Basic_Pursuit_denoising.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import numpy as np\n",
    "import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "import scipy.optimize as spopt\n",
    "import scipy.fftpack as spfft\n",
    "import scipy.ndimage as spimg\n",
    "import scipy as sp\n",
    "import matplotlib.image as img\n",
    "import matplotlib.cm as cm\n",
    "import cvxpy as cvx\n",
    "import time\n",
    "start_time = time.time()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def dct2(x):\n",
    "        return spfft.dct(spfft.dct(x.T, norm='ortho', axis=0).T, norm='ortho', axis=0)\n",
    "def idct2(x):\n",
    "        return spfft.idct(spfft.idct(x.T, norm='ortho', axis=0).T, norm='ortho', axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(50, 50, 3)\n",
      "3\n",
      "3\n"
     ]
    }
   ],
   "source": [
    "pic=img.imread('itu.jpg')\n",
    "print(pic.shape)\n",
    "pic_zoom=pic\n",
    "s=0.5 #sampling data ratio\n",
    "row,col,channel=pic_zoom.shape\n",
    "print(channel)\n",
    "print(channel)\n",
    "k=round(row*col*s)\n",
    "ri = np.random.choice(row *  col, k, replace=False)\n",
    "Xa=np.zeros(pic.shape)\n",
    "masks = np.zeros(pic.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "ECOS 2.0.7 - (C) embotech GmbH, Zurich Switzerland, 2012-15. Web: www.embotech.com/ECOS\n",
      "\n",
      "It     pcost       dcost      gap   pres   dres    k/t    mu     step   sigma     IR    |   BT\n",
      " 0  +0.000e+00  -2.913e-01  +5e+05  1e+00  1e+00  1e+00  9e+01    ---    ---    1  1  - |  -  - \n",
      " 1  +1.076e+05  +1.077e+05  +5e+05  9e+01  1e+02  2e+02  9e+01  0.0008  1e+00   2  2  2 |  0  0\n",
      " 2  +2.086e+02  +2.828e+03  +5e+05  1e+00  2e+00  3e+03  9e+01  0.0015  1e+00   2  1  1 |  0  0\n",
      " 3  +8.251e+02  +5.765e+03  +5e+05  2e+00  2e+00  5e+03  9e+01  0.0224  7e-01   2  2  2 |  0  0\n",
      " 4  +9.784e+02  +5.270e+03  +5e+05  2e+00  2e+00  4e+03  9e+01  0.0035  1e+00   2  1  1 |  0  0\n",
      " 5  +2.863e+03  +1.792e+04  +4e+05  2e+00  2e+00  2e+04  9e+01  0.1044  8e-01   2  1  2 |  0  0\n",
      " 6  +3.435e+03  +1.201e+04  +4e+05  1e+00  1e+00  9e+03  9e+01  0.0823  9e-01   1  1  1 |  0  0\n",
      " 7  +3.128e+03  +8.884e+03  +4e+05  7e-01  9e-01  6e+03  8e+01  0.0999  6e-01   1  1  1 |  0  0\n",
      " 8  +1.634e+03  +7.009e+03  +4e+05  4e-01  7e-01  5e+03  8e+01  0.0701  9e-01   2  1  1 |  0  0\n",
      " 9  +1.579e+03  +5.115e+03  +4e+05  2e-01  3e-01  4e+03  8e+01  0.1567  5e-01   2  1  1 |  0  0\n",
      "10  +1.659e+03  +4.565e+03  +4e+05  1e-01  3e-01  3e+03  8e+01  0.0041  1e+00   2  1  1 |  0  0\n",
      "11  +2.835e+03  +8.962e+03  +4e+05  8e-01  3e-01  6e+03  8e+01  0.0925  9e-01   1  1  1 |  0  0\n",
      "12  +6.589e+03  +7.737e+03  +3e+05  3e-01  2e-01  1e+03  7e+01  0.3826  7e-01   1  1  1 |  0  0\n",
      "13  +5.152e+03  +5.734e+03  +3e+05  2e-01  2e-01  6e+02  7e+01  0.1354  7e-01   2  1  1 |  0  0\n",
      "14  +7.289e+03  +7.660e+03  +2e+05  1e-01  1e-01  4e+02  5e+01  0.5736  5e-01   2  1  1 |  0  0\n",
      "15  +6.248e+03  +6.468e+03  +2e+05  9e-02  6e-02  2e+02  4e+01  0.2632  5e-01   2  1  1 |  0  0\n",
      "16  +6.037e+03  +6.141e+03  +2e+05  6e-02  5e-02  1e+02  3e+01  0.3684  6e-01   2  1  1 |  0  0\n",
      "17  +5.934e+03  +6.024e+03  +2e+05  5e-02  4e-02  9e+01  3e+01  0.1148  6e-01   1  1  1 |  0  0\n",
      "18  +6.059e+03  +6.122e+03  +2e+05  4e-02  4e-02  6e+01  3e+01  0.4525  9e-01   1  1  1 |  0  0\n",
      "19  +5.482e+03  +5.492e+03  +5e+04  7e-03  7e-03  1e+01  1e+01  0.7328  1e-01   1  1  1 |  0  0\n",
      "20  +4.626e+03  +4.630e+03  +3e+04  2e-03  3e-03  4e+00  5e+00  0.6851  2e-01   2  1  1 |  0  0\n",
      "21  +4.309e+03  +4.310e+03  +1e+04  7e-04  8e-04  9e-01  2e+00  0.7885  2e-01   2  1  1 |  0  0\n",
      "22  +4.195e+03  +4.195e+03  +4e+03  3e-04  3e-04  3e-01  8e-01  0.6741  1e-01   1  1  1 |  0  0\n",
      "23  +4.151e+03  +4.151e+03  +2e+03  1e-04  2e-04  1e-01  4e-01  0.6119  2e-01   1  1  1 |  0  0\n",
      "24  +4.123e+03  +4.123e+03  +7e+02  5e-05  5e-05  3e-02  1e-01  0.8036  2e-01   1  1  1 |  0  0\n",
      "25  +4.113e+03  +4.113e+03  +3e+02  2e-05  2e-05  6e-03  6e-02  0.8588  3e-01   1  1  1 |  0  0\n",
      "26  +4.108e+03  +4.108e+03  +1e+02  8e-06  9e-06  2e-03  2e-02  0.6613  1e-01   2  1  1 |  0  0\n",
      "27  +4.106e+03  +4.106e+03  +4e+01  3e-06  3e-06  7e-04  8e-03  0.7614  1e-01   1  1  1 |  0  0\n",
      "28  +4.105e+03  +4.105e+03  +1e+01  1e-06  1e-06  3e-04  3e-03  0.6622  5e-02   1  1  1 |  0  0\n",
      "29  +4.105e+03  +4.105e+03  +4e+00  3e-07  3e-07  6e-05  7e-04  0.8070  8e-02   1  1  1 |  0  0\n",
      "30  +4.105e+03  +4.105e+03  +8e-01  5e-08  6e-08  1e-05  2e-04  0.8613  7e-02   2  1  1 |  0  0\n",
      "31  +4.105e+03  +4.105e+03  +9e-02  6e-09  7e-09  1e-06  2e-05  0.8876  1e-02   1  1  1 |  0  0\n",
      "32  +4.105e+03  +4.105e+03  +2e-02  2e-09  2e-09  4e-07  5e-06  0.8257  1e-01   1  1  1 |  0  0\n",
      "33  +4.105e+03  +4.105e+03  +5e-03  4e-10  4e-10  7e-08  1e-06  0.8420  7e-02   2  1  1 |  0  0\n",
      "34  +4.105e+03  +4.105e+03  +9e-04  6e-11  7e-11  1e-08  2e-07  0.9553  1e-01   2  1  1 |  0  0\n",
      "35  +4.105e+03  +4.105e+03  +2e-05  1e-12  1e-12  2e-10  3e-09  0.9798  1e-04   2  1  1 |  0  0\n",
      "\n",
      "OPTIMAL (within feastol=1.5e-12, reltol=4.2e-09, abstol=1.7e-05).\n",
      "Runtime: 315.838547 seconds.\n",
      "\n",
      "\n",
      "ECOS 2.0.7 - (C) embotech GmbH, Zurich Switzerland, 2012-15. Web: www.embotech.com/ECOS\n",
      "\n",
      "It     pcost       dcost      gap   pres   dres    k/t    mu     step   sigma     IR    |   BT\n",
      " 0  +0.000e+00  -2.913e-01  +4e+05  1e+00  1e+00  1e+00  9e+01    ---    ---    1  1  - |  -  - \n",
      " 1  +1.017e+05  +1.018e+05  +4e+05  9e+01  1e+02  2e+02  9e+01  0.0008  1e+00   2  2  1 |  0  0\n",
      " 2  +2.219e+02  +2.692e+03  +4e+05  1e+00  2e+00  2e+03  9e+01  0.0014  1e+00   2  1  1 |  0  0\n",
      " 3  +7.883e+02  +5.410e+03  +4e+05  2e+00  2e+00  5e+03  9e+01  0.0216  7e-01   2  2  1 |  0  0\n",
      " 4  +9.294e+02  +4.944e+03  +4e+05  2e+00  2e+00  4e+03  9e+01  0.0035  1e+00   2  1  1 |  0  0\n",
      " 5  +2.653e+03  +1.662e+04  +4e+05  2e+00  2e+00  1e+04  8e+01  0.1025  8e-01   2  1  2 |  0  0\n",
      " 6  +3.181e+03  +1.131e+04  +4e+05  1e+00  1e+00  8e+03  8e+01  0.0784  9e-01   2  1  1 |  0  0\n",
      " 7  +2.905e+03  +8.360e+03  +4e+05  7e-01  9e-01  5e+03  8e+01  0.1007  6e-01   1  1  1 |  0  0\n",
      " 8  +1.549e+03  +6.717e+03  +4e+05  4e-01  8e-01  5e+03  8e+01  0.0643  9e-01   2  1  1 |  0  0\n",
      " 9  +1.458e+03  +4.640e+03  +4e+05  2e-01  3e-01  3e+03  7e+01  0.1502  5e-01   2  1  1 |  0  0\n",
      "10  +1.569e+03  +4.143e+03  +4e+05  1e-01  3e-01  3e+03  7e+01  0.0285  9e-01   2  1  1 |  0  0\n",
      "11  +3.348e+03  +1.019e+04  +4e+05  8e-01  3e-01  7e+03  7e+01  0.0542  9e-01   2  1  1 |  0  0\n",
      "12  +4.624e+03  +9.113e+03  +4e+05  5e-01  3e-01  4e+03  7e+01  0.0523  9e-01   2  1  1 |  0  0\n",
      "13  +7.863e+03  +1.009e+04  +3e+05  4e-01  3e-01  2e+03  6e+01  0.4833  5e-01   2  1  1 |  0  0\n",
      "14  +8.875e+03  +1.203e+04  +2e+05  2e-01  9e-02  3e+03  4e+01  0.5346  5e-01   2  1  1 |  0  0\n",
      "15  +7.047e+03  +8.550e+03  +2e+05  8e-02  5e-02  2e+03  3e+01  0.3386  4e-01   2  1  1 |  0  0\n",
      "16  +6.515e+03  +7.679e+03  +1e+05  6e-02  4e-02  1e+03  3e+01  0.2357  6e-01   1  1  1 |  0  0\n",
      "17  +5.589e+03  +6.112e+03  +1e+05  2e-02  2e-02  5e+02  2e+01  0.5070  5e-01   1  1  1 |  0  0\n",
      "18  +5.371e+03  +5.790e+03  +9e+04  2e-02  2e-02  4e+02  2e+01  0.1889  4e-01   1  1  1 |  0  0\n",
      "19  +4.568e+03  +4.630e+03  +4e+04  5e-03  5e-03  6e+01  9e+00  0.9539  4e-01   2  1  1 |  0  0\n",
      "20  +4.238e+03  +4.265e+03  +2e+04  2e-03  2e-03  3e+01  5e+00  0.5222  1e-01   1  1  1 |  0  0\n",
      "21  +4.044e+03  +4.050e+03  +9e+03  8e-04  8e-04  6e+00  2e+00  0.7846  2e-01   2  1  1 |  0  0\n",
      "22  +3.938e+03  +3.940e+03  +4e+03  3e-04  3e-04  2e+00  7e-01  0.6923  1e-01   2  1  1 |  0  0\n",
      "23  +3.897e+03  +3.897e+03  +1e+03  1e-04  1e-04  6e-01  3e-01  0.7028  1e-01   3  1  1 |  0  0\n",
      "24  +3.880e+03  +3.881e+03  +7e+02  5e-05  6e-05  2e-01  1e-01  0.5819  1e-01   2  1  1 |  0  0\n",
      "25  +3.872e+03  +3.872e+03  +3e+02  2e-05  3e-05  7e-02  6e-02  0.6956  2e-01   2  1  1 |  0  0\n",
      "26  +3.868e+03  +3.868e+03  +1e+02  1e-05  1e-05  3e-02  3e-02  0.6287  9e-02   2  1  1 |  0  0\n",
      "27  +3.865e+03  +3.865e+03  +4e+01  3e-06  3e-06  6e-03  8e-03  0.7901  1e-01   2  1  1 |  0  0\n",
      "28  +3.865e+03  +3.865e+03  +1e+01  8e-07  1e-06  1e-03  2e-03  0.7634  8e-02   2  1  1 |  0  0\n",
      "29  +3.865e+03  +3.865e+03  +4e+00  3e-07  4e-07  5e-04  9e-04  0.6931  8e-02   1  1  1 |  0  0\n",
      "30  +3.864e+03  +3.864e+03  +1e+00  7e-08  8e-08  8e-05  2e-04  0.8437  9e-02   2  1  1 |  0  0\n",
      "31  +3.864e+03  +3.864e+03  +2e-01  2e-08  2e-08  2e-05  4e-05  0.7954  1e-02   2  1  1 |  0  0\n",
      "32  +3.864e+03  +3.864e+03  +2e-02  1e-09  2e-09  1e-06  4e-06  0.9332  2e-02   2  1  1 |  0  0\n",
      "33  +3.864e+03  +3.864e+03  +1e-03  8e-11  9e-11  6e-08  2e-07  0.9457  2e-03   2  1  1 |  0  0\n",
      "34  +3.864e+03  +3.864e+03  +2e-04  2e-11  2e-11  1e-08  4e-08  0.8224  4e-02   2  1  1 |  0  0\n",
      "35  +3.864e+03  +3.864e+03  +1e-04  9e-12  1e-11  4e-09  2e-08  0.6476  2e-01   2  1  1 |  0  0\n",
      "36  +3.864e+03  +3.864e+03  +1e-05  1e-12  1e-12  3e-10  3e-09  0.9792  1e-01   1  1  1 |  0  0\n",
      "\n",
      "OPTIMAL (within feastol=1.3e-12, reltol=3.9e-09, abstol=1.5e-05).\n",
      "Runtime: 228.512578 seconds.\n",
      "\n",
      "\n",
      "ECOS 2.0.7 - (C) embotech GmbH, Zurich Switzerland, 2012-15. Web: www.embotech.com/ECOS\n",
      "\n",
      "It     pcost       dcost      gap   pres   dres    k/t    mu     step   sigma     IR    |   BT\n",
      " 0  +0.000e+00  -2.913e-01  +7e+05  1e+00  1e+00  1e+00  1e+02    ---    ---    1  1  - |  -  - \n",
      " 1  +1.595e+05  +1.596e+05  +7e+05  9e+01  1e+02  2e+02  1e+02  0.0009  1e+00   2  2  1 |  0  0\n",
      " 2  -7.001e+00  +3.941e+03  +7e+05  1e+00  1e+00  4e+03  1e+02  0.0025  1e+00   3  1  1 |  0  0\n",
      " 3  +1.288e+03  +9.649e+03  +7e+05  2e+00  2e+00  8e+03  1e+02  0.0386  8e-01   2  2  2 |  0  0\n",
      " 4  +1.593e+03  +8.809e+03  +7e+05  2e+00  2e+00  7e+03  1e+02  0.0057  1e+00   1  1  1 |  0  0\n",
      " 5  +5.344e+03  +3.149e+04  +7e+05  2e+00  2e+00  3e+04  1e+02  0.1488  8e-01   1  1  1 |  0  0\n",
      " 6  +2.339e+03  +6.441e+03  +6e+05  5e-01  1e+00  4e+03  1e+02  0.0869  9e-01   2  1  1 |  0  0\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 7  +9.272e+02  +4.303e+03  +6e+05  9e-02  3e-01  3e+03  1e+02  0.1637  5e-01   2  1  1 |  0  0\n",
      " 8  +2.045e+02  +3.074e+03  +6e+05  4e-02  2e-01  3e+03  1e+02  0.1937  8e-01   3  1  1 |  0  0\n",
      " 9  +2.019e+03  +6.037e+03  +5e+05  2e-01  1e-01  4e+03  1e+02  0.4066  9e-01   2  1  1 |  0  0\n",
      "10  +3.665e+03  +5.414e+03  +4e+05  6e-02  6e-02  2e+03  8e+01  0.3360  2e-01   2  1  1 |  0  0\n",
      "11  +4.432e+03  +4.985e+03  +2e+05  1e-02  2e-02  6e+02  4e+01  0.5480  2e-01   2  1  1 |  0  0\n",
      "12  +4.519e+03  +5.086e+03  +2e+05  2e-02  1e-02  6e+02  4e+01  0.0973  8e-01   1  1  1 |  0  0\n",
      "13  +4.178e+03  +4.518e+03  +2e+05  1e-02  1e-02  3e+02  4e+01  0.4318  8e-01   3  1  1 |  0  0\n",
      "14  +4.165e+03  +4.496e+03  +2e+05  1e-02  1e-02  3e+02  4e+01  0.0240  6e-01   2  1  1 |  0  0\n",
      "15  +4.027e+03  +4.209e+03  +2e+05  7e-03  9e-03  2e+02  3e+01  0.5682  7e-01   1  1  1 |  0  0\n",
      "16  +4.009e+03  +4.110e+03  +1e+05  5e-03  7e-03  1e+02  3e+01  0.5049  7e-01   3  1  1 |  0  0\n",
      "17  +3.752e+03  +3.773e+03  +4e+04  1e-03  2e-03  2e+01  9e+00  0.7334  9e-02   2  1  1 |  0  0\n",
      "18  +3.680e+03  +3.688e+03  +2e+04  5e-04  6e-04  8e+00  4e+00  0.6429  2e-01   2  1  1 |  0  0\n",
      "19  +3.624e+03  +3.627e+03  +9e+03  2e-04  3e-04  3e+00  2e+00  0.6530  1e-01   2  1  1 |  0  0\n",
      "20  +3.594e+03  +3.595e+03  +4e+03  9e-05  1e-04  9e-01  7e-01  0.6832  1e-01   2  1  1 |  0  0\n",
      "21  +3.581e+03  +3.581e+03  +2e+03  3e-05  4e-05  2e-01  3e-01  0.7383  2e-01   2  1  1 |  0  0\n",
      "22  +3.575e+03  +3.575e+03  +7e+02  2e-05  2e-05  7e-02  1e-01  0.7103  3e-01   2  1  1 |  0  0\n",
      "23  +3.573e+03  +3.573e+03  +4e+02  1e-05  1e-05  2e-02  9e-02  0.6629  4e-01   2  1  1 |  0  0\n",
      "24  +3.570e+03  +3.570e+03  +1e+02  3e-06  4e-06  7e-03  3e-02  0.6960  1e-02   2  1  1 |  0  0\n",
      "25  +3.569e+03  +3.569e+03  +4e+01  8e-07  1e-06  9e-04  7e-03  0.8913  2e-01   2  1  1 |  0  0\n",
      "26  +3.569e+03  +3.569e+03  +2e+01  4e-07  5e-07  3e-04  3e-03  0.6935  2e-01   1  1  1 |  0  0\n",
      "27  +3.569e+03  +3.569e+03  +4e+00  8e-08  1e-07  5e-05  7e-04  0.8981  1e-01   1  1  1 |  0  0\n",
      "28  +3.569e+03  +3.569e+03  +9e-01  2e-08  2e-08  1e-05  2e-04  0.8015  4e-02   1  1  1 |  0  0\n",
      "29  +3.569e+03  +3.569e+03  +8e-02  2e-09  2e-09  8e-07  2e-05  0.9432  4e-02   2  1  1 |  0  0\n",
      "30  +3.569e+03  +3.569e+03  +8e-03  2e-10  2e-10  7e-08  2e-06  0.9298  3e-02   1  1  1 |  0  0\n",
      "31  +3.569e+03  +3.569e+03  +1e-04  3e-12  3e-12  9e-10  2e-08  0.9869  9e-04   2  1  1 |  0  0\n",
      "32  +3.569e+03  +3.569e+03  +5e-06  7e-11  4e-13  4e-11  9e-10  0.9571  1e-03   2  1  1 |  0  0\n",
      "\n",
      "OPTIMAL (within feastol=6.7e-11, reltol=1.3e-09, abstol=4.7e-06).\n",
      "Runtime: 197.641891 seconds.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for j in range(channel):\n",
    "    xx=pic_zoom[:,:,j].squeeze()\n",
    "    \n",
    "    X = 255 * np.ones(xx.shape)\n",
    "    X.T.flat[ri] = xx.T.flat[ri]\n",
    "    masks[:,:,j]=X\n",
    "    b = xx.T.flat[ri].astype(float)\n",
    "    A = np.kron(spfft.idct(np.identity(col), norm='ortho', axis=0),spfft.idct(np.identity(row), norm='ortho', axis=0))\n",
    "    \n",
    "    A = A[ri,:] # same as phi times kron\n",
    "    lamda=0.1 #penalty factor\n",
    "    vx = cvx.Variable(row * col)\n",
    "    cost=(0.5)*(cvx.norm2(b-(A @vx))**2)+(lamda*cvx.norm(vx, 1))\n",
    "    objective =cvx.Minimize(cost)\n",
    "    prob = cvx.Problem(objective)\n",
    "    result = prob.solve(verbose=True)\n",
    "    \n",
    "    \n",
    "    x=np.array(vx.value).squeeze()\n",
    "    x=x.reshape(col,row).T\n",
    "    Xa[:,:,j] = idct2(x)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--- 757.8510985374451 seconds ---\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABZUAAAHWCAYAAAAPTmd7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nOzdd5xcV3n/8e8zs129uMhVbmAsq8uSiwADDoSEFpIApgRsAzYYQjEESEgMP2oSA6YYbONGcIAQAgESQjDFgLtXVrNkGwtb7pZkdWm1dc7vj3M3e72eec7dnd1V4fN+vea1s/Oce8659dx7brMQggAAAAAAAAAAKKK0tysAAAAAAAAAANh/0KkMAAAAAAAAACiMTmUAAAAAAAAAQGF0KgMAAAAAAAAACqNTGQAAAAAAAABQGJ3KAAAAAAAAAIDC6FQGRomZzTSzkH2u29v12RuYBgAA/GEys4/l9gHO3Nv1AQAA+zYz+2Ru32Hp3q4P0uhUxgHPzGab2cfN7GYze9TMOs1sq5ndZ2bfMrM3mVnr3q4nAACjwcxuzO2gpz7rh5DvH5vZv5nZQ1nbujFra99nZuMKDF82swvNbJmZdZjZDjP7rZm9tmD5x5vZHjOrmNlpResNAMBoSLSvO83sQTP7oZm9rUg7idFlZguyE6AfM7M5e7s+o83MXp0b34l7uz44MDTs7QoAo8XMZkj6nKTXSbJB4WZJkyU9S9LZkj5tZh8OIfzr2NYSAID9i5k1S7pWsf3MOyj7nC7pQjN7dQhhVY08TNK/S/qzQaGlkpaa2aIQwgcTVfmapBZJV4QQbh3iaAAAMJbGZ5+Zkl4h6aNm9jrar71qgaSLs+/rJFXdZzmAvFrSG7LvV0nasRfrggMEnco4IJnZLEn/I+nI7KduST+T9EtJT0hqk/RsxQ3r8ZKOkHS9mc2X9MEQQqi3DiGE9XpmZzYAAHvT4E7cwToK5PENSf1XE2+WdKWk1ZKmS3qjpMWSjpP0UzNbEkJ4pEoeb8/V5VZJ1yl2EF+oeML3A2b2sxDCDdUqYGZvkHSWpCclfbhAnQEAGEuD29uJkuZLepOkaZKOkvQTM5sXQnhorCsHACOBTmUccMzsEEk/l3Ro9tNtkt4SQrivStqPSHqXpEskNUq6SNJ2SZ8Ym9oCADB2Qgj/Wc/wZvZKDXQoPyzpuSGEh3PxyxSvfjlH0gxJn5f0l1Wyekf2d6Wk54cQerLh/0XS7xSveH6npGd0KpvZlCxfSXpfCGFbPeMEAMBIq9He/ouZfUbSryWdqHjn7EclvW0s6wYAI4VnKuNA9A0NdCjfKumsah3KkhRCqIQQvqR4C2//1ckXm9npo19NAAD2Ox/LfX9HvkNZiu2q4tXG/b//hZmdnE9jZm2S+p9deEV/h3I2/DbFdlySTq1Rh3+SdLCk/w0hfGc4IwEAwN4QQtgo6QO5n16xt+oCAPWiUxkHlOwNoS/J/u2Q9IYQwu7UcCGE/1C8fVeSynr6QXN/3v/3oqPs/7KZvcXMbjCzx8ysN/+CIzObmXsxw3WJeh9iZv9kZmvNbLeZbTGzO8zsouzgW2a23nuJkpldlytvZvbbS8zsP7MXFHaZ2eNm9u9mtiQ1TczsRDP7oJn9yMweyF6i1GVmT5jZT83snWbWksoHAHBgMLMTJM3L/r0/hPCTaulCCHskfT3302sGJZmigcdDPVgliweyv9Oq1OEMSedJ2qN4JfOoMLMzzewbFl/qu8vMus3sSTO728y+Z2ZvNbNDawx7UBa/Pku/w8x6zOwpM7vNzD5Ra9hB+fS36Tdm/082s4+a2Qoz257tK9xiZm8ws9KgYeeY2dVmdr/FlxluNLPvm9nCRJmD93UazOwdZnaTmW3K8vqdmX3RzI708hoKM2sys/OyfY5HLL74cZuZrTKzz/Xv1wDAAeK3ue8Hm9mk1ABZ2/LRbHu8IWuXNpnZb8zsA2Y2vmjhZnaWmV1lZvdk29qerJ34TdZGPScx/Glm9vWsPdiVHb+uy45Hz0wM25Br336e/TYuO+5cltVnd9Z+fsrMJhcYnxea2Tdz9ek/Zr07O/Y9z+LdzP3p35q1c/l9lW/aM1+uuG5QOdfnYkdkv/2Fmf04a7t6zKx3cDnZ542JcRhK2kYze3O2P7I+m15dZvawmf2Xmb3HzA4aXG8NPE9Zkh6pMr5XOWW+1OJ+0TqLL53sn+fXWsEL8iz6KzP7pcV9oo4sj6+lljnsw0IIfPgcMB9J31W84jhIumyIwx4mqSc3/EmD4jfmYlMVdwbCoM/6XPqZud+vc8p9gaQtVfLq/6xSfDb0+sFlDMrnutwwx0r6qpNnn6TznDr9lTNs/rNO0nOcfApNAz58+PDhM3qffPtVZz7vym3Tv5xIuyiX9o5BsSm52BuqDPvRLLZt0O+Nku7OYh8ZpWlVUjzILNIGXlpl+GMl9RYYdpekVyfq0p/2RkknKXa218rvGkmWDXeBnr4/k//0SHpVkWUlm083OWXukPQnTl4fy6U9M7GseOMWJHVJOn9vr0t8+PDh433y261EuuZB27hDEunPk7QzsZ18XNLiRD6HSPpFgTaqp8bwjZKuLjD8tyS11MijIZfu54rvN7rHyesBSUfVyKuctX9F2uxLcsO9teAw6waVd30udrykH1YZprdGOW9MzJtCaRXfW5FqM4OkG2rU2/tcVWOZ+VWBYa+Q1ODUe3w2v2sNv0fx7vFP5n5burfXaT7pD89UxgHDzEzSi3I//ctQhg8hPG5mv9DAlc5nSVpbI/n1im+oXynp24odvhM1cDtvIWZ2kqT/UnxxoCQtz/J+VHED/lpJZ0j6Nw3tGeifVNwo/05xOqyTNEHxxYQvVTxo/qqZ3RxCuLfK8G2KG/Jlkn4j6T5JWxXH8eisXs9SfBHT/1h8wQTPtASAfZyZ/bfi286nKR6gPqJ4kvTqEMKKxOD5x1gsS6RdoXgCsyzpJDOzkB1VhBC2mtkTis9c/lNJ/zpo2Jdlfwe3wR+UNEvSGsV3IYyGdyse2EnSNsU2ebni+xbaFE+WLlE8IVxNk+I4P6B44H63pE2K7e5RivsWL5Q0TtJ3zGxpCOGORJ0mKR64Hq24P3CDYqf0AsVHjYxTfIb1b81sh6SvSdqgeKC9WvEFiH8h6U8U9yWuzdr/TYlyr1HcB1mr+EiShxQfL3Z2Ng0mSPp+Ng7tibyqMrPTFA8y+/eDfqH4ouVHsnqfpniiu03S5WbWFUK4bjhlAcA+ZFbue5ekjbUSmtlFGmjzOiV9T/GE32bFF+S+VNLLFdvUX5jZKdWO78zsYEm3K7YlUmzjvi2pXfEk4TTFFwm+TAOPkhzsXzXwnoQ9im3DrZIqkk6RdK5i5+HZkiaZ2cv62/4aJkv6iWIH7X9K+qniMedxiu9eOFLSMYoXT72wyvDvVWz/lA33TcXj8/42+xhVb7NvUHyR4lmK7agkXar4rOu8XU7dv6Q47ddl5f5OsT1e6gxTFzN7vuI06r9beJ3iRXX3SupWvEhuiWJ7b7lBv6C43LxX0vOz396quAzlrR9U3nTF+XtM9tMaSf8u6X7FeX6y4vQ/TPEFzGUN7EMN9j0N9NXsUDw5sUzxRMWZii96vlZxPwD7k73dq82Hz0h9JD1HA2e1OiU1DiOPi3N5fGdQ7EY9/WzaFySVnLxm5tJeVyNN/mrny6rlp6df6RNU7ErloNjIP+NsoaQv5tJ8tUZesyQd44xbSfFZYP35XDzcacCHDx8+fEb3U6X9qvW5RlKrk88vc2nPLFDuQ7n0RwyKXZr93ifpI4rPSD5C0pdzw7wnl/5YxcdaVSSdMYrTqv9K6G2SnuWkmyhpXpXfp6bqp3hwuysr51dOusFX8JxVJc3SbJoExUeJPKV4ADi5Str81WV/U3BZ+VcN2p9SPFD9p1yaVcqukh6ULr//8ozlRbFT+uEsvkvSS2vU6fjcsrRL0vS9vU7x4cOHT7VPfvuZSPfvubQ3OumWaODul7WSjquR7hUauEPl5hppfpYr838lTamRrqQqd7QoPjqhf/jHJZ1YJc0xGri7NqjKHSZ6+pXK/cftz9j+K3aY5/NaUCXNvVlsi6Tjnek4SdLcKr8XvpI4Sz/4it9vS2py0o/YlcqKdw89mUvzadW4Mlixc/vFifof4dUnS//jXPoP12jrJ+jpVyD/UZU0b8nFH5R0dJU0z5O0e9D05Url/eDDM5VxIDki9/3BkHvxzxDkX+h3uJNumaSLQnwh0bCY2SINnMlcI+mvq+UXQviYnnnWNOVeSW8LIfRWiX1U8eBUGrgqe3CZa0IID9bKPMQXHF6ieBWzJL1piPUDAIytzYq3o16keAXR6yX9raRbcmnOkfRjM6t1Z0z+uYZPFSyz2rBSvKPmMcWD108rXln7iOIjNiTpLkmX59J/VVKrpK+HEG4uUPZwHZ/9/XUI4Xe1EoUQdoQqV3aHELak6hdC+JWkz2X/nmnFnk388RDCz6vkdZPiFVdSPJE7XtJrQvW7hy5WPEiTarT/g6xXfFTW0/anQjz6+5Bi57UkzZb04gL5DfY2xavQpPjSx/+pliiEsE4DV6KNU7waCgD2K2Y2wcyeZ2Y/Urx7pN9nncEuVrz6s1PSy0IIv6+WKITwI0n/nP17upktHlT28yT9UfbvfZL+LISwtUZelRDCf1YJfSj3/c2hytXQ2fHj2Rpoa/7GBj3zv4qPV9v+hxCe0tOnTbV267js76+ytqKqEML2EMLKRD2G6iFJ54YQukc431rerXgnsyR9M4TwtzWO9RVC2B1C+Fk9hWXLUP+dY1eGED6btf+Dy9op6XWKd79J0vuqZHdR7vsbQwgPVcnnN4r7pdjP0KmMA8nU3PfhPoohP9wzXhCUc1k9HcqZV+a+fyWE0Oek/eIQ8/5arQYu2/D336Z6jNX3sr3+zojjsttjAAD7no9ImhFCeEMI4fMhhO+EEL4dQvhMCOEMxUcjdWRpX6SnHzjm5V8C1Fmg3D257xPygexg8bmKV8Y+LaR4K+dZIYQuSTKzsxUPJjfk62Zmr7D4Yrkd2Qtj7jSzc7PHYQ1X/3Q4wcwa68gnJd+Zv7hmqqhPT+9gHyzfif3jEMIj1RKFEB7VwK2tJ6UqqLivU3U+ZweWX8j99GcF8hus/4T0E3rmI1AGl/dLxSvjpOF1YAPAmBr8EjTFW/5/rfioin7vDyH8tMbw0yX9cfbv90MID1RLl3N97vvg7WT+BW2fCCF0aAjM7HjFE4iStDyEcEOttCGEWzVw4dGxGnjBbzW9iieNa/ll7nu1dqt/P+NZzgnx0XJ1iC8mHiv987BP8SKx0Za/aMx95Fi2T9e/HJ+Z33+y+JLn/sen3Z448X6l4nqC/QjPVAaeruiB6G/TSZIW5b7/KpH2xiHmfVsi/lj21xSvHnuyWiIzO0vxzOMpis+CnKB4tryaw1XsyjUAwBjKDvC8+A/M7G0a6Nj7oJld0t+pW2uwEajXg5JeYGZHS3q24mMcVoUQ/u/Zkhbf+v757N/39V+Ba2Yf0sAVTB2KB6aLFB/xcIrisxiH4wZJr1F8pNbPzewSST8f6oGjmc2S9GbFZxKfoHjrbVON5EfU+L3ffTWuPO63Ifc99XzmDYq3J09JpJPSzzXMx08pkN//MbNJGngPxROSXlHgXED/sy15QzyA/d1ySX8VQrjbSbNUA8em3Wb2qkSezbnvg7eT/XfHBsX3+QxV/uRnkStgf6aBZ/cuUbz7qJp7QgjbnXwey32v1m7doHhi/GRJN5jZ5xTb7CInvus1Ev0BhWTPw35W9u/KEMLDY1Dsc7O/nZJmZfs1nv6O5FbFO6fuz/7P7x+4+xUhhD1mdosGTqZgP0CnMg4kW3LfB99mW9Sk3PfBD67Pe8yJFXVY7rt75jnElxptU/HxSnXu5jsKnnGlcnaw910N7WqgiUNICwDYh4QQvmVm/6DYuTtJsTP0l4OS5V9Y01og23yanbUSZbdBPuNWyMxnFV8Y9LMQwrclycwWKj4yQ5I+Jen/KXYqX6D4foILzOx/a9y+m/IhxYPvwxSf7/c8SV1m1q54RfAvJf2y1iO2squkP6v43oGidwSm2k9vf0R6epteNG2zmyqqeSuxFB/1kds3OcxLW8WRGpg+CyT9YAjDFukQB4C9LX8HR39H2xsU310zX9K7zOydzt2vM3Pf35J9ihq8new/efl4ohO3lhm57zUfDVUjzYyaqeo8ZpX0N5JOV9xPODP7dJnZnYpt9i8UH41R9TERdRqJ/oCi8o/lvGeMypyZ/W3R0Npo6enLX37/wN2vGEIa7EPoVMaB5NHc95lm1jiM5yo/K/e9ZkMxQre6jMv+9has524V71Su99Ec31N8G64UOwJ+LGmF4tVE/S9KkuJVzK/Nvte6ghkAsH+4UbFTWdnfwZ3KRR8RVS3NkB9LZWanKj4/t1PSO3Ohdyl2SN4j6e9zz/j7qpm9UvGE6F8rvkl+SEII681svqS/V3wT+WTFDtgzss/fSNpoZp+R9MUqzxf82yyNFG9R/bnioy4eVmzH+9v7kyV9Ivueaj+H0qbX2/7nFbk9un/fZHwq4SCT0klqGs3HkgDAiKh2YtPMPqv4WMN3Szpf8aKoWs+RrWc7OfjOmP5HUO0anLCg/COsdhdIny9nQs1UdbZZIYTfm9k8Sf+g2GE/SbHNXpp9PiRpg5l9SvFxk3XfZZUzlo++yJ98Hu48rKfMocovf/n9g6L7FdiP0KmMA8m9ig3zVMUzavOVvg10sNNy30fzRUDSwAazoWAH+LhEfERkL3Lo71BeqfgG10010p4xFnUCAIyJ/FWu1a4G/Z2kF2TfZ8p5iWz2bMP+K2t2a4hX9GTDX6F46+8nBr2c6HnZ3x9UOUD8nmKn8ulm1jCcq5Oyx2+828zer/hIjdMVD05fqHiQdbDi84RPUu6lcWbWqvj8aimekH1BCGFZjfEbzsuEx1qbnCvMM/37JkM9yM2nvy6EcE7NlABwgAghBDN7n+Ix5yJJHzKzH4YQbq+SPL+dfGMIwX32fMJOxfZrqCcA88P3K3JMmi8n1Y7UJYSwQdKFZvZexUct5NvsCYovt/uS4iNB3lkrn32Ad3dT/jnDw52HQ7U7K2tdCOGEOvLJL8dtBdKPSZ8HRg4v6sMBIzuwzD+n50210lZjZjM00JkqxauLRtPjue/HegnNbIqG/0iPocpPg7+r1aGcOXq0KwMAGDOpK4vzz35cVCWeN08DV+CuHcbVQRcpPnN3jQbeaN+vv7P6UT1T/2/NKnY1dU0hhJ4Qwq0hhM+FEP5M0kGSztXA7bhvM7PZuUFO08DB0BW1OpQz+0P7ebwXNLOpGtg3edxLW0X+JEPqOY0AcMDIXs7+/uzfkmq/BG0kt5P9beNh2WMOh+qJ3PciHYz5NENtH4Yla7NvCSFcEkJ4lWKb/TZJ/S+vf4eZjfUz+fOP76j1XoV+3kvv8/s7YzUO/cvf0WZWT0dvfv67+xVDSIN9CJ3KONB8Kff9nOzlP0V9VANX798QQhjt5xW1576/oGaq6MxRrMdgh+S+/75WIjNr0tjWCwAwup6f+17tmYn/m/v+kkRe+ZesVH2zfS1mNlPxVtYg6QLnTp5qBzmjdgVPCKE7hHCtpC/nfs7fsVOo/cykpt++4IVDiN85lIyzN8Wvzf5daGZHDmV4ANifhRB+q4EXtS81sz+pkix/N9CrrMDbTB39L5UzSS8bxvD5u3//qED6/Ht5hnrn8IgIIXSFEK6S9NXcz4Pvss0/fqOe6VtL/gR96t0DS2oFsou87sv+nWtmRw2zPkMZ3/7lr1HDW2b65ee/u19hZi2KV5pjP0KnMg4oIYSbNHDQO07S9WaWvM0iewZj/5vi+yR9bFQq+HQ/zH2/0My8Zyq+Z7Qrk5N/1tFxTrp3KJ4BBgDs58zsbEknZv/ulHTT4DQhhPsV31gvSSeY2Utr5NWieHVQv+8OsTqXKd4ieXXWrg/Wf9XLSVVi/b91Kf3SuuFan/uef5RcofbTzBaovgO0sfJOM/Ne6Pe+3PfvDyP/b2R/S5I+M4zhAWB/lt/uXTw4GEJ4QtIN2b/P0dBe1DfY9bnvf589rqmwEMI6SauyfxeaWc3OQTNbooGT1A8ovpdnb1qf+z748a/5RzOMxmMX1ua+e9PseEnVTizk9c/DsqRPDrM+Qxnfb+S+X1ykT6WabNlZnf17qpmd5iR/m+p7ljP2AjqVcSB6s6Qns+9LJd1gZlVv0zGzkpldqHjA23+27uMhhFtGu5IhhHYNHLSfLOlLZvaMddLMPqanXz022vJXG/1DtQNKM3u54tvtAQD7MDP76+wAz0vzKklX5X76XAihs0byj+e+f23w1TJZO3aZpP7fvxdCyD82I1Xf1ygeWG3UwAvvBuu/4uq1ZvZ/VwdnBzznZf/eOtTnKZvZDDO7xMyOcdK0Ke5n9FuZ+56/A+mt1fLJ9kf+Q/vHPvixkr6ePd/6/1j0aQ1cTbRKAx0fQ3GZpIey728wsy9kd0FVZWYTs+X5rFppAGB/EUK4QVL/Y5IW17ha+e808ILXr2YngGsys6PN7HNm9rTHPw268OrZkn6QPV6xWh5mZq+oEvrH3Pd/MbNnDU6Q3SX8bQ0cV/9TCGEkXyCbL+sIM/sn787k7LENf5X7aeWgJA/mvi8YyfpJ8UWCku7P/n1BtXlsZgcrvg8i9b6zyzTQx/EmM/v04PY5l2ebmVW7orzw+Gb9If0XwT1H0g/NrOYFZWbWYGavNrMLqoQ/l/t+fbW7k7J3NXGCeT/Ei/pwwAkhbMg2oj+RdKTiQc/dZvZTxduMnlC8AurZkl6tpz/z6fMa/pm/4ThfsRO3TfHFAaeZ2fWKz006RNJrFW/TuVXxAP1wjeyb3av5geIzlA6XtFjSWjO7WvFM82TFg/2XK16R9X3FaQgA2De9UNIXzew+xfcOrFG8gtcUX7b3cj39VsNfydmpDyH80Mz+TbF9OlrSMjO7QvF5y9MUD94WZ8mf0MBzI5PMbKKkS7N/3x9C2Foj6VcV35vQJunXZvYFSZ2K7Wj/gcqXawzraVZ8lvNFZnanYuf1PYq3r05S3G94vQae6fxb5a7oDiE8Zmb97eJkSSuzabNKsRP5dMXp0yLpX/T0A9190X8qTuf5ZvYNSQ8r7pucrYEXG3dJOm8Yz8xWCGF3dkLj14pXJr1X0mvM7LuK02yH4kuWjlFcpl6gOI+G9M4MANiHfUaxQ1GKd8r+JB8MIdxpZu+SdLli2/EtM/uApB8pPmapS7G9eY7ixVSnZIMOfheBFNucOxTb7pdI+r2ZfVuxY3uH4svu5yjuFxyuQX1FIYRvZdvsv8ziy83sOkm3KR6fnqL43oEJ2SA/kXTlUCbGELVI+qCkD+Ta7Hs10GafqNhm9z924lchhFsH5bFScZ9omqS3mNlmxWm0J4vvzh5VUo9LFF88LMXO/KsV9x1MsWP3HMU28N8Vp21VIYStZvY6xZMDzYovBv7LbJ/sXsVnRx+q2F6+TPFE9+ATvvn3T12SnZi/X1L/SfhHQghrcmneIulmxbvAzpL0oJl9T3GeP6U4Dw7LxuPFii95vkKDhBC+kZ0QeYniCevVZnaVpLsUH69xpqQ3Kt4x/hOlr9rGviSEwIfPAfmRNEPSdxQbuZD4PCrpTYn8buxPX7D8mbn8r3PSvUDSVqduqxUPkh/N/l9ZI5/rcsPMTNTNTSvpVElbnDptVdzYfyz325nDnQZ8+PDhw2d0Poodg6k2MGRt5RWS2grk2ax4JZKX3zpJc4dY18uyYW8okPZip+yvD3NaHV1wWgVJv5Q0rUoe0xQ7RGsN16f4Doczc799rEZ9+uM3Jur9llzatyTS3tifNhVX7Ki4yRmXHZL+1CnL3UfIpXu24oFlkeneKemP9/Z6xYcPHz7VPvntVcH0JcUOwf7hqm5TJb1S8Q6eItvJTZKm1MjnUEm/KZBHd43hGyVdU2D470hqqZFHQy7dzxPTp2ZaxcdMFW2zf+5Mk3c6w60blPb6XOyIgvPYBg1XrV07R9Jbc7+90cnvdMWTvKlx/lmN4b/rDHNVlfSTFDu8i0zniqR/qFHueMVO7VrDdkh6neIFfv2/Ld3b6zSf9Gd/uPUOGJYQwhMhhNdJmivpE4pn1B5XPIu3XfGs3L8p3sZ6Qgjhm3upnr9SPJN6ieID+PconmFtl/QBxYf2P6p49liKnb2jXafbFKfbVxTPgvdPs7sVb32aG0L4Se0cAAD7iIsUn1F3jWK78rDijnu34gHqTYqPMzoxhHB+CKGjVkb9Qnz5zdmSXqp4oPGI4tVSTyneWfN+xXZi8G2mNZnZYkkXKB5cvSORXCGEj0v6c8Urk3Yptp3tigdlby9a7qA8H1K8K+gcxSuJVyu2fX2K0+x+xc70l4cQXhhCeMYzm7PfTlW8gmh5NlyHYlt6raTTQwhjeUfUsIUQtime+L5Q0i2KV3N1KY7LlyXNCiH89wiUc5+khYqdJt9QfEnkDsXpvk3xSrJ/Uew8nxFCGNKLHwFgXxXioyHyj5W4uEa6HyperPNOST9WbHf3aKAtv0XxhfUvk3RYqHGnTwjhyRDC87J01ys+DqFD8REbGxTvHLlY1d9ZoBBCTwjhXMWOzWsUTyB3ZHV5UHFb/YIQwutC7cdojYgQHy1xtOLV0d9UPE7tb7N3K7Yl35L0shDCWc40+arixVI/VDzm7hrhegbFO2zeoHg32NasjPWKjx5bFOJLgIvmd4vindbnK17V29+/0an4SKkfK7bbr62RxdlZ/NeK+23uo8JCCNtDCH+p2E5/SfEZ2VsUp/MuxX2jHyi+Z+G4EML/q5HPLsWrnd+Slb01q/PvFS9qWBhC+E5q/LHvsbiMA9iXmdlsDbwc4UshhLF8cR8AAPgDYGY3KnuPQwgh9WZ4AAAA/AHjSmVg//Cu3Pdf7bVaAAAAAAAA4A8encrAXmZmzzWzmuuimV2ogVt5H5P0X2NSMQAAAAAAAKCKhnQSAKPsWkktZvY/is9f3KT4EoTjJP2ZpPm5tBeEENznHgEAAAAAAACjiU5lYN9wuOLLhWrZI+ltIQSuUgYAAAAAAMBeRacysPe9RvENvC+SdKSkaZLaFHIqnp8AACAASURBVN+I+jtJP5f0tRDChr1WQwAAAAAAACBjIYThD2z2x5K+KKks6aoQwme99A0NLaGxcfywyytUpzqHT02NsXkN9miPxb4iNZ51jocl8i+UfSJRnbMqjMASlc5hH1iqU9uZ1LyCpPrXGBuReb1vb1+6uneqt7eTBaqKobTZ06dPDzNnzhx2WWvWrE+mmTVr+PkXKaPIQnBSnXWo19oC06leIzGOa9c8lEgxuu31SScdXV/+I2DNWn8azCpQx7WJPOodz1T+Y1XGaJY/EnUYCanxuCdRx+ckh394yHV6Jn+9TNchPZ1TeXjWr1+vp556iva6iqEeYzc1NYSW1uaa8UpPi1teXyhyNOG/iqmx7C9vvX1+vMiCUEos06FU8eMqu/HuHj//xnK6lnUv0BX/KYehL52FNTYmUtTXZqeG7gup8qWyJUakNxFvqr28S5JCjxtuLnDs2Z04fu2t+Mubev3hGxr95VGSShU/j3KDP549oSlZhqfS15VMU6ok5lVjmx8fg66KUmLdDYntU7qS6VfVhUQeqdHs6fWvB25oSD8htSexyDY11B6P7q4d6unZU7Waw75S2czKki6T9EeSHpV0p5n9KISwttYwjY3jdfwxLx9ukUXrVdfwqU72evMvpt73JyaWFqXHo56TDUXylyQLqfFM7BhYIl7yN9SlIqOYaJBSsyqYv4pVlG50U0ryN+Ql+RuYSmozENINXnIfq+JPx0o5XYanyPJarrNFGpttg79Mp9dbPx432/VKb1/2pnt/9/29XYV90lDb7JkzZ6q9vX3Y5c0+6Zxkmvb2a4edf5EyiqyT7e3X1FWHes2ZdW4yTb3t9UiM47xZb0ukGN32ur39qkT5o2/23AvceHv75ck85s71p2N7+9eHVKdn5D/Hr2MsI11Pz7x5b08ncsu/sq7hJWnBXL8OY9Fep8Zj4bz65vXCee8Ycp2eyV/v2tuvSNTh/GQJqTw8ixYtGvawB7LhHGO3tDZr8enPqZlnx+MnumXu6E4fj/Q0jHPjM8b5+/pP7fA7p0oF1ssJoduNd7f6ZXSXJ7jxRx718z98crqDrlTyDw5DopNQu7e44Z7d6f3w5oMP9hOYX4dKqs2u/X55SdLm7oP88iVNbtjtxktbNvsZHHm8G7aujW782Jb0MdH6Tn952NyZ6HDd5A8/7fBJyTq07fbXq4nTHnfjG3qOdOOpQ+g929InF1u7t7rx0kHz3XgIiePfxDolSZYYkbYp/kmIPdv9eWklv78lVBInOVRg3yTR6fzE1ulu/NApm5J1eCxx3dXMabVPANy96ls1Y/X0Xi6WtC6E8EAIoVvSdyS9so78AADA6KDNBgBg30d7DQDYb9TTqXy4pEdy/z+a/QYAAPYttNkAAOz7aK8BAPuNejqVq107/Yxrts3s7WbWbmbtfb2ddRQHAACGKdlm59vrTZvSt1ABAIARN+Rj7J7u9LM0AQAYDfV0Kj8qKf+QliMkPeOhLiGEK0MIi0IIi8oN/ksCAADAqEi22fn2+qCD0s/DAwAAI27Ix9iNTcN+TRIAAHWpp1P5TkknmNkxZtYk6XWSfjQy1QIAACOINhsAgH0f7TUAYL8x7NOaIYReM3uXpP+VVJZ0TQhhzYjVbJjqfePzSLwROlWH5FthE2/AHAmpOo4FS7x5NqTeeG/+W1tT8zIUOKcSLP2WX0/J/NvRGpV4a2wBQf4dAMESb3ROLG9ljcAtdaXUW6UTbxkegTe5p962XO96OxIssUzXPx2KbFvqG8+RmFcYeftimz37pHPceGpZWb322lGvQ7q9rq8tXbXmmmSaObPOrSuPeoeXpBVrvu7G581+qxtfubq+eTV3zvnJNEHpN7l7SqVEe23+G9gLlWH1XU04d94Fbryc2OeQpAVz/XmVOjRYsfLKZBmeefPenkyzYoVfxl2JOiyYmy5jtC1b4a8zaen2etmKK+osY+/mj+qG015XKm3as3tezfie5sTxb096+9bYu9uNb97hD79z83Y3Pml6X7IOOzTNjZe6mt14qs0+Yoo/fNv4dJu/e6e/7k6e6u9XbLepbnzqjPSx4z2r1rvxo4492o+P8+f1reu73fjhh6TbumSbPX2yGy51PeHGU+3h1j0H++VLMvMX6rkn+svLIy3+elXkGLtrvL/ubur252Wpa49fQJu/vLZN9vOXpMYJE934ka3+Mr9+Y6KAIvvaiWOGdTdvc+MTDxvvxg99Vpsb37UpMZ0llXf7y3zPDL8M293hF9Dqj6MkTW/yt6E7N9Xu+6r01p7Gde3dhhB+Iukn9eQBAABGH202AAD7PtprAMD+YvQvvQMAAAAAAAAAHDDoVAYAAAAAAAAAFEanMgAAAAAAAACgMDqVAQAAAAAAAACF0akMAAAAAAAAACiMTmUAAAAAAAAAQGENe7sCYy2EUNfwZlZ3mnrrUG/5I1GHUEqMY4E8TBU/D2tMVCJRivnxItNJSqQplf14xR/HUqIOZunzPn2pOsqvY6oIqxRZ5v0y+oJfiNU7rxL5xyR9iTISeVRSdUwv9SHUt20otszWy19mU+cizVLDY3+wZs16zT7pnGEPv3rttXXXoZ7yi0qtUyvvvnrU61CvObPOdeOr1lzjxmfPPi9ZxurV/nRYsfqqZB71WLnqilHNX5Lmzb3Ajd+1sv5levmKr9U1fKmU2P6OQHvdvvzyIdVpsPnzz3fjK1Zcmcxj3ry3u/HUepuaCnetTNdhwVy/Dqn42LTX9Vk4z59Xy1ak17uRyAP16+sN2vFU77CH311O77uVe/1tR0ti+HHjJrnxrqYtyTo0dfvr1biD/a6NDRv8/eztD/vlH3tsuuvErNONb9viT+ut2u7GK7smJ+tw9Al+PQ8a59eho+lwN94W1rrxbR1T3LgkTW9u9stofsyN7+rzx/GI42e48e0rNrtxSSo1T3TjodNfJ8rlbX4BFf/YVJJKJX88O7p2uvGubn+7MLHVX56OG7fHjUvS73dNdeMP7k5sXxLHv+PLiT4hSbv6/PGccVKrGy8ljuN3PeVPhyJN/uYdiXnZ0+XGD5vsj2NPos9Hkpq7/XmxpaV23NvF5EplAAAAAAAAAEBhdCoDAAAAAAAAAAqjUxkAAAAAAAAAUBidygAAAAAAAACAwuhUBgAAAAAAAAAURqcyAAAAAAAAAKAwOpUBAAAAAAAAAIXRqQwAAAAAAAAAKKxhb1dgKMwsmSaEMOpl7G1m/jiGUCmQhz+eFavvfEOplB4+BD9NkTzqUVKR6dSTSOHHQ7nJj1urG6+EvkT5Usn88bDQ69eh5C8LQf44xDS+RBEKoZzIIZFBkdU2UcnkepVYHC3UX4nUepnavo3E9itdh/R6g/3frFkz1d5+7V6tw+q1fvlzZp3rxletuabuOqTKSBmJOqS2Tan2evbs89z46tVXD7lOQzV/3jvqGn75iq/VXYdF89/uxlesvNKNL1z4PjceRqS99vNYflf90yFl4YIL3XiqHVq+/Ao3vmC+n78krVjhz4t6LZx3fjLNXanlIZHHshX+dChSh5RF8y9w4+3LL3fjqToWMRJ5oH4WKir17a4Zb2vz97PHTU1vv4LGufHyhk43vrOlzY1PbT0sWYcd3d1ufNcG/5inLXEs0N3l7wPv6k4dF0oNpdTxhN9mT9ZkN378QTuTdVjd4ZfR3eN3ATV1+PNy2oxj3HiltdmNS5LteNCNd4YWvw4nnujGH7t/vRuvTEj3M5RslxvfuMlv01un+vOyZcLEZB26+rrceOVRf91uafaXx6mTGt34A9tSx+iSNd3txqfJX166Ov1x7LD0dOre4c/P5sn+dOjr9KdDqTnR55PsDZGmHOZvv46q+NufzZUJbnxcJb0dt0T/1/Rxtbehm8q1x5ErlQEAAAAAAAAAhdGpDAAAAAAAAAAojE5lAAAAAAAAAEBhdCoDAAAAAAAAAAqjUxkAAAAAAAAAUBidygAAAAAAAACAwuhUBgAAAAAAAAAU1jCWhZkkM6sZDyG4w6fiSuR/oAjBH0ez9LmCUErkkRjeVE4UkKyCVErVMzWefryUGscCdUyVkV7e/HhvpbuOoaNyan6bv5pXUgUUqEQpMS8t+KX0FVleHEW2DekRqa8SwRLrhCRLluFPp3qXxyLTqdi0HN06/CFsx/d3s086x42vXnttMo85s84dqerst+aefF4yTSj525bVq66qrw5zzk8nSmzjV674mhtfnoiPhfblV9Y1/LJlX6i7DksWvdtPkGivUxYtTOQvqX3Zl934srsuq7sMz13L/fzHQpH2OmXZiivc+MJ5/npV7/CS1L788r1eh3ra63vue2jYw+LpytajyaUna8Y3dx7nZ/C4fzwipff1g1qTeXgmNTcn0+yQX8/ULmZqP7ztaH8/vCF57CqplDjuatvixlsTZTy4eVqyCuNLfpopk/z1duO9vW78oR1dbnzRszvcuCR19k1w420bd7rxHrX5BXT78/r+tb/zh5d00oITEin8ef34Q31ufNqEx5N12No91Y1PavHndTn4y9sjjzW58ekH++MgSeVdB7nxidP9bcPvOvx51VigzZ5+qL88bW70l/mmsNWNt2zd5sa7pk5x45K0s9LpJ7AWN7xp9y433lQusH1K9SV07akdrNTePnKlMgAAAAAAAACgMDqVAQAAAAAAAACF0akMAAAAAAAAACiMTmUAAAAAAAAAQGF0KgMAAAAAAAAACqNTGQAAAAAAAABQGJ3KAAAAAAAAAIDCGsaysCAphDDs4c1s5CqzD6t3PCsFBk8lMZX9eKKORcYhmcaa6hq+pB43Hkrpxb9HE/06JKZTQ+/Dbnz+sS1u/MlNG924JG3ePc2Nd5VmuPFyyZ9OjWFrsg4K/vmpXhvnxkvW7Wef2G6UCszLSqU3kcIfhyC/DkW2bSG5zCfWq0r9dUipd/uTqkOR/EdiPHDgW7XmGjc+Z9a5e70Oc08+z42vvPvqkazOsJRK6esLUuvtggXvrWv4Zcu+4MZPWfRuNy5JvYn2enn7p9z44nlvceN3rLjOjb/sua9245JUqtTXXqc0JtrzkVBkefEsPuV9yTSp9rp92Zfd+MIFFw6pTtUsmP9ON37X8q+68WUrrnDjC+edX9fwkrRo/gVuPLXepepAe73/CKGsrjCldoLEats8tS1ZRs/WvkQl/HBra6cbf3jDCOzDJuKHNfntxBOlXW78occT00DScTP9Y8NdXf607uuZ7sbbGvzpKEkyf4Zv39HoxlsPb3bjs2f4U7qjvMeNS1LP+ElufMEJ4934skdvceMTGvx25Myl/rGpJD2125/WqTZ78hR/WWis7EjW4b5HN7nxP150rBvfuX2bG2+duNONd3f5fRWSFMr+dHh091NuvKlxqp9/sgbSzi5/PHZ1+PPiyIP9UjrkbF8lbdvkrzOSNPmg+o6xDxtfceOW3AJKia4EKfhl1MKVygAAAAAAAACAwuhUBgAAAAAAAAAURqcyAAAAAAAAAKAwOpUBAAAAAAAAAIXRqQwAAAAAAAAAKIxOZQAAAAAAAABAYXQqAwAAAAAAAAAKa9jbFRiKEEIyjZmNQU32rmBlN14qNA38NKnpWCrVfz6ibL1uvM8Si2dqOoTOuoaXpG41+Vkk8ij3bnTjZy041o0vX7HFjUvSTfc3uvFKw7PduGmPG2/QpmQdrORPh55KsxsvlXr8/BPLYwgVNx7LqG+ZTZVRqaTrIEukCYllMjEKltxGputYZDvr1mEEtsF/CNvxfd2aNes1+6Rzhj18kWFT83nVmmvc+JxZ5w6pTqNRxtyTz3PjK+++2h9+zvluPPKn08pVlxfIw8nd0tvGu5Zf5sYXLPyAG1+27JIh1WmwsqW3S6n2et7iT7rxFSuuc+Pf+eS73PisQ0a/vV64+BNuvE3+fpUkPW+xv8x1VA5y4+3tX3TjSxa/143fceelblySFp/ynmQaT6q9L9JeL7vLX+brldp/LKJ9ub/uL5znz+uRaGuXrbjCjS+af0HdZSCtzxq0qzS1ZtwS7UjX5vS2I7W8bG+Z4A/fWf/ytu1J/3hhygx/G/t4z0433pdoR44+It110uNXUa1lf/uzs+SPw7hSd7IOhx6y1Y0/uekwNz6zebM/fHeXGw+a5sYlKZg/rW/d6G8jzzv7GDc+vXu1G1++4lE3Lkk3r/Pnd0i02eXEMXZjoi9Ekl54gj8d9mzb5sZ3VfxxmFj228uGKel9xN4tfh07KrPceEtpgxuvVNL7gD0Vf5mstPW58Y4wzo1byV9vdzWl2/SmLf78bktM61Jif73QEfyONj8+qaNILs/AlcoAAAAAAAAAgMLoVAYAAAAAAAAAFEanMgAAAAAAAACgMDqVAQAAAAAAAACF0akMAAAAAAAAACiMTmUAAAAAAAAAQGF0KgMAAAAAAAAACqNTGQAAAAAAAABQWMPersBICyHUNbyZjVBNagulVBllN5quY3ocSiV/1pv50zFVhyLTsWT+OY3UnLRECjN/OpbVlyhBag1b3HhLQ4cbn3diqxs/fLo/Dt9/YJUbl6TmhjPdeJd1u/H0dGxO1iG13llDYnkJ9a13RZa3+rcN9a6XUgipZd6vY73jMBJS45lcFhLTMaoMoUYYDbNmzVR7+7U147NPOqfuMkZ7eV615pq680i116tWX+3G5845P1FCeruxctXlbnzB/Av9EhLrbKmUvr5gyaJ3uPG7ln0tmYfntEXvcuMNlm6vV955sRs/87S3ufH3/sWfu/FLv/cfbvz7C6a6can+9rr99r9PljHaTl3yPjd++x2X1l1GvfvipVL97XXKooX+Mtu+7Ct1l5GycJ6/fVm24opRHV6SFs2/IJkGoy9YSZ0t42vGW/bscocvtA+biB890Y9v3+TnsKenJ1mH5vH+/mGwRn/4Nr+96+5M7IcXaLObmv1jw0rfZDfe3LjHjaeOnyVp81OT3PghjZvd+KaOLjdeafDbu8YCbfa2dSvd+IxnT3Pj6+5Y4ca3l59w4w8+eK8bl6Tmhue78U75bXZ3Zbsbb+xK7wd37vHXi65JfnxSg7/O9No4N17u2+3GJckS+8qhw18nSq1+m12y9HHhuF3+dGgqT3Djm8f582JaYladfGK6jlsf99fdTTua3Pj0if449o3vTdahUf78LLKNq4YrlQEAAAAAAAAAhdGpDAAAAAAAAAAojE5lAAAAAAAAAEBhdCoDAAAAAAAAAAqjUxkAAAAAAAAAUBidygAAAAAAAACAwuhUBgAAAAAAAAAU1rC3KzDSzGxvVyFdh0TclBoHP14uNyaGTzPzzzeUSvWfj6hYkxtvtB43bsHPv1fj/PLV52cgaWrTY278VWdMdOO7Nz7sxh96vNmNP7xtqhuXpPKUNjfeoM1uvGR+HXrlj6MkySp+HazbH1z+shAKzKsUS9QxBH+BSsXNimxO/fGwxEJdqSTGIVl+OZkiWKKOiUL2hW0w9g+pZWXOrHPHqCa1rV59tRufO+f8RA7+OK5cdfkQa1SlBPPX65For29v/3Jdw59+ynvc+K3tX3HjS075QLKMl55+thu/8dZvJ/PwXP8Pr3fj371rS135S9KpSy6qa/jTT/1IMs0tt33GjZ922ofceCmx7zYSbr/j0lEdfsni9yfzWJxYZtuX+cvsooXv8odf/lU3vnDBhW5ckkKpvvY2tQ1eNP+ddeUvSe3La2/jFi1qrzt/RBZ61dj5VO24tfoZpHcgZYnlbdumnXWVMXlSug7a7dehq9nfh+3pTB0j+/k/8diexPDSzCMSxwtl/3ihlFgvQ4HV/sgF/sTsWv2QG+9tPNiNj5/oH2Pv2OEf90nSsSf6x5+TWne48dkzT3LjzW1+/nc+2eLGJenw6f506HjUP8aecuJR/vDr0n02bd1+Ga2NiRUr+G329kb/2HLC1vQyP366318xqcVfLzfuftyNTysfkaxD56TJbjzYLjc+fau/jQydW9341ienuHFJUmIbOmGyv15Z2O7GGzv8ZV5KH+d3bqi97lZ6ay9rXKkMAAAAAAAAACiMTmUAAAAAAAAAQGF0KgMAAAAAAAAACqNTGQAAAAAAAABQGJ3KAAAAAAAAAIDC6FQGAAAAAAAAABRGpzIAAAAAAAAAoLCGvV2BPDNLpEj3gYfQV2cZvkqBwVNFlFRODO9nYFb/uYB0GfVNJ0uMoyT12ng33lx50o2XE1XcY4f65Yf0OB7V8rAbf/G84MZ//t+r3fh9v2tx46Wpi924JO2uTHLjjd3b/XizPx96zc8/8pfJBu104xVrcuOh4k9nWcWPSwqJ+V0q+ctspZIuIyVdhr/9KpX86ZyqY0hMRim97pZKfibpOvjjKNW//cHoW732Wjc+Z9Zbk3msWnNVIo9zE8NfkywjZfbs89x4yfzdpHrb6wXz3+3Gi5VR3/pyZ/uldQ0vSUsXne/Gb2m/wo3PX/wZN778zkuSdfjLF77GjT/4r29348e84Uo3ft/vHnHjpy96vRuXpI7KFDfeGLqSeXhuuc2fjpJ02mkfd+O33vqPbvyM0/52SHUajlOXXOTGb7v9c3Xlf/sdn0+mOe3U99dVRqq9Tkuv13fddXmdZfhGor1eNP+CmrF77ntoyHVCdX0haHdPb814s+12h29pnJAso7e5w403dLa68Ymde9z4rqltyTpUuvx9UOtOdG0kVqtyos0+4sj0dOpLtdmJ4VOHp+WWHck6NHR0uvEne8a58Ue2+GVMt0PceFs5vf0b3+ofY3/63MlufPXd/vLYfr/fZm/qONKNS9L6B/3lqVH+dJqxzV/veo7w1xlJesgOd+Pdff5xW3PJX+/06INuuKttqj+8pKbt3W683OTPq+7xR7jxXR3+8JI0oexvP0rd/nQIDf460ZPoiyiyL95Y8ft9ujuf8DNo8ccx1Y8gSaWSX8+WgxtrD7ut9rBcqQwAAAAAAAAAKIxOZQAAAAAAAABAYXQqAwAAAAAAAAAKo1MZAAAAAAAAAFAYncoAAAAAAAAAgMLoVAYAAAAAAAAAFEanMgAAAAAAAACgsIaxLMwkmdnwh7eQTBPSSfzhS379LKT74U2JPBLTIBUvlVJ1SE/jeutgKtc1vCQp+HmkxtNScTUmyk/Py85d2/z4tu1u/M9fvNCNf/H7T7rxvo5xblyS1DjRj5f73HBX9y5/8ObWZBVKpRY3bol5nVpnSqXUpqqSiCu5WoTExqNk/vIS5E/nWIZfz/R6k5pOfjw1jjGNX8dUPKWeNgBjZ+2a9Zoz69xhD79qzTXJNLNPOseNr1577bDLL6retiwVX77iq2584YK/duNFymhf9sVkHp4lp3wgmeb2Oy9x4ze1X+HGzzz1Qje+/I7L3PjsJZ9345K0+vbvuvF7LntJMg9PqcNvr29p/20yj3lzL3Djd668akh1Go5bb724ruHHYht+2+2fq2v400/9kBu/5bZ/TOZx623pZc5jiX2GlPS+fv3al19edx6L5vvLNMZGg5U1OXU84Fi/ZWMyzdFTprvx7bv8/cPu8Ue68Qkd/vGIJHWp142n9tWV2H61zfAH37MhvV6WUvsNe/zhd0zyj6kmdSaObyVt2uTv2zyw/lE3fsJxk9z4rtatbnxHx6FuXJImN01w453bHnLjExqOduOpNvukhUe4cUlqv80/tmtqPMqNP9XpL6+9uzuSdXjysQfceONRJ7rxlsThabkyzY0fMjmxUkjaUvL7ChqfeNyNHz490SdUnpKuw8Yn3Hjrof72sbHXX7d79/h9Ms0T/HGQpImNPW78yR1+HuMmNrnx3ZY+Rm/u8usw3H285JbRzK4xs41mdnfut6lmdoOZ3Z/9Tc9pAAAwqmizAQDY99FeAwAOBEVOg18n6Y8H/fZhSb8IIZwg6RfZ/wAAYO+6TrTZAADs664T7TUAYD+X7FQOIfxG0pZBP79S0jey79+Q9KoRrhcAABgi2mwAAPZ9tNcAgAPBcB/YdUgI4QlJyv4ePHJVAgAAI4g2GwCAfR/tNQBgvzLqL+ozs7dLerskNTYWeOkYAAAYc7TXAADsH57WZjfQZgMA9o7hXqm8wcxmSFL2t+YrY0MIV4YQFoUQFjWU/beZAgCAEVeozaa9BgBgrxreMXYDbTYAYO8YbqfyjyS9Ofv+Zkk/HJnqAACAEUabDQDAvo/2GgCwX7EQgp/A7NuSzpQ0XdIGSRdL+k9J35V0lKSHJf1lCGHwiwaeoa11ejj+2FcMu7Kpumb1deN98vMomf9EkFT+RdKUSn5ffmp4M3/4Uqn+p5ok66ByYng/LklBfhmN5s+rUPLL6O7pdeMnHtXkxiWpvOE/3Pgr52134y9ZeqIbf6rnUDd+7f887sYlad2WaW78wY0VN95XmuzGu0P6lrqW1kluvNH8OvRW6lsnimwbiqTxh+/zEyTGsUgeBbbHbrySqEJyHIrUIfiF1DuORXjTYe29/6HdHRvTG+oD1Ei12an2etWaa0aoxrXNmnWOG0+116vv/nqyjLlzzvfLGOX2etldX3HjRSw+5X1u/I47v1B3GSlLl3zIjd90+z+Oeh3ecsZz3Ph1N9/jxju+/UI33nb2L4dcp6FaOv/P3PhNy3/gxhcueFeyjJFY5jzPPeOjbnwk2uubb/30kOo02OmnfbBAHfy27NbbPl9XHZYsfr8bv/2O+vKXpMUL/eVhLNrr9uWX14wtWrRI7e3ttNdjcIxdOqjbHb6h0z8WkKSw9Sk3vidxh1O51OjGS60Fjl+7/H3QUinVJtfXZveqzY1LUu/2Djc+bqp//Nnd50+Hpil7knUYX5noxrs6/GPk3uAf3x4y7XA3vmvng25cko4Kt7nxpc9uduNnnzXVjV9zk79M3749PS+7E8vT46t3ufHnTPeP0e/a4o+jJJU2+utV37jdbrzcMN6NN09K1aFIm+3HbftWN76t7Ndxgh+OKv4yfdRB/nh2btzmxp8I/vLSW9npxiWp3/BeswAAIABJREFUb1erG28anzjGToxjKDCvkg1uS+1+n3tW/6t279pQNYvk1juEcHaN0ItSwwIAgLFDmw0AwL6P9hoAcCAY7uMvAAAAAAAAAAB/gOhUBgAAAAAAAAAURqcyAAAAAAAAAKAwOpUBAAAAAAAAAIXRqQwAAAAAAAAAKIxOZQAAAAAAAABAYXQqAwAAAAAAAAAKa9jbFcgLIbhxM0vmUUkkMfkJUmUUqUOpVE7kkcqh/jokS0iNp1LjUH8dW7TLje/SYW68VGr18+9td+MvOt4fR0k67MRmN75h/RNuvLvvJDc+pXGTGz/zBD9/SXrWHn81/q9l/rxYt2WcG7eG9Gaiu2unG29oaXTjpVKbGx+JbUO9zPzpEEJfgVz883hWqrjxSsWPm/nx1LZFkkLw8wiJaZ2qgT8no3Kinv7yUKQEpJw0a6ba26+pGZ990jnu8KvXXlt3Hdas8fOYM/vtbnzunPOTZYx2e73srq+48VMWvTdVQHL7dsedX0jm4TltyYeTaW69/bNu/Kbb/9GNLz7Dr+MdN78vWYeU1z/fb6+/+Xp/Or78T17kxv1Wqphvf/AsN370FL+OS+a90Y1bw/Qh12mw5y+9yI3/+qbP1ZX/Tbd8Kplm6el/V1c8VUZqHzca3etubr/j8258yeL0tuH2Oy5146n2+s5lX3PjCxZckKzDXXdd7sYXza+dxz33PZTMH8WUShVNGLenZnz7Bn/72Nmc3j8st7S48bbEfnBXxT8WCJ2pPch0m50U/PHc2edPp4mNPckimqYkjnkm+tPRVHs+SlJ5l3/cJkkbGvw0480fj0r5UDe+Nfh1/NP5HW5ckuaP96f1xodWuvHuPr/NPnRclxuf1Zte3o6e6PdV/Nej/vCrnkodY6fn5eEztrjxLU1T3HhTS2Ld92elJk1JH1dt25poLyf5ddz9pD8vJo1P16GiJje+7tFtbnxyor9j285ONz5lUnovsWfSZjce+hLz0u/6Umdnejve2eZP69bdu2sHnX4IrlQGAAAAAAAAABRGpzIAAAAAAAAAoDA6lQEAAAAAAAAAhdGpDAAAAAAAAAAojE5lAAAAAAAAAEBhdCoDAAAAAAAAAAqjUxkAAAAAAAAAUFjDWBYWJIUQhj18X7BkGjM/f7N0HqM5fJaLGy2VUn39ftxULlADvw6p8TTzyzBLn68ohx4/j3KLG6+EVjfeaB1ufGrDNjcuSUvnjHfjHcfMdeOVnl1uvEWdbvyg1o1uXJImHTLTjf9i5W43Xqn0uvEiy3wqRVenPx0am5vdeKnkb6qKbFdS41GpVJJ5+PkX2Zz6ZaRGI7VpSI1DkXlZKvnrdl9f/ctLSqXPHw9/GzkS22iMhZNPPteN3333NW581eor3fi8uRcMuU7P5C9Pdy2/zI0vWvgeN96+7ItDrtFQnbbkw2781ts/W3cZp5z+eTce1FZX/j/62xck05w1yu115/Xz3PgDjzzkxiXp7H/e6sZ/8eKlbjzVXre3X5qsQ0qqvX7uGX/rxlPt9fOW/n2yDqn94Hrb65tvrX+ZTzn9tA+68dS8vP2O+udlPcdc0si0534etNcjJYSSurpr70un5uUxh01IlvHIQ/7xhPmHddoyxd/Xn96ZyECSKbVM+9uODTv8+KGT/fUylKclypce6fGPX6fu3u7GJ1US02nK5mQdWrb7x8i7Sv78Th1jP6tnnRt/zvj08evSOePceL1t9svmbHLjDzzysBuXpJ/ed5Ibb2tMHVv6y9OL5h6TrMOOLQ+68Sce8adDJfjLY6rN3rGtwLFjIkklsdpOPmqKG9+859FkHQ4e76/bpb4j3XhHg79/NmWSPy/HNXW7cUkaN2miG9/+pL9vsy2xL92iPck6PHlfkxs/5kh/PGvhSmUAAAAAAAAAQGF0KgMAAAAAAAAACqNTGQAAAAAAAABQGJ3KAAAAAAAAAIDC6FQGAAAAAAAAABRGpzIAAAAAAAAAoDA6lQEAAAAAAAAAhTWMdYGhZMMe1kJ6WEsksUQCUyKeKqBAmlIp1Zfvx0vmzzazciL/AtOhwHjWO3woNbnxcujxy2jo9fNvGO/G73t8qxuXpDPmHeLGW5t2u/HGvg1+AT3+OBx9/Gx/eEm3PTLNjT/w5DY3Xm5uc+MV8+eDJFkluPGQ2NT0dPnTsanVr2MpsSxJUqVSSeThr3chJMbRD0sqsl748VQRpZI/nVPjENP4y2S57G9f+vr8/ItsWSyxifTHo8CMQNKaNet18snn1k5QR1ve7+67r3Hjc+a81Y2vWnWVGx+L9nrRwvf4wyfa6yWn/I0bl6Tb7/wnN37akg8n86jXC09/txsvhyPceKq9fsEpb3Ljv7rzV268CL8VkbqvOcaNN73xQTd+9L8/L1mHX3z+9W78gSf99jbVXhexZNGFiRT+Mvvbmz/txp/33I+48d/c9JlE+dLzlv69G7/plk+58eee8VE3XqS9vumWT6YTOW659Z/rGr6IJYvf68bbl32lrvxN6eOJFNrrsVEJ0p4ep70yf1o//HB6XpcnHuTGO3c95cYbOprduJX9diJL5UZLiX2TGVMSw1ujX3rY5cYl6aBKixufEPzjkVRnxqZeP39JaittduPl0OrHy/7O/O937HTj655It4e7tm9047OPuN+NP/vQ37vxX95/uBt/7snpY+y5zX4e/738ATdebvan8xOP3pysw32b/fVOwW+zGw6b6MYrG7a78e2dfj+DJE1u8fsaukpHufFxPY+48bayPw6StGGtv/908DF+v09Dxw43vkPdbjw0pdsz2zjFjU87aIIb37DDr0PJupJ1mDbV3/6kNk81yx7eYAAAAAAAAACAP0R0KgMAAAAAAAAACqNTGQAAAAAAAABQGJ3KAAAAAAAAAIDC6FQGAAAAAAAAABRGpzIAAAAAAAAAoDA6lQEAAAAAAAAAhVkIYcwKa22dHo477hXDzyCk+8BL5eFnL0kmP4NyuaG+AiSVSv54lEqJMhLTITl8AWaWKCM1L/zhJamp1OvG+0rj/RKs2Y03l7e68XG9y924JL16cY8bf+WcHW58YnmdX0Dwl7fbH5nhDy/p0p/40/rBzuf4GTRNd8OlPn8aSJLJX+Z6Kv68KqkrUYKff1PzhMTwUqnU6MZT28JKpS9ZRkqqjPT2uOIPL7+ORbb36TrWV0YI/jhIkir+tsGz9t7/0O6OjekNEFyLFi0K7e3te7sarrlzznfjK1ddkcxjwfwL3Xi97fUdd34hWYeUUxd/yI3X217ffOunh1ynkfbiF3zEjRdpr3/w25+68R3fOMWNJ9vrN/r7FCPh1EWvdeO3tf/bqNdh8Snvc+P1tte33fmlIdZo71h6+t+58VRbNxbr1alLLnLjqfb69jsudeOnLHp3uhK01/uE1tYZ4bjj3lwzvsM2ucMfpCnJMsq9/n50eWKHG+/Ys8eNt07w45LU3THRjVuivSsnOgoe2LbTjR991EFuXJIadvr7uZY4Rk612Z1lfz5IUuh92I1PbDjYjU+f4NdhR6IZGNezwk8g6fAjD3Xjn37JGjeearMvv3epG3/oQf/YVJIea/W3b/evbHLj5SZ/G3z4hMnJOtz7xBY33tx0tF8H82dWkL88ldpSbb506MSpbvz/s3fncZZV5b3/n3XOqbm6uqp6bhpoUaaeh+qBQfSn3kRjEoxD1KhBJEwOESUgaIjXGBVBBadWMUxiFAdIuHqNGr0q0PRUPVZPQAM9z93V1TXXOWev3x+0iTeX+j672VUNrZ/365VXlO/eaz17WmvvVUW5b7teLxk9Ws+nTx32r1VjdY2zhe6jOuoaj+TqZF474K+XFPP6uYrOWsOBvL7fmov+GJplzt70+OBzNr+pDAAAAAAAAABIjUVlAAAAAAAAAEBqLCoDAAAAAAAAAFJjURkAAAAAAAAAkBqLygAAAAAAAACA1FhUBgAAAAAAAACkxqIyAAAAAAAAACA1FpUBAAAAAAAAAKkVTnSHMQmDZiEX5b4hV/bbj3qdPBcqdJ7T+8eoazQzy+fzzhZOH+IcmZnlcjo3S5zcLARdYwheH1o+7/+8oq80IPOB7u1OC7qPI1Yr81CY67Rv9pl/PyTzRb/slPmohiky7+ztl/nOPb0yNzPLV41wNijJuJTo8+zezmZWXVEj8/q6Rpkn0blWUR9DceCIzM3MqirrdR+5apnncvpEJIk/PoXgjHFB39OJ82i7T20oeluYWbYay2V9rSz441N0xriQ+OMwstmwYatNnXLZ4PnGO4e9hpkzrpK590ymsWr1VzO3kcXC+de52yxdfqvMz194Y6YaLrrwJnebhx/9pMznzrpc5ivXfFPm+9r7ZL52zU9lbmY2ZYq+XzZuXOG2obxi3ttk/psV97ttnPXSd8r8iae+J/MzXvZnMn96y4/cGqZP+0uZZ52vly2/3a0hq1de9HGZ//rhT2TuI+t8fSIMe41DMF+3ti4aNGtpyfZM4ncVLUkODJ4W9bfCvnKP28PEsbqN7m49judyetlhTG6MW0MyQX83hWr9Ln9kh35mJo/U6wT5o847rpn1DnTJvL6myW1DqUvxnV/VME7m7QcOy7xQM17mB9v1MZ5+6itkbmb2q7XbZL5o5HkyHz3mlTK//+c7ZN7QrO9XM7PHt+hvx+Yzm2W+Z8szMh9Rpe9nM7OYOPdkXj+70ZmzzfmGb05O1fubWcfAUzKvbBgr82fW7JV548v8GnI5vSbjrRGOHjFK5u3JPpnXVfnrAB29DXoDbx2yS6+dheoUa6XOp9PAiIli38HvxRf+jQgAAAAAAAAAcNJgURkAAAAAAAAAkBqLygAAAAAAAACA1FhUBgAAAAAAAACkxqIyAAAAAAAAACA1FpUBAAAAAAAAAKmxqAwAAAAAAAAASK1wwnsMiQqdnf018BC8NrLJ5fwaYtQ15HP6tIeQP66a/t/9s58nL8/n9TH09xXdGs6derbMF86fI/OCc5qKVpJ5e8dB3YCZ5WKfzMv9h2Xe131Et1+olHl1bZPMzcwS5zGurx8h87r6RpmXitGtYfv23TL/5S9+I/PgXCvvuUvUsHJMf3+/zKurKnQDOX3DpRkb0tSZpY/E7SDNzxHLMo1R9zEU1yo69wOG39Spk6219c5B82nT3iP3X7/+rsw1rF33dZnPnvVemc+d8363j5WrvnJcNR2vBfOul3nOGVfMzM5feKPMT8R87bn88ssz7X/VVW/JXMPGjfp++dQndO7N179Zcb/MP/NPd8vczOySS/9Y5l+8/d9k/vSWH8n861/7lVtD2/rvu9soC+f5z5Xe/2/dbYIzVy1ZcXumGtJ4ZPGnZf7yC/4+U54kep5bvORmmaexZOkXMu0fzB+fmK9fHJJS2foODj6GjRqj3+Vr8rVuH8Uj+rupoa9T5l0jR8l8d/eAW0MY0MdR2aD33xk6ZD4+6m+iNHN2XY1uw1uq8NYJ+vv9Z+6lc3bJ/E0Xf0DmG9c8LfNz33SaLqD7cZ2bWe/r9HnqO6BP1HZnzv7LP5kg86MVun8zswUXZvvGfurMZplPOfcit4bFa1fLfN/OjTLv3K/vl3PHVsl8084dMjczC33VMj97nB4bek/X5zmW29wacqYf/nJ+tMx3d+sP1JpE3y8He3tkbmZW0KfJBpzX8fG1XTIvJ/44HqO+H8rdewffNxm8QH5TGQAAAAAAAACQGovKAAAAAAAAAIDUWFQGAAAAAAAAAKTGojIAAAAAAAAAIDUWlQEAAAAAAAAAqbGoDAAAAAAAAABIjUVlAAAAAAAAAEBqhRe6gCEX9Tp5cJbRQwi6+ahzM7N8XncSY3Ra0Hkul/1nAd5x+udB11hVVeXWsGvXLpk/urhL5lOnvkTmr3vty2U+fc75MjczC86lyvm3QyburWJm5tSwb39Z5j/58a9l3rZuvVvC1q37ZO4dRwiJs79zkLm8zs0scYro7T8q85rqeqd9vwbvufLEqK+l236KGr37KUbvWjnNe4OwmQWnCDnMD/MziXSmT7/M3SaYvh+9uW7N2q/LfM7sD7g1zGu5RuYrWm9321CyPvNp2li85NMyf/kFN8l8Wetnjrum/+7RxQ/L/Kr3zZf51Vfp+bicYi705uuPffwqv5EMbvjopf5GGedrT9u6dSm2+v8y9WGm5wDP0hVfyti/2cJ5fytzb77+1SP6mTEzu+hC/dw8svif3DaUC8+/MdP+ZmYh6DF04QI9vi1dpse3NPN164pFMm9pee+g2aZN2932kU5IzAq9g+feENqbiJ1/20dB32+dI5tk3lTul3lHqHVrKJQqZL5zk36Xrx6hlz7CCO/7V8ZmZpZznptDRT2GjnE+oSeM0+fAzGzXupEyX3xIz9lJoVnmD69fLvNvf1mPPWb+nL1iyWqZN9fr+bIi1yjz06e+TBdglnnO/oeb2mS+vk3nZmaxQ6+HNE2ok3nn/pJuv7tH5h0D/tjQlXTqPvaNkXlNrb5fa4r+kmV3RYPMew4OyPyojm3cOP1cTxrvr33tO6Lzygp9rTr79THWVvjvkF0D+lzWWXHQTJ0BflMZAAAAAAAAAJAai8oAAAAAAAAAgNRYVAYAAAAAAAAApMaiMgAAAAAAAAAgNRaVAQAAAAAAAACpsagMAAAAAAAAAEiNRWUAAAAAAAAAQGqFE91hCOF5ZWZmFv018JDTbYSQbR09l/P3j4muIZfLy9w9D450+3vbeOfRyxO3gv7igMw3bn5C5suWPyzze+79hswvOH+mzM3MPvHJG2R+6qmjZF4s9co8F5xHMFTo3Mxu+8LdMr/vWz+UeV+PvpaVFXVuDZVVepuKQrXMY4wy9+6nJJZl/mwb3viia+jt65J5tXMOzMyC6We/7D6X3viT7Twe28ppQ9eQc8bgctm/Vp6sYySy865BW9udbhszZ1wp8zVrv35cNf13QzFfz2+5Tub+XJgtP7ZVim0G98jiT2ba38zsold8ROZdne2Z2j/7rD+SeZr5+q57bs1UQ1ZpXi8/9/ls8/Xatf9b5l/92gfdGubOfo/Ms87X581/n8zLzjxlZrZ8+SJni5JMvfk6jYcf1c/NRRfeJHPvPHnz+YXn3+jsb/bYkmz3/PnnXSvzZctvd9uYN09fb+brEyPk8lZVN3LwDZzxKeb965RzXt+SOELm7bUHZV5RTjFnR13n+PG6huC94zq584r7n60oZ47T3wudps9T74D+ljAz64+6j42bn9QNNDbKuHvnGplfcelu3b6Z1TedKfMv3na5zDc/8SOZ79y1T+arvrde5mZmT+/W7zbfvu8BmTfWz5D5zJe4Jdjmx7fJvFCo0g04U9Hmdj1f1lf5axGj83q9o1jSg0dff7fMa0Y5x2hm1U36uYixUua1znmKUX9DH+jwl1VHna7Pw76DusYRQb/77N+hz6OZWdMEfS7D8/ydY35TGQAAAAAAAACQGovKAAAAAAAAAIDUWFQGAAAAAAAAAKTGojIAAAAAAAAAIDUWlQEAAAAAAAAAqbGoDAAAAAAAAABIjUVlAAAAAAAAAEBqhRPdYYxx0CxYXu4bQpo18OC0kS1/MTgRNWbtI83+IVTJvKl5nMwbRjbJvK+3W+ZLlmyTuZnZNe//pMzvve82mdc36GP0fq5zy613O/ubfX3R92U+fsJLZT5yhK6hXC67NcREX++y91wmuo8klnQBKe43MfSYmVku6PEnOjX093W6NVRV18nceya8Y/BOQ5I4DZhZNL1NLqenjXK5KPNUY0PU1yJJErcNDK+2tjszt7F23Tcy7T93zvtlHpxn2sysdZUew+e3XCfzZStukfl5C25waxhuF134DzJ/+NF/dNt4+DeflfmrX3PTcdX03zU1nyXzNPP1my6+RuYPPHS704I3rmSfr6+/7lKZ/921Oh8KVZW1Mn9s2VeGtf/586/K3MbSFYtkvnDeezP34Xn4Uf1+eCKcf54en7z59rEln5f5wgUfcmtYseKrMm+Zq8dpDI1Czmx09eDvb9v26G+ihlMa3T5GjHG+uypPkXmyv0/m/huqWdJ7SOa5uvEyb3eeiWavgFTft/pIdnXpd5P6coPMu5r99ZC6UTqvOKi/Nw60H5R5Y/OZMl+y1J+zp5yjvxe+fte9Mr/qPe+S+Tln6fN06+fukbmZ2W8e2yTz0HiezGtzu2W+o10/E2b+nL11oFrmeecT+uzcEZl3VPnLhb05/eS8ZJouYnubfmZGVvrv8yOqdA1PVOyVeWePfiZ6O/VaRe0Y/zwd2tYr80pneOkZOCrz0aeMcGuoLG3Rfdhpg4fiMrmjUgjh1BDCr0IIm0IIG0IIHzz2z5tDCP8RQnjy2P/XMw0AABg2zNcAAJwcmLMBAL8P0vzqb8nMro0xnmtmC83sfSGEKWZ2g5n9MsZ4ppn98th/BwAALwzmawAATg7M2QCAk567qBxj3BNjXHXsP3ea2SYzO8XMLjaz3/57Cfea2RuGq0gAAKAxXwMAcHJgzgYA/D44rv+hvhDCZDObbWbLzGxcjHGP2bOTopmNHeriAADA8WO+BgDg5MCcDQA4WaVeVA4h1JvZA2Z2TYxR/5Xo/3u/K0IIrSGE1nLZ/2PkAADg+RuK+frAgQPDVyAAADCzoZmzi3xjAwBeIKkWlUMIFfbsZPcvMcYHj/3jfSGECcfyCWa2/7n2jTHeEWNsiTG25PP6f50SAAA8f0M1X48ZM+bEFAwAwB+ooZqzK/jGBgC8QNxF5RBCMLM7zWxTjPELvxP9LzO75Nh/vsTMHhr68gAAQBrM1wAAnByYswEAvw8KKba5wMzeZWZtIYQ1x/7ZR83sZjP7fgjhMjPbbmZvGZ4SAQBACszXAACcHJizAQAnPXdROcb4qJmFQeJXH2+Hz/5QdviEkB/W9tPVoI8xa26WOPv7f9XE68Kt0Sp0A9G/zjHqNsrlosy9P6dSUVEv8+amJpmbmW3c+KTMf/SjX8j87X/1P2S+c9cRmT/wg5/J3Mxs/ISz9QZJjYyL5bLMo+nrYGYWor4nc849G732zXmuE/9+8+5prwZvixj0MZqZ9fb1yLyy2hsb9DOTOCWkGh+dExGjvh/c8xz9M+0Z7nnkZDWU8/WGDVtt2rT3DJp7z2Tb+m8eT3fPac7s98n8RNwHy1tvzdiCfigfW3qL28KF538sUwXufD0EfvmLT8r81a/5lMwrKsbJfNPjX3FrOPfMt8n8u//yHzLPOl9ff92lMjczO//8a/QGznwd41DM1/rZPW/+B2S+ZPmXZb5w3nsz9T8Ulq5YNOx9vPzCG2TuPXcPP6qfmTTP/WNLso5P2tJlt2Vug/l6cEM5Z5dyZgfqBs/r6hvk/iOa/HezmlK3zLvbH5e5//3rlmBVFbUyr3O+gXd2lGTe3KT3H1nnj1+dvTov9D3nXzP5T6HQrPM+/89u9/Tob+Qpk5wPhniKjAcG9P5nnDlJt29mGzeulfmb3jJRNxCda+3M2f1Jv27fzNr36Tm1NtHPxJGo1yJijz9nW6KP8/QKfRydugTr7K2UeS7FN/akhgkyr+zQ16K+rPPuw85DZWY9Azovl/Rx1lXosaGn2CnzmpIYgH9bQ73eprKyS+ZHO0fIfCDqGs3MTq3U7ya9veJ+E7dC6v+hPgAAAAAAAAAAWFQGAAAAAAAAAKTGojIAAAAAAAAAIDUWlQEAAAAAAAAAqbGoDAAAAAAAAABIjUVlAAAAAAAAAEBqLCoDAAAAAAAAAFIrnPguE5Hlh7l9sxiDzEPIlpuZBfO3ySIE7zzFYe3/2Rr0MeZzFW4bSdDb5HL6Zx7BOc6qSt1+ki/J3MwsX6iW+TPPbNMNOJdi964DMh8Y0PubmdVV6vNUTooyD3m9f878axmSssyjkyc5fS1iou/5fN4fymLUNXhjR5Loi5lmbPD6GOjrlnlVZa1Tg3PP6+6HyPCPPzEOfx9/6KZOnWytrXcNmk+fdrncf8b0K90+vDF+zdqvuW1kNb/lWpkvb/28zBfO/4jMc7ns7zWPPvYpmV904T/I/OFH/1Hmr37FZ90avPn6V7/+sMx/+YuPuX0or3rlre42m568P1MfzvBskyY1ynz2rLe5Xaxek63G88+/XuZDMV8/ulTf856lKxbJ/PwF12RqP40FLVfLfFmrP7YsaHmfs4Uev5a1ftntQ0kzzV1w3g26DdPvf48tcca3BR9ya0i89z/m6xMiXy5Y45HmQfOOpqNy/45D/jtsZ9Dv6sVaPYgWjtTLvKbOe083SyrqZL7b+QY/rVHPycF5rsdMGClzM7P6vj6Z79urv1lq+ztl3p8f7dYQne/wiaP09+3R3TtlvrXUI/PerlNkbmZ2+rSxMn90zRaZv/3t/0PmjyxeKfMHv/trmZuZTZwwRuZ74iiZVx/dL/NciqU47/tyfJO+1sVCjcx37Tos88ox/tgweoyea3pLTTLvyut54tAhfx6p7PMWTfT4NapmhMw7RjfIPEb/Izt06bGhz/mGbhil1wlKB/UYa2Z2qKzvBzlni4jfVAYAAAAAAAAApMaiMgAAAAAAAAAgNRaVAQAAAAAAAACpsagMAAAAAAAAAEiNRWUAAAAAAAAAQGosKgMAAAAAAAAAUmNRGQAAAAAAAACQWuGFLuD/EhIdh+xr4CFkbuIk4J+n4JwI7zwlsVfmPd3tbg25QraLUVVVJ/OKipF6/5pqt4/qulqZF8tR5tE5xJK+5d3rZGbmncaS9cm8p6dL5kniFGlm1YUamVc6uUV9z0YryrzzqH+/WSjJOF+olHlVlT6GGPN+DebcL7Es877+TplXVdbLPJfTx2hmluJyO7x7Vp+DVD38YQzkL2pt678p85kzrhz2GubOeb/Mczn/FSeYfm4XzLte75/xXrzgvI+62yxe8mmnhkwlWGf3fneb5a2fz9THK17xcZl78/X/+fV1mfo3M/vEJ/Q9+/GPX56p/dVr7s+0v5nZwoX6nl669CuZ+/BcMP/RnxK4AAAgAElEQVQamSfOfJxEPdf2dOl3DjOz6VPeKvO2jd+T+bLWr7l9eJa1fjVzG8rLz9NjS5r5+pHFN8v8/POulfl5Cz/s9uFZvuJLz3vflpalmfvHs2IoWX/lwUHzygr9PVMq6/dPM//tLReaZF6o7ZZ5Ofjv0d5012D6OKLzYZbL6XzLlg6nArOBfv3uUVenz2RFk97/0OEDbg1z5utvlo7GsTLf06/Hn/ra0TKf3OC/V5RGTJL52BH6PPxs9WqZ57ufkvl50/Q5MDPr7dbz1aE9W2XeNEV/g5/dpM+BmVk5r2to/XW/3t/0t2N+lH5mOo/6c/ZjrXrenz6nWealHv39WpdP8aJbocePXNUEmW88ulvmp1bqY7RQoXMzq+7X57qnUo8NySF9niz6H/FHOw/LfNSIweeKEAavj99UBgAAAAAAAACkxqIyAAAAAAAAACA1FpUBAAAAAAAAAKmxqAwAAAAAAAAASI1FZQAAAAAAAABAaiwqAwAAAAAAAABSY1EZAAAAAAAAAJBa4cR3Ofg6dgjhBPR/IvrIJvt5SLO/3qZUGpB58+g6mV84Y4ZbQW/vXpnX1jTKfPv2ozLfsaNb5hWFapmbmdVUN8g8KWf7uUyxWJR5zOXdNkqJrmHM2GaZn3nmWU4PZbeGndsPy3zr0+26gZweiurr9TGef8E83b6ZJVGf60OH9P2yacNWmRcKVW4NHu/Rj7Ek8/6BLplXVvr3vAVnWki8acO7X/z7CS+8DRu22rRp7xk0z+X0MxnMH7vWrP2azOfMfp/uYwjeGZa3fj7T/uctuCFjBdmP4TeP/KPM/+zPPinz1/zx/Mw1eE47bbTMvfn6da9d5PbR1bVd5o88erPbhvKbhzfK/JoPfdpto656vMyXLv3KcdU0HIplPc8sX/nVYa/hgx+8VebvfOf/lPlQzNfLVnzD3UaZ33KFzJe33pGp/TQeW6LHt/MW/p3Mly77XOYa5rV8YNBs0yb9zCK9JGfWUzv4vJwv6+86i/6c3dcwTub1PQd1A+47rluClWv6ZD7lrLEy37Je7x+dItK8dxztOyTzpKpS5iMqm2T+6j+a6dbQXb1N5n09vTI/9dRRMt+xU8/ZO3tOk7mZWfeBHTIvVI+QeeWA/h4Z97ILZb5k0SMyNzOrrdL3/Oypep1g4pl6/65dzvexmQ1U6O+mYlm3kfRskXmM/TI/5SXnydzM7Kwz9VrD4cP6fomVW2VeSjFn5+v0tTi4abfMx43XaxlNL50m82cO7ZG5mdnIAf3c5Tr1+JKM0Oehq6zXOszMYv9ImRdHDD6PRDGI85vKAAAAAAAAAIDUWFQGAAAAAAAAAKTGojIAAAAAAAAAIDUWlQEAAAAAAAAAqbGoDAAAAAAAAABIjUVlAAAAAAAAAEBqLCoDAAAAAAAAAFIrvNAF/K4Yo8xD8NsIaTbKJEmxTX6Yaxh+FRUVMj94cK/Mx4yd5/bxwQ+8Vea/+tUGmf/Pf/iazKurJ8s8Jv51ilGfh1JJ37OeUqkk86Sc4uc+oVLGPT09Mr/22qtkPnGCbt/M7E1v/rDMS4k+T9UV1TLvOLJd5i1zXy9zM7O3/tVrZf72t31E5gNlfa0KhSq3hiQpyzyEbD/nS+KAzPsHim4blRX1eoPgTBvOI+GN82b+OJ6mDWQzdepka22963nvP3PGle42s2e9V+ar1yx63v2bmc1r+Vt3m/kt18o8BD1P5HL6mX1s6WfcGrJ65UU3yby/eFjmaebrrO677wPD3sdwe8VFU2S+euW33TYuvPCjQ1XO8/bGN+n5evnKr2Rqf+7st8v8wx++xG3ji1+8TuZveUu2+Xrl6nvcGrJa3nrHsPdxwcIPyXzx0tsytb9g/jXuNsuW356pDwyNJATrFt9utd36vaqixu+jtvuAzMeP199M+w836g7K7W4N+V79vfBkW7fMgzNn15e3yrzQPEnmZmanTGyQ+f4D+r3i0KF9Mk8zZ59SP0HmRw42y/w3i5fKvN30eSgU9TqBmVkx7JR54yh9PzXU62v5i0fXyLxQk2IZLOpv4AO7Jsp81itHy7yyyl9XevJR/dydM0V/tx3snC/zfduekvk73/4KmZuZDeT0WsO/P7Rc5sWy/j6u7tRzupnZoT19Mt8zoM/j2X2jZP74uqdlfs6p+pkyMzuY1zVWFvSz3xf1c12f99cR8s16rWAgPr+1CH5TGQAAAAAAAACQGovKAAAAAAAAAIDUWFQGAAAAAAAAAKTGojIAAAAAAAAAIDUWlQEAAAAAAAAAqbGoDAAAAAAAAABIjUVlAAAAAAAAAEBqLCoDAAAAAAAAAFIrnPguE5Hl5Z4x+q1HZ6MQgt9IVkEdo5l3nNmlOFHuNs55ilUy/tm/P+pW8KG/fa3MFz/ya5nv3bND5qecMl7mhZoamZuZBefHLsVi0W0jy/4x+vdrPq/vp6Pth2Xe2XFA5r0jx7g1HNy/X+b5wjiZh6CPYWBgQObLli+RuZnZq171Kpk//vhTMq+qbJZ5knjPvT8+xajbCM7Y4Y1vXvtmZgMDPTLPV+jnJperkLl3rc3MkiTbc4XsNmzYatOmvXvQ3LuOIcXrxeo1i2Q+Z/b7dB/O/Z7Lpfi5+TDP1+cvvFF3n+KdxDuOfF63sWTJV2X+2tf+vVuDN197rwxnnnGxzJ98+iG3hqyuv+5rMv/sLVfL/Gc/X6nbv/5Lbg1NjafJfN7c98j8vm9/UubnnHuKW4M3X3suOO+jMj8R8/UPfvBZmbe0XOH24WmZc5nMW1fdmbmPrLz52rNk6eeGqJLBrWj98qBZS4t/LyCdXIxWXxr82QvVei5LEn+uqyrruWjvvpJuIByUcc776DKzkNNzdoz6OHLOd1VX/iV6/06/xq5uvY33alLf3SHzn/90sVvDv/9EzxX3332PzO/d2Sbzs8/S7/odHbtkbmZWiNUyz5UmynzOtKkyP7C7T+bf6fy5zM3Makbq+2nbAT2GXTT3H2XeuvpXbg3enJ2bcK7MO/Z1y3ygv1/my5YvlbmZ2Y036PeCX/x4j8wrK/VaxfbDlW4N48bq+XCW6TYO9+vzVCrqe37Lzi6Zm5nVFnSNuYIeQyc26PGr45A/Pq3f0ynzsfW1g2ZJefB1DH5TGQAAAAAAAACQGovKAAAAAAAAAIDUWFQGAAAAAAAAAKTGojIAAAAAAAAAIDUWlQEAAAAAAAAAqbGoDAAAAAAAAABIjUVlAAAAAAAAAEBqhRPfpVjHjs4adxjaSp6PGKO7TfDW6p3j8PoIIfuJ8I8jL9NcqJZ5Rb7Rr8HLnQ0qK3UN/QNdMk/cCsyC9cu8VBrQDThdFIslvXss6wbMLFgi85y+lJZz7qcQ/J89edvEqPtw73nnfsznKmVuZlZ2TmUhX6FrSJwTmYq+X2LU19I7j54Ul9KSWNR5v66xUFmra0hRhHc/pBmHMRQGv1ZtbXfKPWdMvzJz76tWf1Xmc+e8X+YrWr/k9jG/5Vq9Qcb5esmym2V+wXkf1R2Y2SOL/8ndJksfzc3NbhveE+eNTN587bno5Z9MsVWnTB9+5JZMNbz2j+c6+b2Z2k/jic27M7fx8CPfzrT/4iWfzrT/hz74KXebCaf4c7rS2nqHzOfNuTpT+2Zmc2e/R28Qs70zhJyea8388WfurMtlPhTzdbms32NXtH7ZbQNDI4rv6HxFld635H9vVOW8NvQ921fQ75dVQd9LZmbFxBkbnFs2cZ6Z0186Tua7th7UHZhZdeMImceODpl3154q81H5kW4N3pzdtnafzL05+2ixT3eQ18dgZhacObt+pD6KtjUbZL53+1aZz54ySeZmZh1H9D1/xpSZMve+sYvdegw2M5vbcqbMVy5vknnSvkfmp47S++dz+vvYzOyxDU/KvDRePxP7lun5csK4NPOhPtfdzrJnVaWTW4/MQ9DrTmZmSaLv6Z7YIPOdHXoMzVX4c/apE/U7f2PP4HPBNnE/85vKAAAAAAAAAIDUWFQGAAAAAAAAAKTGojIAAAAAAAAAIDUWlQEAAAAAAAAAqbGoDAAAAAAAAABIjUVlAAAAAAAAAEBqLCoDAAAAAAAAAFIrvNAF/K4QgrNFPAF9ePx1+Bh1nV4JXo1++1mP0SxJsp3rJEn8jZwyi8WyzL3zEK0o84Fipy7AzIrFdpmXSmPdNuT+RX2e0pxHbxsv9650Yn4N7j3pdOJey4y5mdlAqSTzclnfb/khGS29Z9s717pGT0zxXLvjj1NDX3+XzKuqalLUoMfZNNcbL6x1bd9wt5k182qZr1n7taEqZ1DevbS89ZZh7X/xkk+721x4/sdk/uhjn8rUx+tf5x9j1tcK7zwvXHCtzJcu+7zbx9y5lx9XTSejs86ZOOx9zJv7dzJfsfJzMp8+5S0yb9v4A7eGp5/ukfkZZ9S6bSgrVvljS8ucv5F51vl61Zq73Bo8c2e/R+Yvhvm6Ze77B802bd7uto+hMWZCnczz/fVuG3sP7pX5qJzuoz/o77JnairdGk7p1XNJxUh9z/ZX9Mv80IC+n8dOGC1zM7MDew/JfKCgPygKuw/KPGlqcmu4+77/kHlvvkHm3pzd3zkg892rNsjczGzszHF6g4Ien+LBfTLft32zzPv6/RebHVv1tWgq6xr3bH1G5sUe/1oe3dsn896+p2Xe2az76GrX5/HPpkySuZnZ7Je9ROa3Pv6wzCtH6j6Kzd1uDflD+2XuTtlBz5e58bqB3nZ9nczMqvr1+JR1zq6urHZr8ObsI7WD52WxK7+pDAAAAAAAAABIjUVlAAAAAAAAAEBqLCoDAAAAAAAAAFJjURkAAAAAAAAAkBqLygAAAAAAAACA1FhUBgAAAAAAAACkxqIyAAAAAAAAACC1wontLpjFwdexYyjrvUNI0YdeJ48xDkEfWtY2ymV9HvL5vMydQzQzv0Yvj1HXGK3fL8JRiiWnD0+i4+jkZlYa0MfR39vrtqEUi0WZJ4lfo3dPx8S7ls7+KW6o6JzLXIp7UgreefA7KCcDTgu6jxB0HykulUWnTu9cl50a8+aMPc5za2YWgx5fQtBjbM6pcaCv062hoqre2SL7OA1t6tTJ1tp61zD3knVgyG7Fyi/IfMG862XuPbPLW2+V+fkLPypzM7NcLtvP/89f+BGZNzZ5z5vvhhv1edz4+Pcztb9wwYfcbbz5+sq/+aTMv37HTTJ3hj47d8rb9AZmNrp5sswfefRmmW/etEvm55x7iluDx5uvz1twg8zbNv4gcw3efD1jxl/LvLpitMyXr/y6W0Prqn+W+eyZ75a5N197Zs94l7+RM1+vXntPphqGW0vL0he6hN8fIViucvD7oS4elrsf6fHnmeDM2R2jnXfcQ7r96iP6fjYzO1rZLfPGDr20kXfm7M7+J2TeWzdS5mZmuZw+jh2H98h88oQmmcfY59ZghxtlXFWt25hYr+eSwyVn/zlnyNzMrK9X3xC7tug5fcZ1r5L5I23bZL5u889lbmY2qvl0mR9JemQ+7vTJMt/Xsd2t4eC2Kpnnol5LaCrp+bS9p1bmu9r1/mb+nN276//IfFRBz9kV+QluDYc79TduTZ3ev5To/fvX6TGybnSF7sDMEue9YMvWDpmfcZZ+LvOja9waOo7qMa46VA+axTD42hy/qQwAAAAAAAAASI1FZQAAAAAAAABAaiwqAwAAAAAAAABSY1EZAAAAAAAAAJAai8oAAAAAAAAAgNRYVAYAAAAAAAAApMaiMgAAAAAAAAAgNRaVAQAAAAAAAACpFbwNQgjVZvawmVUd2/6HMcaPhxBeYmb3m1mzma0ys3fFGAd0a9EsJKo3Z/cUa+BeEzH6bajmnfbNzKKVvQ2kXE4fp3cMuZxfpNdGkuhjCLls5zENt8ZYGvYazNT9alYsFjO1XiplP4bEOU/e/RicxyqkuNQh43PlnWe/AH//Ukmfh6TsnMeCd4z+OfDGj+BskHe70OchpjhP3mEkiW7DOwaL/jPT19cj8+qKStH+8I9NL1ZDO19r06dfJvO2tjuzNJ/KylVfkXnL3A+6bbSu/KLMl624ReYL51/n9qG4z4v5c+HLL/h7mT+29LPHVdNz1uDlTo1nn/VGmT/+xIPHWdH/a23btzLtHzNOQ5s23p+tATObO+evZf6d735G5t1d/W4fr/vjS2W+bNV33DaU6VPfJPNXv2aG28aVV79f5t58vXzd12XeMvsKt4bW1XfIfPXae2Q+e+a7nVxf69Xr7pO5mdmsGe9yt8GL15DO2aFsMd85aPz0zjq5++iqruOq/bmUnj4o86MF8e5mZs31/nf+2Eb9Lv/MMx0yHzmmQeZVoytkXur15+zEmUxeNnKczLc+/aTMa182263h0mvmyfzGGx+R+fbObTJ/3UWvk/nSdRtlbmbW23lY5mfPO8NtQ1n+1E6Zj2g+3W0jOAtL2w/r+837xn796+e4NXz1y/pd99SzTpH546sOybxQq5/9rr17ZG6W/Rt75yG9HnJK3OXWUJ/vk3nhcJVuoFlfrMJofYy1fXpsMTPrqtL3y+6j+hv5T1q6ZT5hbN6tYfv2epkPjBx8DEzE7yOn+U3lfjN7VYxxppnNMrPXhhAWmtlnzey2GOOZZtZuZvoLEwAADCfmawAATg7M2QCAk567qByf9dsfYVQc+79oZq8ysx8e++f3mtkbhqVCAADgYr4GAODkwJwNAPh9kOpvKocQ8iGENWa238z+w8yeMrMjMf7n3x/YaWb6d+8BAMCwYr4GAODkwJwNADjZpVpUjjGWY4yzzGySmc03s3Ofa7Pn2jeEcEUIoTWE0Fou6791AgAAnr+hmq8PHDgwnGUCAPAHb8i+sUv+31MHAGA4pFpU/q0Y4xEz+7WZLTSzxhDCb/+H/iaZ2e5B9rkjxtgSY2zJ56uz1AoAAFLIOl+PGTPmxBQKAMAfuMzf2AXnf4QKAIBh4i4qhxDGhBAaj/3nGjN7jZltMrNfmdmbj212iZk9NFxFAgAAjfkaAICTA3M2AOD3QcHfxCaY2b0hhLw9uwj9/Rjjj0MIG83s/hDCP5nZajO7cxjrBAAAGvM1AAAnB+ZsAMBJz11UjjGuM7PZz/HPn7Zn//bTcYmxLNK8t3eK9vU2IXhtBKd9Vf+zcrkKZ4tEpzq2fF6fpyTxawxBH+d//VtXg/RRkrHlQqVbgydxToR3DFbOuL95d4NZsZjtb5gNDAzIPM39ZkFvkwR9sfxnJs150v/SQ5LoPvJO7t0LMXhjh1m55Dz7iX+csgbnPD67jfNwZ+5D59H8+8mrMZge3xJncMgF/xxE0230Dwx+HFnP8clsqOdrpa1Nf+NOn3a520Yup8eN2bOulnkIev9Vq7/q1jB/3ofcbZTlK27LtP/iJZ/KtL+Z2UUX3pRp/z/9k8+62/zoJx+R+WdvvjZT7lm6zD/Ps6a/U+Zr2r6dqYZ77vmpzG/+rL/uM2rU6TJfuepbMt+8eafM6+r9fwX+4cXfkfm8WfparVjzeZl78/XtX/y4zM3M1q8/qPMND8h83pwrZZ5mvp4z61KZr1pzt8xXr71H5rNmvMutwZNE52X8BJg2XR/H+rb7TlAlJ5+hnLOTgcS6tncNmg/U1utaEv97xpuz8w21Mm90vqo6y/67fK5zhMxHjNL7t/d1y3zKxEaZ19cWdQdm1lutx/kD+/bqBsrtMq7WcSoju2pkXt/fIPMn9z6pO+jV18nMrKa+SeZdBw+5bSivmjdP5it+dZfbRnPzaTI/c/JMmXtzzb33+P8SwpRxz/Vn1v/Lys36hjjar9dk6nPOd1NZjx1m/jf2Ba/8c5mvfGyxzL11BDMzq9H3dH6EXnPJ14yWeanjsMy3Ff35uLlSf4ePGqmfu7PqnG/wKl2jmVmjs2ayd+e+QbPSwODj33H9TWUAAAAAAAAAwB82FpUBAAAAAAAAAKmxqAwAAAAAAAAASI1FZQAAAAAAAABAaiwqAwAAAAAAAABSY1EZAAAAAAAAAJAai8oAAAAAAAAAgNQKJ7rDXBTr2DHqnfNJmh5k6nVhpjcIIbgVJImu02sj5+RJUtL757Jf1uicKC+34F8rb4tgeacB51oE5xica51GuZytjYEBfS1D4pwDM8uVdR50F1YsFmVeqKx0awj5Gp1Hncfg7B9qZZ7L+T8fKw30yryc6PMQgnct/Hs+ONtE557OOeNbOXo3g46fpWt0xx93/PIrCKaPI4rxZSiea/imT7lU5m0b787cx6yZV8t89ZqvynzunPe7faxc9RWZz2u5RuYL5v2dzJet+Jxbg+f8hTfK/LGln5H5wvnXyXzp8luPu6b/7vrrvyDzW275cOY+PGvavp1p/+BMI+9+92sz5WZmLz9P3y9zZv6VzO/99qfcPjyvvOhyma9Y881M7W/Y9K8y/9CHP+22cdsXPpqphqGYr1etyT6GKWvW3TcErejjmD5V30/efL12/b+4Faxv08cxfdrbBs22PPW02z7SScysT7zkjanV71VNo6rdPnoP6ueq6HyQNFY738cDA24N9c43TWdefwM3VenviYOH9TFU5v3ztOfQLpnXN+j9J54+Q+ZJOOTW8KlP/VTmnTV9Mq8fr9+lN615SuYVpdEyNzMbdYY+Efv1Z5s9suwXMm8/sEPmF8yYpDsws8e36fyJp1bKvFj8C5k31Pv30+aD22U+esJLZR6C3v9In/NxWKHvFTOzLY+vk/mBjftknsvpsSVG58E3s3079HMxcfI4mfcOHJX59h36hhxTkeL7s16PgWefol9E87lumY+M/hh6YECf64kNIwfNeg4Mvi+/qQwAAAAAAAAASI1FZQAAAAAAAABAaiwqAwAAAAAAAABSY1EZAAAAAAAAAJAai8oAAAAAAAAAgNRYVAYAAAAAAAAApMaiMgAAAAAAAAAgtcILXcDvCiHIPInRbSOa3iaX03247aeowTmMzH3kcvqypavRKzLRaeyXeb5QcmvIOX3kQtFpoCzjqJtPSTdSLOrz4CmV9HlK4oDfSNDXe2BA97Fz+wGZT5v2EreEl7xkpMwf+XWrzBsaR8v84OHHZf7Kiy6XuZnZM89slXl/n77fGkY4z1Xwbzjv0Yym7+kkDv/4lfVnjeXEOQ/O/XqsFR2naQKZbNiw1aZPuXTQvG3j3XL/adPe4/axfv1dMl+z9mtuG0q6+z2brH0snP8Rd5tcLtszuXT5rTK/+OJPum089NBNMr/llg/L/Jyz/0LmI+tPlfmylV+SeRpveeMHZf6DB78o80WL/k3mX7z9W24NY0fr+fREzNe/fvib7jbKnJnvlPnufZtl/o2vf87t418feETmf/Gml8t8+cpFMp87+zK3Bs/sGe+SuTdfr23z7xefHhvaNnxH5tOmvF3nU9963BXhhVEoFGz06FGD5o2n6Pvx0B7/e6axtl7mI6srZH6k2OP24dld1Pd8gzNdemsJe9YclnnpnCbdgflz9r5n2mVerNM1js9PcmtojkdlXq0/22z/Eb3/mLG6hoM7xukOzOzQjr0y37fxKZmfeumFMt9Q3yXzR1f/VOZmZmNGTZZ51jn7HZf8uVvDti36nty8eZ/M9yX6Wh5t3y7zV170AZmbmSXlvMyf3KmvxYF2PTZMnOQvWSbF/TIfGNA3fd4ZAkc36hrqU6wDeHP2EX0prPE0PTbs6NDXwcyssqDbiMnzW2vgN5UBAAAAAAAAAKmxqAwAAAAAAAAASI1FZQAAAAAAAABAaiwqAwAAAAAAAABSY1EZAAAAAAAAAJAai8oAAAAAAAAAgNRYVAYAAAAAAAAApFZ4oQs4LjHFGnjQcZKUZZ7P54+joOcWo+7DzOkj6P2TRO+ey/mXNUmibiOvOxko9si8vs4/j8F0H42N9TLv7euS+ciR+mYol73r5F9L91rr02yJczFDdC62mZlzT+dCpcwfePBnMn/t6+a7JXzqM38v89s+f7vMDxxsl/lNF98s81e84hUyNzN7x9uvkHl9wwiZl5OizGN0LnaKbRJz7idnfPPaD8F/Lr02YsYavXvezCwE/1wOLsu+GCrr19+VuY2ZM66UuTdfr1q9yO1j7pz3y9x7Zlas/LzM57dcK/PlrXp/M7OF829wt8nioYduytzGzZ/5qsy9+Xrz41+SectcfZ3MzIr9B2R+5plj3DaU9773DTL/0m33uG088ph/vZW/ueyTMk8zX+/ec0TmEyc2ynzV2m/L/F8f/IXML37DK2VuZvb6114u879408vdNpSVq+/MtL+Z2ep192VuQ5k5468ytzFt6ltlvn7j92Q+5dy3uH1s3PQDmU+d8maRMl8PqTj4C9iTeyvkrpOaG9zma5PtMt9/dKTM83n9fVrb59dgNUdlnDhrBV3FKpk3n6m/BSY06v3NzPYc0d8LFbU1Mh/VpN+zR7brbyYzs6O5Dr1B0MfhzdmTqnX7yWn+s13s03N2jHrOfuyxjTI/tPmwzIPz/WxmtnvPXplPm1Qt8wf/9ecyTzNnX+q8e3zrc9+VeX3pXJn/0T9cIvN/+d4SmZuZ9Rx5RuajR42TeX+dfiZCvX8/VY86XeadR/UY2HekU7c/Xu/vPFJm5q+/jTxNfyM3O2t8B0r+PV3s3CXzqpETRTp4/fymMgAAAAAAAAAgNRaVAQAAAAAAAACpsagMAAAAAAAAAEiNRWUAAAAAAAAAQGosKgMAAAAAAAAAUmNRGQAAAAAAAACQGovKAAAAAAAAAIDUWFQGAAAAAAAAAKRWeKEL+F0xRpnnQvY2QtCNJEni7O+vw3t9mOk+zPJOB3p/7xyYmeVy+jhKxbLMG0c0yPx1r3udW8O+fd0yv+Dlr5H58uU7Zf7klv0yr6qqkrmZWXQupXe/ePLOtU5zLctJv8xra+tkvnxpm8yvuupTbg1XXvFOmd9440dl3ttXkvlTW7bK/B3veK/Mzcy2bjsg89q6RpmXirrGFJfKzFIMYrIP3UnZKyLN2HA8BT1nF3rscKIieKsAACAASURBVIfHFG3w89AX3vQpl8o8BGceM7Po3AzePLVq9SK3D8/KVV/J3IbkzNdpLF1+s8wXzr/O2f/WzDXs29ch8xtufJ/Mvfl6+vTLZN7WdqfM03jjxf48kUWa+Tqrf77zJplfeYU/X3/jjo/J/PDBAZk3j66U+ajmiW4Nnv/9029m2n/2zHfLfPXae9w2Zs24ROZr1t17HBUdv7XrvuNuM33a22S+fsP3MtWwcdMP3G2mnPsXzhbM1ydKsMHHoDFl/a0wcKjdbX+vjZB5Lq/n9KryOJlvP6THHjOz+pG1Mu+v1O8ejVV9uoOcnrOrXzZe729mZ/To78+nNup33K6+PTIfaDjLreHSy/5S5t/97kMynzt7rsyf3KK/qbqO6G8mM7Pzpuu5oq+sr1WznSHzjtMPyjzNjF13ZJfM9+RnyPyUSr3WcfXVn3Zr+KM3XiTzKz6q35/u/9wdMvfm7Ioqfwyf8LIWmW/dq69l7RP6HbF9a9GtoaG5WuYDPXr/nfUVMp/SpGvobdfrLWZmhUKvzIsH9F25r0mPweWo2zczqxihx+GYiBpExEwPAAAAAAAAAEiNRWUAAAAAAAAAQGosKgMAAAAAAAAAUmNRGQAAAAAAAACQGovKAAAAAAAAAIDUWFQGAAAAAAAAAKTGojIAAAAAAAAAILXCC13A7wohOFskbhs5020kThP5fN7twxNjzLR/CLrIGPXPArz9zfzzYM55tFgh40VfudOt4TM3b5N5fX2DzAv5GpnnnB+ZpLlO3jaJd8s68nl9HtNITNdYSsoyr64ZKfPFj653a3j419fIvH6EvlblqG/I9sNHZV5Xq4/BzKy2tlnmpWJJ5iHosSE6x/DsNs49595Pug9v/EvDqzFGfT9lbT/dNqKGjOMvnjV16mRrbb37ee8/Y+p73G3a1t/1vNs/UebOeb/Mc85Es6L1S5lrWDDvemeL7M+951WveqfMN2z4kcwf/NfPZOp/7tyr3W0G+g/JvG399zPVcMfXfiLzx7c8lKn9NOa3XC7z5a3fzNxH8+jKTPtf8u6/k3ma+Xr9xu/KfPbMv5b56rXfkvmsGZe4NWR9l58x/a1O+8M/X0+d8uZM7W/Y+MPMNWza/MCgWUtLy3HXhOcWS8F62wf/rM+9tEnuX3VEj59mZuVwUOaNNY0yHyhul/nkRv/9cn//aJmPrO+SubfWcHRgt8z37PG/23p3uptIufw5Mj98tNdt49WvfpfM//Q1r5R5e6e+H3r275F506TTZG5m1rZ1l8zf8PoZMi+bXkco1Jwp81PO1tfazKyzUd/z3Tt0vvJon8zP2ae/Pc3M1rc+KfPepU/LfOdOfa0+d+l1Mk8zZ48/dYTMD+7R5yFXqb9vRzT5S5ZJoueiIzm9FnFWU7fMS116HaBQ8J9LT8UY7xtcP5cVeX0dzMzKJd3Hvs7Bj7MoFr74TWUAAAAAAAAAQGosKgMAAAAAAAAAUmNRGQAAAAAAAACQGovKAAAAAAAAAIDUWFQGAAAAAAAAAKTGojIAAAAAAAAAIDUWlQEAAAAAAAAAqRVOfJeDr2PHmJyA7nUfSRJkHnT8bBe5bGv1SaJrDCHq/VP0EazC6UMfaE9fSe8f/VurqvJU3Udvl8zz1i/znFXqAmKa66S3cS+1c78UKvTVilZ2OjCLUd8PXl4u6SJra5pS1KDbKJWde9q5lqObRzr962M0MysOOPesczG95zJG/1p5ojP+RO+Gcp7+NOfJ440NWcevZ9t4/nVmP0KYmW3YsNWmT7ls0Lxt451y/3Ub7nL7mDH1PTKPOX2vtbXpGmbOuMqtwZuvs97vnnkt17jbrGi9PVMfngVz/Rq8+Tqr2TMGv9fMzFav09fazGz6tLfJ/I1vfK/MH/jhIplfcfWfyPysl/2pzM3MmppOkfmyFd+QeVlPYyfEnFn6uX1m609k3jL7CrePWdN1H9587Vmz7l53m5nT35mpD++9qG39/TKfNvUtbh/rN/wgUxve/lOnvNGtYdPmf5P5OWdfPGi2ddsWt32kEwrRqpsHfwMq9/bK/RtqD7t9HO0eLfMj3c63QLEo88qqGreG6ppumUfn2857v+zv1HN69w59DGZmVfX6+7W701lr6Nsp8yPJWL8GZ85euV3X2NnRIPMR4/Q3z6Eufb+ZmTVU6mu14YndMv/Epz8m8/u+q8f5nY9vlLmZWWPjRJnX1x+ReblUL/PYpHMzs/KRfTJ/4JlamW97aq/Mm/O6/1Dl13hwd7Zv7P6BPpnnKye5NYT+/TKvrdJjnPeJPKFBH8OeDn9s6C3ptamagh5/9hxul/mEZn8toqukn80x9YOfiO5Dg9fHbyoDAAAAAAAAAFJjURkAAAAAAAAAkBqLygAAAAAAAACA1FhUBgAAAAAAAACkxqIyAAAAAAAAACA1FpUBAAAAAAAAAKmxqAwAAAAAAAAASK1w4rtMhnnfINMYvXX0qFsPOn+2D68NXaMv+88CYiw7W+gavWPI5SvdGioSvU05KeoG4oDbhxKc62Rm1tfXJ/NSWdfg9VBRoa9lT0+304JZ08gsz5SvVPLPcxKy3ZPeE+Ge5xTXMuR0L0nitaGfmTQ1WPCuVcZn2zmGNCNP4ty1qY5TtZ/492uWMTLr6Irf9fzHlulTLnG3adt47/NuP53hn69XrvqKzFvmfjBT+yfCspW3u9ssmHf9CahkcC2z3+tu483XDz64KFMN9937E5k/seXHbhsL5l4u87mzL3Na0HPEnJl/49awau0/u9vI/dfclWn/1tV3ZNp/KMyY9g53m3Xr/yVzG1ms3/ADd5upU94sc298m3Lum5z9/Tng3HPeIHM1xr3wo9/vm8Gv14Hdh+We+aaRbuuFQpfMm0Y0yLynQ4/RvVbt1mDOPZ1zXqPbK/X+oxpP82twjM/rc93XrM91qbZZ5ps373drGNswSeajOkbIvK92lMy7B9plPrq2TuZmZke7OmQ+skrfT4u++0uZN5TGyHz8aafL3Mys74j+/iyNmChzb61hw+qdbg2TF46XefuT22VecAbac+bo87Bt0zrdgJl1HtDnYdS5Z8l8f/9Bmcej+9waQtTrRjW5Ct2HM98dLtfIvHbAv+djhR4bvE/k8c44nZS99T2zurw+DvVZom4lflMZAAAAAAAAAJAai8oAAAAAAAAAgNRYVAYAAAAAAAAApMaiMgAAAAAAAAAgNRaVAQAAAAAAAACpsagMAAAAAAAAAEiNRWUAAAAAAAAAQGosKgMAAAAAAAAAUiu80AUMtRij3iBk2z9JEreGfF6v1Xtt5HLe/uVM+z+7jbdFXqYx6hOZ4jRZCLqN2so6mff1606i6fNUTnpkbmY2anSDzN/whjfI3LndbMGChTKf2zLTacFsz64jMveulZlzsbzdzSzEUqY+onu/eTeUf8P5Q4PewBsbvPvt2T70yfT70Ll3qdzxMUUf7v4p+hjONrL3DjOzqVMnW2vr3YPm06dcmrmP6VMukXnbxnsztb923TfcbWbNvFrmaeZ8pXXlFzPtfyIsmHe9u82yFbfI/Lx518h8yYrbZb563Z0ynz3z3TI3M3tyy7/J/K67fyzzSy/5U5l78/XFf67PgZlZtKLMV66+R+ZzZunnbtXawZ/ZtGbP/Gtni+Gfr9esu8/dRpk57R0yTzNfe7x5qm3D/Zn7yFrDxk0PDHsNnnPOvnjQjPl66BSTsu3tHfx7YHzTSN1A8K9GuVwj856B0TIvxgqZ5/P+2JAk3venfrZr9BBs5QGd19T26Q3M7GjlBN1Hrl/m4eh+mY9rmOzWkC/qOrd3bpf5qNFTZL65q0rmp5b8b+zJZzfL/MI/ukDml7791TJ/8onDMn/gwf+QuZnZno5umb/0dH2tn9q2V+ZTxupnysys75B+NjuCrtGiXvRZ97izfzhD52Y24vSdMm+qb5L5/gNdMo/Bn7NHVIyS+ZFih8wrCnr8GuUMoYfa9XNt5q+PjR+h17529Oo++jsOujUUG2tl3lgefJxWdyK/qQwAAAAAAAAASI1FZQAAAAAAAABAaiwqAwAAAAAAAABSY1EZAAAAAAAAAJAai8oAAAAAAAAAgNRYVAYAAAAAAAAApMaiMgAAAAAAAAAgtcILXcDxyb4GHmPUPYSQuY8kKcs8BH0cXo1DUKIlSSLznFOjma4xjeAcSAwVMq+sHinz/oGjMu/tPSRzM7PTJk+W+cYNW2R+1dWfkHltbb3Mx44dK3Mzs21bd+sNYqWMQ04PAzHq+3kolEv6fsw7I1U0v0b3uXJuaW9/S/Fces+dRWdscJ67rLlZiuN06WN0z4H5YwOG34YNW236lEsHzds23i33V/umbcMzfdrluv3133Tb8ObrdW13yHz2rPfKfPWaRW4NWc2b+2GZr1j5BZlnf+b9+XrBfF3jsuW6xt6+DreGiy/+kMwfeug2mV955fDP1w/9r9tlPmv6O2W+pu3bbh9ZrV77rUz7z5pxiczTzNeemdPervtw7um2DfdnrqFtw3dkPnXKXzot6LkwzXPpbXPuOW+U+abND7p9eM45+2KZb378oUGzlpaWzP3jWRW5vI2vaRw07y7Xyv3rcj1uH6Wj7TLvd/YvO/NERdm/5xPnPXlMY5XMD3fq567fumU+dkyNzM3Majr0GLerZ4TMm8ZXyzx2+uep1LlT5rlRL5P5/oObZD6qR5+n3tgpczOz2vzpMl/8w5/JfPmyVbp9Z84eM2aMzM3Maur2ynzz+h0yb56or2VHu/9NFOOTMp875lSZbzmsn8wjJb1eEq0kczOznh69JvP4Vq8PreOQX0Os08c5aXyTzHccOiDzXTudOTufYs52voH3durnZnSlHt86xjf7NXTpGmpPHRg0y+0Z/BhTr9KGEPIhhNUhhB8f++8vCSEsCyE8GUL4XghBr1wBAIBhx3wNAMCLH/M1AOBkdzy/+vtBM/vdH1t91sxuizGeaWbtZnbZUBYGAACeF+ZrAABe/JivAQAntVSLyiGESWb2ejP752P/PZjZq8zsh8c2udfM3jAcBQIAgHSYrwEAePFjvgYA/D5I+5vKt5vZ9fZffwBslJkdiTH+9g+c7DSzU55rxxDCFSGE1hBCa7nUl6lYAAAgMV8DAPDi97znazPmbADAi4O7qBxC+FMz2x9jXPm7//g5Nn3Ov9wcY7wjxtgSY2zJF/QfKwcAAM8P8zUAAC9+WedrM+ZsAMCLQyHFNheY2Z+HEP7EzKrNrMGe/clqYwihcOynqZPMbPfwlQkAABzM1wAAvPgxXwMAfi+4v6kcY7wxxjgpxjjZzN5mZv8nxvgOM/uVmb352GaXmNlDw1YlAACQmK8BAHjxY74GAPy+SPObyoP5iJndH0L4JzNbbWZ3Dk1Jwys36L9EdMx//hmr5xZC3u3D7SKWnT68HtwNUtA/T0hMnweLev9czv9z3eWyPg+5UCXzECp0bgMyP+fsaTI3M8vlizJfsWydzEc01Ml83742mY8ZM07mZmYzZkyX+VNbtsv8aGevzNNcS0+M+qmodv6tva6eTplXVOh7wcwsOA9WTGTsShK/Aa+Gsulnwm3//2/vzoPrvM77jj/n3osdxEKAi2gttBQROyWKsiIn7SST6Uzc1pbkLXYnsUVJqTJp3SSTdHEz7TjuMm0mE1edJul4kUWO4zZ2ZUmm3dWT2FEnGaumRJEgQIKUKJGUuAAgCWJf7r2nfwCZaGTj97zA+wIXkL+fGU8I/O57znnPPe/73PcAgZx5/us/26foNrz30uPNAdbEhqzXfd2PyLx/8CndQNmpUwkc7/+CzPt6/77M+098MfUYPPfs+0cyT3tNeZ9JzMzes/83Zf6DFz8n8/vf89sy33f3p2R+auhZmZuZfehDn3Zfo7z5xqjMs6jXn/ylfy3zl/v/2G1Duavvk6mON/Pv8cdPfEXmLx8/lHoMd+39uMxTlqFEens+JvMTA1+T+cDg12Xe3fVhZwR+vT55yr8u0ujseHBN28ePtPp6LS6MUs2IPrZY7zafb27V3U/pZ6bYoGvVzmr93GZmNrygHxhGxvTflt75Ln382MkxmY8O+fsAtsV5Pg36s8vYhPPclaBmx+bdMh8u6fPckdsm8/qaJpmX4y6Zm5nlnKkcuHJF5tuczz5+zd6uB2Bm7W3dMm+qvSzz7UV9kidnbrhjsBm9ZmMclvmeO/R1u31U15rjY80yNzN3g817xvZq+vbtej2amc0sXJf57FU9T9ub6mR+5YYzyJDgGbtar9mWWn3vOHVZH7+j3h9DDPo85kvLbw/HuHz/K9pUjjF+z8y+t/Tvs2Z230qOBwAAa496DQDAxke9BgBsZul/BREAAAAAAAAA8GODTWUAAAAAAAAAQGJsKgMAAAAAAAAAEmNTGQAAAAAAAACQGJvKAAAAAAAAAIDE2FQGAAAAAAAAACRWqPQANpsYS/5rQpB5cPJyjDLPBednAfrwpTG4r3DisoznF4ruGGpqqnUb87qP2poGmeeCbt9ijc7NLJ/Ly3z37ttlXlWlj29rb9cDcObZzKy5sUk34bzZCwsLMq+trXPHUIy6jeoqfXxn13aZT07oc3z93LDuwMwWigkujDUWnWvbe7+944OlOz7pa9Lxf5aZ5D6LtdXTs9uOHHlq1cf3D/rH9nU/kir3+vCONzOLuXT1eu/eX5b58eNfcsfg8+qAHuP+ez4l8xdf+oMVjueH7bvrN2R+9NgTqdp/6AP/xn3Nc9/69zJ/+BO6jW//9/+0ojH9UPsHPuO+xqvXaQXv86GZvXz8YKo+PvGLvyXzyYl5mT972J/nY8f/ZEVjerveno+lOt4sfS3s7P6wzE+d/IbMuzof8vvoeFD3MfRNtw2Ner1ZlEKw8drln2vai7pO1O701/v4JZ2XG/SH/cLcrMwvmP9cFp0lGYI+j8sX9RgKjTfJ/Nq0ft4xM9sadc32PleUJnX71QmeDWONfk3rtH6G3nWLM0/Vep68Z3gzs+vXr8i8c49+xp66rs/hjlv0PkB1kz7ezK/Znbfq85ye0uslhAl3DG1bdN6Q1+9VbYPe7psst8i8vclZkGZWLrTJvHFEn+eb43Myv/DKmDuG9lv1tX/R2drKjetadsftuv2zZxM8Y8/rPoqNepAtzg0wRneDz7xnivyN5fsIYvj8pjIAAAAAAAAAIDE2lQEAAAAAAAAAibGpDAAAAAAAAABIjE1lAAAAAAAAAEBibCoDAAAAAAAAABJjUxkAAAAAAAAAkBibygAAAAAAAACAxArr2VkwsxDCsnmMcR1GUU55vL8Pn3NOIy4/BYmUS07/3gAWW5FpCPo8i8WizPfdc5c7goX5MZlfeGNY5qMjl2Xe3Nys279wTuZmZvPOGMtxTualkn6zCtVVMi8u6PbN9DVlZhZM97FlS5vMb0xOuGNobMjL/O59XTK/7eZGmdfXt8v8whv/S+ZmZnNzes3ncnrNl6Ne8yH41513jys7eS7qc/COz0La+7SzXLFBDA68bnt7Hl0299ZB/+BTbh/ea/q6H3bbSDsGT1/fYzI/fvxLMt/b97g+vv8L7hheOvqH7mvSOHDgj9zXHDz4D2S+RZdb13vv/02Zz8/6daiv+6My9+q1p7f3IZknqdenhv5nqjG8Z//y16SZ2fTcVKr2zcwefeyzMv/KV39f5r/3u4dSjyGtEwNfk3lP90fcNgYGn5Z5R9eHZO7Va8/6PBNpp4aedV/TseeBdRgJPIUYrWVuftl8wTl++kaC59umVpnnJ0ZlPllo0HneX/M36yZsdkq3sd15/h3Jzcp8quRf17u2Vsu8Oa/neseOLbqD/M3uGOaKF2V+8YKuFTe3dcv8B0ODMk9Ss71n7JEr52W+o/YnZH5l4VWZFxeWv17+yi1bnbmOLTK+sXBV5klqdmjVfdx6x249hqkZmf/UT++V+UuvvCBzM7MzF/R7ZcUmGVfn9V5F2236HMzM5vJbZZ5f0HfB4NwbBl/Xx9dEvRey2IfWkNsu8xs7R2Re6+xlmJnNjunzbKhb/v6Tyy1/jvymMgAAAAAAAAAgMTaVAQAAAAAAAACJsakMAAAAAAAAAEiMTWUAAAAAAAAAQGJsKgMAAAAAAAAAEmNTGQAAAAAAAACQGJvKAAAAAAAAAIDE2FQGAAAAAAAAACRWWM/OokUrl4vL5iHk12EUaffRy+4rYvRyfZ7RaSBYkHm57I8xBN2GxxvjGxcuum289723y/wnf3K/zEeGb8j86NGjMp+dm5O5mVlT060yn5iYkPnUpO5jampa5oUa/5ooVOn83e++Q+ZXhq/JfG7eHYL19nbKvG1rk8zfvKTfyxe+/+cyLxb9W1nI6deU4/L3JjOzEPSa966JxdeUZO7dnZL0sZbHm/n3Dq8Pbw6wMXT37LYjR768bN7X/cg6jGLtf+7d1/2wzPsHD8m8t/dRmQfT9/C9fY/L3My/5o4d/3yqPrZv3+6O4V/+i+XXgpnZ888/4bahtG2tkXkW9fqZZ39f5j91v14L+aCLbZJ6fe/+D8v8yIvfkPkPXtTvQxa+/ORnZP74478j8xe+/5LM/8k/0/OcRHfXR2U+ePK/yXxg8Gm3j67Oh2Q+dOo5t400sqjXaXXsecB9zdDpw+swErhCsFBTvWxcntUf5ucvOw8TZlbfoJ8XLK9r1c46fY+crfdr/uyYfm4an7yqG2jS9c57Or6p2R/j9Ih+Dh8e1/N48bJ+r7Zt82uN94z9+rkhmR+9eE7mba3LrzUzs7l5nZv5NbutoV7mA6+cl3mhqMdQVeM/O47N6/dqanJB5u13bpF568iMO4Yzr+lacGxOj3F/X6vMv/pfnpH5+OyozM3MWlu2yXxsSq/peG1W5iHB80ChdUTmW25ulnlxVo+hra5B5iPn9J5QEtfm9WfdRr3cbGpG72WYmdU779Xo5eXXZFH0z28qAwAAAAAAAAASY1MZAAAAAAAAAJAYm8oAAAAAAAAAgMTYVAYAAAAAAAAAJMamMgAAAAAAAAAgMTaVAQAAAAAAAACJsakMAAAAAAAAAEisUOkBrLcYY6rjQ0iyD192xlBy+sg7x6c7hyRKJT3GXE7Pw/DwiNvHd/7PazKfn9dj+Ik7O2Q+evWizPffe7fMzczee/99Mn/t7BsyP3HijMwvXx6W+b57emRuZvbmm6/KvLe3W+ZzLx2X+eTkpDuG4StXZP4Xz39X5gtWL/O62gaZhxBkvkhfNznnui2X0193a33trse9YT36QOUNDLxmvV2fXDb36lQW+gefknlv1wGZJ7svaD09D8t8YOBQ6j7W2vH+L1R6CPYLH/2MzL/9P/5d6j7++Ct/ker4v/y+fi9/5m/+msyT1Osn/uOvyPyzv3NQ5i869frw4c+5Y3jfz/+qzM8MDco8bb3Ohq7XXZ0fkvnJU8+4PXi1rrPjQbcN5dTQN1MdD7xVLJetODOzbJ5znl/v2O7Xy9HXamU+1nhe5tG2yTzM+2MIzmf5xsatOq/W944zo6My31mqkbmZ2WiV7qOtoJ9va5paZN6U9z9/ff25IzIvNjTLfP9OvV6OvlIl8327dsnczOyBD/6MzP/8yDdkvqtH9zFyekrmd+/Tz8dmZqMT0zK/89ZWmXs1++JUtTuGPZ16u+7M6WMy/9+X09XseH63zM3Mxu11mVfnb9J9tOhzXCjOu2PIl/V1MXF+TI/B2yeY0+vJavQ8L9L3uCqblfmEMwSneTMzqx2+JvPG9vZls0tXlr8v8JvKAAAAAAAAAIDE2FQGAAAAAAAAACTGpjIAAAAAAAAAIDE2lQEAAAAAAAAAibGpDAAAAAAAAABIjE1lAAAAAAAAAEBibCoDAAAAAAAAABILMcZ166yurj3esfv9qz4+hJDhaFanZP58eePMxXKqMYSQd17h/6wgppxL973I+WPI26zThD7P2Zl5mVdV6+MLBW8ezaqrq2U+OTEj8xiqZF5XXyPzhsaCzM3M6mr0e3H+wjmZe/eAUtFfrwsLCzKvr6vTfZg+z1Kp5I7BkzN9HjHqPrw1n8W9dD3vx+9U584/b7OzY5UvFpucV69PnDy45mPo635E5v2DT8m8u/uA28fg4EFnDA87Yzjk9qHs7fll9zVp63X/iS+mOn493L33gMyzqNd/+f3/vJIh/ZD77vuUzNejXp848YzMOzsedMdwauib7muU7p6PyXxw4Gsy7+j8iNvH0KmnVzSmt+vqfEjmJ08957bhzWXaej10+nCq45Po2PPAmvfhnYcaA/U6O17NjiOj8vixuMPto7VKf5YPWydlXtOmj5++Ou2OYX5Un0dde5vOd7fIfOqifq7LzSd4hs85SzrqvGrrVpnPTelzMDObadD5jUtvyPy2ljmZl+eLMs/n/X2Amhr9DGy14zKeeU2PobB9m8wbGv3PFXW1+r26cOG8zGsKt8p8cvo1dww3ZvVnm5myPo9dTXrNXq1ulHnLjH4fzMx2der9julX9Hs9XhyT+cx1vedjZlbb6nwGK+uaHU2/1zsaWmV+Icm94ZKOa9r1PLifOhI8L+wq6vO4VLi+bKZqNr+pDAAAAAAAAABIjE1lAAAAAAAAAEBibCoDAAAAAAAAABJjUxkAAAAAAAAAkBibygAAAAAAAACAxNhUBgAAAAAAAAAkxqYyAAAAAAAAACCxwnp3GEJY9bExxjVtP0kfIeQTtFJONQZPjCWZJ5kCfya9451Oyn4P0epknivrua6tbnD7UEp6Gs3MbHJa51XV+hwsFGW8sDAn89GRCd2+mc07bRRCtW7AWU+5nL+ea2p0H/MlvV6izcg87/z8K8m9Iab8GVq5rOehnO7WY2YZ/JQvpL2yzSxmcCJCknt0kvcTa6unZ7cdOXJw1cf3dh1wX+Othf7Bp1L1kaxea/2Dh1K3oRwf+FLqNnp7H9V532MyP9H/ZOoxpJW3JplnUa89++/5NZmX44LMv/tnf5BuAGbW0/XxVMfnStDXFwAAC01JREFUcuk/1nd0/oLMo82maj8482hm1tnxQZmfGnpW5l69vrPjAXcMHq9eD50+LPOOjg+kHsPQ0LfSjWGPnoe0z1TeGO69997U7WNRiDmrjluWzW/Ub5XHt9Rf9fvITcl8Yr5G5uVRfZNOUrN7u/fKfCq+KfPrC/r+VVWjz+Hs0FmZm5lVN+6Q+U4d2/zYqMznitfcMYzP68/RNzXr40ev6/diW4N+/nVuwWZmNjmt7y9V460yD1v18d4z9pat/rPG9Wldr2JJP/+OvqnXS/0Ov2bPF2/IfMcW/WYWS3oe2vK1Mvc++5iZXTqp9ztyJb2eYl4vmKoWf568JZdz6lmTc00Mj+vrribJPmWNvvibW2+X+djYa7r9BDU7NI3LvGqycfljbfn3kd9UBgAAAAAAAAAkxqYyAAAAAAAAACAxNpUBAAAAAAAAAImxqQwAAAAAAAAASIxNZQAAAAAAAABAYmwqAwAAAAAAAAASY1MZAAAAAAAAAJAYm8oAAAAAAAAAgMQK691hjHHVx4YQUrfvteHlsVx0x5DLeXv1ld/Lz63+bTAzM/+dKPuvCPMyL3lrJVTJ2F0LCc4i56yHYmlO9+GuWZ0X8vXO8WY1hQaZLyzoMRZLszKPUR+/+Bp9XfizoF9Rjv568pUyaGN5Sa4p9/7n3X9Srulk999091C39QRjSHsfR+WdOHnQfU1v1wGZ93U/IvMs6rXXR//gU24ba93+3p5HZe59ovCulrt6H3PHUHYa6e9/Uvex91dkvh712huDt55ePvZ5me+7+1dlbmaWd/oo5HU998Toz1NnxwdlPjT0bKoxdOx5QLd/+nDqNrzck6Ree+NMOwZz3qsk87TW0jyz/RU1T+fOv5K6fSzK580atyyfXxgflcfPFvwtgZai/izeWOU9Lzif9Uv+5/RXhy/LPNfcLPP5Ul7m+fExmbfvbJG5mVlDnfMCp6B6H3Eb8v4zUa3zWb5qQZ9nS0O7zBecZ8Mkn9LdZ2xnSQZnHtsLZ2R+4dVO3YGZ5U3vNRTyej1N1W2Veb1dccewrVHvBTTfpD/r5mb0c/71685aePeCzM3Mrp/Vb1Y5f0Pm0VkxeWc9m5nloh5n2ap1H8FZcM5ex/Zmf56ujk3L/PQFfZ7bxT3eLFnNHrvWJvOFmuXXZBR7KZXf3QQAAAAAAAAAbBpsKgMAAAAAAAAAEmNTGQAAAAAAAACQGJvKAAAAAAAAAIDE2FQGAAAAAAAAACTGpjIAAAAAAAAAIDE2lQEAAAAAAAAAiRUqPYCViDG6rwkhpG5DyeX8fXivD2+MaaU9x/XqI+b0PHjzVC6XU4/Bk/Y8veO9PMl683jz6PVRLqUeQoJ5XPs166n0dbtZME8/HgYGXrfergOrPv7EyYPua7y10j/4lMy98Q2cOuSOIc05ZqGv+5HUbXjzlIXe3kdl3tf3mG4g6jrTf+KLKx3Siu3te1zmx45/PlX7WXwmOXbiSZn39XxC5idPPe320dnxQZl37HnIaWHtP3sNnT4s8z13fkDmp898K/UY0vax1scnaSMt730w88fJZ4L1USyX7drkxLL5jmb9PjRsybt9TI3pa3/h9csyr9q9U+ZXtvj3lp2TOq+9viDzLbdNy3zsqm6/obZavyCBy8PNMt9WVy/z+jZnkGZWXJjTeWzRDUT9XlwZe13mF+e7dPtm9pEOvR7OOn10dF2T+dh0rczjOf8B11uR10avyHznVj3PI7lZdwwtC3rN3bjoPUPreTDnue70i21O+2ZtW6ZknisUZd5eapB5sbXVHcPY+AWZN7bre8O1ET0Ps4V5mV8eq5K5mVldo27j9kZ93U7M6PYbGvx5mirrNbvams1vKgMAAAAAAAAAEmNTGQAAAAAAAACQGJvKAAAAAAAAAIDE2FQGAAAAAAAAACTGpjIAAAAAAAAAIDE2lQEAAAAAAAAAibGpDAAAAAAAAABILMQY16+zEEbM7NxbvtVuZqPrNoB3LuYxG8xjNpjHbDCPq3NbjHFbpQex2VGv1wzzmA3mMRvMY3aYy5WjXmeEmr1mmMdsMI/ZYB6zwTyuzrI1e103lX+o8xCOxBjvrdgA3iGYx2wwj9lgHrPBPGIjYT1mg3nMBvOYDeYxO8wlNhLWYzaYx2wwj9lgHrPBPGaPP38BAAAAAAAAAEiMTWUAAAAAAAAAQGKV3lT+QoX7f6dgHrPBPGaDecwG84iNhPWYDeYxG8xjNpjH7DCX2EhYj9lgHrPBPGaDecwG85ixiv5NZQAAAAAAAADA5lLp31QGAAAAAAAAAGwiFdlUDiG8L4QwFEJ4JYTw6UqMYbMKIXw5hDAcQjjxlu9tDSF8J4RwZun/tlZyjBtdCOGWEMJ3QwgnQwgDIYRfX/o+87hCIYTaEML/CyEcW5rLzy59/90hhBeW5vJrIYTqSo91owsh5EMIR0MI3176mjnEhkDNXh3qdTao2dmgXmeLmo2NiHq9etTs9KjX2aBeZ4t6vfbWfVM5hJA3sz80s79tZt1m9vdCCN3rPY5N7KCZve9t3/u0mf1pjPFOM/vTpa+xvKKZ/VaMscvM7jezf7i0BpnHlZszs5+LMd5lZneb2ftCCPeb2e+a2X9YmsvrZvZYBce4Wfy6mZ18y9fMISqOmp3KQaNeZ4GanQ3qdbao2dhQqNepHTRqdlrU62xQr7NFvV5jlfhN5fvM7JUY49kY47yZ/YmZPViBcWxKMcbnzeza2779oJkdWvr3ITN7aF0HtcnEGC/FGF9a+veELd5k3mXM44rFRZNLX1Yt/S+a2c+Z2dNL32cuHSGEm83s75rZl5a+DsYcYmOgZq8S9Tob1OxsUK+zQ83GBkW9ToGanR71OhvU6+xQr9dHJTaV32VmF97y9RtL38Pq7YgxXjJbvJmb2fYKj2fTCCHsNrN9ZvaCMY+rsvT/UvKymQ2b2XfM7FUzG4sxFpdewjXue8LM/qmZlZe+bjPmEBsDNTtb1JkUqNnpUK8zQ83GRkS9zh51ZpWo1+lQrzNDvV4HldhUDj/ie3HdR4EfeyGERjP7hpn9RoxxvNLj2axijKUY491mdrMt/pZE14962fqOavMIIbzfzIZjjC++9ds/4qXMISqBtYgNgZqdHvU6PWo2NjDWITYE6nV61Ov0qNfrp1CBPt8ws1ve8vXNZnaxAuN4J7kSQrgpxngphHCTLf5EC0IIocoWi91XY4zPLH2beUwhxjgWQvieLf4NrZYQQmHpp4Bc49pPm9kDIYS/Y2a1ZtZkiz9VZQ6xEVCzs0WdWQVqdrao16lQs7FRUa+zR51ZIep1tqjXqVCv10klflP5B2Z259J/dbHazD5uZocrMI53ksNm9vDSvx82s29WcCwb3tLf0nnSzE7GGD/3loh5XKEQwrYQQsvSv+vM7G/Z4t/P+q6ZfWTpZcylEGP85zHGm2OMu23xfvhnMcZfNOYQGwM1O1vUmRWiZmeDep0NajY2MOp19qgzK0C9zgb1OhvU6/UTYlz/3/Ze+mnBE2aWN7Mvxxj/7boPYpMKIfxXM/tZM2s3sytm9hkze87Mvm5mt5rZeTP7aIzx7f+hASwJIfwNM/u/ZtZvf/33dX7bFv/mE/O4AiGEvbb4B+7ztvhDqq/HGP9VCOF2W/wPhGw1s6Nm9ksxxrnKjXRzCCH8rJn94xjj+5lDbBTU7NWhXmeDmp0N6nX2qNnYaKjXq0fNTo96nQ3qdfao12urIpvKAAAAAAAAAIDNqRJ//gIAAAAAAAAAsEmxqQwAAAAAAAAASIxNZQAAAAAAAABAYmwqAwAAAAAAAAASY1MZAAAAAAAAAJAYm8oAAAAAAAAAgMTYVAYAAAAAAAAAJMamMgAAAAAAAAAgsf8PEGsEMoPYz8MAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1800x1800 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "fig,axis=plt.subplots(1,3,figsize=(25,25))\n",
    "axis[0].imshow(np.uint8(pic))\n",
    "axis[0].set_title('Orignal',fontsize=30)\n",
    "axis[1].imshow(np.uint8(masks))\n",
    "axis[1].set_title('50% sample',fontsize=30)\n",
    "axis[2].imshow(np.uint8(Xa))\n",
    "axis[2].set_title('Reconstructed',fontsize=30)\n",
    "\n",
    "\n",
    "print(\"--- %s seconds ---\" % (time.time() - start_time))\n"
   ]
  }
 ],
 "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}