{
 "metadata": {
  "name": ""
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "heading",
     "level": 1,
     "metadata": {},
     "source": [
      "Nejlep\u0161\u00ed v\u00fdsledky ka\u017ed\u00e9ho t\u00fdmu"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Zdrojov\u00e1 data pro tuto tabulku jsou dostupn\u00e1 [zde](https://raw.githubusercontent.com/mjirik/ZDO2014semestralky/master/ZDO2014evaluation.csv)"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import pandas as pd\n",
      "orig_data = pd.read_csv(\"ZDO2014evaluation.csv\")\n",
      "#orig_data.sort(column=['score'], ascending=False)\n",
      "#print orig_data\n",
      "gb = orig_data.groupby('team')\n",
      "#print gb.max().sort(column=['score'], ascending=False)\n",
      "\n",
      "gbsort = gb.max().sort(column=['score'], ascending=False)\n",
      "gbsort['score']\n",
      "#tt.tolist()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 8,
       "text": [
        "team\n",
        "RakHejdovaJuna      0.940804338003\n",
        "RakHejdovaJuna2      0.78640776699\n",
        "nedvedj             0.699029126214\n",
        "cervenym             0.68932038835\n",
        "sample M. Jirik     0.300970873786\n",
        "nedvedj2           0.0194174757282\n",
        "Name: score, dtype: object"
       ]
      }
     ],
     "prompt_number": 8
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Bodov\u00e1n\u00ed za technickou \u010d\u00e1st funk\u010dnosti"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "K t\u011bmto bod\u016fm jsou pak v hodnocen\u00ed funk\u010dnosti je\u0161t\u011b p\u0159id\u00e1ny body za dokumentaci v k\u00f3du, demo re\u017eim, git atd."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%pylab inline --no-import-all\n",
      "import matplotlib.pyplot as plt\n",
      "\n",
      "print np.power(gbsort['score'].astype(np.double),2)*36\n",
      "\n",
      "## vizualizace progrese\n",
      "te = np.arange(0,1.1,0.1)\n",
      "y = np.power(te,2)*36\n",
      "plt.figure(figsize=(2,1.5))\n",
      "plt.plot(te,y)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Populating the interactive namespace from numpy and matplotlib\n",
        "team\n",
        "RakHejdovaJuna     31.864061\n",
        "RakHejdovaJuna2    22.263738\n",
        "nedvedj            17.591102\n",
        "cervenym           17.105854\n",
        "sample M. Jirik     3.261005\n",
        "nedvedj2            0.013573\n",
        "Name: score, dtype: float64\n"
       ]
      },
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 9,
       "text": [
        "[<matplotlib.lines.Line2D at 0x7f52e824e590>]"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAJQAAAB1CAYAAACoJMjzAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f52e816b810>"
       ]
      }
     ],
     "prompt_number": 9
    },
    {
     "cell_type": "heading",
     "level": 1,
     "metadata": {},
     "source": [
      "V\u0161echny experimenty"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "#pd.set_option('display.max_columns', None)\n",
      "print orig_data.tail(40)\n",
      "#print dir(pd)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "                team     score        date             time\n",
        "132          nedvedj  0.323995  2014-06-10  10:27:20.666130\n",
        "133   RakHejdovaJuna  0.940804  2014-06-10  10:28:26.399551\n",
        "134  RakHejdovaJuna2  0.000000  2014-06-10  10:28:26.485093\n",
        "135         nedvedj2  0.000000  2014-06-10  10:28:26.662503\n",
        "136  sample M. Jirik  0.000000  2014-06-10  10:30:10.339569\n",
        "137         cervenym  0.618445  2014-06-10  10:32:06.897540\n",
        "138          nedvedj  0.344937  2014-06-10  10:32:59.784266\n",
        "139   RakHejdovaJuna  0.939873  2014-06-10  10:34:05.136954\n",
        "140  RakHejdovaJuna2  0.000000  2014-06-10  10:34:05.223635\n",
        "141         nedvedj2  0.000000  2014-06-10  10:34:05.401491\n",
        "142  sample M. Jirik  0.000000  2014-06-10  10:40:05.238037\n",
        "143         cervenym  0.629295  2014-06-10  10:42:01.262574\n",
        "144          nedvedj  0.322785  2014-06-10  10:42:54.287133\n",
        "145   RakHejdovaJuna  0.940778  2014-06-10  10:43:58.248363\n",
        "146  RakHejdovaJuna2  0.000000  2014-06-10  10:43:58.333504\n",
        "147         nedvedj2  0.000000  2014-06-10  10:43:58.510197\n",
        "148  sample M. Jirik  0.300497  2014-06-10  10:53:00.830229\n",
        "149         cervenym  0.617262  2014-06-10  10:54:55.937289\n",
        "150          nedvedj  0.323995  2014-06-10  10:55:49.244489\n",
        "151   RakHejdovaJuna  0.940804  2014-06-10  10:56:52.098387\n",
        "152  sample M. Jirik  0.298373  2014-06-20  09:47:57.243949\n",
        "153         cervenym  0.618445  2014-06-20  09:49:51.402369\n",
        "154          nedvedj  0.000000  2014-06-20  09:49:51.403685\n",
        "155   RakHejdovaJuna  0.939873  2014-06-20  09:50:55.445074\n",
        "156  sample M. Jirik  0.250000  2014-06-23  09:26:35.657323\n",
        "157         cervenym  0.673077  2014-06-23  09:26:42.847864\n",
        "158          nedvedj  0.673077  2014-06-23  09:26:44.600640\n",
        "159   RakHejdovaJuna  0.759615  2014-06-23  09:26:49.038464\n",
        "160  sample M. Jirik  0.223301  2014-06-23  09:50:27.159108\n",
        "161         cervenym  0.592233  2014-06-23  09:50:33.911572\n",
        "162          nedvedj  0.699029  2014-06-23  09:50:35.571370\n",
        "163   RakHejdovaJuna  0.718447  2014-06-23  09:50:39.678260\n",
        "164  sample M. Jirik  0.223301  2014-06-23  09:55:08.823339\n",
        "165         cervenym  0.592233  2014-06-23  09:55:15.655003\n",
        "166          nedvedj  0.699029  2014-06-23  09:55:17.301895\n",
        "167   RakHejdovaJuna  0.718447  2014-06-23  09:55:21.398045\n",
        "168  sample M. Jirik  0.174757  2014-06-23  09:57:23.405620\n",
        "169         cervenym  0.543689  2014-06-23  09:57:30.615811\n",
        "170          nedvedj  0.689320  2014-06-23  09:57:32.268525\n",
        "171   RakHejdovaJuna  0.854369  2014-06-23  09:57:36.362442\n",
        "\n",
        "[40 rows x 4 columns]\n"
       ]
      }
     ],
     "prompt_number": 10
    },
    {
     "cell_type": "heading",
     "level": 1,
     "metadata": {},
     "source": [
      "Logovac\u00ed soubor"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "V pr\u016fb\u011bhu vyhodnocov\u00e1n\u00ed je vytv\u00e1\u0159en logovac\u00ed soubor. Nahl\u00e1dnout do n\u011bj m\u016f\u017eete [zde](https://raw.githubusercontent.com/mjirik/ZDO2014semestralky/master/znacky.log). Z\u00e1kladem je modul logging. Ten pak pou\u017e\u00edv\u00e1te nam\u00edsto printu. "
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Tyto \u0159\u00e1dky p\u0159idejte na za\u010d\u00e1tek souboru k ostatn\u00edm import\u016fm"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import logging\n",
      "logger = logging.getLogger(__name__)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 11
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Tento \u0159\u00e1dek p\u0159ijde do funkce main"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "logging.basicConfig(level=logging.DEBUG)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 12
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "A takto si vypisujete libovoln\u00e9 zpr\u00e1vy pomoc\u00ed loggeru."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "logger.debug('muj vypis cislo ' + str(1))"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 13
    },
    {
     "cell_type": "heading",
     "level": 1,
     "metadata": {},
     "source": [
      "Matice z\u00e1m\u011bn (confusion matrix)"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "Matice z\u00e1m\u011bn ukazuje, kolikr\u00e1t byl label $i$ klasifikov\u00e1n jako $j$. V ide\u00e1ln\u00edm p\u0159\u00edpad\u011b jsou tedy nenulov\u00e1 \u010d\u00edsla pouze na diagon\u00e1le. Tam jsou spr\u00e1vn\u00e9 klasifikace. Jak to m\u011blo dopadnout ukazuj\u00ed data pojmenovan\u00e1 *reference*\n",
      "\n",
      "Zdrojov\u00e1 data lze naj\u00edt [zde](https://raw.githubusercontent.com/mjirik/ZDO2014semestralky/master/ZDO2014classifs.csv)"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%pylab inline --no-import-all\n",
      "import sklearn\n",
      "import sklearn.metrics\n",
      "import matplotlib.pyplot as plt\n",
      "\n",
      "df = pd.read_csv(\"ZDO2014classifs.csv\")\n",
      "\n",
      "cols = df.columns\n",
      "\n",
      "\n",
      "print cols\n",
      "\n",
      "# z\u00edsk\u00e1me seznam v\u0161ech label\u016f zna\u010dek, kter\u00e9 jsou v testu\n",
      "unlabels = np.unique(df['reference'].values)\n",
      "# p\u0159evedeme do \u010d\u00edsel (index\u016f)\n",
      "y_true = np.searchsorted(unlabels, df['reference'].values)\n",
      "print unlabels\n",
      "\n",
      "for col in cols:\n",
      "    y_pred = np.searchsorted(unlabels, df[col].values)\n",
      "    cmat = sklearn.metrics.confusion_matrix(y_true, y_pred)\n",
      "    plt.matshow(cmat)\n",
      "    plt.title(col)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Populating the interactive namespace from numpy and matplotlib\n",
        "Index([u'RakHejdovaJuna', u'cervenym', u'nedvedj', u'reference', u'sample M. Jirik'], dtype='object')\n",
        "['A11' 'A22' 'A24' 'B13' 'B16' 'B2' 'B20a' 'B21a' 'B24a' 'B24b' 'B28' 'B29'\n",
        " 'B4' 'C4a' 'C4c' 'E13' 'E2b' 'E3a' 'E4' 'E7b' 'E9' 'IJ4a' 'IJ4b' 'IJ7'\n",
        " 'IP11a' 'IP15a' 'IP15b' 'IP18b' 'IP19' 'IP5' 'IP6' 'IS12a' 'IS12b' 'IS18a'\n",
        " 'IS18b' 'IS23' 'IS24b' 'IS2c' 'IS3a' 'IS3b' 'IS3c' 'IS3d' 'IS5' 'IS6a'\n",
        " 'IS6b' 'IS6e' 'IS6f' 'IS6g' 'IS7a' 'IS8b' 'IS9a' 'P1' 'P2' 'P4' 'X1' 'X2'\n",
        " 'X3' 'Z3' 'Z4a' 'Z4b' 'Z4c' 'Z4d']\n"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAP0AAAEHCAYAAABlS0A3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAGdBJREFUeJzt3XtQVNcdB/DvRcj4Qh4qu1bUazWoIAKtSuJowOCiViFY\njUENIWqdOGk7atsE00k1th3dxDiJaWynD8VN2xjNdKrEGFNQIRpbaQO+YpQmsoCG3VGXBUGtAqd/\nqJvdFXYX2Ad6vp+ZO8Pu3r33B/rlnLPncK8ihBAgImkEBboAIvIvhp5IMgw9kWQYeiLJMPREkmHo\niSTD0Hcz27dvx5QpU3x2/NDQUBiNxoCcm7oHht5HVFVF7969ERoaCq1Wi5ycHDQ0NHTpmO2FUlVV\nHDhwwKNjXL16FaqqdqmOrgoKCsL58+cDWoPMGHofURQFe/fuxdWrV3HixAmcOnUKv/71r312LkVR\nfHJsevAw9H6g0WiQnp6Ozz//HACg1+sxcuRI9OvXD3Fxcdi9e3e7733hhRcwZcqUDvUStm3bhtjY\nWERGRmLGjBmorq62vWbfyl65cgWZmZkICwtDcnIyvvrqK4fjHD16FBMmTEB4eDgmTpyIf/7znwCA\nnTt3YsKECQ77vvHGG3jiiScAAB9++CGSkpIQFhaGoUOHYt26de3W+uyzz+IXv/iF7XFxcTGGDBli\ne6yqKjZt2oSEhASEh4cjOzsb//vf/wAAVqsVs2fPRlRUFCIjI5GRkYGLFy96/HOSFUPvQ3dXOF+4\ncAH79+9HcnIyAGDkyJE4cuQIGhoasHbtWjz99NMwm833vHfZsmU4ffo0CgsL0a9fP4/OuWfPHmzY\nsAF///vfcfnyZUyZMgULFixoc98f/vCH6N27N0wmE7Zt24b8/Hxbj8FisWDWrFlYuXIlLBYLfvKT\nn2DWrFmoq6tDRkYGzp07hy+//NJ2rHfffReLFi0CAPTt2xd/+ctfUF9fjw8//BC/+93vsGfPnjZr\ncNdLURQF77//Pj7++GNUVlbi5MmT2L59OwCgtbUVS5cuRXV1Naqrq9GrVy/86Ec/8ujnJDVBPjFs\n2DDRt29fERoaKhRFEVlZWaKlpaXNfRMTE8WePXuEEELk5+eL5ORkMX/+fDFv3jxx69Yt2375+fki\nODhYhIeHO2xBQUHiwIEDQgghZsyYIbZu3Wp7T0tLi+jdu7eorq4WQgihKIr46quvRHNzswgJCRHn\nzp2z7fvzn/9cTJ48WQghxDvvvCOSk5Md6nz00UfF9u3bhRBCPP300+KXv/ylEEKIiooKERoaKq5f\nv97m97dixQqxatUq2+O7NQghxLPPPitefvll22uHDh0S0dHRtseqqoq//vWvtscvvviiWL58eZvn\nKS8vFxEREW2+Rt9gS+8jiqJgz549aGhoQHFxMQ4ePIj//Oc/AIB33nkHSUlJiIiIQEREBE6fPo0r\nV67Y3vvll1/igw8+wJo1axAcHOxw3EceeQR1dXUO29ChQ22vV1VVYcWKFbZj9+/fHwDu6fZeunQJ\nzc3NDl1p++N8/fXXDo8BYNiwYfj6668BAAsXLsSOHTsA3G7l58yZg549ewIAjh07hqlTpyIqKgrh\n4eH4/e9/7/D9dZRWq7V93atXLzQ2NgIArl27hueeew6qqiIsLAwpKSmor6+39bCobQy9Hzz22GP4\n8Y9/jLy8PFRXV2PZsmXYsmULLBYL6urqMHbsWIf/qGPGjMG2bdswc+ZMVFRUdOhcQ4cOxR/+8AeH\nXwpNTU145JFHHPYbOHAggoODHcb79l8PHjwYVVVVDu+pqqrC4MGDAQDTpk3DpUuXcOLECbz33ntY\nuHChbb+FCxciKysLFy5cgNVqxfLly9Ha2tpmvX369MG1a9dsj00mk8ff66ZNm1BRUYHS0lLU19ej\npKQEQgiG3g2G3k9WrlyJ0tJSXLhwAUFBQRgwYABaW1uRn5+P06dP37N/dnY21q9fj2nTpnVoemv5\n8uVYv349zpw5AwCor6/H+++/f89+PXr0wPe//3288soruH79Os6cOQODwWAbX9/9hbNjxw40Nzdj\n586dOHv2LGbPng0ACAkJwZNPPomf/exnqKurg06nsx27sbEREREReOihh1BaWop333233XF7YmIi\n9u3bh7q6OphMJrz55psef6+NjY3o1asXwsLCYLFYXH5gSN9g6P1kwIAByM3NxcaNG/HTn/4Ujz76\nKLRaLU6fPo3Jkyfb9rP/YOuZZ57BmjVr8Pjjj6OqqsqjqbmsrCzk5eUhOzsbYWFhiI+Px8cff9zm\nvm+//TYaGxuh1WqxZMkSLFmyxPZa//79sXfvXmzatAkDBgzA66+/jr179yIyMtK2z8KFC3HgwAE8\n+eSTCAr65r/Sb3/7W6xZswb9+vXDr371Kzz11FP3nPvu95GTk4OEhASoqooZM2YgOzvb7Qd7d19f\nuXIlrl+/jgEDBmDSpEmYOXMmpy49oAj2haTR2tpq69JHR0cHpIaGhgaEh4fDarV6PCNB3sWWXiKn\nTp1Cz549HT4Y87edO3fa1ihQYPg09Pv378fo0aPx8MMP49VXX/Xlqdq0ZMkSaDQaxMfH256zWCzQ\n6XSIiYlBeno6rFar3+qpqanB1KlTERcXh7Fjx+Ktt97yW01/+9vfkJaWhtdee802I3Djxg0kJycj\nMTERsbGxeOmll3xaz6RJk7B582b86U9/crlfS0sLkpKSkJGR4dN6PKGqKsaNG4ekpCRMnDgx4PVY\nrVbMmzcPY8aMQWxsLI4dO9bxenw1F9jc3CxGjBghKisrxc2bN0VCQoI4c+aMr07Xpk8++USUlZWJ\nsWPH2p574YUXxKuvviqEEEKv14u8vDy/1VNbWyvKy8uFEEJcvXpVxMTEiDNnzgS0pqamJiGEELdu\n3RLJycni8OHDAa1HCCE2bdokFi5cKDIyMoQQgf03U1VVXLlyxeG5QNbzzDPP2NZh3Lp1S1it1g7X\n47PQHz16VEyfPt32eMOGDWLDhg2+Ol27KisrHUI/atQoYTKZhBC3Qzhq1Ci/13TXE088IQoLC7tF\nTU1NTWL8+PHi9OnTAa2npqZGpKWliYMHD4rZs2cLIQL7b6aqqrh8+bLDc4Gqx2q1iuHDh9/zfEfr\n8Vn3/uLFiw4LP6Kjo7vFumiz2QyNRgPg9pp45+Wv/mI0GlFeXo7k5OSA1tTa2orExERoNBrb0COQ\n9axatQobN250mA0IZD2KomDatGkYP348/vjHPwa0nsrKSgwcOBCLFy/Gd77zHSxbtgxNTU0drsdn\nob8fpk4C9ddpjY2NmDt3LjZv3ozQ0NCA1hQUFITjx4/jwoUL+OSTT3Do0KGA1bN3715ERUUhKSmp\n3QU2/v75fPrppygvL8dHH32ELVu24PDhwwGrp7m5GWVlZXj++edRVlaGPn36QK/Xd7gen4V+8ODB\nqKmpsT2uqakJ2DSRPY1GY1v1VVtbi6ioKL+e/9atW5g7dy5ycnKQlZXVLWoCgLCwMMyaNQufffZZ\nwOo5evQoCgoKMHz4cCxYsAAHDx5ETk5OQH8+gwYNAnB7BeOcOXNQWloasHqio6MRHR1t+wvHefPm\noaysDFqttkP1+Cz048ePx3//+18YjUbcvHkTO3fuRGZmpq9O57HMzEwYDAYAgMFgsAXPH4QQWLp0\nKWJjY7Fy5cqA13T58mXbJ73Xr19HYWEhkpKSAlbP+vXrUVNTg8rKSrz33nt4/PHH8ec//zlg9Vy7\ndg1Xr14FADQ1NeEf//gH4uPjA1aPVqvFkCFDbEuzi4qKEBcXh4yMjI7V44PPG2z27dsnYmJixIgR\nI8T69et9eao2ZWdni0GDBomQkBARHR0ttm3bJq5cuSLS0tLEww8/LHQ6nairq/NbPYcPHxaKooiE\nhASRmJgoEhMTxUcffRSwmk6ePCmSkpJEQkKCiI+PF6+99poQQgT0Z3RXcXGx7dP7QNVz/vx5kZCQ\nIBISEkRcXJzt/3Agfz7Hjx8X48ePF+PGjRNz5swRVqu1w/VwRR6RZLgij0gyDD2RZLoU+kAvsyWi\njuv0mL6lpQWjRo1CUVERBg8ejAkTJmDHjh0YM2aMt2skIi/qdEtfWlqKkSNHQlVVhISEIDs7u92L\nHxJR9xHsfpe2tbXM9tixYw77DEsdiuqSGue3EpHPDYMQxjZf6XToPVl6WF1SgylrJ6OquArDUofh\n8LoRANTOntKLigGkBrgGe8VgPa4Ug/W4UozbuTLaPVfS7t6d7t57usw25ZUpGJY6DCmvTEH3CDzR\ng0jF7V9Ed7f2dbqlt19m+61vfQs7d+60XRLZ3q+VhwD0wOF1D2Etvrlw4Tqs7eypiagLOh364OBg\nvP3225g+fTpaWlqwdOlSF5/cq509jY+ogS7AiRroApyogS7AiRroApyogS7AidqhvTsdeuD2ZZJn\nzpzpwZ5qV07jA2qgC3CiBroAJ2qgC3CiBroAJ2qgC3CidmjvLoW+o+y79PZdfefXiMh3uAyXSDIM\nPZFkGHoiyfh1TG/PeQzP6Twi/2BLTyQZhp5IMgHr3juz79K/LG46vHZ7VR8ReQNbeiLJMPREkmHo\niSTTbcb09pzH8FyyS+Q9bOmJJMPQE0mGoSeSTLcc0ztzHsPbz+NzDp+oY9jSE0mGoSeSzH3RvXdm\n36XndB5Rx7ClJ5IMQ08kGYaeSDL35Zjenqsr8LT1OpHs2NITSYahJ5IMQ08kmft+TO+MV9klcs1t\nS79kyRJoNBrEx8fbnrNYLNDpdIiJiUF6ejqsVqtPiyQi73Eb+sWLF2P//v0Oz+n1euh0OlRUVCAt\nLQ16vd5nBRKRdylCCOFuJ6PRiIyMDJw6dQoAMHr0aJSUlECj0cBkMiE1NRVnz5699+CKAnSjLjWn\n80ge69BetDv1QZ7ZbIZGowEAaDQamM3mztdGRH7V5Q/yFEW506K3p9juaxXd797eRA8C453NvU6F\n/m63XqvVora2FlFRUS72Tu3MKYioQ1Q4Nqgl7e7ZqdBnZmbCYDAgLy8PBoMBWVlZnTmM33E6j8iD\nMf2CBQswadIknDt3DkOGDEF+fj5Wr16NwsJCxMTE4ODBg1i9erU/aiUiL3Db0u/YsaPN54uKirxe\nDBH53gO3Iq8j7Lv0nM4jWXDtPZFkGHoiyTD0RJKRekxvj1fgIVmwpSeSDENPJBmGnkgyHNO3gzfN\npAcVW3oiyTD0RJJh995D9l16+66+82vucJhAgcaWnkgyDD2RZBh6IslwTN8JzmPxjozTOY6nQGNL\nTyQZhp5IMuzee4GrLntXpveIfIEtPZFkGHoiyTD0RJLhmN7HeigbnJ7hFXgosNjSE0mGoSeSDENP\nJBmO6X2sI1fg4Z/dkj+4belramowdepUxMXFYezYsXjrrbcAABaLBTqdDjExMUhPT4fVavV5sUTU\ndW5DHxISgjfeeAOff/45/vWvf2HLli344osvoNfrodPpUFFRgbS0NOj1en/US0Rd5LZ7r9VqodVq\nAQB9+/bFmDFjcPHiRRQUFKCk5PaN73Nzc5Gamsrge8BbV+Ah6qwOfZBnNBpRXl6O5ORkmM1maDQa\nAIBGo4HZbPZJgUTkXR5/kNfY2Ii5c+di8+bNCA0NdXhNURQoitLOO4vtvlbvbETkXcY7m3sehf7W\nrVuYO3cucnJykJWVBeB2624ymaDValFbW4uoqKh23p3qUSFE1BUqHBvUknb3dBt6IQSWLl2K2NhY\nrFy50vZ8ZmYmDAYD8vLyYDAYbL8MyDVOy1GgKUII4WqHI0eO4LHHHsO4ceNsXfgNGzZg4sSJmD9/\nPqqrq6GqKnbt2oXw8HDHgysKuNbcEUNP/rEO7UXbbUs/efJktLa2tvlaUVFR1+oiIr/jijw/87R1\nX4t1Do+dV/YRdRbX3hNJhqEnkgxDTyQZjum7KVd/nQfwk3/qPLb0RJJh6Ikkw9ATSYZj+vuE8xje\nfh6fc/jUEWzpiSTD0BNJht37+5SrLj2n98gVtvREkmHoiSTD0BNJhqEnkgxDTyQZhp5IMgw9kWQ4\nT/8A4o0xyRW29ESSYeiJJMPuvQTsu/S8yi6xpSeSDENPJBmGnkgyHNNLxnkMzyvwyMdlS3/jxg0k\nJycjMTERsbGxeOmllwAAFosFOp0OMTExSE9Ph9Vq9UuxRNR1LkPfs2dPHDp0CMePH8fJkydx6NAh\nHDlyBHq9HjqdDhUVFUhLS4Ner/dXvUTURW6797179wYA3Lx5Ey0tLYiIiEBBQQFKSm7f9D43Nxep\nqakM/n3KvkvPK+7Iwe0Hea2trUhMTIRGo8HUqVMRFxcHs9kMjUYDANBoNDCbzT4vlIi8w21LHxQU\nhOPHj6O+vh7Tp0/HoUOHHF5XFAWKorg4QrHd1+qdjYi8y3hnc8/jT+/DwsIwa9YsfPbZZ9BoNDCZ\nTNBqtaitrUVUVJSLd6Z6egoi6jQVjg1qSbt7ugz95cuXERwcjPDwcFy/fh2FhYVYu3YtMjMzYTAY\nkJeXB4PBgKysLK+UTYHl6oYaAKf0HhQuQ19bW4vc3Fy0traitbUVOTk5SEtLQ1JSEubPn4+tW7dC\nVVXs2rXLX/USURe5DH18fDzKysrueT4yMhJFRUU+K4qIfIfLcIkkw2W41C4u2X0wsaUnkgxDTyQZ\ndu/JY1yy+2BgS08kGYaeSDIMPZFkOKanTnG1ZJfTed0bW3oiyTD0RJJh9568wr5Lz7/O697Y0hNJ\nhqEnkgxDTyQZjunJ65zH8Fyy272wpSeSDENPJBmGnkgyDD2RZBh6Iskw9ESS4ZQd+ZyrKTp303n2\nr3OqzzvY0hNJhqEnkgxDTyQZjukpoNzeNFPhn+V6m0ctfUtLC5KSkpCRkQEAsFgs0Ol0iImJQXp6\nOqxWq0+LJCLv8Sj0mzdvRmxsLBRFAQDo9XrodDpUVFQgLS0Ner3ep0USkfe4Df2FCxewb98+/OAH\nP4AQAgBQUFCA3NxcAEBubi52797t2yqJyGvcjulXrVqFjRs3oqGhwfac2WyGRqMBAGg0GpjNZt9V\nSA8cV3Pvrv4sl/P03uEy9Hv37kVUVBSSkpJQXFzc5j6Koti6/W2zf596ZyMi7zLe2dxzGfqjR4+i\noKAA+/btw40bN9DQ0ICcnBxoNBqYTCZotVrU1tYiKirKxVFSPa2aiDpNhWODWtLunoq4O1B3o6Sk\nBK+//jo++OADvPjii+jfvz/y8vKg1+thtVrb/DDvdg+AUy7kHbzKbkesQ3vR7tDinLvd+NWrV6Ow\nsBAxMTE4ePAgVq9e3fUaicgvPF6ck5KSgpSUFABAZGQkioqKfFYUEfkOl+ESSYbLcOm+4TyG500z\nO4ctPZFkGHoiybB7T/ct+y49b6jhObb0RJJh6Ikkw9ATSYZjenoguL0CD6f0bNjSE0mGoSeSDENP\nJBmO6emBxCW77WNLTyQZhp5IMuzekxRaxEvfPHB1SUcJsKUnkgxDTyQZhp5IMhzTk/Rku6EGW3oi\nyTD0RJJh956k4Krbbv+av67AE8ghBVt6Iskw9ESSYeiJJMMxPZEdf12BJ5BTgx6FXlVV9OvXDz16\n9EBISAhKS0thsVjw1FNPoaqqCqqqYteuXQgPD/d1vUTURR517xVFQXFxMcrLy1FaWgoA0Ov10Ol0\nqKioQFpaWpu3qiai7sfjMb3zva4LCgqQm5sLAMjNzcXu3bu9WxkR+YQi2rtzvZ1vf/vbCAsLQ48e\nPfDcc89h2bJliIiIQF1dHYDbvxAiIyNtj20HVxRA8quUELXF91fyWXdPQ32XR2P6Tz/9FIMGDcKl\nS5eg0+kwevRoh9cVRbkT8LYU232t3tmIyLuMdzb3PAr9oEGDAAADBw7EnDlzUFpaCo1GA5PJBK1W\ni9raWkRFRbXz7lSPCiGirlDh2KCWtLun2+79tWvX0NLSgtDQUDQ1NSE9PR1r165FUVER+vfvj7y8\nPOj1elit1ns+zGP3nh5k3fuGGl3o3pvNZsyZMwcA0NzcjEWLFiE9PR3jx4/H/PnzsXXrVtuUHRF1\nf25DP3z4cBw/fvye5yMjI1FUVOSToojId7gMl0gyXIZL1En36w012NITSYahJ5IMu/dEXmLfpe/O\n03ls6Ykkw9ATSYahJ5IMx/REPuA8hvfXVXY9wZaeSDIMPZFkGHoiyXBM7wWdvVtJdxrnkW91p39b\ntvREkmHoiSTD7r0XdLbr1p26fBQ4/l6yy5aeSDIMPZFkGHoiyXBMTxRgrq7A09brXcWWnkgyDD2R\nZBh6IslwTE/Uzfj6Krts6Ykkw9ATSYbde6Juzr5L742/zPSopbdarZg3bx7GjBmD2NhYHDt2DBaL\nBTqdDjExMUhPT4fVau3wyYnI/zwK/YoVK/C9730PX3zxBU6ePInRo0dDr9dDp9OhoqICaWlp99ym\nmoi6J7f3p6+vr0dSUhLOnz/v8Pzo0aNRUlICjUYDk8mE1NRUnD171vHgvD89kVd53r3vwv3pKysr\nMXDgQCxevBgnTpzAd7/7Xbz55pswm83QaDQAAI1GA7PZ3LHqiajDnEPemek8t9375uZmlJWV4fnn\nn0dZWRn69OlzT1deUZQ7rXpbiu02o0dFEVHHGAE4Zq19bkMfHR2N6OhoTJgwAQAwb948lJWVQavV\nwmQyAQBqa2sRFRXVzhFS7TbV3emIqBNUAI5Za5/b0Gu1WgwZMgQVFRUAgKKiIsTFxSEjIwMGgwEA\nYDAYkJWV5eIoRnen8TNjoAtwYgx0AU6MgS7AiTHQBTgxBroAm3VYi3XIxTqsxcvipm1zxaN5+t/8\n5jdYtGgRbt68iREjRiA/Px8tLS2YP38+tm7dClVVsWvXLhdHMKJ7tfJGsB5XjGA9rhhxP9fjUegT\nEhLw73//+57ni4qKPD4REXUPXIZLJBvhQykpKQIAN27c/LylpKS0m0u3i3OI6MHC7j2RZBh6Iskw\n9ESSYeiJJMPQE0nm/4eVSh8FLS+GAAAAAElFTkSuQmCC\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f52e803e110>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAP0AAAEHCAYAAABlS0A3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nPZFk/g/JmCSutkLhrgAAAABJRU5ErkJggg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f52e82fc5d0>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAP0AAAEHCAYAAABlS0A3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f52e8179090>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAP0AAAEHCAYAAABlS0A3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAFv9JREFUeJzt3WtQVOf9B/DvQflPxgwoWPasFXS9ZMVFhE1RRydVzGbX\nXAqF4hBtyhBi0mQyzphMp8W0yVinGdyYdBJNfdFk1Nn2hQ15Y9BRErZc6mWUJkBpEix1ZAUVdmJx\ntd5GWZ7/C3W7LLAX2Bs838/MmYFzzp7zy7Zfn+c55+EcRQghQETSSIh1AUQUXQw9kWQYeiLJMPRE\nkmHoiSTD0BNJhqGfpJxOJ1avXo3k5GT88pe/jHU5FEemxroAioyPPvoIGo0G165di3UpFGfY0k9Q\nAwMDfrefP38eixcvjsixaWJj6CcQnU6HnTt3YunSpUhKSsKJEyewatUqpKSkIDc3F01NTQCA559/\nHn/605+wc+dOJCUlob6+HkIIWK1WLFy4EN/73vfw7LPP4sqVKwAAh8OBhIQE7Nu3D3PnzsUTTzwB\nANi3bx8MBgNSU1Px5JNPoru721NLQkIC/vjHP0Kv1yMlJQWbN28eUuvHH38Mg8GA5ORkZGVlobW1\nFQBw6dIllJSUQKPRYP78+fjwww+j8dWRN0ETxty5c4XRaBQXLlwQFy9eFDNnzhRHjx4VQghRV1cn\nZs6cKS5fviyEEOL5558Xb731luezH3zwgVi5cqW4ePGiuHPnjnj55ZfFxo0bhRBCdHV1CUVRRHl5\nubh586a4deuWOHjwoFi4cKE4c+aMcLvd4u233xarVq3yHE9RFFFQUCCuXr0quru7RVpamqitrRVC\nCFFdXS1mz54tvvzySyGEEGfPnhXnz58XbrdbPProo+J3v/uduHv3rjh37pyYP3+++Pzzz6Py/dE9\nDP0EotPpxP79+4UQQlitVlFWVjZk+7p164TNZhNC3Av9m2++6dm2ePFi8de//tXz+6VLl0RiYqJw\nu92e0Hd1dXm2P/nkk2Lv3r2e391ut5g2bZro7u4WQtwL/YkTJzzbS0tLxTvvvCOEEMJisYjdu3cP\nq//UqVNizpw5Q9ZVVVWJioqKUL4GGideyJtgMjIyANwbs3/66ac4dOiQZ9vAwAAef/zxET/ncDhQ\nXFyMhIT/jeimTp0Kp9M57NgPjr9lyxb84he/GHKcixcvevbTarWe9dOmTcP169cBABcuXMCCBQuG\n1XD+/HlcunQJKSkpnnVutxurV68O/B9OYcPQTzCKogAA5syZg7KyMnz00UdBfW7OnDnYv38/Vq5c\nOWybw+EYcuwH+7/11lvYuHFjyDVmZGTg7NmzI9Ywb948dHZ2hnxMCh9eyJugfvazn+HQoUP44osv\n4Ha7cfv2bTQ2NuLixYsAAOHzF9OvvPIKfv3rX3suxn333XeoqakZ9fivvPIKqqqq8O233wIArl69\nik8//XTU/cW9oSIA4MUXX8R7772HlpYWCCFw9uxZdHd3Y/ny5UhKSsLOnTtx69YtuN1ufP311/jy\nyy/H9V1QaBj6CSo9PR2fffYZqqqqoNFoMGfOHPz+97/3BE9RlCEt95YtW1BYWAiLxYLk5GSsXLkS\nzc3Nnu3e+wJAUVERKisrsWHDBkyfPh3Z2dn4/PPPR93f+3zr16/Hb37zG/z0pz9FcnIyfvKTn+DK\nlStISEjA4cOH0dbWhvnz5yMtLQ0///nPOZcgyhTh2yQQ0aTGlp5IMhENfW1tLTIzM/HII4/gnXfe\nieSpRvTCCy9AVVVkZ2d71vX398NsNkOv18NiscDlckWtnp6eHqxduxZZWVlYsmQJdu/eHdOabt++\njRUrViA3NxcGgwFvvPFGTOt5wO12w2g0oqCgIOb16HQ6LF26FEajEcuXL495PS6XC+vXr8fixYth\nMBhw+vTpkOuJWOjdbjc2b96M2tpafPvttzhw4AA6OjoidboRVVRUoLa2dsg6q9UKs9mMzs5OmEwm\nWK3WqNWTmJiI999/H9988w1OnTqFPXv2oKOjI2Y1PfTQQ2hoaEBbWxva29vR0NCA48ePx/Q7AoBd\nu3bBYDB4rhHEsh5FUdDY2IjW1lbPNZBY1rNlyxY8/fTT6OjoQHt7OzIzM0OvJ1ITAE6ePCnWrVvn\n+X3Hjh1ix44dkTrdqLq6usSSJUs8vy9atEj09fUJIYTo7e0VixYtinpND/z4xz8WdXV1cVHTjRs3\nRF5envj6669jWk9PT48wmUyivr5e/OhHPxJCxPZ/M51O55nl+ECs6nG5XGLevHnD1odaT8Raeu9J\nHMC9q80PbifFktPphKqqAABVVYdMTokmh8OB1tZWrFixIqY1DQ4OIjc3F6qqeoYesazn9ddfx7vv\nvjtkElEs61EUBU888QTy8vLw8ccfx7Serq4upKWloaKiAo8++iheeukl3LhxI+R6IhZ631s68cj3\ntla0XL9+HSUlJdi1axeSkpJiWlNCQgLa2tpw4cIF/O1vf0NDQ0PM6jl8+DA0Gg2MRuOweQaxqAcA\nTpw4gdbWVhw9ehR79uzBsWPHYlbPwMAAWlpa8Oqrr6KlpQUPP/zwsK58MPVELPSzZ89GT0+P5/ee\nnh6kp6dH6nRBU1UVfX19AIDe3l5oNJqonv/u3bsoKSlBWVkZioqK4qImAJg+fTqeeeYZfPXVVzGr\n5+TJk6ipqcG8efOwceNG1NfXo6ysLKbfz6xZswAAaWlpKC4uRnNzc8zqSU9PR3p6OpYtWwbg3nyI\nlpYWaLXakOqJWOjz8vLw73//Gw6HA3fu3MEnn3yCwsLCSJ0uaIWFhbDZbAAAm83mCV40CCGwadMm\nGAwGvPbaazGv6fLly54rvbdu3UJdXR2MRmPM6qmqqkJPTw+6urrwl7/8BY8//jj+/Oc/x6yemzdv\n4r///S8A4MaNG/jiiy+QnZ0ds3q0Wi0yMjI805jtdjuysrJQUFAQWj0RuN7gceTIEaHX68WCBQtE\nVVVVJE81og0bNohZs2aJxMREkZ6eLvbt2yf+85//CJPJJB555BFhNpvFlStXolbPsWPHhKIoIicn\nR+Tm5orc3Fxx9OjRmNXU3t4ujEajyMnJEdnZ2WLnzp1CCBHT7+iBxsZGUVBQENN6zp07J3JyckRO\nTo7Iysry/H84lt9PW1ubyMvLE0uXLhXFxcXC5XKFXA9n5BFJhjPyiCTD0BNJZlyhj/U0WyIK3ZjH\n9G63G4sWLYLdbsfs2bOxbNkyHDhwYMxPYCWi6BhzS9/c3IyFCxdCp9MhMTERGzZswGeffRbO2ogo\nAsb8uKyRptmePn16yD5z8+egu6nH96NEFHFzIYRjxC1jDn0wUw+7m3rww22P4XzjeczNn4tj2xcA\n0I31lGHUCCA/xjV4awTr8acRrMefRtzLlcNrXdOoe4+5ex/sNNs1v/0h5ubPxZrf/hDxEXiiyUiH\ne/8QPVhGN+aW3nua7fe//3188sknOHDgwLD93lb+D8AUHNv+f9iG7Z7127FtrKcmonEYc+inTp2K\nP/zhD1i3bh3cbjc2bdrk58q9bqyniRBdrAvwoYt1AT50sS7Ahy7WBfjQxboAH7qQ9h7Xc++feuop\nPPXUU0HsqRvPaSJAF+sCfOhiXYAPXawL8KGLdQE+dLEuwIcupL2j+rIL7y69d1ffdxsRRQ6n4RJJ\nhqEnkgxDTySZmL3A0ncMz9t5RNHBlp5IMgw9kWTi5v303l36N8WdIdvuzeojonBgS08kGYaeSDIM\nPZFk4mZM7813DM8pu0Thw5aeSDIMPZFkGHoiycTlmN6X7xje+z4+7+EThYYtPZFkGHoiyUyI7r0v\n7y49b+cRhYYtPZFkGHoiyTD0RJKZkGN6b/6ewDPSdiLZsaUnkgxDTyQZhp5IMhN+TO+LT9kl8i9g\nS//CCy9AVVVkZ2d71vX398NsNkOv18NiscDlckW0SCIKn4Chr6ioQG1t7ZB1VqsVZrMZnZ2dMJlM\nsFqtESuQiMJLEUKIQDs5HA4UFBTgn//8JwAgMzMTTU1NUFUVfX19yM/Px5kzZ4YfXFGAOOpS83Ye\nyWM7Rov2mC7kOZ1OqKoKAFBVFU6nc+y1EVFUjftCnqIo91v00TR6/axD/L3bm2gycNxfAhtT6B90\n67VaLXp7e6HRaPzsnT+WUxBRSHQY2qA2jbrnmEJfWFgIm82GyspK2Gw2FBUVjeUwUcfbeURBjOk3\nbtyIVatW4V//+hcyMjKwf/9+bN26FXV1ddDr9aivr8fWrVujUSsRhUHAlv7AgQMjrrfb7WEvhogi\nb9LNyAuFd5eet/NIFpx7TyQZhp5IMgw9kWSkHtN74xN4SBZs6Ykkw9ATSYahJ5IMx/Sj4EszabJi\nS08kGYaeSDLs3geJL82kyYItPZFkGHoiyTD0RJLhmH4MeDuPJjK29ESSYeiJJMPufRh4d+m9u/q+\n24jiAVt6Iskw9ESSYeiJJMMxfZj5juE5ZZfiDVt6Iskw9ESSYeiJJMMxfYTxpZkUbwK29D09PVi7\ndi2ysrKwZMkS7N69GwDQ398Ps9kMvV4Pi8UCl8sV8WKJaPwChj4xMRHvv/8+vvnmG5w6dQp79uxB\nR0cHrFYrzGYzOjs7YTKZYLVao1EvEY2TIoQQoXygqKgImzdvxubNm9HU1ARVVdHX14f8/HycOXNm\n6MEVBWAXdlScskuRsx2jRTukC3kOhwOtra1YsWIFnE4nVFUFAKiqCqfTOf46iSjigr6Qd/36dZSU\nlGDXrl1ISkoask1RlPut+kgavX7W3V+IKLwc95fAggr93bt3UVJSgrKyMhQVFQGAp1uv1WrR29sL\njUYzyqfzgyqEiMZDh6ENatOoewYMvRACmzZtgsFgwGuvveZZX1hYCJvNhsrKSthsNs8/BhQ83zE8\nn8BD0RDwQt7x48exevVqLF261NOF37FjB5YvX47S0lJ0d3dDp9OhuroaM2bMGHpwXsgLCUNP4TP6\nhbyALf1jjz2GwcHBEbfZ7fbx1UVEUccZeXGEL9SgaODceyLJMPREkmHoiSTDMX2c8vdCDYBX92ns\n2NITSYahJ5IMQ08kGY7pJwh/T9nlPXwKBVt6Iskw9ESSYfd+gvLu0vN2HoWCLT2RZBh6Iskw9ESS\n4Zh+EvD3BJ6RtpPc2NITSYahJ5IMQ08kGY7pJyE+ZZf8YUtPJBmGnkgy7N5LgE/ZJW9s6Ykkw9AT\nSYahJ5IMx/SS8R3D8wk88vHb0t++fRsrVqxAbm4uDAYD3njjDQBAf38/zGYz9Ho9LBYLXC5XVIol\novHzG/qHHnoIDQ0NaGtrQ3t7OxoaGnD8+HFYrVaYzWZ0dnbCZDLBarVGq14iGqeA3ftp06YBAO7c\nuQO3242UlBTU1NSgqeneS+/Ly8uRn5/P4E9QfAKPfAJeyBscHERubi5UVcXatWuRlZUFp9MJVVUB\nAKqqwul0RrxQIgqPgC19QkIC2tracPXqVaxbtw4NDQ1DtiuKAkVR/Byh0etn3f2FiMLLcX8JLOir\n99OnT8czzzyDr776Cqqqoq+vD1qtFr29vdBoNH4+mR/sKYhozHQY2qA2jbqn39BfvnwZU6dOxYwZ\nM3Dr1i3U1dVh27ZtKCwshM1mQ2VlJWw2G4qKisJSNsWWvxdqALylN1n4DX1vby/Ky8sxODiIwcFB\nlJWVwWQywWg0orS0FHv37oVOp0N1dXW06iWicfIb+uzsbLS0tAxbn5qaCrvdHrGiiChyOA2XSDKc\nhkuj4pTdyYktPZFkGHoiybB7T0HjlN3JgS09kWQYeiLJMPREkuGYnsbE35Rd3s6Lb2zpiSTD0BNJ\nht17CgvvLj3/Oi++saUnkgxDTyQZhp5IMhzTU9j5juE5ZTe+sKUnkgxDTyQZhp5IMhzTU8T5juG9\nx/gc30cfW3oiyTD0RJJh956izrtLzym70ceWnkgyDD2RZBh6IslwTE8x5e+FGiNtp/ELqqV3u90w\nGo0oKCgAAPT398NsNkOv18NiscDlckW0SCIKn6BCv2vXLhgMBiiKAgCwWq0wm83o7OyEyWSC1WqN\naJFEFD4BQ3/hwgUcOXIEL774IoQQAICamhqUl5cDAMrLy3Hw4MHIVklEYRNwTP/666/j3XffxbVr\n1zzrnE4nVFUFAKiqCqfTGbkKSSp8aWbk+Q394cOHodFoYDQa0djYOOI+iqJ4uv0j8/6c7v5CROHl\nuL8E5jf0J0+eRE1NDY4cOYLbt2/j2rVrKCsrg6qq6Ovrg1arRW9vLzQajZ+j5AdbNRGNmQ5DG9Sm\nUfdUxIOBegBNTU147733cOjQIfzqV7/CzJkzUVlZCavVCpfLNeLFvHs9AHbJKDx4Oy8U2zFatEOa\nnPOgG79161bU1dVBr9ejvr4eW7duHX+NRBQVQU/OWbNmDdasWQMASE1Nhd1uj1hRRBQ5nIZLJBlO\nw6UJg7fzwoMtPZFkGHoiybB7TxOWd5eeL9QIHlt6Iskw9ESSYeiJJMMxPU0KvmN4TtkdHVt6Iskw\n9ESSYeiJJMMxPU1KnLI7Orb0RJJh6Ikkw+49SYFTdv+HLT2RZBh6Iskw9ESS4ZiepONvyq4Mt/PY\n0hNJhqEnkgy79yQ97y69DH+dx5aeSDIMPZFkGHoiyXBMT+TF31/njbR9Igoq9DqdDsnJyZgyZQoS\nExPR3NyM/v5+PPvsszh//jx0Oh2qq6sxY8aMSNdLROMUVPdeURQ0NjaitbUVzc3NAACr1Qqz2YzO\nzk6YTKYRX1VNRPEn6DG977uua2pqUF5eDgAoLy/HwYMHw1sZEUWEIkZ7c72X+fPnY/r06ZgyZQpe\nfvllvPTSS0hJScGVK1cA3PsHITU11fO75+CKAkyCMRDRAxNnyu72YQ31A0GN6U+cOIFZs2bhu+++\ng9lsRmZm5pDtiqLcD/hIGr1+1t1fiCi8HPeXwIIK/axZswAAaWlpKC4uRnNzM1RVRV9fH7RaLXp7\ne6HRaEb5dH5QhRDReOgwtEFtGnXPgN37mzdvwu12IykpCTdu3IDFYsG2bdtgt9sxc+ZMVFZWwmq1\nwuVyDbuYx+49TWbxfTtvHN17p9OJ4uJiAMDAwACee+45WCwW5OXlobS0FHv37vXcsiOi+Bcw9PPm\nzUNbW9uw9ampqbDb7REpiogih9NwiSTDabhEYzRRX6jBlp5IMgw9kWTYvScKk4nyBB629ESSYeiJ\nJMPQE0mGY3qiCPAdw8fTSzPZ0hNJhqEnkgxDTyQZjumJoiCeXprJlp5IMgw9kWTYvSeKgVhO2WVL\nTyQZhp5IMgw9kWQ4pieKsWi/NJMtPZFkGHoiyTD0RJLhmJ4ozkT6Kbts6Ykkw9ATSYbde6I4592l\nD8cTeIJq6V0uF9avX4/FixfDYDDg9OnT6O/vh9lshl6vh8VigcvlCvnkRBR9QYV+y5YtePrpp9HR\n0YH29nZkZmbCarXCbDajs7MTJpNp2GuqiSg+BXw//dWrV2E0GnHu3Lkh6zMzM9HU1ARVVdHX14f8\n/HycOXNm6MH5fnqisAq+ez+O99N3dXUhLS0NFRUV+Mc//oEf/OAH+OCDD+B0OqGqKgBAVVU4nc7Q\nqieikIXjCTwBu/cDAwNoaWnBq6++ipaWFjz88MPDuvKKotxv1UfS6LU4giqKiELjADA0a6MLGPr0\n9HSkp6dj2bJlAID169ejpaUFWq0WfX19AIDe3l5oNJpRjpDvtegCnY6IxkAHYGjWRhcw9FqtFhkZ\nGejs7AQA2O12ZGVloaCgADabDQBgs9lQVFTk5yiOQKeJMkesC/DhiHUBPhyxLsCHI9YF+HDEugCP\n7diG7SjHdmzDm+KOZ/EnqPv0H374IZ577jncuXMHCxYswP79++F2u1FaWoq9e/dCp9OhurrazxEc\niK9W3gHW448DrMcfByZyPUGFPicnB3//+9+Hrbfb7UGfiIjiA6fhEslGRNCaNWsEAC5cuER5WbNm\nzai5DDg5h4gmF3bviSTD0BNJhqEnkgxDTyQZhp5IMv8PM8GsNGsOGEgAAAAASUVORK5CYII=\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f52e808f190>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAP0AAAEHCAYAAABlS0A3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nwvWUKN05uQAAAABJRU5ErkJggg==\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f52e81e9d10>"
       ]
      }
     ],
     "prompt_number": 14
    }
   ],
   "metadata": {}
  }
 ]
}