Recognizer.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import cv2\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "from keras.models import load_model\n",
    "model = load_model('./training_models/NN_digit.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAhgAAAJCCAYAAACCkb6BAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAKC4/w/Phf7k/xUIUAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd7004f0950>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Points = np.load('./Points.npy')\n",
    "\n",
    "img = cv2.imread('./Data/S_03.jpg')\n",
    "\n",
    "gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)\n",
    "\n",
    "r, c = gray.shape\n",
    "out_row = 500\n",
    "gray = cv2.resize(gray, (int(out_row * float(c)/r), out_row))\n",
    "\n",
    "gray = cv2.GaussianBlur(gray, (3, 3), 0)\n",
    "\n",
    "im_th = cv2.adaptiveThreshold(gray, 255, 1, 1, 11, 2)\n",
    "\n",
    "plt.figure(num =0, figsize=(10, 10))\n",
    "plt.imshow(im_th, 'gray')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[6 6 6 7 6 6 6 7 2]\n",
      " [6 6 6 6 6 6 6 7 2]\n",
      " [6 6 6 2 6 6 6 7 7]\n",
      " [6 6 6 6 6 6 6 6 2]\n",
      " [6 6 6 6 6 6 6 2 2]\n",
      " [6 6 6 6 6 6 6 2 6]\n",
      " [6 6 6 6 6 6 6 6 6]\n",
      " [6 6 6 7 6 6 2 7 7]\n",
      " [2 7 2 2 6 6 7 3 3]]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nqSnjl5r6P5KJwSuN6z/bAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c83abc10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nFf9PKAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cae66350>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c8d8f950>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n8EtFGX6pqP8DVkXfXq3uxlQAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c87f1c10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c872ca10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c957be90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c7e7ac90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSRE/kBTxA0kRP5AU8QNJET+Q1P8BVnqthp/5N7UAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c88e2e50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c8ea0b90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c87413d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nlmzJpwTzAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c7de91d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEBThB4Ii/EBQ/wX6MC6AXOa/6wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6d0e10910>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/EBS/wNDQtn0NuyOYAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c84a3710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c8288650>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c9a31c50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nDyRF/EBSxA8kRfxAUv8HGLZ2tg/OiJ0AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6f9fa53d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cb9f91d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cb8ec910>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nKD+Q1P8DKQ33CrbCEyAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cb741710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n5CxrOgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cb6f1e50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAADW1JREFUeJzt3W2IXPUVx/HfMTX4EEHTdNcQkyaNUpEgWoboi1hSrGLE\nPIggikLEh1XRUKNg1QrmZShV8UUV1xoTHxIVG3EFMWoQNVCCq6zZmLTVStTE3Y0aIUaiqe7pi7mx\nq+78ZzJzZ+7snu8Hws7cc+/cwyW/vXPnf3f+5u4CEM9hRTcAoBiEHwiK8ANBEX4gKMIPBEX4gaAI\nPxAU4QeCIvxAUD9r5c7MLHk74aRJk5Lbz549u2LtnXfeqa8pYJxxd6tlvYbCb2bnSbpP0gRJf3P3\nlY28XqlUStaffvrpirWOjo5Gdg2EU/fbfjObIOmvkhZIOkXSpWZ2Sl6NAWiuRq7550p6390/cPcD\nkp6UtDiftgA0WyPhnybp4xHPd2bLfsDMusys18x6G9gXgJw1/QM/d++W1C1V/8APQOs0cubfJWn6\niOcnZMsAjAGNhP9NSSeZ2SwzmyjpEkk9+bQFoNnqftvv7t+a2Y2SNqg81LfK3d9NbXPEEUfoxBNP\nrFifNWtWcp/HHntsxdrtt9+e3HZgYCBZv+6665J1YCy44ooral63oWt+d39B0guNvAaAYnB7LxAU\n4QeCIvxAUIQfCIrwA0ERfiAoa+WMPaVSyXt7ucUfSBkcHEzWV66s/Jfz69at09DQUE1/z8+ZHwiK\n8ANBEX4gKMIPBEX4gaAIPxBUS4f6JkyY4EceeWTF+llnnZXc/rHHHqtYmzlzZr1tAW1leHg4Wd+/\nf3+yXutXd3PmB4Ii/EBQhB8IivADQRF+ICjCDwRF+IGgWjpF9/DwsL766quK9a+//jq5feqehNTr\nAvgpzvxAUIQfCIrwA0ERfiAowg8ERfiBoAg/EFRD4/xmtkPSl5K+k/Stu5caeb2FCxc2sjmAQ5DH\nTT6/c/fPcngdAC3E234gqEbD75JeMrO3zKwrj4YAtEajb/vnufsuM+uQ9LKZ/dPdXx+5QvZLgV8M\nQJtp6Mzv7ruyn7slPStp7ijrdLt7qdEPAwHkq+7wm9nRZnbMwceSzpW0Na/GADRXI2/7OyU9a2YH\nX2etu7+YS1cAmq6l39tvZsmd9ff3J7fv7OysWOvo6KivKWCc4Xv7ASQRfiAowg8ERfiBoAg/EBTh\nB4Ii/EBQhB8IivADQRF+ICjCDwRF+IGgCD8QFOEHgiL8QFCEHwiK8ANBEX4gKMIPBEX4gaAIPxAU\n4QeCymOWXoxjZ599drK+YsWKZH3atGkVaz09Pcltb7311mT9wIEDyTrSOPMDQRF+ICjCDwRF+IGg\nCD8QFOEHgiL8QFBVx/nNbJWkCyTtdvc52bLJkp6SNFPSDkkXu/sXzWsT9TrssPTv96uvvjpZf/DB\nB5P1vr6+ZP2jjz6qWLv++uuT286YMSNZv/POO5P1bdu2JevR1XLmXy3pvB8tu03SRnc/SdLG7DmA\nMaRq+N39dUl7frR4saQ12eM1kpbk3BeAJqv3mr/T3Qeyx4OSOnPqB0CLNHxvv7u7mXmlupl1Sepq\ndD8A8lXvmX/IzKZKUvZzd6UV3b3b3UvuXqpzXwCaoN7w90hamj1eKum5fNoB0CpVw29m6yT9Q9Kv\nzWynmV0laaWkc8zsPUm/z54DGEOqXvO7+6UVSuk/9EZbOPnkk5P1auP4W7ZsSdYXLFiQrA8ODlas\nVfsugLvuuitZr9ZbtdePjjv8gKAIPxAU4QeCIvxAUIQfCIrwA0Hx1d1I2rRpU7KeGsqr5v7770/W\nly1blqxfdNFFyTpDfWmc+YGgCD8QFOEHgiL8QFCEHwiK8ANBEX4gKMb5UZgzzzwzWT/qqKOS9bVr\n1+bZTjic+YGgCD8QFOEHgiL8QFCEHwiK8ANBEX4gKMb5x7lq01Rffvnlyfrjjz+erA8NDSXrn3zy\nScXawoULk9sODAwk6+vXr0/WkcaZHwiK8ANBEX4gKMIPBEX4gaAIPxAU4QeCqjrOb2arJF0gabe7\nz8mWrZB0jaRPs9XucPcXmtUkmufVV19taPtq9wl88803FWsdHR3Jbfv6+pL1nTt3JutIq+XMv1rS\neaMsv9fdT8v+EXxgjKkafnd/XdKeFvQCoIUauea/0cy2mNkqMzsut44AtES94X9A0mxJp0kakHR3\npRXNrMvMes2st859AWiCusLv7kPu/p27D0t6SNLcxLrd7l5y91K9TQLIX13hN7OpI55eKGlrPu0A\naJVahvrWSZovaYqZ7ZR0l6T5ZnaaJJe0Q9K1TewRQBOYu7duZ2bJnfX39ye37+zsrFirNmYc1fHH\nH5+sb9iwIVnfu3dvsr548eJkfc+eygNFZ5xxRnLbV155JVlfvXp1sr5s2bJkfbxyd6tlPe7wA4Ii\n/EBQhB8IivADQRF+ICjCDwTFV3ePc2vWrEnWTz311GT9yiuvTNZTQ3nVbN68OVnfv39/sj5//vy6\n9w3O/EBYhB8IivADQRF+ICjCDwRF+IGgCD8QFOP849zEiRMb2v6RRx7JqRO0G878QFCEHwiK8ANB\nEX4gKMIPBEX4gaAIPxAU4/zjXE9PT7Je7W/iFy1a1NDro31x5geCIvxAUIQfCIrwA0ERfiAowg8E\nRfiBoKpO0W1m0yU9KqlTkkvqdvf7zGyypKckzZS0Q9LF7v5Flddiiu4WmzFjRrL+2muvJeupYy5J\ny5cvT9b37dtXsXbPPfckt50yZUqyvmTJkmT9+eefT9bHqzyn6P5W0i3ufoqkMyXdYGanSLpN0kZ3\nP0nSxuw5gDGiavjdfcDd384efylpu6RpkhZLOjgdzBpJ6V/DANrKIV3zm9lMSadL2iyp090HstKg\nypcFAMaImu/tN7NJkv4u6SZ332v2/8sKd/dK1/Nm1iWpq9FGAeSrpjO/mR2ucvCfcPf12eIhM5ua\n1adK2j3atu7e7e4ldy/l0TCAfFQNv5VP8Q9L2u7uIz+e7ZG0NHu8VNJz+bcHoFlqGeqbJ+kNSf2S\nhrPFd6h83f+0pBmSPlR5qC85XzNDfe1nzpw5yXq1Kbovu+yyPNv5gZtvvjlZf/HFF5P1zz//PM92\nxoxah/qqXvO7+yZJlV7s7ENpCkD74A4/ICjCDwRF+IGgCD8QFOEHgiL8QFBVx/lz3Rnj/EDT5fkn\nvQDGIcIPBEX4gaAIPxAU4QeCIvxAUIQfCIrwA0ERfiAowg8ERfiBoAg/EBThB4Ii/EBQhB8Iqubp\nujA+TZw4MVk/cOBAizpBq3HmB4Ii/EBQhB8IivADQRF+ICjCDwRF+IGgqo7zm9l0SY9K6pTkkrrd\n/T4zWyHpGkmfZqve4e4vNKtRNMeiRYuS9WeeeaZFnaDVarnJ51tJt7j722Z2jKS3zOzlrHavu/+l\nee0BaJaq4Xf3AUkD2eMvzWy7pGnNbgxAcx3SNb+ZzZR0uqTN2aIbzWyLma0ys+MqbNNlZr1m1ttQ\npwByVXP4zWySpL9Lusnd90p6QNJsSaep/M7g7tG2c/dudy+5eymHfgHkpKbwm9nhKgf/CXdfL0nu\nPuTu37n7sKSHJM1tXpsA8lY1/GZmkh6WtN3d7xmxfOqI1S6UtDX/9gA0S9Upus1snqQ3JPVLGs4W\n3yHpUpXf8rukHZKuzT4cTL0WU3QDTVbrFN21fNq/SdJoL8aYPjCGcYcfEBThB4Ii/EBQhB8IivAD\nQRF+IKi2+urutWvXJuvLly9vUSfA+MeZHwiK8ANBEX4gKMIPBEX4gaAIPxAU4QeCqvr3/LnuzOxT\nSR+OWDRF0mcta+DQtGtv7dqXRG/1yrO3X7r7L2pZsaXh/8nOzXrb9bv92rW3du1Lord6FdUbb/uB\noAg/EFTR4e8ueP8p7dpbu/Yl0Vu9Cumt0Gt+AMUp+swPoCCFhN/MzjOzf5nZ+2Z2WxE9VGJmO8ys\n38z6ip5iLJsGbbeZbR2xbLKZvWxm72U/R50mraDeVpjZruzY9ZnZ+QX1Nt3MXjWzbWb2rpn9IVte\n6LFL9FXIcWv5234zmyDp35LOkbRT0puSLnX3bS1tpAIz2yGp5O6Fjwmb2W8l7ZP0qLvPyZb9WdIe\nd1+Z/eI8zt3/2Ca9rZC0r+iZm7MJZaaOnFla0hJJV6jAY5fo62IVcNyKOPPPlfS+u3/g7gckPSlp\ncQF9tD13f13Snh8tXixpTfZ4jcr/eVquQm9twd0H3P3t7PGXkg7OLF3osUv0VYgiwj9N0scjnu9U\ne0357ZJeMrO3zKyr6GZG0TliZqRBSZWnMSpG1ZmbW+lHM0u3zbGrZ8brvPGB30/Nc/ffSFog6Ybs\n7W1b8vI1WzsN19Q0c3OrjDKz9PeKPHb1znidtyLCv0vS9BHPT8iWtQV335X93C3pWbXf7MNDBydJ\nzX7uLrif77XTzM2jzSytNjh27TTjdRHhf1PSSWY2y8wmSrpEUk8BffyEmR2dfRAjMzta0rlqv9mH\neyQtzR4vlfRcgb38QLvM3FxpZmkVfOzabsZrd2/5P0nnq/yJ/38k/amIHir09StJ72T/3i26N0nr\nVH4b+F+VPxu5StLPJW2U9J6kVyRNbqPeHlN5NuctKgdtakG9zVP5Lf0WSX3Zv/OLPnaJvgo5btzh\nBwTFB35AUIQfCIrwA0ERfiAowg8ERfiBoAg/EBThB4L6H/VwW4bl6wSXAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cb487c50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\npIgfSIr4gaSIH0iK+IGk/g/O5oCQzu2KcwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cb2073d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cb09b1d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nhB9IivADSf0fzogmecY7dZ8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c9ace910>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAC0RJREFUeJzt3V2IXPUZx/Hfr77cqGjSTEOIsWslFEKhsQ6hoBSDVWJA\no16IuZAUhFVUiOBFxV7Uy1Cj0osixBqSFqsUXzAXoTUNQhCKuEqabEzStbLBhLg7Sy4SL8REn17s\niazJzoszZ+ZM8nw/sMzMOTM7D4Nf5+VM9u+IEIB8flD1AACqQfxAUsQPJEX8QFLEDyRF/EBSxA8k\nRfxAUsQPJHXpIO9s0aJFMTIy0nT/l19+2fL2U1NTTfddffXVLW97zTXXtNwPXAwmJyc1MzPjTq7b\nU/y210j6o6RLJP05Ija1uv7IyIjGxsaa7j98+HDL+9u8eXPTfXfddVfL2959990t9wMXg3q93vF1\nu37Zb/sSSX+SdKekFZLW217R7e8DMFi9vOdfJemTiPg0Ir6S9JqkdeWMBaDfeol/qaTP5lw+Wmz7\nDtujtsdsjzUajR7uDkCZ+v5pf0RsiYh6RNRrtVq/7w5Ah3qJ/5ikZXMuX1tsA3AB6CX+DyQtt329\n7cslPSBpRzljAei3rg/1RcQZ249L+qdmD/VtjYgDvQyzdOl5Hxl8x8aNG7v+3ePj413fFhikZ599\ntuX+VofEDx061PH99HScPyJ2StrZy+8AUA2+3gskRfxAUsQPJEX8QFLEDyRF/EBSA/33/O20+ue+\nkrR69eoBTQJc/HjmB5IifiAp4geSIn4gKeIHkiJ+ICniB5IifiAp4geSIn4gKeIHkiJ+ICniB5Ii\nfiAp4geSIn4gKeIHkiJ+ICniB5IifiAp4geSIn4gqZ7+dLftSUmnJH0t6UxE1MsYCkD/lfF3+1dH\nxEwJvwfAAPGyH0iq1/hD0ju2P7Q9WsZAAAaj15f9t0TEMds/krTL9qGI2DP3CsX/FEYl6brrruvx\n7gCUpadn/og4VpxOS3pL0qp5rrMlIuoRUa/Var3cHYASdR2/7StsX3X2vKQ7JI2XNRiA/urlZf9i\nSW/ZPvt7/hYR/yhlKgB913X8EfGppJ+XOAuAAeJQH5AU8QNJET+QFPEDSRE/kBTxA0kRP5AU8QNJ\nET+QFPEDSRE/kBTxA0kRP5AU8QNJET+QFPEDSRE/kBTxA0kRP5AU8QNJET+QFPEDSRE/kBTxA0kR\nP5AU8QNJET+QFPEDSRE/kBTxA0m1jd/2VtvTtsfnbFtoe5ftieJ0QX/HBFC2Tp75t0lac862pyTt\njojlknYXlwFcQNrGHxF7JJ04Z/M6SduL89sl3VPyXAD6rNv3/Isj4nhx/nNJi0uaB8CA9PyBX0SE\npGi23/ao7THbY41Go9e7A1CSbuOfsr1EkorT6WZXjIgtEVGPiHqtVuvy7gCUrdv4d0jaUJzfIOnt\ncsYBMCidHOp7VdK/Jf3U9lHbD0naJOl22xOSfl1cBnABubTdFSJifZNdt5U8C4AB4ht+QFLEDyRF\n/EBSxA8kRfxAUsQPJEX8QFLEDyRF/EBSxA8kRfxAUsQPJEX8QFLEDyRF/EBSxA8kRfxAUsQPJEX8\nQFLEDyRF/EBSxA8kRfxAUsQPJEX8QFLEDyRF/EBSxA8kRfxAUsQPJNU2fttbbU/bHp+z7Rnbx2zv\nLX7W9ndMAGXr5Jl/m6Q182x/ISJWFj87yx0LQL+1jT8i9kg6MYBZAAxQL+/5H7e9r3hbsKC0iQAM\nRLfxvyjpBkkrJR2X9FyzK9oetT1me6zRaHR5dwDK1lX8ETEVEV9HxDeSXpK0qsV1t0REPSLqtVqt\n2zkBlKyr+G0vmXPxXknjza4LYDhd2u4Ktl+VdKukRbaPSvq9pFttr5QUkiYlPdzHGQH0Qdv4I2L9\nPJtf7sMsAAaIb/gBSRE/kBTxA0kRP5AU8QNJET+QVNtDfUBVXn/99Zb7H3300Zb7p6enyxznosMz\nP5AU8QNJET+QFPEDSRE/kBTxA0kRP5AUx/kxtGZmZlruP3369IAmuTjxzA8kRfxAUsQPJEX8QFLE\nDyRF/EBSxA8kxXF+DK1HHnmk6hEuajzzA0kRP5AU8QNJET+QFPEDSRE/kBTxA0m1jd/2Mtvv2v7Y\n9gHbG4vtC23vsj1RnC7o/7gAytLJM/8ZSU9GxApJv5T0mO0Vkp6StDsilkvaXVwGcIFoG39EHI+I\nj4rzpyQdlLRU0jpJ24urbZd0T7+GBFC+7/We3/aIpBslvS9pcUQcL3Z9LmlxqZMB6KuO47d9paQ3\nJD0RESfn7ouIkBRNbjdqe8z2WKPR6GlYAOXpKH7bl2k2/Fci4s1i85TtJcX+JZLmXRUxIrZERD0i\n6rVarYyZAZSgk0/7LellSQcj4vk5u3ZI2lCc3yDp7fLHA9AvnfyT3pslPShpv+29xbanJW2S9Hfb\nD0k6Iun+/owIoB/axh8R70lyk923lTsOgEHhG35AUsQPJEX8QFLEDyRF/EBSxA8kxZ/uRks33XRT\ny/1Hjhxpub/dMtuoDs/8QFLEDyRF/EBSxA8kRfxAUsQPJEX8QFIc50dLExMTLfffd999Lfdv27at\nxGlQJp75gaSIH0iK+IGkiB9IiviBpIgfSIr4gaQ4zo+WTp482XI/x/EvXDzzA0kRP5AU8QNJET+Q\nFPEDSRE/kBTxA0m1jd/2Mtvv2v7Y9gHbG4vtz9g+Zntv8bO2/+MCKEsnX/I5I+nJiPjI9lWSPrS9\nq9j3QkRs7t94APqlbfwRcVzS8eL8KdsHJS3t92AA+ut7vee3PSLpRknvF5set73P9lbbC5rcZtT2\nmO2xRqPR07AAytNx/LavlPSGpCci4qSkFyXdIGmlZl8ZPDff7SJiS0TUI6Jeq9VKGBlAGTqK3/Zl\nmg3/lYh4U5IiYioivo6IbyS9JGlV/8YEULZOPu23pJclHYyI5+dsXzLnavdKGi9/PAD90smn/TdL\nelDSftt7i21PS1pve6WkkDQp6eG+TAigLzr5tP89SZ5n187yxwEwKHzDD0iK+IGkiB9IiviBpIgf\nSIr4gaSIH0iK+IGkiB9IiviBpIgfSIr4gaSIH0iK+IGkHBGDuzO7IenInE2LJM0MbIDvZ1hnG9a5\nJGbrVpmz/TgiOvp7eQON/7w7t8ciol7ZAC0M62zDOpfEbN2qajZe9gNJET+QVNXxb6n4/lsZ1tmG\ndS6J2bpVyWyVvucHUJ2qn/kBVKSS+G2vsX3Y9ie2n6pihmZsT9reX6w8PFbxLFttT9sen7Ntoe1d\ntieK03mXSatotqFYubnFytKVPnbDtuL1wF/2275E0n8l3S7pqKQPJK2PiI8HOkgTticl1SOi8mPC\ntn8l6QtJf4mInxXb/iDpRERsKv7HuSAifjsksz0j6YuqV24uFpRZMndlaUn3SPqNKnzsWsx1vyp4\n3Kp45l8l6ZOI+DQivpL0mqR1Fcwx9CJij6QT52xeJ2l7cX67Zv/jGbgmsw2FiDgeER8V509JOruy\ndKWPXYu5KlFF/EslfTbn8lEN15LfIekd2x/aHq16mHksLpZNl6TPJS2ucph5tF25eZDOWVl6aB67\nbla8Lhsf+J3vloj4haQ7JT1WvLwdSjH7nm2YDtd0tHLzoMyzsvS3qnzsul3xumxVxH9M0rI5l68t\ntg2FiDhWnE5LekvDt/rw1NlFUovT6Yrn+dYwrdw838rSGoLHbphWvK4i/g8kLbd9ve3LJT0gaUcF\nc5zH9hXFBzGyfYWkOzR8qw/vkLShOL9B0tsVzvIdw7Jyc7OVpVXxYzd0K15HxMB/JK3V7Cf+/5P0\nuypmaDLXTyT9p/g5UPVskl7V7MvA05r9bOQhST+UtFvShKR/SVo4RLP9VdJ+Sfs0G9qSima7RbMv\n6fdJ2lv8rK36sWsxVyWPG9/wA5LiAz8gKeIHkiJ+ICniB5IifiAp4geSIn4gKeIHkvo/ozCNyEHG\n/GcAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c9520710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAC7xJREFUeJzt3VuIXIUdx/Hfr8bIYkRicyHE2FiNlVBoLEMMNBSrVWJe\noi/SPJQUhERUUPChYh/qgw+hRKUPQYg1mFarFFTMQ2hNg5AKNThKmotJGxtjLsRcCMSoizbrvw97\nImvcuWTmzJzZ/L8fWHbmnDN7/gz5Zi5ndo8jQgDy+U7VAwCoBvEDSRE/kBTxA0kRP5AU8QNJET+Q\nFPEDSRE/kNSkfu5s2rRpMXfu3H7uEkjlwIEDOnnypNvZtqv4bS+R9HtJl0j6Q0Ssbrb93LlzVa/X\nu9klgCZqtVrb23b8tN/2JZLWSrpT0nxJy23P7/TnAeivbl7zL5T0QUTsj4gvJb0saVk5YwHotW7i\nny3p0Jjrh4tl32B7pe267fqJEye62B2AMvX83f6IWBcRtYioTZ8+vde7A9CmbuI/ImnOmOtXF8sA\nTADdxP+OpHm2r7U9WdIvJG0sZywAvdbxob6IOGv7QUl/0+ihvvURsbvZbY4dO6Y1a9Y0XH/rrbc2\n3ef999/fwaRAHnv37m17266O80fEJkmbuvkZAKrBx3uBpIgfSIr4gaSIH0iK+IGkiB9Iqq+/z//F\nF19o//79Dde3+nXEbdu2lT0SkBaP/EBSxA8kRfxAUsQPJEX8QFLEDyTV10N9rcyYMaPp+muuuabh\nuoMHD5Y9Di5i119/fdP1rf7k3OnTp8scpxI88gNJET+QFPEDSRE/kBTxA0kRP5AU8QNJDdRx/vnz\nm5/n8+abb264juP8uBCLFi1quv7tt99uup7j/AAmLOIHkiJ+ICniB5IifiAp4geSIn4gqa6O89s+\nIOmMpBFJZyOi+d/ebuHDDz9sun7nzp3d/Hjgay+88ELVI1SujA/5/CwiTpbwcwD0EU/7gaS6jT8k\nvWH7XdsryxgIQH90+7R/cUQcsT1D0mbbeyNi69gNiv8UVkrSlClTutwdgLJ09cgfEUeK78clvSZp\n4TjbrIuIWkTUhoaGutkdgBJ1HL/ty21fce6ypDsk7SprMAC91c3T/pmSXrN97uf8OSL+WspUAHqu\n4/gjYr+kH5U4i2bPnt10/Q033NBw3d69e8scBbjocagPSIr4gaSIH0iK+IGkiB9IiviBpAbqT3dP\nnjy56frLLrusT5MAFz8e+YGkiB9IiviBpIgfSIr4gaSIH0iK+IGkiB9IiviBpIgfSIr4gaSIH0iK\n+IGkiB9IiviBpIgfSIr4gaSIH0iK+IGkiB9IiviBpIgfSIr4gaRaxm97ve3jtneNWXaV7c229xXf\np/Z2TABla+eR/3lJS85b9qikLRExT9KW4jqACaRl/BGxVdKp8xYvk7ShuLxB0l0lzwWgxzp9zT8z\nIo4Wlz+WNLOkeQD0Sddv+EVESIpG622vtF23XR8eHu52dwBK0mn8x2zPkqTi+/FGG0bEuoioRURt\naGiow90BKFun8W+UtKK4vELS6+WMA6Bf2jnU95Kkf0r6ge3Dtu+VtFrS7bb3Sfp5cR3ABDKp1QYR\nsbzBqttKngVAH/EJPyAp4geSIn4gKeIHkiJ+ICniB5IifiAp4geSIn4gKeIHkiJ+ICniB5IifiAp\n4geSIn4gKeIHkiJ+ICniB5IifiAp4geSIn4gKeIHkiJ+ICniB5IifiAp4geSIn4gKeIHkiJ+ICni\nB5JqGb/t9baP2941Ztnjto/Y3l58Le3tmADK1s4j//OSloyz/OmIWFB8bSp3LAC91jL+iNgq6VQf\nZgHQR9285n/Q9o7iZcHU0iYC0Bedxv+MpOskLZB0VNKTjTa0vdJ23XZ9eHi4w90BKFtH8UfEsYgY\niYivJD0raWGTbddFRC0iakNDQ53OCaBkHcVve9aYq3dL2tVoWwCDaVKrDWy/JOkWSdNsH5b0W0m3\n2F4gKSQdkLSqhzMC6IGW8UfE8nEWP9eDWQD0EZ/wA5IifiAp4geSIn4gKeIHkiJ+ICniB5IifiAp\n4geSIn4gKeIHkiJ+ICniB5IifiAp4geSIn4gKeIHkiJ+ICniB5IifiAp4geSIn4gKeIHkiJ+ICni\nB5IifiAp4geSIn4gKeIHkiJ+IKmW8dueY/tN2+/b3m37oWL5VbY3295XfJ/a+3EBlKWdR/6zkh6J\niPmSFkl6wPZ8SY9K2hIR8yRtKa4DmCBaxh8RRyPiveLyGUl7JM2WtEzShmKzDZLu6tWQAMp3Qa/5\nbc+VdJOkbZJmRsTRYtXHkmaWOhmAnmo7fttTJL0i6eGI+GTsuogISdHgditt123Xh4eHuxoWQHna\nit/2pRoN/8WIeLVYfMz2rGL9LEnHx7ttRKyLiFpE1IaGhsqYGUAJ2nm335Kek7QnIp4as2qjpBXF\n5RWSXi9/PAC9MqmNbX4i6ZeSdtreXix7TNJqSX+xfa+kjyTd05sRAfRCy/gj4i1JbrD6tnLHAdAv\nfMIPSIr4gaSIH0iK+IGkiB9IiviBpNo5zo8Bd+ONNzZc9/nnnze97cGDB8seBxMEj/xAUsQPJEX8\nQFLEDyRF/EBSxA8kRfxAUhznvwgcOnSo4bqRkZE+ToKJhEd+ICniB5IifiAp4geSIn4gKeIHkiJ+\nICmO818EPvvss6pHwATEIz+QFPEDSRE/kBTxA0kRP5AU8QNJET+QVMv4bc+x/abt923vtv1Qsfxx\n20dsby++lvZ+XABlaedDPmclPRIR79m+QtK7tjcX656OiDW9Gw9Ar7SMPyKOSjpaXD5je4+k2b0e\nDEBvXdBrfttzJd0kaVux6EHbO2yvtz21wW1W2q7brg8PD3c1LIDytB2/7SmSXpH0cER8IukZSddJ\nWqDRZwZPjne7iFgXEbWIqA0NDZUwMoAytBW/7Us1Gv6LEfGqJEXEsYgYiYivJD0raWHvxgRQtnbe\n7bek5yTtiYinxiyfNWazuyXtKn88AL3iiGi+gb1Y0j8k7ZT0VbH4MUnLNfqUPyQdkLSqeHOwoUmT\nJsWVV17ZcP3atWubznLfffc1XHf69OmmtwWyiAi3s1077/a/JWm8H7bpQocCMDj4hB+QFPEDSRE/\nkBTxA0kRP5AU8QNJ9fVPd4+MjOjUqVMN10+fPr3p7efNm9dwXb1e73guDKZardZ0/apVq/o0ycTx\nxBNPtL0tj/xAUsQPJEX8QFLEDyRF/EBSxA8kRfxAUi1/n7/UndknJH00ZtE0SSf7NsCFGdTZBnUu\nidk6VeZs34uI5h+YKfQ1/m/t3K5HRPNPclRkUGcb1LkkZutUVbPxtB9IiviBpKqOf13F+29mUGcb\n1LkkZutUJbNV+pofQHWqfuQHUJFK4re9xPa/bX9g+9EqZmjE9gHbO4szD1f6e8LFadCO2941ZtlV\ntjfb3ld8H/c0aRXNNhBnbm5yZulK77tBO+N135/2275E0n8k3S7psKR3JC2PiPf7OkgDtg9IqkVE\n5ceEbf9U0qeS/hgRPyyW/U7SqYhYXfzHOTUifj0gsz0u6dOqz9xcnFBm1tgzS0u6S9KvVOF912Su\ne1TB/VbFI/9CSR9ExP6I+FLSy5KWVTDHwIuIrZLO/+snyyRtKC5v0Og/nr5rMNtAiIijEfFecfmM\npHNnlq70vmsyVyWqiH+2pENjrh/WYJ3yOyS9Yftd2yurHmYcM8ecGeljSTOrHGYcLc/c3E/nnVl6\nYO67Ts54XTbe8Pu2xRHxY0l3SnqgeHo7kGL0NdsgHa5p68zN/TLOmaW/VuV91+kZr8tWRfxHJM0Z\nc/3qYtlAiIgjxffjkl7T4J19+Ni5k6QW349XPM/XBunMzeOdWVoDcN8N0hmvq4j/HUnzbF9re7Kk\nX0jaWMEc32L78uKNGNm+XNIdGryzD2+UtKK4vELS6xXO8g2DcubmRmeWVsX33cCd8Toi+v4laalG\n3/H/r6TfVDFDg7m+L+lfxdfuqmeT9JJGnwb+T6Pvjdwr6buStkjaJ+nvkq4aoNn+pNGzOe/QaGiz\nKpptsUaf0u+QtL34Wlr1fddkrkruNz7hByTFG35AUsQPJEX8QFLEDyRF/EBSxA8kRfxAUsQPJPV/\nNkexSUPSgwQAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c8f92e50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c8ca8c50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nasr4paaMX2rq/wBrjZ/765quAwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c8b693d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nB5Ii/EBS/wcWOO5QZP0rOQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c8a7c1d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nH0iK+IGk/g/F811+ukoRVQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c896d910>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c7d41710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c7d34e50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQFKEH0iK8ANJEX4gKcIPJPX/dFURCs8k1L4AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c7c48c50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nH0iK+IGkiB9IiviBpIgfSIr4gaT+Dx1EhO0B9G8eAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c7bcb3d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c7b5e1d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n+YGkKD+Q1P8BHOGiY8WoktYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c7ad0910>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c7a63710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nP2jRl5SS9bTgAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6fa0c3390>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n8ANJEX4gKcIPJEX4gaT+D5gACdMB1A/pAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cacfc8d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cbebac90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6d037d610>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6d079f210>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6d0d45250>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nUlGWXyrK8ktFWX6pqP8HfEq5rt0oBdoAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6d083f790>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nB5Ii/EBShB9IivADSRF+ICnCDyT1fwsaMOJJuy0dAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6d06da4d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n8UtNGb/UlPFLTRm/1JTxS039P9eggElH88REAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c86adf50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nDyRF/EBSxA8k9X/nv3RuH9ytQwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd7004a85d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAC11JREFUeJzt3V+InXeZwPHvs1HpH71odTMMtbVuWoQQMC6HIFhEcZVa\npKlQir2QCMJ4YUHBiw3uxZZeFfEPXoh0tMW4ZKtbtCQXZdduWOgKi2Raaie1us2WBBOmyUoC1lLQ\nTh4v5k0d2zl/cs57znvS5/uBYc553/fMeTj0m3PeM2f6i8xEUj1/0/UAkrph/FJRxi8VZfxSUcYv\nFWX8UlHGLxVl/FJRxi8V9ZZZ3llEDPw44c6dOwfefm1tre++8+fPjzeU9CaTmTHKcTHJx3sj4lbg\n28A24PuZef+Q4wfe2erq6sD7u++++/rue+SRRwbeVqpi1PjHftkfEduA7wCfBHYCd0fE4KduSXNj\nknP+PcDxzHwhM/8I/AjY285YkqZtkvivA3676fqpZttfiYiliFiJiJUJ7ktSy6b+hl9mLgPLMPyc\nX9LsTPLMfxq4ftP1dzfbJF0GJon/KHBzRLw3It4GfAY43M5YkqZt7Jf9mflqRNwD/Acbv+p7KDOf\nHXSbK664gptuuqnv/iuvvHLgfd5www199+3atWvgbaUKjh8/PvKxE53zZ+ZjwGOT/AxJ3fDjvVJR\nxi8VZfxSUcYvFWX8UlHGLxU10Z/0Xqper5crK37EX5qWXq/HysrKdP+kV9LlzfilooxfKsr4paKM\nXyrK+KWiZvqrvm3btuWgP9s9evTowNufPHmy774777xz7LmkN4tXXnmF9fV1f9UnqT/jl4oyfqko\n45eKMn6pKOOXijJ+qaiZLtF94cIFXn755b7719fXB97+0KFDffcN+rmS3shnfqko45eKMn6pKOOX\nijJ+qSjjl4oyfqmoif6ePyJOAC8B68CrmdkbcvzAO1tdXR14fwsLC333bd++feBtpSoyc6S/52/j\nQz4fzczftfBzJM2QL/uloiaNP4GfRcSTEbHUxkCSZmPSl/23ZObpiNgOPB4Rv87MJzYf0Pyj4D8M\n0pyZ6Jk/M083388CjwJ7tjhmOTN7w94MlDRbY8cfEVdHxDsuXgY+ARxrazBJ0zXJy/4F4NGIuPhz\n/jUz/72VqSRN3djxZ+YLwPtbnEXSDPmrPqko45eKMn6pKOOXijJ+qSjjl4oyfqko45eKMn6pKOOX\nijJ+qSjjl4oyfqko45eKMn6pKOOXijJ+qSjjl4oyfqko45eKMn6pKOOXijJ+qSjjl4oyfqko45eK\nMn6pKOOXijJ+qSjjl4oaukR3RDwEfAo4m5m7mm3XAj8GbgROAHdl5vnpjbnh4MGD074LqYxRnvl/\nANz6um37gSOZeTNwpLku6TIyNP7MfAI497rNe4EDzeUDwB0tzyVpysY951/IzLXm8ovAQkvzSJqR\noef8w2RmRkT22x8RS8DSpPcjqV3jPvOfiYhFgOb72X4HZuZyZvYyszfmfUmagnHjPwzsay7vAw61\nM46kWRkaf0Q8DPwP8L6IOBURnwfuBz4eEc8D/9Bcl3QZicy+p+vt39mA9wYAVldXB95+YaH/+4rb\nt28fbyjpTSYzY5Tj/ISfVJTxS0UZv1SU8UtFGb9UlPFLRRm/VJTxS0UZv1SU8UtFGb9UlPFLRRm/\nVJTxS0UZv1SU8UtFGb9UlPFLRRm/VJTxS0UZv1SU8UtFGb9UlPFLRRm/VJTxS0UZv1SU8UtFGb9U\nlPFLRQ2NPyIeioizEXFs07Z7I+J0RDzdfN023TEltW2UZ/4fALdusf1bmbm7+Xqs3bEkTdvQ+DPz\nCeDcDGaRNEOTnPPfExHPNKcF17Q2kaSZGDf+7wI7gN3AGvCNfgdGxFJErETEypj3JWkKxoo/M89k\n5npmXgC+B+wZcOxyZvYyszfukJLaN1b8EbG46eqngWP9jpU0n94y7ICIeBj4CPCuiDgF/DPwkYjY\nDSRwAvjCFGeUNAVD48/Mu7fY/OAUZpE0Q37CTyrK+KWijF8qyvilooxfKsr4paKMXyrK+KWijF8q\nyvilooxfKsr4paKMXyrK+KWijF8qyvilooxfKsr4paKMXyrK+KWijF8qyvilooxfKsr4paKMXyrK\n+KWijF8qyvilooxfKsr4paKGLtEdEdcDPwQWgASWM/PbEXEt8GPgRuAEcFdmnp/eqHqzuf322wfu\nX1xcHLj/gQceaHOcckZ55n8V+Epm7gQ+CHwxInYC+4EjmXkzcKS5LukyMTT+zFzLzKeayy8BzwHX\nAXuBA81hB4A7pjWkpPZd0jl/RNwIfAD4BbCQmWvNrhfZOC2QdJkYes5/UUS8HfgJ8OXM/H1EvLYv\nMzMiss/tloClSQeV1K6Rnvkj4q1shH8wM3/abD4TEYvN/kXg7Fa3zczlzOxlZq+NgSW1Y2j8sfEU\n/yDwXGZ+c9Ouw8C+5vI+4FD740malsjc8tX6Xw6IuAX4b2AVuNBs/iob5/3/BtwAnGTjV33nhvys\ngXe2uro6cJarrrqq774dO3YMvK1URWbG8KNGOOfPzJ8D/X7Yxy5lKEnzw0/4SUUZv1SU8UtFGb9U\nlPFLRRm/VNTIH++dB/v3+4eDUlt85peKMn6pKOOXijJ+qSjjl4oyfqko45eKMn6pKOOXijJ+qSjj\nl4oyfqko45eKMn6pKOOXijJ+qSjjl4oyfqko45eKMn6pKOOXijJ+qSjjl4oaGn9EXB8R/xURv4qI\nZyPiS832eyPidEQ83XzdNv1xJbVllEU7XgW+kplPRcQ7gCcj4vFm37cy8+vTG0/StAyNPzPXgLXm\n8ksR8Rxw3bQHkzRdl3TOHxE3Ah8AftFsuicinomIhyLimj63WYqIlYhYmWhSSa0aOf6IeDvwE+DL\nmfl74LvADmA3G68MvrHV7TJzOTN7mdlrYV5JLRkp/oh4KxvhH8zMnwJk5pnMXM/MC8D3gD3TG1NS\n20Z5tz+AB4HnMvObm7Yvbjrs08Cx9seTNC2jvNv/IeCzwGpEPN1s+ypwd0TsBhI4AXxhKhNKmopR\n3u3/ORBb7Hqs/XEkzYqf8JOKMn6pKOOXijJ+qSjjl4oyfqko45eKMn6pKOOXijJ+qSjjl4oyfqko\n45eKMn6pqMjM2d1ZxP8DJzdtehfwu5kNcGnmdbZ5nQucbVxtzvaezPzbUQ6cafxvuPOIlXn9f/vN\n62zzOhc427i6ms2X/VJRxi8V1XX8yx3f/yDzOtu8zgXONq5OZuv0nF9Sd7p+5pfUkU7ij4hbI+I3\nEXE8IvZ3MUM/EXEiIlablYc7XWKsWQbtbEQc27Tt2oh4PCKeb75vuUxaR7PNxcrNA1aW7vSxm7cV\nr2f+sj8itgH/C3wcOAUcBe7OzF/NdJA+IuIE0MvMzn8nHBEfBv4A/DAzdzXbvgacy8z7m384r8nM\nf5yT2e4F/tD1ys3NgjKLm1eWBu4APkeHj92Aue6ig8eti2f+PcDxzHwhM/8I/AjY28Eccy8znwDO\nvW7zXuBAc/kAG//xzFyf2eZCZq5l5lPN5ZeAiytLd/rYDZirE13Efx3w203XTzFfS34n8LOIeDIi\nlroeZgsLzbLpAC8CC10Os4WhKzfP0utWlp6bx26cFa/b5ht+b3RLZv498Engi83L27mUG+ds8/Tr\nmpFWbp6VLVaWfk2Xj924K163rYv4TwPXb7r+7mbbXMjM0833s8CjzN/qw2cuLpLafD/b8TyvmaeV\nm7daWZo5eOzmacXrLuI/CtwcEe+NiLcBnwEOdzDHG0TE1c0bMUTE1cAnmL/Vhw8D+5rL+4BDHc7y\nV+Zl5eZ+K0vT8WM3dyteZ+bMv4Db2HjH//+Af+pihj5z/R3wy+br2a5nAx5m42Xgn9h4b+TzwDuB\nI8DzwH8C187RbP8CrALPsBHaYkez3cLGS/pngKebr9u6fuwGzNXJ4+Yn/KSifMNPKsr4paKMXyrK\n+KWijF8qyvilooxfKsr4paL+DJkSqh2vigMkAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6d0831650>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAADO5JREFUeJzt3V2sVfWZx/Hfb5jWN7hQmQEiZig9ZqIxgY5HnKQ4dlLb\nWNOIeGHgojKJzmliSSRp4vgSM0a9UDNt04ux8TCS0rFjGdMa0JgZHNLEYGrDkQBiwZEhNOUEOCU2\nVHyrwjMXZ2kOevba273X2mvD8/0kJ2fv9ey115Mdfue/3th/R4QA5PNnTTcAoBmEH0iK8ANJEX4g\nKcIPJEX4gaQIP5AU4QeSIvxAUn/ez43ZLr2d8MILLyxdf8GCBS1re/fuLV337bffLq0DZ4qIcCev\n6yn8tq+T9ENJMyT9W0Q83Mv73XDDDaX1devWtawtXbq0dN2XXnqpq56AM1XXu/22Z0j6V0nfkHSZ\npJW2L6uqMQD16uWYf4mkfRGxPyL+JOlnkpZV0xaAuvUS/osk/W7K84PFslPYHrE9Znush20BqFjt\nJ/wiYlTSqNT+hB+A/ull5B+XdPGU5/OLZQBOA72Ef5ukS2x/wfbnJa2QtKmatgDUrevd/oj40PZq\nSf+tyUt96yLitbJ1zj77bA0NDbWsz58/v9t2tHDhwtL6sWPHun7vQTd37tyWtYceeqiPnaAKx48f\nL61v3bq1Ze3xxx/veDs9HfNHxPOSnu/lPQA0g9t7gaQIP5AU4QeSIvxAUoQfSIrwA0m5nzP2DA8P\nx9gYt/gDdRkeHtbY2FhH/5+fkR9IivADSRF+ICnCDyRF+IGkCD+QVF8v9c2YMSPOOeeclvVbbrml\ndP3HHnusZe3aa68tXffll18ubw44A7z77rs6ceIEl/oAtEb4gaQIP5AU4QeSIvxAUoQfSIrwA0n1\ndYrukydPlk6V/d5773X93u3WZYpu4FSM/EBShB9IivADSRF+ICnCDyRF+IGkCD+QVE/X+W0fkPSW\npBOSPoyI4SqaAlC/Km7y+fuIOFrB+wDoI3b7gaR6DX9I2mz7FdsjVTQEoD963e1fGhHjtv9S0gu2\n90bEi1NfUPxR4A8DMGB6GvkjYrz4PSHpGUlLpnnNaEQMczIQGCxdh9/2ebZnffRY0tcl7a6qMQD1\n6mW3f46kZ2x/9D7/ERH/VUlXAGrXdfgjYr+kRRX2AqCPuNQHJEX4gaQIP5AU4QeSIvxAUoQfSIrw\nA0kRfiApwg8kRfiBpAg/kBThB5Ii/EBShB9IivADSRF+ICnCDyRF+IGkCD+QFOEHkiL8QFKEH0iq\nill6cRobGhoqre/bt6+n9589e3bL2qJF5d/8vnPnztL60aNMDt0LRn4gKcIPJEX4gaQIP5AU4QeS\nIvxAUoQfSKrtdX7b6yR9U9JERFxeLLtA0gZJCyQdkHRzRPyhvjbRrVmzZpXWN23aVFrfsWNHT9uf\nP39+y9rVV19duu6GDRtK6ytWrOiqJ0zqZOT/saTrPrHsLklbIuISSVuK5wBOI23DHxEvSnrzE4uX\nSVpfPF4v6caK+wJQs26P+edExKHi8WFJcyrqB0Cf9Hxvf0SE7WhVtz0iaaTX7QCoVrcj/xHb8ySp\n+D3R6oURMRoRwxEx3OW2ANSg2/BvkrSqeLxK0sZq2gHQL23Db/spSb+S9Ne2D9q+VdLDkr5m+w1J\n1xbPAZxG2h7zR8TKFqWvVtwLanDWWWeV1i+99NLS+ubNm3va/sREyyNCnXvuuaXrtvv//OgNd/gB\nSRF+ICnCDyRF+IGkCD+QFOEHkuKru1FqzZo1tb333LlzS+uHDx+ubdtg5AfSIvxAUoQfSIrwA0kR\nfiApwg8kRfiBpLjOj8ZwHb9ZjPxAUoQfSIrwA0kRfiApwg8kRfiBpAg/kBTX+c9wV155ZU/r33bb\nbRV18mnPPvtsaf3IkSO1bRuM/EBahB9IivADSRF+ICnCDyRF+IGkCD+QVNvr/LbXSfqmpImIuLxY\ndr+kf5T0++Jl90TE83U1ie6Nj4+X1o8ePVpaX7t2bZXtnGL37t2l9QceeKC0/vTTT1fZTjqdjPw/\nlnTdNMt/EBGLix+CD5xm2oY/Il6U9GYfegHQR70c86+2vcv2OtvnV9YRgL7oNvw/kvRFSYslHZL0\nvVYvtD1ie8z2WJfbAlCDrsIfEUci4kREnJS0VtKSkteORsRwRAx32ySA6nUVftvzpjxdLqn8tC2A\ngdPJpb6nJH1F0mzbByX9s6Sv2F4sKSQdkPTtGnsEUIO24Y+IldMsfqKGXlCDXbt2ldaXL19eWr/i\niiuqbOcUd9xxR2n9wQcfLK1znb833OEHJEX4gaQIP5AU4QeSIvxAUoQfSIqv7k5u69atPdV7MTQ0\nVFq/6aabSusLFy4sre/fv/8z95QJIz+QFOEHkiL8QFKEH0iK8ANJEX4gKcIPJMV1/jPc7NmzS+sf\nfPBBaf3YsWNVtnOKzZs3l9ZXr15dWr/mmmtK61znL8fIDyRF+IGkCD+QFOEHkiL8QFKEH0iK8ANJ\ncZ3/DPfII4+U1l9//fXS+qOPPlplO6e48847S+vtpg/fvn17le2kw8gPJEX4gaQIP5AU4QeSIvxA\nUoQfSIrwA0m1vc5v+2JJP5E0R1JIGo2IH9q+QNIGSQskHZB0c0T8ob5WUYf77ruvtL53796e3n/R\nokUta1dddVXpus8991xpfefOnV31hEmdjPwfSvpuRFwm6W8lfcf2ZZLukrQlIi6RtKV4DuA00Tb8\nEXEoIrYXj9+StEfSRZKWSVpfvGy9pBvrahJA9T7TMb/tBZK+JOnXkuZExKGidFiThwUAThMd39tv\ne6akn0taExF/tP1xLSLCdrRYb0TSSK+NAqhWRyO/7c9pMvg/jYhfFIuP2J5X1OdJmphu3YgYjYjh\niBiuomEA1Wgbfk8O8U9I2hMR359S2iRpVfF4laSN1bcHoC6d7PZ/WdK3JL1qe0ex7B5JD0v6T9u3\nSvqtpJvraRG9OH78eGl95syZpfWNG3v7m/7++++3rG3btq103dtvv72nbaNc2/BHxFZJblH+arXt\nAOgX7vADkiL8QFKEH0iK8ANJEX4gKcIPJMVXd5/h7r777tJ6u/sAerVnz56WtSeffLLWbaMcIz+Q\nFOEHkiL8QFKEH0iK8ANJEX4gKcIPJMV1/jPcO++8U1q/9957+9QJBg0jP5AU4QeSIvxAUoQfSIrw\nA0kRfiApwg8kRfiBpAg/kBThB5Ii/EBShB9IivADSRF+ICnCDyTVNvy2L7b9S9u/sf2a7TuK5ffb\nHre9o/i5vv52AVSlky/z+FDSdyNiu+1Zkl6x/UJR+0FE/Et97QGoS9vwR8QhSYeKx2/Z3iPporob\nA1Cvz3TMb3uBpC9J+nWxaLXtXbbX2T6/xTojtsdsj/XUKYBKdRx+2zMl/VzSmoj4o6QfSfqipMWa\n3DP43nTrRcRoRAxHxHAF/QKoSEfht/05TQb/pxHxC0mKiCMRcSIiTkpaK2lJfW0CqFonZ/st6QlJ\neyLi+1OWz5vysuWSdlffHoC6dHK2/8uSviXpVds7imX3SFppe7GkkHRA0rdr6RBALTo5279Vkqcp\nPV99OwD6hTv8gKQIP5AU4QeSIvxAUoQfSIrwA0kRfiApwg8kRfiBpAg/kBThB5Ii/EBShB9IivAD\nSTki+rcx+/eSfjtl0WxJR/vWwGczqL0Nal8SvXWryt7+KiL+opMX9jX8n9q4PTao3+03qL0Nal8S\nvXWrqd7Y7QeSIvxAUk2Hf7Th7ZcZ1N4GtS+J3rrVSG+NHvMDaE7TIz+AhjQSftvX2X7d9j7bdzXR\nQyu2D9h+tZh5uNEpxopp0CZs756y7ALbL9h+o/g97TRpDfU2EDM3l8ws3ehnN2gzXvd9t9/2DEn/\nK+lrkg5K2iZpZUT8pq+NtGD7gKThiGj8mrDtv5N0XNJPIuLyYtmjkt6MiIeLP5znR8Q/DUhv90s6\n3vTMzcWEMvOmziwt6UZJ/6AGP7uSvm5WA59bEyP/Ekn7ImJ/RPxJ0s8kLWugj4EXES9KevMTi5dJ\nWl88Xq/Jfzx916K3gRARhyJie/H4LUkfzSzd6GdX0lcjmgj/RZJ+N+X5QQ3WlN8habPtV2yPNN3M\nNOYU06ZL0mFJc5psZhptZ27up0/MLD0wn103M15XjRN+n7Y0Iv5G0jckfafYvR1IMXnMNkiXazqa\nublfpplZ+mNNfnbdznhdtSbCPy7p4inP5xfLBkJEjBe/JyQ9o8GbffjIR5OkFr8nGu7nY4M0c/N0\nM0trAD67QZrxuonwb5N0ie0v2P68pBWSNjXQx6fYPq84ESPb50n6ugZv9uFNklYVj1dJ2thgL6cY\nlJmbW80srYY/u4Gb8Toi+v4j6XpNnvH/P0n3NtFDi74WStpZ/LzWdG+SntLkbuAHmjw3cqukCyVt\nkfSGpP+RdMEA9fbvkl6VtEuTQZvXUG9LNblLv0vSjuLn+qY/u5K+GvncuMMPSIoTfkBShB9IivAD\nSRF+ICnCDyRF+IGkCD+QFOEHkvp/3C4OvoY8ChUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6d03faf90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cb26c0d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cb80d590>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6d028b910>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nFOEHgiL8QFB/BxrBWiGr1eJIAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd700443210>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6d08c8ad0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cbf76f10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cab2eed0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngaSIH0iK+IGk/g+4gK+wh/SR1gAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6d04e8290>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngiL8QFCEHwiK8ANBEX4gKMIPBPV/DI0XOmI+4GoAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cb0dbc90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6caf47050>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTWvKX1PAmwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6caac5f50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6d02d3d50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n4QeC+h9P4COpFqvFeAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cac47b10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cb1d6f10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\npAg/kBThB5Ii/EBSfwIX1xHrvLc6PgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cb0e8190>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAC4pJREFUeJzt3V+IXYWdwPHvT21lMHnQzRpCImu2iBAKmy5DWKgslW6L\nlYVYEWkeShZk04cKrfRhxUVWfJJl29CHpTBdQ9Ola3clFfMgu3XDghSX4CipiXW3upLSDGPSEKEW\nhtQxv32YYxl17h/vPfeeO/19PzDMvefcm/Pjku+ce++5MycyE0n1XNX1AJK6YfxSUcYvFWX8UlHG\nLxVl/FJRxi8VZfxSUcYvFXXNNDcWEX0/Trhnz56+919eXu657q233hptKOn3TGbGMLeLcT7eGxF3\nAN8Grgb+KTMfG3D7vhs7ffp03+09+uijPdc9+eSTfe8rVTFs/CM/7Y+Iq4F/BL4A7AEORET/Xbek\nmTHOa/59wOuZ+UZm/hb4IbC/nbEkTdo48e8Efrnu+rlm2ftExKGIWIyIxTG2JallE3/DLzMXgAUY\n/Jpf0vSMs+dfAm5ad31Xs0zSJjBO/C8At0TE7oj4OPAl4Hg7Y0matJGf9mfmakTcD/wHa4f6jmTm\nK/3us2vXLh544IGe62+99da+29y3b1/Pdaurq33v2+8wIcDWrVv7rpdmxbFjx3quO3z48ND/zliv\n+TPzGeCZcf4NSd3w471SUcYvFWX8UlHGLxVl/FJRxi8VNdXf519dXeXixYt91/dz991391x3++23\n973v5cuXx1ovzYpxGlrPPb9UlPFLRRm/VJTxS0UZv1SU8UtFjfXXez+qq666Kq+99tqe60+dOtX3\n/g8//HDPdf71XmnNxP96r6TNzfilooxfKsr4paKMXyrK+KWijF8qaqq/0rtt2zbuueeenusH/elu\nSe1xzy8VZfxSUcYvFWX8UlHGLxVl/FJRxi8VNdZx/og4C7wNvAusZuZ8G0NJmrw2PuRze2b2/kPi\nkmaST/ulosaNP4EfR8SLEXGojYEkTce4T/tvy8yliLgReDYi/iczn1t/g+aHwiGALVu2jLk5SW0Z\na8+fmUvN9wvAU8C+DW6zkJnzmTk/Nzc3zuYktWjk+CPiuojY+t5l4PPAmbYGkzRZ4zzt3w48FRHv\n/Tv/kpn/3spUkiZu5Pgz8w3gT1qcRdIUeahPKsr4paKMXyrK+KWijF8qyvilooxfKsr4paKMXyrK\n+KWijF8qyvilooxfKsr4paKMXyrK+KWijF8qyvilooxfKsr4paKMXyrK+KWijF8qyvilooxfKsr4\npaKMXyrK+KWijF8qyvilogbGHxFHIuJCRJxZt+yGiHg2Il5rvl8/2TEltW2YPf/3gDs+sOxB4ERm\n3gKcaK5L2kQGxp+ZzwGXPrB4P3C0uXwUuKvluSRN2Kiv+bdn5nJz+U1ge0vzSJqSsd/wy8wEstf6\niDgUEYsRsbiysjLu5iS1ZNT4z0fEDoDm+4VeN8zMhcycz8z5ubm5ETcnqW2jxn8cONhcPgg83c44\nkqZlmEN9TwD/DdwaEeci4j7gMeBzEfEa8BfNdUmbyDWDbpCZB3qs+mzLs0iaIj/hJxVl/FJRxi8V\nZfxSUcYvFWX8UlHGLxVl/FJRxi8VZfxSUcYvFWX8UlHGLxVl/FJRxi8VZfxSUcYvFWX8UlHGLxVl\n/FJRxi8VZfxSUcYvFWX8UlHGLxVl/FJRxi8VZfxSUcYvFWX8UlEDT9EdEUeAvwQuZOYnm2WPAH8N\n/Kq52UOZ+cykhlR/O3fu7Lnu8uXLfe978eLFtsfRJjHMnv97wB0bLD+cmXubL8OXNpmB8Wfmc8Cl\nKcwiaYrGec1/f0S8HBFHIuL61iaSNBWjxv8d4BPAXmAZ+GavG0bEoYhYjIjFlZWVETcnqW0jxZ+Z\n5zPz3cy8AnwX2NfntguZOZ+Z83Nzc6POKallI8UfETvWXf0icKadcSRNyzCH+p4APgNsi4hzwN8B\nn4mIvUACZ4GvTHBGSRMwMP7MPLDB4scnMItG1O9Y/jvvvDPFSbSZ+Ak/qSjjl4oyfqko45eKMn6p\nKOOXihp4qE+zz1/L1Sjc80tFGb9UlPFLRRm/VJTxS0UZv1SU8UtFeZxfm9aNN97Yd/3u3bt7rjt5\n8mTb42w67vmlooxfKsr4paKMXyrK+KWijF8qyvilojzOr01r0N8xuHTJ88v2455fKsr4paKMXyrK\n+KWijF8qyvilooxfKmrgcf6IuAn4PrAdSGAhM78dETcA/wrcDJwF7s3MtyY3qvR+V65cGWt9dcPs\n+VeBb2TmHuDPgK9GxB7gQeBEZt4CnGiuS9okBsafmcuZ+VJz+W3gVWAnsB842tzsKHDXpIaU1L6P\n9Jo/Im4GPgWcBLZn5nKz6k3WXhZI2iSGjj8itgDHgK9n5q/Xr8vMZO39gI3udygiFiNicWVlZaxh\nJbVnqPgj4mOshf+DzPxRs/h8ROxo1u8ALmx038xcyMz5zJyfm5trY2ZJLRgYf0QE8DjwamZ+a92q\n48DB5vJB4On2x5M0KcP8Su+ngS8DpyPiVLPsIeAx4N8i4j7gF8C9kxlR0iQMjD8zfwJEj9WfbXcc\nSdPiJ/ykooxfKsr4paKMXyrK+KWijF8qyvilooxfKsr4paKMXyrK+KWijF8qyvilooxfKsr4paKM\nXyrK+KWijF8qyvilooxfKsr4paKMXyrK+KWijF8qyvilooxfKsr4paKMXyrK+KWijF8qamD8EXFT\nRPxXRPwsIl6JiK81yx+JiKWIONV83Tn5cSW15ZohbrMKfCMzX4qIrcCLEfFss+5wZv7D5MaTNCkD\n48/MZWC5ufx2RLwK7Jz0YJIm6yO95o+Im4FPASebRfdHxMsRcSQiru9xn0MRsRgRiysrK2MNK6k9\nQ8cfEVuAY8DXM/PXwHeATwB7WXtm8M2N7peZC5k5n5nzc3NzLYwsqQ1DxR8RH2Mt/B9k5o8AMvN8\nZr6bmVeA7wL7JjempLYN825/AI8Dr2bmt9Yt37HuZl8EzrQ/nqRJGebd/k8DXwZOR8SpZtlDwIGI\n2AskcBb4yrjDnDnT/+fH888/P+4mJDWGebf/J0BssOqZ9seRNC1+wk8qyvilooxfKsr4paKMXyrK\n+KWihjnOPzUXL17su35paWlKk0i//9zzS0UZv1SU8UtFGb9UlPFLRRm/VJTxS0VFZk5vYxG/An6x\nbtE2oP/B/e7M6myzOhc426janO2PMvMPh7nhVOP/0MYjFjNzvrMB+pjV2WZ1LnC2UXU1m0/7paKM\nXyqq6/gXOt5+P7M626zOBc42qk5m6/Q1v6TudL3nl9SRTuKPiDsi4n8j4vWIeLCLGXqJiLMRcbo5\n8/Bix7MciYgLEXFm3bIbIuLZiHit+b7hadI6mm0mztzc58zSnT52s3bG66k/7Y+Iq4GfA58DzgEv\nAAcy82dTHaSHiDgLzGdm58eEI+LPgd8A38/MTzbL/h64lJmPNT84r8/Mv5mR2R4BftP1mZubE8rs\nWH9maeAu4K/o8LHrM9e9dPC4dbHn3we8nplvZOZvgR8C+zuYY+Zl5nPApQ8s3g8cbS4fZe0/z9T1\nmG0mZOZyZr7UXH4beO/M0p0+dn3m6kQX8e8Efrnu+jlm65TfCfw4Il6MiENdD7OB7c1p0wHeBLZ3\nOcwGBp65eZo+cGbpmXnsRjnjddt8w+/DbsvMPwW+AHy1eXo7k3LtNdssHa4Z6szN07LBmaV/p8vH\nbtQzXreti/iXgJvWXd/VLJsJmbnUfL8APMXsnX34/HsnSW2+X+h4nt+ZpTM3b3RmaWbgsZulM153\nEf8LwC0RsTsiPg58CTjewRwfEhHXNW/EEBHXAZ9n9s4+fBw42Fw+CDzd4SzvMytnbu51Zmk6fuxm\n7ozXmTn1L+BO1t7x/z/gb7uYocdcfwz8tPl6pevZgCdYexr4DmvvjdwH/AFwAngN+E/ghhma7Z+B\n08DLrIW2o6PZbmPtKf3LwKnm686uH7s+c3XyuPkJP6ko3/CTijJ+qSjjl4oyfqko45eKMn6pKOOX\nijJ+qaj/B1ikvg4K9s1lAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6f9de9190>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSIr4gaSIH0iK+IGkiB9I6v+TRcaz/moPkwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6fa0cc550>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nxA8k9X8GaYTdvFN23gAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cad2f990>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c979f610>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c86519d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nC6LzAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6d0afab50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6d0a56110>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n4QeSIvxAUoQfSIrwA0kRfiCp/wHXluuI8HlvOAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cb0ce0d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n9X+lzHYeYHpNAgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6d084e650>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6d0ca5610>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIn6gKOIHiiJ+oCjiB4oifqCo/wDhm3/YOfib5wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cbdbbb90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n0gAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cbbeeb50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\npIgfSIr4gaSIH0jq/6J1s2nUMgYrAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6c9258110>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAADSpJREFUeJzt3X+MVfWZx/HPZ12QkWLiryUTMTu1MasGs3SdkA01hk1t\ntQbFGiXlj8ImuhhTkm3CHytsov5pNgtNE1YSuiI/0rVd06pjNF1YssQ0bIqjcVFR1x+hGQgyECoF\nM4aVPvvHHJpR5p473HvuPRee9yuZzL3nOed+n9zwmXPv+V7u1xEhAPn8Sd0NAKgH4QeSIvxAUoQf\nSIrwA0kRfiApwg8kRfiBpAg/kNSfdnOwK6+8MgYGBlo+/tNPP21YmzlzZsuPC/SS06dPl9Y/+eST\nhrWjR4/qxIkTnso4bYXf9h2SfizpIkn/GhFPlO0/MDCg4eHhlsfbs2dPw9r8+fNbflyglxw/fry0\nPjQ01LD26KOPTnmcll/2275I0r9I+o6kGyUttX1jq48HoLvaec8/X9IHEfFRRJyS9DNJi6tpC0Cn\ntRP+qyWNTLh/oNj2BbZX2B62PXzkyJE2hgNQpY5f7Y+IjRExGBGDV111VaeHAzBF7YT/oKRrJtyf\nU2wDcB5oJ/yvSrrO9ldtT5f0PUmNL0MC6CktT/VFxOe2V0r6D41P9W2KiLfLjhkdHdX69esb1m+9\n9dbSMe+7776GtRtuuKH0WOB8cfLkydL67t27KxmnrXn+iHhZ0suVdAKgq/h4L5AU4QeSIvxAUoQf\nSIrwA0kRfiCprv5//s8++0z79u1rWJ87d27p8SMjIy3VAJyNMz+QFOEHkiL8QFKEH0iK8ANJEX4g\nqa5O9U2bNk1z5szp5pAAGuDMDyRF+IGkCD+QFOEHkiL8QFKEH0iK8ANJdXWev7+/X2vWrGlY37Vr\nV/eaAZLjzA8kRfiBpAg/kBThB5Ii/EBShB9IivADSbU1z297v6QTkk5L+jwiBsv2P3jwoFavXt2w\nfvvtt7fTDoBzUMWHfP4mIo5W8DgAuoiX/UBS7YY/JG23/ZrtFVU0BKA72n3Zf0tEHLT9Z5J22H43\nIl6ZuEPxR2GFJF166aVtDgegKm2d+SPiYPF7VNJzkuZPss/GiBiMiMFLLrmkneEAVKjl8NueaXvW\nmduSvi3praoaA9BZ7bzsny3pOdtnHuffIuJXlXQFoONaDn9EfCTpL8/lmNOnT+v48eOtDgmgQkz1\nAUkRfiApwg8kRfiBpAg/kBThB5Ii/EBShB9IivADSRF+ICnCDyRF+IGkCD+QFOEHkurqEt3oPc2+\nWu2uu+4qrS9btqzKdr7gwQcfLK2PjIx0bOwMOPMDSRF+ICnCDyRF+IGkCD+QFOEHkiL8QFLM81/g\nrrjiitL60NBQaX3BggVVtnNO1q5dW1pfsmRJlzq5MHHmB5Ii/EBShB9IivADSRF+ICnCDyRF+IGk\nms7z294kaZGk0YiYW2y7XNLPJQ1I2i9pSUT8rnNtolU333xzab3ZPP66detK61u2bDnnns54+OGH\nS+vNPqOA9kzlzL9Z0h1f2vaIpJ0RcZ2kncV9AOeRpuGPiFckHfvS5sWSzvzJ3yLpnor7AtBhrb7n\nnx0Rh4rbH0uaXVE/ALqk7Qt+ERGSolHd9grbw7aHx8bG2h0OQEVaDf9h2/2SVPwebbRjRGyMiMGI\nGOzr62txOABVazX8Q5KWF7eXS3qhmnYAdEvT8Nt+RtJ/S/oL2wdsPyDpCUnfsv2+pNuK+wDOI03n\n+SNiaYPSNyvuBR2wevXqto5ftWpVRZ2cbdu2baV15vk7i0/4AUkRfiApwg8kRfiBpAg/kBThB5Li\nq7svAGXLaDf7L7vtTgW2Y/fu3bWNDc78QFqEH0iK8ANJEX4gKcIPJEX4gaQIP5AU8/wXgBkzZjSs\nTZ8+vfTY48ePV90OzhOc+YGkCD+QFOEHkiL8QFKEH0iK8ANJEX4gKeb5k3vyySfbqnfSokWLSusv\nvfRSlzq5MHHmB5Ii/EBShB9IivADSRF+ICnCDyRF+IGkms7z294kaZGk0YiYW2x7XNLfSTpS7LYm\nIl7uVJPonPXr15fWX3zxxbYef9asWQ1rmzdvLj32oYceKq3v2LGjtH7q1KnSenZTOfNvlnTHJNt/\nFBHzih+CD5xnmoY/Il6RdKwLvQDoonbe86+0vdf2JtuXVdYRgK5oNfwbJH1N0jxJhyStbbSj7RW2\nh20Pj42NtTgcgKq1FP6IOBwRpyPiD5J+Iml+yb4bI2IwIgb7+vpa7RNAxVoKv+3+CXe/K+mtatoB\n0C1Tmep7RtJCSVfaPiDpMUkLbc+TFJL2SyqfkwHQc5qGPyKWTrL5qQ70ghrs27evtL59+/aOjX3T\nTTeV1h977LHS+oIFC0rru3btOteWUuETfkBShB9IivADSRF+ICnCDyRF+IGk+Opu1CYi6m4hNc78\nQFKEH0iK8ANJEX4gKcIPJEX4gaQIP5AU8/wXgL179zasffjhh6XH3nvvvaX1DRs2tNTTVNju2GOj\nOc78QFKEH0iK8ANJEX4gKcIPJEX4gaQIP5AU8/wXgPfee69hbWRkpPTY2267rbR+/fXXl9bffffd\n0vr06dMb1q699trSYw8fPlxaP3aM9WPbwZkfSIrwA0kRfiApwg8kRfiBpAg/kBThB5JqOs9v+xpJ\nWyXNlhSSNkbEj21fLunnkgYk7Ze0JCJ+17lW0YqtW7eW1hcuXFhaf/7550vre/bsKa3PmDGjYe3+\n++8vPfbZZ58trZd9jwGam8qZ/3NJqyLiRkl/LekHtm+U9IiknRFxnaSdxX0A54mm4Y+IQxHxenH7\nhKR3JF0tabGkLcVuWyTd06kmAVTvnN7z2x6Q9HVJv5E0OyIOFaWPNf62AMB5Ysrht/0VSb+Q9MOI\n+P3EWowvujbpwmu2V9getj08NjbWVrMAqjOl8NuepvHg/zQifllsPmy7v6j3Sxqd7NiI2BgRgxEx\n2NfXV0XPACrQNPwe/4rVpyS9ExHrJpSGJC0vbi+X9EL17QHolKn8l95vSPq+pDdtv1FsWyPpCUn/\nbvsBSb+VtKQzLaIdTz/9dFv1ZlOFd999d2n94osvbljbtm1b6bErV64sraM9TcMfEb+W1OgL1r9Z\nbTsAuoVP+AFJEX4gKcIPJEX4gaQIP5AU4QeS4qu7UWrZsmWl9QULFpTWy766e9euXa20hIpw5geS\nIvxAUoQfSIrwA0kRfiApwg8kRfiBpJjnR1t2795ddwtoEWd+ICnCDyRF+IGkCD+QFOEHkiL8QFKE\nH0iK8ANJEX4gKcIPJEX4gaQIP5AU4QeSIvxAUoQfSKpp+G1fY/u/bO+z/bbtvy+2P277oO03ip87\nO98ugKpM5cs8Ppe0KiJetz1L0mu2dxS1H0XEP3euPQCd0jT8EXFI0qHi9gnb70i6utONAeisc3rP\nb3tA0tcl/abYtNL2XtubbF/W4JgVtodtD4+NjbXVLIDqTDn8tr8i6ReSfhgRv5e0QdLXJM3T+CuD\ntZMdFxEbI2IwIgb7+voqaBlAFaYUftvTNB78n0bELyUpIg5HxOmI+IOkn0ia37k2AVRtKlf7Lekp\nSe9ExLoJ2/sn7PZdSW9V3x6ATpnK1f5vSPq+pDdtv1FsWyNpqe15kkLSfkkPdaRDAB0xlav9v5bk\nSUovV98OgG7hE35AUoQfSIrwA0kRfiApwg8kRfiBpAg/kBThB5Ii/EBShB9IivADSRF+ICnCDyRF\n+IGkHBHdG8w+Ium3EzZdKelo1xo4N73aW6/2JdFbq6rs7c8j4qqp7NjV8J81uD0cEYO1NVCiV3vr\n1b4kemtVXb3xsh9IivADSdUd/o01j1+mV3vr1b4kemtVLb3V+p4fQH3qPvMDqEkt4bd9h+33bH9g\n+5E6emjE9n7bbxYrDw/X3Msm26O235qw7XLbO2y/X/yedJm0mnrriZWbS1aWrvW567UVr7v+st/2\nRZL+V9K3JB2Q9KqkpRGxr6uNNGB7v6TBiKh9Ttj2rZJOStoaEXOLbf8k6VhEPFH84bwsIv6hR3p7\nXNLJulduLhaU6Z+4srSkeyT9rWp87kr6WqIanrc6zvzzJX0QER9FxClJP5O0uIY+el5EvCLp2Jc2\nL5a0pbi9ReP/eLquQW89ISIORcTrxe0Tks6sLF3rc1fSVy3qCP/VkkYm3D+g3lryOyRtt/2a7RV1\nNzOJ2cWy6ZL0saTZdTYziaYrN3fTl1aW7pnnrpUVr6vGBb+z3RIRfyXpO5J+ULy87Ukx/p6tl6Zr\nprRyc7dMsrL0H9X53LW64nXV6gj/QUnXTLg/p9jWEyLiYPF7VNJz6r3Vhw+fWSS1+D1acz9/1Esr\nN0+2srR64LnrpRWv6wj/q5Kus/1V29MlfU/SUA19nMX2zOJCjGzPlPRt9d7qw0OSlhe3l0t6ocZe\nvqBXVm5utLK0an7uem7F64jo+o+kOzV+xf9DSf9YRw8N+rpW0v8UP2/X3ZukZzT+MvD/NH5t5AFJ\nV0jaKel9Sf8p6fIe6m2bpDcl7dV40Ppr6u0Wjb+k3yvpjeLnzrqfu5K+anne+IQfkBQX/ICkCD+Q\nFOEHkiL8QFKEH0iK8ANJEX4gKcIPJPX/YRcSdqP+RTUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cbb640d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6d02db650>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nqH8A8lprm35dBi0AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6cb7bf610>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\npQJSAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd6d0808b90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "S = np.zeros((9, 9), dtype = 'uint8')\n",
    "kk = 0\n",
    "for i in range(0, 9):\n",
    "    for j in range(0, 9):\n",
    "        \n",
    "        y1 = int(Points[j+i*10][1])\n",
    "        y2 = int(Points[j+i*10 +10][1])\n",
    "        \n",
    "        x1 = int(Points[j+i*10][0])\n",
    "        x2 = int(Points[j+i*10+1][0])\n",
    "        \n",
    "        x = x1\n",
    "        y = y1\n",
    "        w = x2 - x1\n",
    "        h = y2 - y1\n",
    "        \n",
    "        # cropped cell\n",
    "        cropped_cell = im_th[y:y+h, x:x+w]\n",
    "        \n",
    "        #Region of Interest\n",
    "        roi = cropped_cell\n",
    "        # Resize the image\n",
    "        roi = cv2.resize(roi, (28, 28), interpolation=cv2.INTER_AREA)\n",
    "        \n",
    "        roi = roi.flatten()\n",
    "        roi = roi[np.newaxis]\n",
    "        plt.figure(kk)\n",
    "        kk += 1\n",
    "        plt.imshow(roi.reshape((28, 28)), cmap='gray')\n",
    "        \n",
    "        S[i][j] = model.predict_classes(roi,verbose=0)\n",
    "print S\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}