUv+CsXewTHqrEZNlBSi2aHc3+K05K8c7NiheFUuuwx271Y8LI7iQki9oB9MmWv7PCG2HfawtED42WK3jQND5YmWRFLmsVcGuXB5c5b5FaDBZsQbipW86dkfjk3oaQGlcqJWribys6g4LcayjbiReIIN33mCu57cn3NfutKSFi0NNiM2kz6vaPEWaA+plRafJloGABgZyxZ7YTE9JKgyUxYtKY/KtcDfgd3Ab2VZ3ilJ0o2SJL0pddhPAbckSfuA/wRuSD12J/BbYBfwN+AaWZYTUz0ngaCajKY+1fsj8ZImMgqxuWeYaDzJz54+wEigeKVgokpL94Biwp0o4bYlNQ6dNTm0dSt8//vK12edBbt2we9/D0uWlPRzLKy3otdJ9A4HiSWSfPLXW0kmZT550bK8x58830WfN1xWdeTF3lFGgzE2pALcMqmzmUjKpU/t+MLxouPOmZyZquqMN/8WwmkxlO1p2dXnwxeO5zV29wyNsbDemlX9kCSJtgZrXgHrC+efjFKNub5wnEPDAfYNKpk+w+Ped8FoXIgWQVWpSE6LLMt/lWW5S5blk2RZvil125dkWX4w9XVYluUrZVleJsvyWbIs92Q89qbU45bLsvxwJc5HIKgm3gzTZ/8UWkSbe4YxGXQEYwnuerKn6LHFKi2yLLN3YIzOCfwskG4PzauzwMAAfPjDsGYNfP3r4PMpB61YUdbPYTLoWFBv4dBIkG/9fQ8vHBrlG287vWClZeV8xZOyq99X8mts3D2AQSdxXldu61gLmCtRBPnDubt5CqH6WkppD4GS3VLu9JC6KXpXX+7vo+d4IKs1pNLWYONw3vZQ/skotR3mC8W01tCSJnuOaAmJcDlBlRGJuAJBhcm8OE6lRbS5Z4SzlzTyxtMXcM+mgwyPRQoeq25nPjIayhlbHQ5E8YZiLCsgEjJpdpgxJmK8eeOvobMTfv5zuP56eOUVcLkmfHwhFjXaeWLPIHc92cN71nfwxjMWFDx25YKUaMlzkS7Exl0DrF/qzntBbrCrm54nFgvReJJwLInTXFql5XWnzudDr17C+qXuiQ8mVWkpsz2kLl0c9Ec0cQqKGD0wFGBpU65oaS9Sacm3CDLtaYmzcdcAXa0OVrXX57znwrGkEC2CqiJEi0BQYTzBmOZxODpJ0TI8FmHPgJ9zTnLzHxd3Ep6g2qJezCLxpCZgVPKZNQvR7DTTHBjlgl/9AM47D15+Gb7zHaivn/Cxxehw2/CF45y60MV/v3589mQ2TQ4zrS5zyaLl4PEA+4cCXLyiJe/9dVbFUFxKpcVfIMekEHVWI//9hpUlX8gd5vLbQ9sOe7QKWGb1aWgsgj8Sz/KzqLQ12PCH4zlTU4rJOPdnU1cWHBkN8fzBES5e0YrbbspqS8YTSaKJpGgPCaqKEC0CQYVR81AMOmnSlZZnD4wAsH6pm2UtDt50xgLu2XSI4wWqLUP+iOZDGe9rOXBc8ScU2voLKOLkC19gaZOd0LyF7PrnM/DQQ7B8+aTOfzxrOhpocpi4/arVJU2frJzvKrk9pJpGN6zI9bOAYkyF7LZdIUqJ8J8KjhLj8lWGxyL0jgS5co2Se5Mp5NJiNPe/qzr2PL7a4gvH8woyi1GHUS/xlx19xJMyG1a00OgwEYwmCKWyW8Kp6SUhWgTVRIgWwayk1EmQauAJRXHbTbS6LPR7J+dp2bR/GLtJz2kL6wD4j4s7icQT3PlE7gRJLJFkOBDVTKHjRUvPUCDlK8mTcDs8DNdeC2ecAXfcQePwMbZ+6VJOP2917rFT4G1r2nju8xtK9n6sXOBi3+AY4djEvvyNuwdY3urURn3HU29TKi2lGHu1xNgSpocmg8tiLMvToraGzu9qZmG9lZ19Xu0+dXKokKcF4PBI9nvBX2CcW5IknBYjh0dCuO0mVrU30GRXqjvDqbFnVbxYRHtIUEWEaBHMOvzhGJff9hTf/vueap9KXjzBGPU2IwvrrZNuD23uGWbt4kaMqfHdpc0O3rxqIb/cfEi7eKgMp8ZSVVPo+FCx/UMBFrtt6DNHmGMx+MEPFN/Kj34EH/847NsHixZN6nxLYaLwtUxOWVBHPClrm6kL4Q3GeP7gaMHWEKRNpqWk4k57pSXVHpooLl9l22EPep3EaW11nLIgu/rUMzSG2aBjQZ51C+2NuZUWWZYZDkRzguVU1J/5wpNb0OskGlM5Per7S33fiUqLoJoI0SKYVciyzOce2MErx/y8csxf7dPJIRRNEIknqbeZWFBvmVR7aMgfoXtwjHNOyjZ3nr+8mXAsmVPyV/0si9x23HZT3vbQ0qZxLYRwGG6+WZkMeukl+OEPwV2amXQmSE8QeYse9/jeQRJJmYsLtIZAWVroshhKSsVNh69NT6XFYTGQlNHi8idi22EPy1ud2EwGVi5wceB4gGBUEVY9QwGWNNnzisE6qxGH2ZD1Xnj5qA9vKMbqjoa8r6VWl9Q2m9uhiBbV1xKKCdEiqD5CtAhmFfduPsRD2/sx6KQpZaBMF56Q8ge+3mZkfr2VY95wTqLtvZsP8fzBkYLP8ewBZTv0+IkUreQ/XrSMKS2oFqc5J58jnkjSOxJkSbNdWWJ4zTVKlcXphC1b4B//gFNPneRPO310NNpwmA0TmnE37h6kyWHSqkyFqLeZSjTiTm+lpZz9Q8mkzLbDHlal2n6nLKhDlmF3vyLWDxQYd4bMrJa0aNm4ewCdpFRSCp2bSa/jNZ1NALhT7SHVR6WKFhHjL6gmQrQIZg07jnj56kO7uXB5Mxcsb9Y+FdcSowHlnBpsRhbUW4kn5SzzbCSe4MY/7+JXmw8VfI5N+4dxmA2cuiB7xDhd8s+upKiVlmanmbYGW9aivMOjIawBP2/+5XcVcfLLX8L27cqdCxcq+4JqEJ1OYsV8JzsnEC07j3pZt7gxu/WVhwabUVuvUIxCW5ArhcOcveOnGD3Hx/CH45og00bB+33EUmI0p4KWwXgBu3H3AKs7GrS2z3guO20+Hz1/qbYjKafSonpaRKVFUEWEaBHMCryhGJ+47wXcDhPfeccq6qymmhQtaqWlzmpiYb0yzZPpa9nZ5yOaSBZtVWzuGWbd4oacOPpmhxmzQVdQtDQ5zNqn62RShkSCyB0/4tG7P0rXr38CV18N3d1KS2gWsHK+i939voKma1mWOeoJaZMyxaizmUo04saRJHCYql9pUUPlzkyJlgV1FuqsRnb1eekdCRJPynnHnVXaGmxabk+/N8TOPl/exGCV965fxKcvTU+L2Ux6zAadFjCnmqJFTougmgjRIpgV/Pb5wxweCfHDq86k0W6izmrUNtbWEuqn+fpUpQWyA+a2pS5EhYLOBn1h9g8FcvwsoJT8F+YJDRv0R3BZDFiMetoabUQT6ayW5l/+jP2Nbfif2gR33w2thS9atcbKBS4C0QS9ebYVg1IBiMST+aeixtFgM5YULucLxXCYDWWZhsvBOW4xYTG2HfbgNBu0UXVJkhQzbp9P281ULHunrcHKWETJalFTbjcUMSyPR5IkmhxmzYgbFEZcQQ0gRItgVjDoD2M16lmzqBFQttSOReLEE7mbb6uJKloabCbm1+URLakR1kKZIZsz8lny0dZgyxljHfJHtPCxZYFBvvWXW+k/2Ad6PT/6/I/42Ie+g+vcs6fwU1WHUxYo496FWkTqFu35eaZnxqNsei5temi6xp0h3R4qJRV322EPZ7TXZwmolfNdvHLMT3dqN1Dx9lB62/MjuwdY5LYVz+rJQ6PdlB55FkZcQQ0gRItgVjCaGiNWUS8s5aaLTjeq2bPeZsRlMeAwG7SLK2SIlgKf+p/tGcZpNmgX7PHkW4Q35I/QZkzAF77A2a9/Da/f8xTBpzcBsCNqYmnLxDuHapFlLQ4MOqngBFGfVxFvC0uotNTZTPjCsRxT9Hh84di0mXAhw9Mywfs2FE3wyjF/jsF45QIXkXiSjbsHcNtN1NkKCyy1bbbnmJ9n9g+zYUXrhAszx+N2mHKnh0R7SFBFpu//ToGggijZJ2kDobaVNhTPur3aeEMxLMb0Ft3MsWc13dRq1OMJxZBlOecicng0xNJme0FjaXuDjdFgjLFIXLkAyjKrn/gz1/z9J+A5TvJdV3FR/aW85+SzeBXKhEm+JYKzAYtRz7IWR8EJIvX3uqDekvf+TBpsRmRZaf80FDCiQmpZ4jRWWjJ3/BRjx1EviaScI1pUMfvCoVHWLc4/uqzSnqq03P98L9F4smiWTSEa7SYtKyccFaJFUH1EpUUwK/AEo9q2XsjYSltjY8/KeaYvigvqrVpFQK2ynLvMTSIpE8iT1eEJRouKMPXTszYhJEms3/o4Y83z4JlnMNz3K+ILFnJkNIQ/HGPQHylp51CtsnKBq0h7KITZoCs4DZOJWqWbaOzZF4pPa6XFblYu+BO1h7YdHgXQxp1VljbbMRlSgYNFWkMAdTYjTouB5w+O4rQYWLe4sezzdafaQ7Isa5UWi0FcNgTVQ7z7BLMCTyimbeuFzEpLbYmW8W2s+XVWrT2kppueu0zJwcg3zTISiBa9CLc1WJnnO47rYx+E7m4CkTifuvx6/nznA3DOOdoxR0ZD6Zj3CS5utczK+a6c7cYqfd4wC+qtJbU8tCj/Cd4v/khs2sadQQm6sxr1jEWKn8eWg6N0NNpocpizbjfqdSxvVdp9S0oQo6qv5cLlLVq6cjm4HWbCsaSygyiWwKTX5Uy1CQQziXj3neDEasyoOlk8wai2rRfSZfZaq7R4x4mWhfUWRgJRwrGElm6qLjbM52vxjHt8FsEgXXd+j0d/8lFaHv4TbNnCoD+C32yn2ZX2dai+F1W0nDTLKy0Au/MsT+zzhEpqDQFalW6ipYn+8PRWWmDipYnJpMyzB0ZYvzR/ZeSU1O9kaZFxZxW1MjeZ1hCgCeiRQJRQNIHFKC4Zguoi3oEnMOFYgnU3beTn/zpQ7VOZErIs51zMXdZUeyhUe0bc8e0hUCY41HRTVXyNv4BG40nGInEa8rWHfv97WLEC+81f5cll6/jR7X+Gd70rK1hOpa3BxlFPiH2DY+gk6HDnXyQ4GzhlvuLhyLfxuc8Tyrt3Jx9qpaVYe0iW5WmfHgIlq6WYEXf3MSVuP9/YO8CpqSWana0TG6wXu20Y9RIXdE1OtDSlAuaOj0UIRRPCzyKoOsKIewJzzBvGE4xx26P7uHJtu5Z0OdsIRBPEkzINGaJFXfpWa5WW8W0sVbQ83T2kpZuq5z6+VaG2i/IaRZ9+Ghoa4J57+PYLsMystHzyixYrsYTMpv3DtDXYMBtm74WmzmZknsvC3nF7pqLxJIP+SEkZLYD23ik29hyMJkgk5WmvtDjNhqKels09xcfer1zbxiK3rWiwnMrHzj+J1506v+iUUTEaU1H+I4EooVgC2zSF7gkEpSIqLScwasDYSCDKPZsKx8bXOqOpkcvMCobdZEAnFR4drgZKRSi7jaVWAv768jFASTdVK0bjz300mF4BwMAAfOhD8Pjjyp033wwvvADnn6+0fzzK2POQP713SKW9UamsbD3smdUmXJXOVgd7B7NFy4AvjCyXNjkESqibJKX/n8jHdEf4qyjtocLv2037h1nsthXMnzEb9Lyms7SJMLfDzJpFxaeMij4+Y9NzKJYQEf6CqiNEywmM+im8o9HGXU/ur7lMk1JRL+6Z7SGdTsJpMdaUETcYTRBLyFnn2VpnRpLg+YMjWrppfYFP/aPBKKZ4jFPvvRM6O+Gee+Dll5U7bTbQKxeMtgarFjA3NBZBr5OyWkqqjyExQcz7bKGr1cm+wbGsOP/0uHNplRa9TmLd4kbu3XSIgymvz3ime1miitNsLPj/YiIp89yB4YJVlplG3T80nPJlWYWnRVBlxDvwBEYVLV9+40pGgzF+8czB6p7QJEkHtmW3TVxWQ8Wj/F8+6kWWiweQFUJt92S2scwGPc0OM7KMlm5qNeox6qWcSovhbw/zj59+gsXfuhEuuAB27oRrr815nfYGG95QDF84xqAvQpPDlJWamhm2trTMBNRapKvVQTiWzNpurY6RlypaAL77jjPQ6SQ+8asXtT06magCeLo9LQ5L4fbQ7n4fvnC8oJ9lprGZDFiNeoaFp0VQIwjRcgIz5Fc+hV+wvIULlzdz91M9Je08qTU8wVwxAMrFpZKVlleO+XjDD57mH7sGJvV4tY2V2R6C9IVVDQqTJIk6qwlvKNsUKu3bR1RvZPiBB+HBB5VqSx7UMdajoyGGxiJZfhZQQtnU2046ASotquF0byrkDNIR/qUacUH5vX33HWewq9/HVx/alXP/TFVaHObC00Obe4aBwn6WatBoV1Jxg9GEiPAXVB0hWk5gBv1h3HYTep3EdRu68MzSaotawRhvJnRZjBU14u5JmT13Hs0fGz8R3jyVFkj7LjLTTeusBuKDQ3DNNUobCHj28ndy+ftvw/6Gy4u+jtr+OTIaUvYOjcvyAGhPHVNKlket09miVIv2DqR9LX2eEA02Y9mf/C9e0cpHz1/Kr57t5U/bjmbdN1OeFpfFwFg0nnd79ab9wyxtstPqKs2rMxM0OUwcV9tDwogrqDJCtEyRm/6yizse31ft08jLkD9Ci0u5oJ3RXs/FJ7dw91MHiMZnV3aLJ48RF1LtoQqOPO9Pbc7N/ERfDoXaWGo1QEs3jcW46tk/8ZVPvwXuvBMOKCPpI+EkRot5QrOjKloOjwSV/8bO3AtcW4MNm0nPvBq6+E0Wp8XIgjoL3eNESzmtoUw+c+ly1i5q4HMP7ND26gBaq3EmclpkGYLjWlSKn2WEs2uoygJqpSVCSHhaBDWAeAdOkUd2D/Lk3qFqn0ZehsayP4VfsrIVbyjGgC9c5FG1hycUw27Sa/HlKpWutPQMKWJl/KRKqRRqY71n/SK++bbTlXTTJ56AM87gg7+7lT3ty+Gll+DLXwaU6aHxj81Ho92EzaSndyTIcCCa0x4CZdT1W28/o+wFebVKZ6szpz00WdFi1Ov4zGuXE4wm2H7Eo90+Y54Ws7p/KPu9u7PPiz9SO34WFbfDrE0PifaQoNoI0TJFRoPRmp3KGfJn+x3USPDhQPH9K7XGaIF9PHXWynpa1ATZQ8NBIvFco+ZEqDkr49sLi5vsvGNdu/KN1wuxGD/9zPf45NXfgFNOyXp8KcsfJUmircGqLdXLJ1pWLnDx+tPnl/0z1CpdrQ72D41pW5r7vKGStjsXQo3C784QQv5wHJNeN+1jvWolZ7wZV/OzLCl/R9B0ouwfUjwtFmHEFVQZIVqmQDIp4w3FJtzYWg2SSZnjY9mfwptSXx/Ps8ellhkfja/ishqV4LkKrCqQZZkDxwO0OM0kkjI9Q/nHYovhCcawGvXZFz2PB/7zP+FrX1O+f+MbYedOjpx3Cd5x75uRYPG9Q5m0NdjYkfLe5BMtJxqdrU4i8SS9I0F8YeX/OXUdwmRosJtocpizfDK+cExLWp5OHCnRMj4Vd9P+YU5qttNSYy09t8NENJ4kGk+KSoug6gjRMgX8kThJeeI189VgNBhVPoU7Mist6Uju2YRSackjWrRNz1P//R/zhQlGE1x6SiuQbfosFU8oo72TSCh+lc5OuPVWGBxUbpckMJmosypZHZm7oYruHRpHW4NV8ybNBdHSpU0Q+elXJ4emUGlRntPB3sHsSotzmltDoCTiqq+nEk8kef7gaE1NDamoqbiAEC2CqiNEyxRQ2wH+cGzS2R7TxaAW757+1Ka2h2abaPGEYnnbJpXc9HwgVVm5eEUrep2U1TYoFU8wSp3NBFu2wJo18LGPwYoVSpLtbbdlHVuf59xHy6i0tDek9wnlmx460VAniLoH/GUHyxWiq9XJvgG/9v+uLxTThPB0ogqjzPbQK8f8jEXiNSla1IA5AJtoDwmqjBAtU0A1XsYSMpEam8hRg+XU6SFQ8jscZgPHx2aXp8UbjGkX+Uwquel5f8rPcvI8J4vdtslVWgJRpdJiMIDfD7/9rWK8PfPMnGPrxkX5J1KtxlI8LZCeIIK5UWmxmw0srLeyd2BMC5abiqcFlPUAgWiCoykR5A/HZqTSoraHxiLp9213yvy9Yv7ESxBnGneGkBYx/oJqI0TLFMjcGDuTi/u+v7Gbx14ZLHqMtkhv3KfwJodpVlVaZFlOVVrye1qgvE3P+wbH+NwDO7LaMqBMDlmNyohwV6uT7sEyKi1+P3z+87zvnq8r57lqFezdC1deqbSD8qCOb6sZNN5QDFnOnTwqhBowZzfpZ+0izHLpanWwN1VpMeikKYu1rnFmXF84PjOeljztoQNDAWUjd2Pt5eq4M/6GiERcQbURomUKZMawz5Sv5fhYhFsf2csftx4tepy6GG78H/Ymh3lWiRZ/JE4iKWft1lFRLzDlCMbHXhnk18/18uKh0azbDxwPsKTJjiRJdLY6OTQcyBv1nkUyCb/4BSxfDl//OkQi1Kt/1PXF/7iPr7SoArh0I65SZZgLVRaVrlYnPUMBekdCtLos6HVTG+fuakn7ZCBVaTHPQKUlj2jZfzxAe6MtZ6y/FsistAhPi6Da1N7/IbOIzIV3MyVaHn1lEFnOrvLkY8gfwZbnU7giWmZPe8gTSKXhFmsPleFp8aSi8zf3jGTd3jMU0DYid7U6SMqwf6hItWX3bli/Ht73PujoQN60iesuu576PEFv+VB/Hm/qPaRtsi6xPVRvM+IwG+aUaOlsdRJNJNncMzzl1hAowrHFadbyX/wzVGnR6yTsJn1WVELPUIClNbpywWLUY0+JcSFaBNVGiJYpkCkcZmqnz8bUXpzhCYTH+IwWFbfDxPAsqrSoIiNfpUW98JdTaRlNiYRNPce12yLxBEdGg9pywfFtgyxUw3V9Pfh88MtfwjPPMLZqDfGkXPL0j+rRUc3cowWC6QohSRKnLnRp5zoX6GpV/vsM+SPaaoSpP6eT7kE/sUSSYDQxI54WUHwt6t+MZFLm4PFATS+3bEyZcUVOi6DazI1m+DSRWWkptLW1koRjCZ7qVi62E1VaBv1hWvKIliaHmdFgjFgiiVFf+5pV/R3nEwM2kx69TirL06JWNl7s9RCOJbAY9fQOB0nKaJ90F7vtGHRSthk3GIRvfxueeQYefhjmz4ddu0CnS51nOHWepVVKVD+ON3Xu6n/PfOKsEL/4wFnoTpDE21JY1pK+qM+vQKUFFDPu/c8dzkjDnZk/iU6LUau0HPNPHA12AAAgAElEQVSFCcUSLKnRSguA227m8EhITA8Jqk7tX7VqGE8wilGvXDRmoj20af8woViCk+c5GQ5Ei45ZF6q0qAFzI7MkFbfQPh9Qqg0ui6HMSksUnQTReJKtvUqEu7pzSG0PmQw6FjfZlbaBLMP998PJJytx+y4XBFLBc7r0/z6auCpx2Z5Rr8NhNmiVJLU91FCipwXAbNDPCuFZKWwmA+2NiliZ6rizSmeLk1AswSupZZkzVmnJ2PTcM+79V4uovhbRHhJUm7nzF28a8IRiWm99JqaHNu4ewGbSc/lp84nGlXJ2IQpt/21OlXmHZkkqrmpULdR2cZUZ5e8JxlizqAGdlI5N7zmutIEyP+l2tTrw7N0Pr3kNvOtd4HYr48u//S04csv4qvgotdICSnsrbcSNYdRLmndAkB/VPLuwYu0h5b/lloOKMXu6NzyrOC1p0XIg9f5b2lS77SE1q0WIFkG1EaJlCowGY9ro6XRXWmRZ5pHdg5zX2cy8VHx5oWpJOJbAF47nr7TMsv1Do0WMuKCYcb1liBZvKEZ7o41TFtSxKSVaDgwFaHaa05+yk0k6W5zsjBhIhiNw991KYNx552nPMxKIcvtj+7QJo3I9KerPpLarPMEoDTbTCbPgcLroTHl4KlZpST3fC72KaJnuDc8qTotBaw/tHwpgN+lpddWuqVpNxRWeFkG1EaJlCnhTCaZ2k37aRcvOPh/HfGEuXtFCY+rTfCHRcrzAuDOkMxdmy/4hTyiK02wo2AZxWQ1lxfiPpsTBOSe52ZbytfQcT01uRCJwyy2wejUn1xsJGSzs+r9/woc+lDPC/OvnevnW3/dw40O7AOW9AOlR5lKot6UF10ggWpafZa6yYUULZ3bUs9hdmVZKndVIq8vM1tQI/HRveFZxmA2aD67neIAlzfaaFqyvOsnNucvc2E3CBimoLkK0TAE19MxpMU779NDG3QNIElx4covmexgpYMbVguXyVlpm1/4hTzBWVAi4LKW3hyLxBMFognqrkfVLG4kmkrx4aJSeQT9v6HkOVq6EG26Ajg6WWxW/0N4CIXMbdw9g0Enc92wvf9p2VKu0qKFxpVBnNWrhcuXsHZrLrF3cyB8/cW5Fk1m7Wp3a8sKZqrQ4zOm/GQeOj7GkhltDAOd1NfOrD62fcjaOQDBVhGiZJFrsutWY1Z+eLh7ZPcjqjgaaHGbNFDdaoNKi7h1qyZMZ4jAbMBt0s6Y9pLZNCuGyGEv2E6mtmHq7iXWLG9FJ8PimPfzwf/+L937zOrBY4B//gAcfpH3FYox6KW8y7pA/wrbDHj5x4TLWLW7gcw/sYGvvKHaTvqxwsHqbUTPwlrN3SFBZOlvSY+MzVWlxWgwEoglC0QRHRkM1m9EiENQaQrRMEmVJomK8dFoM+CPTV2k55g2z46iXi1e0AOkJk0LtoWKVFkmSlIC5WdMeKl6BqLMZSx55Vqsa9VYjToPEaQvruL/bR0RvYs8XboKXXoJLLgGU6Z4lTXa68+wgeiwV8PfaU1r5wbtWYzHqeWzPUFkmXEibiGVZTm2yFqKlGqhmXEjvBZpu1IrOzj4vslzbk0MCQS0hRMskGc3ID3FajNOa0/Joas/QhhWtgJIlYdBJRUWLJBWOhG9ymLSY/1rHE4wVNOGC8rsIxRJES1hY6QnGMCTirPzd/0JnJxe5wRdN8oErv4Lpuk8piw4z6Gx1srvfnzNavnH3AAvqLKyc72JenYXv/dsqJKnwhFMh6q0moqlQM08wVpaJV1A5VDOuw2yYsfaHGuX/0hEvUNuTQwJBLSFEyyTxZISBTXd7aO+AH4fZQGcqXEuSJBrspoIBc0NjERptpoLm1dkU5T9heyglaErxFOn//jB/+9m1nHTTf0NXF+talQkUg07K2pqscn5XM0c9IR7bk15OqQb8XbyiVTNOnt/VzE1vPo2rX7W4nB9NEzlHPSHiSVm0h6pEZ6rSMlPBcpDOg9l+RMkKWiIqLQJBSQjRMknUVkNdqtJSzgRLuQz5I7S4zFnTBY02U8Eo/0LBcipNDvOsiPJPqr6hCYy4QPHffzQKr389az7+HnRykqH7fg9/+xunn3sGep1Eh9uWV+C95cyFtDdauXVjt1ZtUQP+NqxszTr2qrM7eMfa9rJ+PrWCdOC4Ei4m2kPVwWUxMr/OMmPBcpBuQ20/4qXVZdYqLwKBoDhCtEySzEqLK2OPyHQw4AvTOs5U21is0jKRaHGaGA5ESSYLJ+rWAv5wnKRc/GKubXrON0EUTf1+TCZoa2PTx2/gtR+8HetbrwBJwmE28OplTaxd1JD3uY16HZ+8sJPtR7w8sluptmzcPYDdpGf90sap/XCk03MPpkSLaA9Vj/VL3VlrAqYbVaSo28UFAkFpCNEySTJj2x1mA5F4siRfxWQY8Idzgqca7aaCE0ATiRa33UwiKWvVolpFS5kt6mnJszQxkYAf/xgWLYKdO5Xb7ryTJ990NbLRlJU6+7P3reOWt51e8PnfsnohHY02bn1kL8mkEvD3ms5mzIapj9yqra2DwynRItpDVeM7V57BD686c8ZeL7MVVcuLEgWCWkOIlkkyGowhScqFR50EmI5qiyzLDPgitLqyKy0NdmPekWdZlkuotKQC5ibZInq2Z5hX3/LotGfTaCmz9iKiRVs8mDqXRx+FM8+Ej38currG7QdSJnQy22x6nVQ01Muo1/HJi5bx8lEf33+km2O+cE5raLKoba8DWqVFiJZqoZvgfVBpMqeUxLizQFA6QrRMEm8wistiRK+TtF74dJhxfaE40XgyR4Q02s14QjES41o8vlCcaCKZd++QylQD5p7qPs6R0RB9nvCkHl8qaguurkhgm1ZpCcbgqqvg4ovB54Pf/Q4efxxWrMh4vskFuL3lzIUsdtv4/iPdSsDf8uaynyMfatvr4PEgINpDc4lMD4sYdxYISkeIlkkymnEBTFdaKi9aBvyKMBhfaWm0GZHl9IVdZWhMOb5YpUUVNJOdINqbyi6Z7iWRnoyx8kLUycoxvkhcEShf+xrs3g1vfzuM++TsCcZK3sKciUGv45MXdQKwpqNBW4UwVewmPXqdxDFfGJ00c8FmgupjNxm0t6cYdxYISkeIlkmihJ4pn5TTlZbKX8QHfPlFi+p/GG/GHSwSLKcy1f1DakrsdLeHMs3OOSST8POfY1nRxUUHX1CMuF/8InzhC2DNv0xvKgFuV6xawIYVrbz3nEWTenw+JEnSRFS9zYRORKTPGXQ6CYfJgFGff9xeIBDkZ0qiRZKkRkmS/ilJUnfq3zljGJIkrZIkaZMkSTslSdouSdK/Zdz3c0mSDkiStC31z6qpnM9M4g1GtQuOVmmJTEOlxacIi/FGXHdq6+pIIFs4DBWJ8Feptyptrcm0h8KxBIdSxtFSk2gni+ppycnP+Ne/4Oyz4f3vR1q8mEhDU0lVn4nGp4th0Ov4ydVruWLVwkk9vhB1mmgRVZa5hsNioKPRhqFAnpJAIMhlqv+33AA8IstyJ/BI6vvxBIF/l2X5FOB1wK2SJNVn3P//ZFlelfpn2xTPZ8YYzUgwdU2jp0WttIwXIao5dSSQLTyKRfir6HQSbnvhnJdi7B8aQ7XRTHelxRuKKem/mX/U/+M/4NWvhv5+uPde+Ne/6DtpZUkCStnwXFviQF0G2ShMuHOOeXUWTllQV+3TEAhmFVNNNLoCuCD19S+Ax4HPZh4gy/LejK/7JEkaBJoBzxRfu6p4MloN0zk9NOSP4LQYsJqyR2wbtf1DuZUWk0E3YbqnkopbfqWleyC9QHA6A/Ug43ccDCpZKwYDrF6ttIE++1mwKwZGl8UwYaUlHEsQjiVrLsCtLqM9JJhb3PneNRUZnRcI5hJTrbS0yrLcD5D6d0uxgyVJOgswAfszbr4p1Tb6niRJk3Y4eoMx/vv/dvDDR7sn+xQlE08k8YXj2gXHMZ1GXF84x88CaZ/HeE/LkD9Cs8M84fim22GalGjZO+DHoJMw6qVpN+KOBqK8adcTcPLJcPfdyo3vex/ceKMmWCC9eLAY6kh0rbVh1BZjY5GxbsGJSYvTUnSvlkAgyGXCSoskSRuBeXnu+kI5LyRJ0nzgl8DVsiyrKWyfA46hCJm7UKo0NxZ4/EeAjwB0dHRk3bdx1wCf/+MOBv0R2hutXJua9Jgu1AqD2mow6nVYjLppM+KO97MAWIx67CZ9TovnwHCA9saJjX3NDjM9Q4Gyz2fvwBhLmuyMBqPTum+JLVu44eYPs6Jnh5K7cnrhADiXxUi/t/j4tSru6ouMT1cDtcIiMloEAoFgYiastMiyvEGW5VPz/PMnYCAlRlRRMpjvOSRJcgF/Af5bluXNGc/dLytEgP8FzipyHnfJsrxWluW1zc1KTsZoIMp192/lQ/dsodFuYsOKVga8kZytvJVGuwBmXGicFuM0VVoiORH+KuOXJsqyzL6BMbpSW2uL0eRU2kPl/q66B/10tTqVfUvTlaj71a/CunW0Dh7hvg9/CZ5/Hs49t+DhLqshHS5XAHV8utY8LS7RHhIIBIKSmWp76EHg6tTXVwN/Gn+AJEkm4I/APbIs/27cfargkYA3Ay+X8+Kbe4Z5aHs/123o5MFrX82rl7mJJpKMFIi3rxT58kOmY9Ozmm7bkqc9BChm2oyftd8bxh+J01mKaHGYiMSTjJUx8RSKJugdCdLZ6kjtW6rgzxsOQyBV+TnnHPiv/+IN1/6UPa+/EvTF+/6ltIe0oLoaEy2iPSQQCASlM1XR8g3gEkmSuoFLUt8jSdJaSZJ+kjrmHcB5wPvyjDb/SpKkHcAOoAn4Wjkv/rpT5/HYZy7gug1dmAw65tUpF/eJWgVTxRvKX2mptMfDE4wRTSRpKTAJ1GA3ZUX5q6FvXSUsflNHpssJmNs/NIYsk660VOLnlWX44x9h5Ur4n/9RbtuwgcTXv0E/JupKqEC4LEYi8SThWKLgMWmhWVsVDWHEFQgEgtKZ0vSQLMvDwMV5bt8CfCj19b3AvQUef9FUXl+SJNobbdr38+oUL8cxb5hTF07fKOFoIL0sUcVlMZRVtSiFQmm4Ko02U9Y0j/p1qe0hUKL8S90yq4miVgcuq4FjvimKw+3b4brr4LHH4JRT4NJLtbv6PCFkWUn+nQi1xeIPx7EY81dlRmu0PeROrVRoqlDKrkAgEJzInFCpRvNTlZYpX0wnQN2O3JBVaal8e6hQsJxKo92U1QrbO+CnyWEuaVuwun9ouIwJor0DYxj1EovcdpzmKXpafvxjxWD70kvwwx/Ctm2wYYN2911P9mDQSVy8YuLlhOp4d7HKjycUxaTXYS0gaqrFazqb+dG7V7O6o37igwUCgWCOc0KJliaHWdnlMs3tIU8wik5K57MAOM3Gik8PFYrwV2mwmwjFEoSiSltk7+AYXa2l7TFR9w8NldEe6h7ws7TJgVGvw2WdhEiLRmFkRPn6/PPh2muhuxuuuUbJYEnR5wnxm+cPc+Xa9qxKWiHUSksxEeVN7YqayU2+paDXSVx22vyaOy+BQCCoRU4o0aLXSbQ4zdPuafEEY9RZjVm7Yqaj0jLoK7780J2xf0iZHPKX1BqCdDhdOfuH9g766UyJIqfFSCiWIJZITvCoFH/9K5x2Gnz0o8r3K1bA978PjY05h97+2D7+f3v3Hh9XXed//PVNZjIzuU2SNk1TaC2FUlsu5VIqCIhQkIuu4q6rrPoTWXe9C4qrgroqP3dZBBEV2J+rC6uudxQWdvHyaxEQKyK30pYW2oJA07RNm2SSZibJTGa++8c5J5k0M5OZNOlkZt7PxyOPnJ45Z+bMt6czn36/n+/na7F89Lxj8npqr8crVwK2s+7Q7BoaEhGRwpRV0AJOr8TewzA8dHDiZEPQTyyeZCTfL/E8dB0YJhzyZ83TaK4b+7LeFRkkGk+OBhWT8VVX0Vzrz7vAXCw+ws6eQZbOc4KixnwL6m3dChdfDG98o5N0e/nlOQ/v6I3xsyd28vZVCzmiKb+F5FrT8nOyicQm/p2JiEhpKbugpT0cZHff4Iy+RiTD/9q9qrjTmYybrbCcpyUtaCkkCdfT1hhkVyS/ttrR5T3/WE8L5B6S4Re/cIrC/eEPcPPNsHkzvOlNOV/n9gdfwGD4yLn59bLAWH7Ovhy9RpFYYlzitIiIlJ6yC1rmh4OjCawzJdMXYMMMlPLf2z+cNZ8FxgctY9Od8w9aTl7UzJMv9ebVO7TNDYq8GjDpM3bGGRmBzk5n+5xz4EMfcvJWrr7aWT8oh509Me56YifvOG0hC/LsZQEI+KppqvXTlStoGdTwkIhIqSu/oKUxyMDwyIyuQNybtliiJ58ZLIXq6h+asLpzupba9KBlgHkNgYKKp51x9BwODI+wZXf/pMdu33uAmuoqFs9xEmMzLhL52986Cxq++c2QSsHcufDNb8K8nEtSjfra2m1UGcOHzz067/fgaa0PTNrTolL5IiKlrfyCFm/a8wwm43ozUdJ5wyXT1dOSSlm6DgwzL8fwUDjkp8o4QZRXXr8Qpx/lJME++kL3pMdu23uAJa11+KqdW6YhPUh74QV461thzRo4cAA++1kocDbML57s4J6nd/HBc5bQHs6/l8XT2pA9aBmMJxkeSc26argiIlKYsgtavC+8marVkkimODA8MmHhPe9LfGCagpbeWJyRlKUty8whgKoqQ3NtDfsHnJyWfJNwPfMagxzdWscfX8wnaBkYtzxAoxuk1fxhvVPNdu1auP56J/H2L/+yoKBl294DfP6/NvOao1q4cs3UFrtsbQiwL0sibsStYKyeFhGR0lZ2Qcv8xpkt5e8tzNdcl6WnZXh6hofGCstlHx5yrqOGTbsiDCaSBfe0AJy+ZA6PT5LXkkim2BUZHKucm0oR3r0TgJeOOR6uvBK2bYNrr4Vg7us9WHR4hA//8CnqAtXc+jcnj/bkFCrX8FCmCsYiIlJ6yi5o8YZTDmV4yFqbdfVjbw2b8Awn4nol/LMtluhpqa1hS6eTk5JvYbl0Zxw9h4HhETZ3Zs9r8dY3am0IwPr1sHo1DReuIZgYoi9ZBTfdBAsWFPza1lr+8b8288K+Ab5x2cmTvtdcWhsCxOJJohlmb0UyrBUlIiKlp+yClqC/mpa6mkMaHvrKr5/nHf/2x4yPeasFT6zTMr1BS9doNdzca9K01NWQcuOrYwqYOeR5zVFzAHIOEe0fiLOgv4s1110FZ50Fe/ZgbrwRX13dlBOP+2IJPvmzZ7j76V1ctWYpZx4zd0rP4/GC1UwziDKtyi0iIqWn7IIWcIaIDqWnZWNHhG1dBzI+Fsmy8F7AV01NddW0zR7qcoeHslXD9XgF5uY3Bif0/uSjtSHA0nn1OZNxBzc+ywPf+RDzH/wNfOEL8Pzz8M530hjyTylIW7tlL+ff8jD3PtPJlecdw8fOm1oeS7rWeqeXJtMQ0djfmXpaRERK2SGt8jxbOQXmph60dEYG6R9MkErZcaX6wUmQBSYk4sL0lvLfe2CI5lo/AV/uBf5a3NyaQpNw052+ZA53P9VBIpnC7+WUWOsk1a5YQUfbQh4+/W385devZfGpx42e1xAsfNHEL9y7me8/+jLL2xv5j/eeNm2rcXvBXcagZXR4SD0tIiKlrCx7WtrCwSkPD6VSls6+IVIWDmTIj/AScZvqJn4BTmvQMklhOU9LnfNlPZUkXM8ZR88hGk+yaVefs+Pxx+HMM+G002D3brqjCb555t8QfvX4HpFCF03c2z/E9x99mbevOpJ7P3LmtAUskB60TPx7j8QSBHxVWZdDEBGR0lCWQUt7Y5CeaJyhRLLgc7ujceIjzkyaTL0IvbE41VWGhsDETqqG4PSt9NzVP5RXYqrX0zKVJFzPa9x6LRv/tAXe+15YvRpefBFuvRXa2uiODlNdZTIkH/sLGg57YGsXAH939hJqfNN76zWF/PiqTMZpz5FYXENDIiJloCyHh7wCc139wyxyK7jmK33dokgswcKDFiH2SvibDHVIprunZWkevSdHza2nysDKhU1Tfq059QFOq0vyjv/zBkiNwGc+4xSIa2wEnIq7zbU1E4bKGoM+dnTl/37Xbd3LwpYQS+dNPcDKpqrKMDfLtOfeDMUARUSk9JRlT4sXtExl4cTOtAUE+zL0tEQGE1krqzYEfdNSXC6VsuwbGJ505hDASQubeOLzF/Dq+Y2Fv5C18NRTAKw4cQlfff17SWzcBDfcMBqwgDN7aE5dphye/HtaBuNJ1u/Yz/nL2zIGfNNhXmMg4+yhTBWMRUSk9JRl0NLulfKfQl7LrsjYOV4CZ7pILJ61SNl0DQ/998ZOkimbV04LjC2cWJBnnoHzzoNVq2DDBs44eg53rLyETaHWCYf2ROPMqZ/4Gl5OS7aaNul+v2M/wyMpzl/eVvi15ilbgbnIYDxj4rSIiJSWsgxavC/7qUx7Tu9p8abKpsu18N6hDg/tHxjmwz98kqt+soHjj2jkkhPap/xcWe3bBx/8oLOw4aZNcPvtcPzxo4m8L+2PTjile2A4Y2DUEPSTTFkG88gdWrdlLw1BH6uPapn02KnKtv5QbywxoYKxiIiUnrLMaWkI+qkP+KY07Xl33yBHNIXYFRnMPDwUS7BsfuZck4agn4H4SMap0pN5ZPs+rvzx00SHk3zqwmV84HVLplzSPqvhYTjpJOjqgo99DL74RWhuBtIWmszQO9UdjTO3fuJQlbf+UP/gCLU1zq2USln+848v86YT25njnpNKWR54rotzjm0dm1I9A1obAnRH4yRTlmq3/a219MUShNXTIiJS8sqypwWcL+G9UxweWtJaR8BXlSVoyT4TpSHgw1oYiBfe23LL2m3UBXzcf+VZfOTcY6Y3YFm/3slfCQTg5pth40b4+tdHAxaA2hof4ZB/Qu/U8EiSA0MjWXpavCrAY+20ZXc/X7zvWT76o6dJuqV6n+mIsH9geEaHhsAJWpIpO1pLB5x8nHgyxbxJivSJiMjsV75BS+PUCsx1RpyelqZaP30HDQ/FR1JE48kcOS1TL+W/s3eQM4+em9eMobxt3QoXX+yU3v/lL519l10Gy5dnPDxTJWFvscFMOS3e+01Pxn2p2xleevTFbr6xbhvgTHWurjK8ftnEfJnp1Or27HjVhMEJogCWt08hUVlERGaV8g1awoWX8h8eSbLvwDDt4RDhkH9CIu5oZdUsia+jKz0XmIw7lHBe98jmUEHnZdXbCx//OJxwAjz6KHzta3DBBZOeNj9DUb79bt2TTLOHGt3grT8tSHu5OwbAm05s59YHd/C7bftYt3Uvq17VPOMLFnrrD6XXanm20ymYt0JBi4hIySvLnBZwZhDtGxhmJJnKe6hlb5/zZbegKUhTqGZCIq7X8zLdPS273OTfI1umIWixFs4910my/fu/hy9/GVrz6+GY3xgc7Znw9LgrPM/JmNPi9rSkDaO90h1jbn2Am962ku17B/jYj5+mbzDB5y7J3LsznTKtP7Sls58jmkJZp6mLiEjpKNuelrbGIMmUZf/AxGnL2XjBg/cld3BOS+8kqwV7QcvBw0qT2dnj9E4c2VxYIbxxHn4Y4nEwBm66yam/8q1v5R2wgNPTsn9gmEQyNbqvO+oEAJlyWhpHe5bSelp6orxqTi2hmmpuf9cpo891/oqZzWcBmNvgXOPBQctxC9TLIiJSDso2aGmfQoE5b7rzgiZneOjgoCXiJnhmS8Rd0lpPXU0192zYVdC1dvQ6r7twKkHLjh1w6aXw+tfDHXc4+y64AFauLPip2sNBrGVcgbZuN+ibWzexp8UbDkvPaXmlO8arWpz3ccy8er552clcceZijppbV/D1FKq2xkd9wDcatESHR/hzd5QVClpERMpC2QYtXq2WQmYQeUHL/HCQpoxBi/Png9fg8YRDft575mJ+uWk3z+85kPfrdvQO4q82hc1w6e93yu2vWAHr1sH118MVV+R/fgZt4Yn1bbqjcXxVhsbQxJHEoL8Kf7UZ7WkZSiTZ3T80bumE81e08cW/OG7CuTOltSEwmtPy3J4DWKt8FhGRclG2QYvXa7Ft70De53T2DTG3voagv5pwyE8snhxdPBHGEnGbc1Sg/fuzl1BX4+MbD2zL+3U7emMc0RQqrLbLZZfBjTfCu94F27fDtddCML8Kutm0ZwhaegbitNTVZCy9b4xxSvm7wV1Hbwxr4VUFrvc0nVrrA3S5geoWNwn3uGlcTVpERIqnbIOWcK2flUeG+e1zXXmf0xkZZEGTkwzr5a2k97ZEYgl8VYa6muqsz9FUW8MVZy7ml5v2sPWgpNZsdvYO5pfP8vvfQ3e3s/3lL8Of/gT/8R/QPj2Vc+c3ThxS645mrobraUyrAuzNHHrVnJkfCsqmtXGsp2XL7n7CIT8LwocWzImIyOxQtkELwPnL29iwM0LXgfyGiDojg6O9DWE3b6Uvbdqzt1rwZAv+ve+so2gI+PjGuu15ve6u3lju6c6vvOL0rJx9tlMcDuDUU+G00/J6/nyFQ36C/qpxQ2rZquF60hdNHA1aWorb0+LltHhJuDO1QKOIiBxeZR20rHErsD6YR2+LtXZcT4uXt5I+7blvMJ5XrRGvt+XXz+4ZrROSzWA8yf6BOAszfdFHo/CFL8CyZXDffU7Z/c9/ftLXnypjDO3h0LiifN3u8FA23qKJAK/0xKgP+Ka2gOM0aW0IcGBohIHhEZ7bc0D5LCIiZaSsg5bl7Q0sCAdZt3XyoKV/aIRoPMkR3vBQaOLwUG80kbVGy8Hed9YSGoKT97bsinjTnTP0tHziE84w0KWXwnPPwZe+BLUz24vR1hgYn9OSZYVnT0NgbGXrl7qjLGqpLWrPRqubzPynP3czPJLSzCERkTJS1kGLMYbzV7TxyPZ9DE2yEnH6dGfI3NMSGUzkXdU1XOvn8jMWs3br3tH6L5nsdKc7jwYtjz/uTGMG+Oxn4ZFH4Mc/hkWL8nrdQ9UeDo1WxR1KJBkYHot1qSAAABVWSURBVMlYDdfTGPLRP+j2tHTHipqEC2NBy0PP7wPguAVKwhURKRdlHbSAM0Q0lEjxhxf25zzOC1q8nJbMibjxrIXlMnnHaQuxFu5+siPrMR1uYblFQxG4/HJYvRquu855cPFiZ92gw8hbaDKVsjmr4Xoagk5PSzJl2dkbK2oSLoytP/TQ8/uo8VWxpLW41yMiItOn7IOW05e0UFdTzdotuYeIOt0hEW94yCucFjlo9lBzAUHLwpZaTl/Sws+f6sBam/GY3XsjXPnYXcw99UT4yU+c2iu33573a0y3+Y1BEklLTyw+FrTknD3kJxpPsqt3kETSFr2nxVt/6JWeGMvaGvBP52rZIiJSVGX/iR7wVfO6Y1v57XN7swYO4PS0+KvN6EyZ6ipDY9BHn1sFdyiRZDCRLHjRv7edupCXu2M8/lJvxsdX/OBbXP3Q9zBveANs2QI33ACNxcvDmJ9Wq2V0scRcOS3u0gWb3YTjYs4cAphTF8Ard6Py/SIi5aXsgxZwhoj29g+zeVf2uinOdOfxBd6aamtGh4e839mq4WZzyQnzqaup5q4ndo7t3LDBqbEC/PA1l/Ivn7wN7r4bjj66oOeeCWO1WobSelqyDw95Kz1v2uUELYuK3NNSXWVoca9XSbgiIuWlIoKWc5e1UmVg7da9WY9Jr9HiCYf8o8NDXkJutnWHsqmt8fHGE9u5f9Nuojs74QMfgFNOcYaBgOeHqul/7esKes6ZNFoVt39odN2hlnx6Wnb14a92pkwXm5eMq+nOIiLlpSKCljn1AU5Z1MwDOYOWodF8Fk9T2krPve4wUSGJuJ6/PqGNd/7+LmqWvxruvBOuugruvpvo8Ag90XjuwnKH2Zz6AL4qw56+QbqjcWqqq2gITFx3yJMetCxsrqW6kKUIZkhrQwBj4NUKWkREykpFBC0AZy9t5dnOfgbjE6c+J1OWPf1Do9OdPY0hP32x8T0tUwlaVq3/JZ9/8E42Lz4eNm2CW26B5ubRqdCzKWiprnIWbtzdN0T3wHDWdYc8jW7Ccm8sUfQkXM+K9kZOXthEfY5gS0RESk/FfKp7s0oig3FCNeODhK4DQyRTlvam8cNDTWnDQ145/7wTcbdudcrvX3gh5j3v4b96qvn4/rn8rnURXsWVjl5nunPGarhF5E17DvqqcybhwljQAsVdcyjdZy5aRip7zrWIiJSoiulpacpQLM7TGXGmOy8IZx4estbSO5rTMklPS08PXHklnHCCMwyUSoHPx+q/ezvGwE8ef2X00I6DC8vNEvPDQXb3DbE/mruEPzjF5TyLZknwZYyZFcNUIiIyvSonaHF7SLzclHTd7tReL4HTEw75SaYsA8MjRGIJaqqrCPmzrPA8MuLUV1m61Pn9/vc7qzJXOU28oCnEJce38/1HXybiXsPOnhgBX9VoQbTZYn5jiD3u8FCuxRKBcUMws2V4SEREylMFBS1uhdsMPS3Z8lWaQt5KzwkisTjhXCs8P/wwfPSjcNJJzpTmf/1XmDt33CFXnb+UaHyE7zzyIuD0tBzRHJp1qxC3h4PE4kl29w1N2tPiq66irsYJ5BS0iIjITKqYoKV5tKdlYtDS4/Z8HDyduTFtSCljNdwdO+BHP3K216yB3/0O1q1zhoYyOLatgTee0M53179ETzROR+8gRzbPvi/6NnfaczJlJ81pAad6sDHMyvciIiLlo2KCFq8XJTI4cXioNxanxldFbU11xnP6BxP0xuKjPS/098OnPw0rVjh5KzEnoZazz4ZJek2uWrOUWCLJdx55kY7eGAtnWT4LMK5eTa4S/p7GkI/2xiDBbENnIiIi06Bigpagv5qAryrz8FDU6UU5eJhmLNBJ0DeYoDlYBXfc4eStfPWr8O53w8aNUJt/D8PStgb+4sQFfHf9S/TGErOyd8Krigu5q+F62sMhls1vmMlLEhERqZwpz+AM/2RKxO2JxTNWug0fNDy0xvTA1e+HM86A+++HVaumdB1XrjmG/97YCcy+mUMwNj0cclfD9dzyjpOYXVk5IiJSjiqmpwWcnpNMU54jWYKWplANR/R1sfBHd9Abi5NYugweewweeWTKAQvAMfMaePPKBcDsDFoCvmrmusHK3Dx6WlrqamjOYxhJRETkUFRUT0u2oKU3luDYtvrxO6NRgv/yLzzw7zdRVV1N+H3foql26SEFK+muufjVtDUGOf6I8LQ833RrawyyfyCeV0+LiIjI4XBIPS3GmBZjzFpjzHb3d3OW45LGmA3uz31p+48yxjzmnv9TY8yMfkM2hWoyJ+JG42OVblMp+MEPYNkyzD//Mw+tOJNr/vmndDXMGUvEnQbt4RCfvWQ5/urZ2dnVHg4S8I1NZxYRESm2Q/3GvAZ4wFq7FHjA/XMmg9bak9yfN6ft/wpwi3t+L/C+Q7yenJrr/BOmPKdSlshgghYvaOnpgQ9/GNrbYf16br78izxtnSTTSavhlpGTFzWzcmHTrKshIyIiletQg5a3AN9zt78HXJrvicb5NjwP+PlUzp+KcKiGvphTlt9zYGiEuX37OPdn3wJrnYJwjz7q5K689rWEQ3529jhTmsMVFLR85Nxj+NkHzij2ZYiIiIw61KClzVq7G8D9PS/LcUFjzBPGmD8aY7zAZA4QsdaOuH/uAI44xOvJqanWTzyZYjDhrvQ8OEjqn/6JB7/zAU76z391pi8DHHfcaPn9plo/I+7qe5mSdUVEROTwmDQR1xizDpif4aHPFfA6i6y1ncaYJcBvjTGbgP4Mx2Vdm9cY837g/QCLFi3KdlhO3vBObzRO7b33wKc+RfPLL/OrY19L0223cMbKlRPO8ariwsQy/yIiInL4TBq0WGvPz/aYMWavMabdWrvbGNMOdGV5jk7394vGmIeAk4FfAE3GGJ/b23Ik0JnjOr4NfBtg1apVWYObXMLeWkKRAY749KchHOap7/6cD20Ncs+ypRnPSU++nc5EXBERESnMoQ4P3Qdc7m5fDtx78AHGmGZjTMDdngucCWyxTmLJg8Dbcp0/bbq6OPHW6wkkhulNVsEDD8CTT/Ln418DZB/68XpXAr4qQppJIyIiUjSHGrTcAFxgjNkOXOD+GWPMKmPMv7vHLAeeMMY8gxOk3GCt3eI+9hngamPMDpwclzsO8Xomisfh5pth6VLa7/x/rO541qnVsmQJ+HyjFXKzFUfzquJqaEhERKS4Dqm4nLW2G1iTYf8TwN+5238AMi57bK19EVh9KNeQ4+Lgf/4HPvlJ2L4dLrmE7uuu55Gfd3BRWq2W3lic6ipDYzBzU3jBipJwRUREimt2VjabLl/5ClRXw69+BfffT/3K4wHGVcXtjSVoCk1cLNHjJeKGQ+ppERERKabyClp6euDqq2HPHjAG7rrLmcZ80UWAs9JzyF9NJG3RxN5oPOe6OU0aHhIREZkVyiNoGRmB226DY46Bb3zDSbIFp6qtf3yw0Vw7vipubyyes9Kt18Oi4SEREZHiKv2gZe1aWLkSPvYxOPlk2LAB3vWurIeHa2vGDQ9FYomcAYm3JlElVcMVERGZjUp/lefvfheGhuCee+Atb3GGhXJoCvnHDQ/1ROOsPLIp6/HhkJ/l7Y2clOMYERERmXmlH7TceivU1UEgkNfhzXV+tu0dAMBaSySWoKkuey9KdZXhV1edPS2XKiIiIlNX+kFLS0tBh4dDY8NDsXiSeDI1tsKziIiIzFqln9NSoOZaZ3jIWktP1C0sp6BFRERk1qu4oMVbtTkaT472uGg6s4iIyOxXgUGL06vSG42PlvBvyVGnRURERGaHygta3LorfYOJ0aClScNDIiIis17lBS1ugBKJJegdzWnR8JCIiMhsV3FBixeg9Mbi9MQSGKN1hUREREpBxQUtXmXbyGCCSCxOY9CPr7rimkFERKTkVNy3dVPIHR6KxumNJTQ0JCIiUiIqLmip8VVRV1NNZDAx6QrPIiIiMntUXNACTjJubyzurvCsoEVERKQUVGjQ4qcvlnDWHdLwkIiISEmo2KAlMpigJxrXukMiIiIlokKDlhr29A0xmEgqp0VERKREVGbQEvLT2TfobGt4SEREpCRUZtBS68daZ1vDQyIiIqWhIoOW9BlDWndIRESkNFRk0JJetr+5TsNDIiIipaAig5b0nhYND4mIiJSGigxa0pNvNTwkIiJSGio6aKkP+KjxVWQTiIiIlJyK/Mb2elc03VlERKR0VGTQ4iXiat0hERGR0lGRQYu/uoqGgE/VcEVEREpIRQYtAO1NQRaEg8W+DBEREcmTr9gXUCx3vvc06gMV+/ZFRERKTsV+ax/ZXFvsSxAREZECVOzwkIiIiJQWBS0iIiJSEhS0iIiISElQ0CIiIiIlQUGLiIiIlAQFLSIiIlISFLSIiIhISVDQIiIiIiVBQYuIiIiUBAUtIiIiUhIUtIiIiEhJUNAiIiIiJUFBi4iIiJQEBS0iIiJSEhS0iIiISElQ0CIiIiIlQUGLiIiIlAQFLSIiIlISFLSIiIhISVDQIiIiIiXBWGuLfQ0FM8bsA152/zgX2F/EyykFaqPc1D65qX1yU/tMTm2Um9oHXmWtbZ3soJIMWtIZY56w1q4q9nXMZmqj3NQ+ual9clP7TE5tlJvaJ38aHhIREZGSoKBFRERESkI5BC3fLvYFlAC1UW5qn9zUPrmpfSanNspN7ZOnks9pERERkcpQDj0tIiIiUgFmZdBijLnTGNNljNmctu8kY8wfjTEbjDFPGGNWu/vDxpj/NsY8Y4x51hhzRdo5lxtjtrs/lxfjvcyELO2z0hjzqDFmk9sejWmPXWuM2WGMed4Yc2Ha/ovcfTuMMdcc7vcxUwppH2PMBcaYJ939Txpjzks751R3/w5jzDeNMaYY72cmFHoPuY8vMsYMGGP+IW1fxd9D7mMnuo896z4edPeX5T1U4L8xvzHme+7+rcaYa9POKdf7Z6Ex5kH3/T5rjLnK3d9ijFnrfietNcY0u/uNe3/sMMZsNMackvZcZfk9NmXW2ln3A7wOOAXYnLbv/wMXu9uXAA+5258FvuJutwI9QA3QArzo/m52t5uL/d5msH0eB85xt/8W+LK7vQJ4BggARwEvANXuzwvAEre9ngFWFPu9FaF9TgYWuNvHA7vSzvkTcAZggF959185/BTSRmmP/wK4C/gH98+6h5xtH7ARWOn+eQ5QXc73UIHt807gJ+52LfASsLjM75924BR3uwHY5n4W3whc4+6/hrHvrkvc+8MApwOPufvL9ntsqj+zsqfFWvs7nOBj3G7A+59NGOhM29/g/g+m3j1vBLgQWGut7bHW9gJrgYtm+toPhyztswz4nbu9Fvgrd/stOB8Yw9baPwM7gNXuzw5r7YvW2jjwE/fYkldI+1hrn7bWevfSs0DQGBMwxrQDjdbaR63z6fF94NKZv/rDo8B7CGPMpTgfmM+mHa97yPEGYKO19hn33G5rbbKc76EC28cCdcYYHxAC4kA/5X3/7LbWPuVuHwC2AkfgvL/vuYd9j7H74S3A963jj0CTe/+U7ffYVM3KoCWLjwM3GWN2Al8FvC7G24DlOEHMJuAqa20K5wbZmXZ+h7uvXG0G3uxu/zWw0N3O1g5qn4n+CnjaWjuM0xYdaY+Ve/tAljYyxtQBnwGuO+h43UOOYwFrjPmNMeYpY8yn3f2Vdg9la5+fA1FgN/AK8FVrbQ8Vcv8YYxbj9Og+BrRZa3eDE9gA89zD9Dmdp1IKWj4EfMJauxD4BHCHu/9CYAOwADgJuM0dS800dlzOU6X+FviIMeZJnO7IuLs/WzuofdIYY44DvgJ8wNuV4TnKuX0gextdB9xirR046PhKa6Ns7eMDzgLe5f5+qzFmDWofr31WA0mcz+ijgE8aY5ZQAe1jjKnHGVb9uLW2P9ehGfZV4uf0pHzFvoACXA5c5W7fBfy7u30FcIPb/brDGPNn4NU4Eenr084/EnjosFxpEVhrn8PppsYYcyzwRvehDsb3KhzJ2NBatv1lJ0f7YIw5ErgHeI+19gV3dwdOm3jKun0gZxu9BnibMeZGoAlIGWOGgCfRPQTOvfKwtXa/+9gvcfI9fkAF3UM52uedwK+ttQmgyxizHliF04NQtvePMcaPE7D80Fp7t7t7rzGm3Vq72x3+6XL3Z/ucrqjvsXyUUk9LJ3COu30esN3dfgVYA2CMacMZV30R+A3wBmNMs5uh/QZ3X1kyxsxzf1cBnwe+5T50H3CZm6dxFLAUJznwcWCpMeYoY0wNcJl7bFnK1j7GmCbgfuBaa+1673i36/aAMeZ0N1/qPcC9h/3CD6NsbWStPdtau9hauxj4OnC9tfY2dA95/8Z+A5xojKl18zbOAbZU2j2Uo31eAc5zZ8jU4SSaPkcZ3z/u3/cdwFZr7dfSHroP5z/guL/vTdv/HreNTgf63Punor7H8lLsTOBMP8CPccY/EziR5vtwul2fxMkwfww41T12Ac7Mok04Y6rvTnuev8VJPN0BXFHs9zXD7XMVTob6NuAG3MKB7vGfw8nSf5602Qs4Gevb3Mc+V+z3VYz2wflwjeIMMXo/89zHVrn31As4uVOmWO+p2PdQ2nlfwp09pHto3L+xd+MkKW8GbkzbX5b3UIH/xupxesefBbYAn6qA++csnGGcjWmfK5fgzCx7AOc/3Q8ALe7xBrjdbYdNwKq05yrL77Gp/qgiroiIiJSEUhoeEhERkQqmoEVERERKgoIWERERKQkKWkRERKQkKGgRERGRkqCgRUQOK7cWxe+NMRen7Xu7MebXxbwuEZn9NOVZRA47Y8zxOLU7TsZZ7XcDcJEdq0g8lef0WWtHpukSRWQWUtAiIkXhLgsQBeqAA9baLxtjLgc+AtQAfwA+aq1NGWO+jVMaPwT81Fr7f93n6AD+DWfl269ba+8qwlsRkcOklNYeEpHych3wFM7Ceqvc3pe3Aq+11o64gcplwI+Aa6y1PW6Z/AeNMT+31m5xnydqrT2zGG9ARA4vBS0iUhTW2qgx5qfAgLV22BhzPnAa8ISzdAshnEX1AP7GGPM+nM+sBcAKnJLwAD89vFcuIsWioEVEiinl/oCz/sqd1tp/TD/AGLMUZ12b1dbaiDHmB0Aw7ZDoYblSESk6zR4SkdliHfB2Y8xcAGPMHGPMIqAROAD0G2PagQuLeI0iUkTqaRGRWcFau8kYcx2wzhhThbOC8AeBJ3CGgjYDLwLri3eVIlJMmj0kIiIiJUHDQyIiIlISFLSIiIhISVDQIiIiIiVBQYuIiIiUBAUtIiIiUhIUtIiIiEhJUNAiIiIiJUFBi4iIiJSE/wUf+/EZfdkrzQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 648x648 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "ax = df.plot.line(y='J-D', figsize=(9, 9), legend=None)\n",
    "ax.plot(series0.index, y_pred0, 'r--')\n",
    "ax.plot(series1.index, y_pred1, 'r')\n",
    "ax.set_xlim([1879, 2018])\n",
    "\n",
    "leg = f\"\"\"Trend in ℃/century (R²)\n",
    "Full: {reg0.coef_[0]*100:0.2f} ({var0:0.2f})\n",
    "30-year: {reg1.coef_[0]*100:0.2f} ({var1:0.2f})\n",
    "\"\"\"\n",
    "\n",
    "ax.text(0.10, 0.75, leg, transform=ax.transAxes);"
   ]
  }
 ],
 "metadata": {
  "_draft": {
   "nbviewer_url": "https://gist.github.com/d1a8ea95faf3ff48a3114593f05a547e"
  },
  "gist": {
   "data": {
    "description": "gistemp.ipynb",
    "public": true
   },
   "id": "d1a8ea95faf3ff48a3114593f05a547e"
  },
  "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}