main.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Video generation using GANs with LSTMs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from keras.datasets import mnist\n",
    "# import pandas as pd\n",
    "# import matplotlib.pyplot as plt\n",
    "# import numpy as np\n",
    "# import cv2\n",
    "import random\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading data from https://s3.amazonaws.com/img-datasets/mnist.npz\n",
      "10985472/11490434 [===========================>..] - ETA: 0s"
     ]
    }
   ],
   "source": [
    "A = mnist.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = A[0]\n",
    "Y = A[1]\n",
    "\n",
    "x_train = X[0]\n",
    "y_train = X[1]\n",
    "\n",
    "x_test = Y[0]\n",
    "y_test = Y[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7fe7a9e40320>"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRfiBoAg/EBThB4Ii/EBQhB8I6v8AG8x2aarNGp8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a9e16940>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(x_train[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "def insert_num_background(location, img, background_dim):\n",
    "\n",
    "    background = np.zeros([background_dim[0],background_dim[1]])\n",
    "    x = location[0]\n",
    "    y = location[1]\n",
    "    \n",
    "    background[x:x+img.shape[0], y:y+img.shape[1]] = img\n",
    "    return background"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.5/dist-packages/matplotlib/pyplot.py:523: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`).\n",
      "  max_open_warning, RuntimeWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0a4f710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nb2dwjQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0526080>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0521d68>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJfR/iWXFdQDGq78AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0f4e9b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nH8B+wEHrWM1VAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0f3ae48>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0ebe128>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a067db00>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nuyTGSD7VAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0a74438>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n234zFAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0a71f60>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n3ywhB98sof8H0SfGRx8Y64AAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0a26828>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n8M0S+n/dnMjPQ+wexQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0640a58>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a06542b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0f0d4e0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0d80320>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQg6+WUL/Dz0zxkex8JwoAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0b25f98>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n3ywhB98sof8H0SfGRx8Y64AAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0b3e588>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0adba58>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAADzhJREFUeJzt3X2QVfV9x/H3x12eFQETCYIpGEFK\nbRWzUalJbCWmaDJiZozRcRLG0uGPJIoPraLtH20609GO40MaTMtoDEmsmqKjjGO1iqTTJIouaovy\nIAQfWEQxBjUhirvw7R/n7Lo4wF527929d7+f18zOnoffuec7v9nP/s45+9t7FRGYWS6HDHQBZtb/\nHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOE+hR8SXMkbZC0SdKiahVlZrWl3k7gkdQEvAicCbQBTwMX\nRsTa6pVnZrXQ3IdjTwY2RcRmAEl3A3OB/QZ/qIbFcEb14ZRmdiDvs5MPYpd6ateX4E8EtnRbbwNO\n+WgjSQuABQDDGckpmt2HU5rZgayKFRW1q/nDvYhYEhEtEdEyhGG1Pp2ZVaAvwd8KHN1tfVK5zczq\nXF+C/zQwVdIUSUOBC4Dl1SnLzGqp1/f4EdEh6dvAI0AT8IOIeKFqlZlZzfTl4R4R8RDwUJVqMbN+\n4pl7Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJ\nOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5\n+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgn1GHxJR0taKWmtpBckLSy3j5P0qKSN5fextS/XzKqhkhG/\nA7gyImYApwLfkjQDWASsiIipwIpy3cwaQI/Bj4htEfFMufxbYB0wEZgLLC2bLQXOrVWRZlZdB3WP\nL2kyMBNYBYyPiG3lrteB8VWtzMxqpuLgSzoUuBe4LCLe7b4vIgKI/Ry3QFKrpNZ2dvWpWDOrjoqC\nL2kIRejvjIj7ys1vSJpQ7p8AbN/XsRGxJCJaIqJlCMOqUbOZ9VElT/UF3A6si4gbu+1aDswrl+cB\nD1S/PDOrheYK2pwGfB1YI+m5ctu1wHXATyXNB14Bzq9NiWZWbT0GPyJ+Dmg/u2dXtxwz6w+euWeW\nkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQ\ng2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCD\nb5aQg2+WkINvlpCDb5aQg2+WkINvllDFwZfUJOlZSQ+W61MkrZK0SdI9kobWrkwzq6aDGfEXAuu6\nrV8P3BQRxwI7gPnVLMzMaqei4EuaBHwJuK1cF3AGsKxsshQ4txYFmln1VTri3wxcBewp148A3o6I\njnK9DZi4rwMlLZDUKqm1nV19KtbMqqPH4Ev6MrA9Ilb35gQRsSQiWiKiZQjDevMSZlZlzRW0OQ04\nR9LZwHBgNHALMEZScznqTwK21q5MM6umHkf8iLgmIiZFxGTgAuDxiLgIWAmcVzabBzxQsyrNrKr6\n8nf8q4ErJG2iuOe/vTolmVmtVXKp3yUifgb8rFzeDJxc/ZLMrNY8c88sIQffLCEH3ywhB98sIQff\nLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98s\nIQffLCEH3ywhB98sIQffLCEH3yyhg/pADbN60HHGpwHY9s3i05f/d9bSrn0nPDEPgKMWDwWgaeUz\n/VxdY/CIb5aQR3xrGHtOnwnAd3/wPQCOHVL8+O7p1ubZWXcAsKFlNwB/M/nU/iuwgXjEN0vII77V\nvfYvtgBw1a0/BmDakOL+fU851m9ub+9q+86eYQDMLL6x66zPADBi5ZrimPffr33BDcAjvllCHvGt\nrjSNHt21vPPz0wG4/KZ/B+DPR/yu3LP3ePXDHX/atbzi1lkA/OLvvwvAo7f9KwAzfvJtAI65+onq\nF92APOKbJeQR3+pK248mdi0//ZnFFR3znSOf7lp++NBi9L/45S8CsHTyYwCMnvFWtUocFDzimyVU\nUfAljZG0TNJ6SeskzZI0TtKjkjaW38fWulgzq45KL/VvAR6OiPMkDQVGAtcCKyLiOkmLgEXA1TWq\n0wa5zmm4d534va5thzB0rzYXvzIbgNbH/hCANfOLtivfG97V5sjW9wDYtKN4MDjkn1YWr6VaVN24\nehzxJR0OfB64HSAiPoiIt4G5QOck6aXAubUq0syqSxFx4AbSicASYC1wArAaWAhsjYgxZRsBOzrX\n92e0xsUpml2Num2Q6JyGe/PSW4EPp+F2d876rwDQdN5OAH7zpeMAeOv4YhiftnhLV9uOLW17Hfvg\n1tUAbNtdXAn85bxLi9capP+8sypW8G78psfrm0ru8ZuBk4DvR8RMYCfFZX2XKH577PM3iKQFklol\ntbazq4LTmVmtVXKP3wa0RcSqcn0ZRfDfkDQhIrZJmgBs39fBEbGE4oqB0Rp34MsLS0Of/iMAfn1F\nMRJ3TsNd3W1sePx3MwB46+6jAThiRzH55vCfPFl8L9t1VHC+8U3FHN63Lvs9AEeu7HXpg0KPI35E\nvA5skXRcuWk2xWX/cmBeuW0e8EBNKjSzqqv0qf4lwJ3lE/3NwMUUvzR+Kmk+8Apwfm1KtMHkkJEj\nAej453cBeHL6fQC81PEBAFdce2VX27H/8yoAR44qLiZ3V+H8J094BYCXq/Bajayi4EfEc0DLPnb5\nSZ1ZA/KUXetX751e3Ns/Mv3Wvbb/1cLLATjs/ie7tlVy72694ym7Zgk5+GYJ+VLf+tWf/ONzABxS\njjmd03BH3P9UTc43RE0AtJd/SG6S/6IMHvHNUvKIbzX39tdndS3/3fgbANhT/gPO6v8qJul8kl/W\n5Nztsbs8X/H+fA+vK843lcE5ZbdSHvHNEvKIbzXXMeLD5cMPKUb6J94vptAe86PXijZVOlfnBKH1\nNxxfbin+SeeizWcBMH3hS0B1JgM1Mo/4Zgl5xLcB8dbuQwHo2Pxyn1+rc5QH2HDdHwOwfm7xJh3/\n+fviX3leW3wsAIfteBLziG+Wkkd8GxB//YuvAjCtvAfvjc438dhe/msvwLqWYqSfveZrAIyasxmA\nw/BI351HfLOEPOJb7XV7I6jOGXu3fPYuABYz7aBf7pXvFPMC7v3GjcCHb+IBcNJTxVtEHPWVtb0q\nNQuP+GYJOfhmCflS32qv2//FdE6dPX1E8ZFWl/2weD/9T91RbB/y+m+72r5x+scBGPe14p1zL/nk\nCgDOGlk8EFy+czwA31gzp+uYj/3bqKqXPxh5xDdLyCO+DYjhKn701p1ZfIz1zz9XfBrOxl2f6Gpz\n8eEv7/PYha99DoCHf3kiAFMX+k91B8sjvllCPX6STjX5k3Ryapr2qa7laXcV73J7/See2KtN55/5\nOp8BdPfsrmLfhf+9oHiNi3s/6Wewq+Yn6ZjZION7fKu53S/+qmt541cnAzDjkksAWHv+v+z3uOkP\nfROA424tPv1m2rMe6avFI75ZQr7HNxtEfI9vZvvl4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl\n5OCbJeTgmyVUUfAlXS7pBUnPS7pL0nBJUyStkrRJ0j2Shvb8SmZWD3oMvqSJwKVAS0QcDzQBFwDX\nAzdFxLHADmB+LQs1s+qp9FK/GRghqRkYCWwDzgCWlfuXAudWvzwzq4Uegx8RW4EbgFcpAv8OxWcP\nvx0RnZ9u3AZM3NfxkhZIapXU2s6u6lRtZn1SyaX+WGAuMAU4ChgFzDngQd1ExJKIaImIliEM63Wh\nZlY9lVzqfwF4KSLejIh24D7gNGBMeekPMAnYWqMazazKKgn+q8CpkkZKEjAbWAusBM4r28wDHqhN\niWZWbZXc46+ieIj3DLCmPGYJcDVwhaRNwBHA7TWs08yqyG+9ZTaI+K23zGy/HHyzhBx8s4QcfLOE\nHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4Qc\nfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8\ns4QcfLOEHHyzhBx8s4QcfLOEFBH9dzLpTWAn8Ot+O2nffIzGqRUaq95GqhUap94/iIiP99SoX4MP\nIKk1Ilr69aS91Ei1QmPV20i1QuPV2xNf6psl5OCbJTQQwV8yAOfsrUaqFRqr3kaqFRqv3gPq93t8\nMxt4vtQ3S6jfgi9pjqQNkjZJWtRf562UpKMlrZS0VtILkhaW28dJelTSxvL72IGutZOkJknPSnqw\nXJ8iaVXZx/dIGjrQNXaSNEbSMknrJa2TNKte+1bS5eXPwPOS7pI0vJ77tjf6JfiSmoDFwFnADOBC\nSTP649wHoQO4MiJmAKcC3yprXASsiIipwIpyvV4sBNZ1W78euCkijgV2APMHpKp9uwV4OCKmAydQ\n1F13fStpInAp0BIRxwNNwAXUd98evIio+RcwC3ik2/o1wDX9ce4+1PwAcCawAZhQbpsAbBjo2spa\nJlGE5QzgQUAUE0ya99XnA1zr4cBLlM+Uum2vu74FJgJbgHFAc9m3f1Gvfdvbr/661O/szE5t5ba6\nJGkyMBNYBYyPiG3lrteB8QNU1kfdDFwF7CnXjwDejoiOcr2e+ngK8CZwR3lrcpukUdRh30bEVuAG\n4FVgG/AOsJr67dte8cO9j5B0KHAvcFlEvNt9XxS/7gf8zyCSvgxsj4jVA11LhZqBk4DvR8RMimnb\ne13W11HfjgXmUvyyOgoYBcwZ0KJqoL+CvxU4utv6pHJbXZE0hCL0d0bEfeXmNyRNKPdPALYPVH3d\nnAacI+ll4G6Ky/1bgDGSmss29dTHbUBbRKwq15dR/CKox779AvBSRLwZEe3AfRT9Xa992yv9Ffyn\nganlk9GhFA9LlvfTuSsiScDtwLqIuLHbruXAvHJ5HsW9/4CKiGsiYlJETKboy8cj4iJgJXBe2awu\nagWIiNeBLZKOKzfNBtZSh31LcYl/qqSR5c9EZ6112be91o8PTc4GXgR+BfztQD/c2Ed9n6W41Pw/\n4Lny62yKe+cVwEbgMWDcQNf6kbr/DHiwXD4GeArYBPwHMGyg6+tW54lAa9m/9wNj67VvgX8A1gPP\nAz8GhtVz3/bmyzP3zBLywz2zhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4T+H6a+u02VJ1Xp\nAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a09fcd68>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAADzZJREFUeJzt3X+QVeV9x/H3h11+KwomEgRTMIKU\n2ipmo1KT2EpM0WTEzBij4ySMpcMfSRR/tIq2f7TpTEc7jkoaTMtoDEmsmqKjjGO1iqTTJIouaovy\nQwj+YBHEGNSEKLLw7R/n7LKYhb3s3nv3br6f18zOveec59zznWf2s885Z597ryICM8tlUH8XYGb1\n5+CbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl1KfgS5olab2kjZIWVKsoM6st9XYCj6Qm4CXgbKANeAa4\nOCLWVK88M6uF5j7seyqwMSI2AUi6B5gNHDD4QzQ0hjGyD4c0s4N5n518ELvUU7u+BH88sLnLchtw\n2ocbSZoHzAMYxghO08w+HNLMDmZlLK+oXc1v7kXE4ohoiYiWwQyt9eHMrAJ9Cf4W4NguyxPKdWbW\n4PoS/GeAyZImSRoCXAQsq05ZZlZLvb7Gj4h2Sd8EHgWagO9FxItVq8zMaqYvN/eIiIeBh6tUi5nV\niWfumSXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl\n5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk\n4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXUY/AlHStphaQ1kl6UNL9cP0bSY5I2lI+ja1+umVVDJSN+\nO3B1REwDTge+IWkasABYHhGTgeXlspkNAD0GPyK2RsSz5fNfA2uB8cBsYEnZbAlwfq2KNLPqOqRr\nfEkTgenASmBsRGwtN20Dxla1MjOrmYqDL+kw4D7gioh4t+u2iAggDrDfPEmtklp3s6tPxZpZdVQU\nfEmDKUJ/V0TcX65+Q9K4cvs4YHt3+0bE4ohoiYiWwQytRs1m1keV3NUXcAewNiJu7rJpGTCnfD4H\neLD65ZlZLTRX0OYM4KvAaknPl+uuB24AfixpLvAqcGFtSjSzausx+BHxU0AH2DyzuuWYWT145p5Z\nQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllC\nDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIO\nvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIVB19Sk6TnJD1ULk+StFLSRkn3ShpSuzLNrJoOZcSfD6zt\nsnwjcEtEHA/sAOZWszAzq52Kgi9pAvAF4PZyWcBZwNKyyRLg/FoUaGbVV+mIfytwDbC3XD4KeDsi\n2svlNmB8dztKmiepVVLrbnb1qVgzq44egy/pi8D2iFjVmwNExOKIaImIlsEM7c1LmFmVNVfQ5gzg\nPEnnAsOAUcBC4EhJzeWoPwHYUrsyzayaehzxI+K6iJgQEROBi4AnIuISYAVwQdlsDvBgzao0s6rq\ny//xrwWukrSR4pr/juqUZGa1VsmpfqeI+Anwk/L5JuDU6pdkZrXmmXtmCTn4Zgk5+GYJOfhmCTn4\nZgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhm\nCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJHdLXZJsdivazPgnA\n1q/v6lz3vzOWAHDSk3MAOGbREACaVjxb5+py84hvlpBHfKu6vWdOB+Db3/sOAMcP3vdrtrd8fG7G\nnQCsb9kDwN9MPL1+BZpHfLOMKgq+pCMlLZW0TtJaSTMkjZH0mKQN5ePoWhdrZtVR6an+QuCRiLhA\n0hBgBHA9sDwibpC0AFgAXFujOm0A2P35FgCuue2HAEwZXNy429t5gg+bdu8G4J29QwGYXjyw65xP\nATB8xerOtnvff7+2BSfW44gv6Qjgs8AdABHxQUS8DcwGlpTNlgDn16pIM6uuSkb8ScCbwJ2STgJW\nAfOBsRGxtWyzDRhbmxKtUTWNGgXAzs9OBeDKW/4dgD8f/puyxe+OK9/f8acALL9tBgA/+/tvA/DY\n7f8KwLQffbOz7XHXPln9og2o7Bq/GTgF+G5ETAd2UpzWd4qIAKK7nSXNk9QqqXU3u7prYmZ1VsmI\n3wa0RcTKcnkpRfDfkDQuIrZKGgds727niFgMLAYYpTHd/nGwgantB+MBeOZTiyre51tHPwPAI4cV\nI/+lr3wegCUTHwdg1LS3qlmiHUCPI35EbAM2SzqhXDUTWAMsA+aU6+YAD9akQjOrukrv6l8G3FXe\n0d8EXErxR+PHkuYCrwIX1qZEayQd03AB7j65mKAziCH7tbn01ZkAtD7+hwCsnvudzm0r3hsGwNGt\n7wGwcUdxf2DwP60oXku1qNo+rKLgR8TzQEs3m2ZWtxwzqwcV9+XqY5TGxGny34qBqGMa7q1Lbutc\n13UqLsB5674EQNMFOwH41ReKq8O3Ttw3jE9ZtBmA9s1t++370JZVAGzd817nur+cc3nxen4DT8VW\nxnLejV/1eN7kKbtmCTn4Zgn53Xl2UPrkHwHwy6uKU/COabgAq8ppGU/8ZhoAb91zLABH7Sgm3hzx\no6eKxy6v197D8cY2De18/tYVvwXg6BW9Kt0OwiO+WUIe8a1bg0aMAKD9n98F4Kmp9wPwcvsHnW2u\nuv5qAEb/z2sAHD2ymMO1p0o1nDruVQBeqdLr2T4e8c0S8ohv3XrvzOLa/tGpt+23/q/mX9n5/PAH\nimv4nq7brfF4xDdLyCO+detP/vF5AAaVY0PHNNzhDzxdk+MNVhMAu7vMJ2uS39NVKx7xzRLyiG/7\nefurxQdk/N3YmwDYW74BZ9V/Ff+r/zg/r8lxd8ee8nj7PqbrkbXFMSfjKbvV5hHfLCGP+Laf9uHF\n4xGDipH+yfeLmXTH/eD1YnuVjtMxT2DdTSeWa4o36Vyy6ZzONlPnvwxUb16A7eMR3ywhB98sIZ/q\n20G9tecwANo3vVKV1+s4xV9/wx8DsG528ek8//nb4q08ry86vrPt4Tueqsox7Xd5xDdLyCO+HdRf\n/+zLAEwpb771Rsen9wBsL9/eu7alGOlnrv4KACNnbQLgcDzK14NHfLOEPOLb/spPa+uYqrvw03cD\nsIgph/xSr36rmAx039du7lzX8UEepzxdfDL7MV9a0+tSrfc84psl5BHf9le+L6Zj6uyZw4tvtrni\n+8Xn6X/izn1Tagdv+zUAb5z5UQDGfKX45NzLPr4cgHNGFPcFlu3c97WKX1s9C4CP/NvImpRvlfGI\nb5aQR3w7qGEqfkXWnl18m+1PPzOsc9uGXR8D4NIjXul23/mvfwaAR35+cue6yfN9174ReMQ3S8jf\npGP7aZryCQCm3F180OWNH9v/O+oHdRkrur6FFuC5XcW2i/97XvEal/b+f//WO/4mHTM7IAffLCHf\n3LP97HnpFwBs+PJEAKZddhkAay78lwPuM/XhrwNwwm3FN99Mec6n+I3OI75ZQr65Z/Z7xDf3zOyA\nHHyzhBx8s4QqCr6kKyW9KOkFSXdLGiZpkqSVkjZKulfSkJ5fycwaQY/BlzQeuBxoiYgTgSbgIuBG\n4JaIOB7YAcytZaFmVj2Vnuo3A8MlNQMjgK3AWcDScvsS4Pzql2dmtdBj8CNiC3AT8BpF4N+h+PaD\ntyOi4/sV2oDx3e0vaZ6kVkmtu9lVnarNrE8qOdUfDcwGJgHHACOBWZUeICIWR0RLRLQMZmivCzWz\n6qnkVP9zwMsR8WZE7AbuB84AjixP/QEmAFtqVKOZVVklwX8NOF3SCEkCZgJrgBXABWWbOcCDtSnR\nzKqtkmv8lRQ38Z4FVpf7LAauBa6StBE4CrijhnWaWRV5rr7Z7xHP1TezA3LwzRJy8M0ScvDNEnLw\nzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDN\nEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0S\ncvDNEnLwzRJy8M0ScvDNElJE1O9g0pvATuCXdTto33yEgVMrDKx6B1KtMHDq/YOI+GhPjeoafABJ\nrRHRUteD9tJAqhUGVr0DqVYYePX2xKf6Zgk5+GYJ9UfwF/fDMXtrINUKA6vegVQrDLx6D6ru1/hm\n1v98qm+WUN2CL2mWpPWSNkpaUK/jVkrSsZJWSFoj6UVJ88v1YyQ9JmlD+Ti6v2vtIKlJ0nOSHiqX\nJ0laWfbxvZKG9HeNHSQdKWmppHWS1kqa0ah9K+nK8nfgBUl3SxrWyH3bG3UJvqQmYBFwDjANuFjS\ntHoc+xC0A1dHxDTgdOAbZY0LgOURMRlYXi43ivnA2i7LNwK3RMTxwA5gbr9U1b2FwCMRMRU4iaLu\nhutbSeOBy4GWiDgRaAIuorH79tBFRM1/gBnAo12WrwOuq8ex+1Dzg8DZwHpgXLluHLC+v2sra5lA\nEZazgIcAUUwwae6uz/u51iOAlynvKXVZ33B9C4wHNgNjgOayb/+iUfu2tz/1OtXv6MwObeW6hiRp\nIjAdWAmMjYit5aZtwNh+KuvDbgWuAfaWy0cBb0dEe7ncSH08CXgTuLO8NLld0kgasG8jYgtwE/Aa\nsBV4B1hF4/Ztr/jm3odIOgy4D7giIt7tui2KP/f9/m8QSV8EtkfEqv6upULNwCnAdyNiOsW07f1O\n6xuob0cDsyn+WB0DjARm9WtRNVCv4G8Bju2yPKFc11AkDaYI/V0RcX+5+g1J48rt44Dt/VVfF2cA\n50l6BbiH4nR/IXCkpOayTSP1cRvQFhEry+WlFH8IGrFvPwe8HBFvRsRu4H6K/m7Uvu2VegX/GWBy\neWd0CMXNkmV1OnZFJAm4A1gbETd32bQMmFM+n0Nx7d+vIuK6iJgQERMp+vKJiLgEWAFcUDZriFoB\nImIbsFnSCeWqmcAaGrBvKU7xT5c0ovyd6Ki1Ifu21+p40+Rc4CXgF8Df9vfNjW7q+zTFqeb/Ac+X\nP+dSXDsvBzYAjwNj+rvWD9X9Z8BD5fPjgKeBjcB/AEP7u74udZ4MtJb9+wAwulH7FvgHYB3wAvBD\nYGgj921vfjxzzywh39wzS8jBN0vIwTdLyME3S8jBN0vIwTdLyME3S8jBN0vo/wGoibhgvvrjfwAA\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0c074e0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0e63048>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a9e16400>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nB98sIQffLKH/B/ikxXVCaOABAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a04f07f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nzRL6f2XvwEHRocylAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a053e6d8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0cb7518>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAADz9JREFUeJzt3XuQlfV9x/H3h90FBG9gIkHQghGk\n1FYxG5WaxI7EFE1GzIwxOk7CWDv8kUTx0iraTqdNZzracbwkwbSMxpKbmqCjjGO1iqSTi6KL2qJc\nhOAFEMQoakoUWfj2j+fZdXEW97B7bsv385rZOee5nec7v9nP/p7nt79zjiICM8tlSKMLMLP6c/DN\nEnLwzRJy8M0ScvDNEnLwzRJy8M0SGlDwJc2UtEbSOknzqlWUmdWW+juBR1IL8AJwBrAReAq4ICJW\nVq88M6uF1gEcexKwLiLWA0i6C5gF7DX4QzUshjNyAKc0s4/yHtt5P3aor/0GEvxxwIYeyxuBkz+8\nk6Q5wByA4YzgZM0YwCnN7KMsiyUV7Vfzwb2IWBAR7RHR3sawWp/OzCowkOBvAo7ssTy+XGdmTW4g\nwX8KmCRpoqShwPnA4uqUZWa11O97/IjolPQt4GGgBfhBRDxftcrMrGYGMrhHRDwIPFilWsysTjxz\nzywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQff\nLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98s\nIQffLCEH3ywhB98sIQffLCEH3yyhPoMv6UhJSyWtlPS8pLnl+tGSHpG0tnwcVftyzawaKunxO4Er\nI2IqcArwTUlTgXnAkoiYBCwpl81sEOgz+BGxOSKeLp//HlgFjANmAQvL3RYC59SqSDOrrn26x5c0\nAZgGLAPGRMTmctMWYExVKzOzmqk4+JIOBO4BLouId3pui4gAYi/HzZHUIaljJzsGVKyZVUdFwZfU\nRhH6n0TEveXq1ySNLbePBbb2dmxELIiI9ohob2NYNWo2swGqZFRfwO3Aqoi4scemxcDs8vls4P7q\nl2dmtdBawT6nAl8DVkh6tlx3LXAd8DNJFwMvA+fVpkQzq7Y+gx8RvwK0l80zqluOmdWDZ+6ZJeTg\nmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCb\nJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl\n5OCbJeTgmyXk4Jsl5OCbJeTgmyVUcfAltUh6RtID5fJEScskrZN0t6ShtSvTzKppX3r8ucCqHsvX\nAzdFxDHANuDiahZmZrVTUfAljQe+CNxWLgs4HVhU7rIQOKcWBZpZ9VXa498MXAXsLpcPA96KiM5y\neSMwrrcDJc2R1CGpYyc7BlSsmVVHn8GX9CVga0Qs788JImJBRLRHRHsbw/rzEmZWZa0V7HMqcLak\ns4DhwMHALcChklrLXn88sKl2ZZpZNfXZ40fENRExPiImAOcDj0XEhcBS4Nxyt9nA/TWr0syqaiD/\nx78auELSOop7/turU5KZ1Voll/rdIuIXwC/K5+uBk6pfkpnVmmfumSXk4Jsl5OCbJeTgmyXk4Jsl\n5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk\n4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyW0T1+aadZMOk//FACbv7Gje93/TF8I\nwPGPzwbgiPlDAWhZ+nSdq2tu7vHNEnKPb4PO7tOmAfCdH3wPgGPaPvg13l0+PjP9DgDWtO8C4G8n\nnFK/AgcB9/hmCbnHt0Fj5xfaAbjq1h8BMLmtuH/f3d3Pw/qdOwF4e/cwAKYVD+w489MAHLB0RXHM\ne+/VvuAm5h7fLKGKgi/pUEmLJK2WtErSdEmjJT0iaW35OKrWxZpZdSgi+t5JWgj8MiJukzQUGAFc\nC7wZEddJmgeMioirP+p1DtboOFkzqlG3JdBy8MEAbP/cFAAuv+mnAJw5YhsAQ8p+q+el/j9sLS7p\nl9w6HYBf/+N39th36o+/BcDRVz9e09obZVks4Z14U33t12ePL+kQ4HPA7QAR8X5EvAXMAhaWuy0E\nzul/uWZWT5UM7k0EXgfukHQ8sByYC4yJiM3lPluAMbUp0bLa+MNxADz16fkVH/Ptw58C4KED/xyA\ni176AgALJzwKwMFT36hmiYNWJff4rcCJwPcjYhqwHZjXc4co7hd6vWeQNEdSh6SOnezobRczq7NK\nevyNwMaIWFYuL6II/muSxkbEZkljga29HRwRC4AFUNzjV6Fm2491TcMFuPOEYoLOEIbusc9FLxfj\nRB2P/jEAKy7+Xve2pe8OB+DwjncBWLetGB9o+5elxWv1efebQ589fkRsATZIOrZcNQNYCSwGZpfr\nZgP316RCM6u6Skf1TwBuA4YC64GLKP5o/Aw4CngZOC8i3vyo1/Govu1N1zTcmxfe2r2u51RcgLNX\nfxmAlnO3A/DmF4u+6I3jPujGJ8/fAEDnho17HPvApuUAbN5VXAn81exLu7ftT2/gqXRUv6KZexHx\nLNDeyyan2GwQ8pRdayh96k8A+N0VRU/cNQ0XYHk5FvzY/00F4I27jgTgsG3F/+AP+fETxWOP1+vs\n43xjWoo5vG9c9ofudYcv7Vfpg5qn7Jol5OCbJeRLfWuIISNGAND5r+8A8MSUewF4sfP97n2uuPZK\nAEb98hUADh9Z/Md4VxXOf9LYl7ufv1SF1xts3OObJeQe3xri3dOKQb2Hp9y6x/q/nnt59/OD7isG\n7/oasLN95x7fLCH3+NYQf/bPzwIfvF22axruAfc9WZPztakFgJ3lfLUW5Z497h7fLCH3+FZXb32t\n+ICMvx9zAwC7yzfgLP+vYpLOUfymJufdGbvK8xUf2vHQqqnd2yax/0zZrZR7fLOE3ONbXXUeUDwe\nMqTo6R9/r5hCe/QPXy22V+k8XfMEVt9wXLmmeJPOhevPBGDK3Be7963GvIDBxj2+WULu8a2h3th1\nIACd61+qyut19fRrrvtTAFbPKj6k4z//ULyV59X5xwBw0LYnqnK+wco9vllCDr5ZQr7Ut4b6m19/\nBYDJ5eBbf3R9eg/A1vJ9/avai0v8GSu+CsDImesBOIjcl/hd3OObJeQe3+qr/DS4rqm6t3zmTgDm\nM3mfX+rlbxeTge75+o3d67o+wefEJ4vPgT3iyyv7Xer+zD2+WULu8a2+yvfGdE2dPe2A4pttLvuP\n4vP0P3nHB9+D17bl9wC8dtrHARj91eKTcy85agkAZ44oxgUWb//gS5y+vmImAB/795E1KX9/4R7f\nLCH3+NZQw1X8Cq46498A+NVnh3dvW7vjEwBcdMhLvR4799XPAvDQb07oXjdprkftK+Ee3yyhir5J\np1r8TTrWMvmTAEy+s/iwy+s/sef31A/p0Rf1/N57gGd2FNsu+O85xWtc1P///e+vKv0mHff4Zgn5\nHt/qatcLvwVg7VcmADD1kksAWHned/d6zJQHvwHAsbcW334z+Rn39APlHt8sIQffLCEP7pntRzy4\nZ2Z75eCbJeTgmyXk4Jsl5OCbJVRR8CVdLul5Sc9JulPScEkTJS2TtE7S3ZKG1rpYM6uOPoMvaRxw\nKdAeEccBLcD5wPXATRFxDLANuLiWhZpZ9VR6qd8KHCCpFRgBbAZOBxaV2xcC51S/PDOrhT6DHxGb\ngBuAVygC/zbF9xG9FRFd33i0ERjX2/GS5kjqkNSxkx3VqdrMBqSSS/1RwCxgInAEMBKYWekJImJB\nRLRHRHsbw/pdqJlVTyWX+p8HXoyI1yNiJ3AvcCpwaHnpDzAe2FSjGs2syioJ/ivAKZJGSBIwA1gJ\nLAXOLfeZDdxfmxLNrNoqucdfRjGI9zSwojxmAXA1cIWkdcBhwO01rNPMqsjvzjPbj/jdeWa2Vw6+\nWUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5Z\nQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllC\nDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQoqI+p1Meh3YDvyubicdmI8xeGqFwVXvYKoVBk+9\nfxQRH+9rp7oGH0BSR0S01/Wk/TSYaoXBVe9gqhUGX7198aW+WUIOvllCjQj+ggacs78GU60wuOod\nTLXC4Kv3I9X9Ht/MGs+X+mYJ1S34kmZKWiNpnaR59TpvpSQdKWmppJWSnpc0t1w/WtIjktaWj6Ma\nXWsXSS2SnpH0QLk8UdKyso3vljS00TV2kXSopEWSVktaJWl6s7atpMvL34HnJN0paXgzt21/1CX4\nklqA+cCZwFTgAklT63HufdAJXBkRU4FTgG+WNc4DlkTEJGBJudws5gKreixfD9wUEccA24CLG1JV\n724BHoqIKcDxFHU3XdtKGgdcCrRHxHFAC3A+zd22+y4iav4DTAce7rF8DXBNPc49gJrvB84A1gBj\ny3VjgTWNrq2sZTxFWE4HHgBEMcGktbc2b3CthwAvUo4p9VjfdG0LjAM2AKOB1rJt/7JZ27a/P/W6\n1O9qzC4by3VNSdIEYBqwDBgTEZvLTVuAMQ0q68NuBq4CdpfLhwFvRURnudxMbTwReB24o7w1uU3S\nSJqwbSNiE3AD8AqwGXgbWE7ztm2/eHDvQyQdCNwDXBYR7/TcFsWf+4b/G0TSl4CtEbG80bVUqBU4\nEfh+REyjmLa9x2V9E7XtKGAWxR+rI4CRwMyGFlUD9Qr+JuDIHsvjy3VNRVIbReh/EhH3lqtfkzS2\n3D4W2Nqo+no4FThb0kvAXRSX+7cAh0pqLfdppjbeCGyMiGXl8iKKPwTN2LafB16MiNcjYidwL0V7\nN2vb9ku9gv8UMKkcGR1KMViyuE7nrogkAbcDqyLixh6bFgOzy+ezKe79GyoiromI8RExgaItH4uI\nC4GlwLnlbk1RK0BEbAE2SDq2XDUDWEkTti3FJf4pkkaUvxNdtTZl2/ZbHQdNzgJeAH4L/F2jBzd6\nqe8zFJea/ws8W/6cRXHvvARYCzwKjG50rR+q+y+AB8rnRwNPAuuAnwPDGl1fjzpPADrK9r0PGNWs\nbQv8E7AaeA74ETCsmdu2Pz+euWeWkAf3zBJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0S+n+3\nb8jPTzjC8wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0652518>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nHHyzhBx8s4T+H2h9yM/6/yTFAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0b36160>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nb5bQ/wOiRbskoCjCkQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a051b6d8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n8M0ScvDNEnLwzRJy8M0S+n+FoLsk52ofxQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0b15e10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQg6+WUIOvllC/w/DMbskVeU0UAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0bd2898>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nHHyzhBx8s4QcfLOEHHyzhP4fpbe61zBEZwgAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0c1db70>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0616ba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nB2bqyM/7V8YQAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0b724e0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nUgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0cdf198>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n9P8LlLhg0VWHAAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0caa400>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nWUIOvllC/wfNpLhggO7pMAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0d7c198>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJfT/Gciyxtuid3IAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a025f320>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n3ywhB98sIQffLKH/BytyxFeYSiivAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a04c8e48>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAAD0ZJREFUeJzt3XuwlPV9x/H3x3O4KwImIoIWDCCl\ntoo5UakxphJTNBkxM9boOMkZa4c/chEvraLtH23a6WjH8ZIG0zIaQy5VU3WUcaxWETtNouhBbFEu\nkeIFEMUoaoKK58i3fzzPOZwoeJZzdvfs+v28Zs7sPpfd33d+8Nnf8zz7211FBGaWy36DXYCZ1Z+D\nb5aQg2+WkINvlpCDb5aQg2+WkINvltCAgi9prqT1kjZIWlitosysttTfCTySWoBfAacCm4EngHMj\nYk31yjOzWmgdwGOPAzZExEYASbcB84C9Bn+ohsVwRg2gSTP7KO+yg/dip/rabyDBnwhs6rW8GTj+\ngztJmg/MBxjOSI7XnAE0aWYfZUUsq2i/ml/ci4jFEdEWEW1DGFbr5sysAgMJ/hbgsF7Lk8p1Ztbg\nBhL8J4BpkqZIGgqcAyytTllmVkv9PsePiC5J3wIeAFqAH0TEM1WrzMxqZiAX94iI+4D7qlSLmdWJ\nZ+6ZJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk\n4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTg\nmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJdRn8CUdJmm5pDWSnpG0oFw/TtKDkp4tb8fWvlwzq4ZKRvwu\n4NKImAmcAHxT0kxgIbAsIqYBy8plM2sCfQY/IrZGxJPl/d8Aa4GJwDxgSbnbEuDMWhVpZtW1T+f4\nkiYDs4AVwPiI2FpuehkYX9XKzKxmKg6+pP2BO4GLIuKt3tsiIoDYy+PmS+qQ1NHJzgEVa2bVUVHw\nJQ2hCP1PI+KucvUrkiaU2ycA2/b02IhYHBFtEdE2hGHVqNnMBqiSq/oCbgbWRsS1vTYtBdrL++3A\nPdUvz8xqobWCfU4EvgaslvRUue5K4CrgZ5IuAF4Azq5NiWZWbX0GPyJ+Dmgvm+dUtxwzqwfP3DNL\nyME3S8jBN0vIwTdLyME3S8jBN0vIwTdLyME3S8jBN0vIwTdLyME3S8jBN0vIwTdLyME3S8jBN0vI\nwTdLyME3S8jBN0vIwTdLyME3S8jBN0vIwTdLyME3S8jBN0vIwTdLyME3S8jBN0vIwTdLyME3S8jB\nN0vIwTdLyME3S8jBN0vIwTdLyME3S6ji4EtqkbRK0r3l8hRJKyRtkHS7pKG1K9PMqmlfRvwFwNpe\ny1cD10XEVGA7cEE1CzOz2qko+JImAV8CbiqXBZwC3FHusgQ4sxYFmln1VTriXw9cBuwqlw8C3oiI\nrnJ5MzBxTw+UNF9Sh6SOTnYOqFgzq44+gy/py8C2iFjZnwYiYnFEtEVE2xCG9ecpzKzKWivY50Tg\nDEmnA8OB0cANwBhJreWoPwnYUrsyzaya+hzxI+KKiJgUEZOBc4CHI+I8YDlwVrlbO3BPzao0s6qq\nZMTfm8uB2yT9A7AKuLk6JVm1dZ3yaQC2fqO4xvI/s5cAcPSj7QAcumj3O7Ety5+sc3U2GPYp+BHx\nCPBIeX8jcFz1SzKzWhvIiG8NbtfJswD47g++B8DUIcU/d/dbM6tm3wLA+rb3ex7zV5NPqF+BNmg8\nZdcsIY/4HzOdX2zruX/ZjT8GYPqQ4hx+VznWb+zsBODNXcXbq7N6vcu687TPADBi+eriMe++W9uC\nbVB4xDdLyCN+k2sZPRqAHZ+bAcDF1/1bz7Y/GfHb8t7vvr7/cPsfA7DsxtkA/OJvv9uz7cGb/gWA\nmT/5FgBHXP5o9Yu2QecR3ywhj/hNbvOPio9IPPGZRRU/5jsHPwHA/fsXI//5z3+xZ9uSyQ8BMHrm\na9Uq0RqQR3yzhDziN6nu2Xi3HlO8R78fH/4elPNfmANAx0O/D8DqC4p9l78zHICDO94BYMP2GT2P\nGfKPy4vnUy2qtkbhEd8sIQffLCFFRN0aG61xcbzm1K29j6PuabjXL7kR2D0Nt9sZ677Sc7/lrB0A\nvP6lIwF47aji+H36ok0AdG3a/KHnv3dL8bULW98vTgP+vP3C4rn84Z2msCKW8Va83ueJmkd8s4R8\nca8J6NN/0HP/15cUI3H3NNyV5beZPfzbmQC8dtthPfsetL2YfHPgTx4rbsv1XfRtfEsxj/e1i94G\n4ODl/SrdGpRHfLOEPOI3sP1GjgSg65/e6ln32Iy7AHiu6z0ALrnyUgDG/veLABw8alvPvrs/bNt/\nx014AYDnq/Bc1jg84psl5BG/gb1zcnFu/8CMGz+07S8WXAzAAXcX5++VnLebdfOIb5aQR/wG9kd/\n/xQA+/V6fe6ehjvi7sdr0uYQtQDQWU7vaFH95nlY/XjEN0vIwTdLyIf6DeiNrxXfjPM3468BYFev\nT96t/M9ios7h/LImbXfG+2Wbxffz3b+2aG8anrL7ceIR3ywhj/gNqGtEcXvgfsVI/+i7u78G94gf\nvVTsU4V2uicIrbvmqF5riw/pnLfxNABmLHgOqM5kIGscHvHNEvKI3wRee3//nvtdG58f8PN1j/Tr\nr/pDANbN+17Ptv94u/goz0uLpgJwwPbHBtyeNR6P+GYJecRvAn/5iz/ruT+9PAfvj+4v8dhWfrR3\nbVsx0s9Z/dWefUbN3QjAAXik/zjziG+WkEf8RlR+cVL3VN0bPntrz6ZFTN/np3vhO8W8gDu/fi2w\n+0s8jn28HYBDv7Km36Vac/KIb5aQR/xGVH4upnv23Mkjdv+qzUU/LL5P/1O3FNuGvPwbAF45+ZMA\njPvq7i/Q/PbhywA4bWRxXWDpjvEAfH31XAA+8a+jalK+NT6P+GYJVRR8SWMk3SFpnaS1kmZLGifp\nQUnPlrdja12smVVHpYf6NwD3R8RZkoYCI4ErgWURcZWkhcBC4PIa1ZnacO3+Z1p7avEz1j8/qfgZ\nrGd3HgLA+Qc+v9fHL3jpJADu/+UxAExb4LfqsutzxJd0IPA54GaAiHgvIt4A5gFLyt2WAGfWqkgz\nq64+f0lH0jHAYmANcDTFpzgWAFsiYky5j4Dt3ct741/SqUzL9E8BMP3W4hturz7k0Q/t0/1WX/cF\nwG6rdu5+LT/3v+YXz3N+/yf9WHOp5i/ptALHAt+PiFnADorD+h5RvHrs8RVE0nxJHZI6OtlZQXNm\nVmuVjPiHAI9FxORy+SSK4E8FPh8RWyVNAB6JiCM/6rk84u+b1iMmA7Du24f0rFtz9j8DHx7xZ9z3\nDQCOvPHtnn1j1TP1KNMaSNVG/Ih4GdgkqTvUcygO+5cC7eW6duCeftZqZnVW0a/lluf5NwFDgY3A\n+RQvGj8DDgdeAM6OiNc/6nk84pvVVqUjfkVv50XEU0DbHjY5xWZNyDP3zBJy8M0ScvDNEnLwzRJy\n8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLw\nzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDN\nEnLwzRJy8M0Sqij4ki6W9IykpyXdKmm4pCmSVkjaIOl2SUNrXayZVUefwZc0EbgQaIuIo4AW4Bzg\nauC6iJgKbAcuqGWhZlY9lR7qtwIjJLUCI4GtwCnAHeX2JcCZ1S/PzGqhz+BHxBbgGuBFisC/CawE\n3oiIrnK3zcDEPT1e0nxJHZI6OtlZnarNbEAqOdQfC8wDpgCHAqOAuZU2EBGLI6ItItqGMKzfhZpZ\n9VRyqP8F4LmIeDUiOoG7gBOBMeWhP8AkYEuNajSzKqsk+C8CJ0gaKUnAHGANsBw4q9ynHbinNiWa\nWbVVco6/guIi3pPA6vIxi4HLgUskbQAOAm6uYZ1mVkWKiLo1Nlrj4njNqVt7ZtmsiGW8Fa+rr/08\nc88sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH\n3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQff\nLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIUVE/RqTXgV2AL+uW6MD8wmap1Zornqb\nqVZonnp/LyI+2ddOdQ0+gKSOiGira6P91Ey1QnPV20y1QvPV2xcf6psl5OCbJTQYwV88CG32VzPV\nCs1VbzPVCs1X70eq+zm+mQ0+H+qbJVS34EuaK2m9pA2SFtar3UpJOkzScklrJD0jaUG5fpykByU9\nW96OHexau0lqkbRK0r3l8hRJK8o+vl3S0MGusZukMZLukLRO0lpJsxu1byVdXP4feFrSrZKGN3Lf\n9kddgi+pBVgEnAbMBM6VNLMebe+DLuDSiJgJnAB8s6xxIbAsIqYBy8rlRrEAWNtr+WrguoiYCmwH\nLhiUqvbsBuD+iJgBHE1Rd8P1raSJwIVAW0QcBbQA59DYfbvvIqLmf8Bs4IFey1cAV9Sj7QHUfA9w\nKrAemFCumwCsH+zaylomUYTlFOBeQBQTTFr31OeDXOuBwHOU15R6rW+4vgUmApuAcUBr2bd/2qh9\n29+/eh3qd3dmt83luoYkaTIwC1gBjI+IreWml4Hxg1TWB10PXAbsKpcPAt6IiK5yuZH6eArwKnBL\neWpyk6RRNGDfRsQW4BrgRWAr8Cawksbt237xxb0PkLQ/cCdwUUS81XtbFC/3g/42iKQvA9siYuVg\n11KhVuBY4PsRMYti2vbvHNY3UN+OBeZRvFgdCowC5g5qUTVQr+BvAQ7rtTypXNdQJA2hCP1PI+Ku\ncvUrkiaU2ycA2warvl5OBM6Q9DxwG8Xh/g3AGEmt5T6N1Mebgc0RsaJcvoPihaAR+/YLwHMR8WpE\ndAJ3UfR3o/Ztv9Qr+E8A08oro0MpLpYsrVPbFZEk4GZgbURc22vTUqC9vN9Oce4/qCLiioiYFBGT\nKfry4Yg4D1gOnFXu1hC1AkTEy8AmSUeWq+YAa2jAvqU4xD9B0sjy/0R3rQ3Zt/1Wx4smpwO/Av4P\n+OvBvrixh/o+S3Go+b/AU+Xf6RTnzsuAZ4GHgHGDXesH6v48cG95/wjgcWAD8O/AsMGur1edxwAd\nZf/eDYxt1L4F/g5YBzwN/BgY1sh9258/z9wzS8gX98wScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0S\ncvDNEvp/bSrDyOUBzOsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a04d1cc0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nLCEH3ywhB98sof8DX0bFdRVoiuQAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0496a58>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nARPqs4TX1/axAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a04056a0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAAD0tJREFUeJzt3XuQlfV9x/H3h93lqgiYSBBMwQhS\nYquYjUpMYioxRZMRM2ON1kkYS4c/kiheWkXbP9L0MrHjeEmDtozGkMSqqTrKOFarSDq5KLqILcol\nUhQBUYyiJqjIyrd/PM+uq4J72D3n7Dl+P6+ZnX2u5/nOA5/z+53f/s45igjMLJdBA12AmdWfg2+W\nkINvlpCDb5aQg2+WkINvlpCDb5ZQv4IvaZakdZLWS1pQraLMrLbU1wk8klqA3wAnAZuBR4GzImJ1\n9cozs1po7ce5xwDrI2IDgKRbgNnAXoM/WENiKCP6cUkz+yBvsoO3Yqd6O64/wR8PbOqxvhk49r0H\nSZoHzAMYynCO1cx+XNLMPsjyWFrRcTUf3IuIRRHRHhHtbQyp9eXMrAL9Cf4W4JAe6xPKbWbW4PoT\n/EeByZImSRoMnAksqU5ZZlZLfX6NHxGdkr4N3Ae0AD+MiCerVpmZ1Ux/BveIiHuAe6pUi5nViWfu\nmSXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCb\nJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl\n5OCbJeTgmyXk4Jsl5OCbJeTgmyXUa/AlHSJpmaTVkp6UNL/cPkbS/ZKeKn+Prn25ZlYNlbT4ncBF\nETENOA74lqRpwAJgaURMBpaW62bWBHoNfkRsjYjHyuXfAWuA8cBsYHF52GLgtFoVaWbVtU+v8SVN\nBKYDy4GxEbG13PU8MLaqlZlZzVQcfEn7AbcD50fEaz33RUQAsZfz5knqkNSxi539KtbMqqOi4Etq\nowj9TRFxR7n5BUnjyv3jgG17OjciFkVEe0S0tzGkGjWbWT9VMqov4AZgTURc2WPXEmBOuTwHuKv6\n5ZlZLbRWcMzxwNeBVZIeL7ddBnwP+JmkucBG4IzalGhm1dZr8CPil4D2sntmdcsxs3rwzD2zhBx8\ns4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyz\nhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOE\nHHyzhBx8s4QcfLOEHHyzhBx8s4QqDr6kFkkrJd1drk+StFzSekm3ShpcuzLNrJr2pcWfD6zpsX45\ncFVEHAZsB+ZWszAzq53WSg6SNAH4MvCPwIWSBJwI/Hl5yGLgO8B1NajR9qDzxE8BsPWbOwH4nxmL\nATjyoTndxxy8sOiEtSx7rM7VWaOrtMW/GrgY2F2uHwi8EhGd5fpmYPyeTpQ0T1KHpI5d7OxXsWZW\nHb22+JK+AmyLiBWSvrCvF4iIRcAigJEaE/tcob3L7hOmA/D9H/4AgMPain/CrmfklTNu7D52Xfvb\nAPz1xOPqV6A1hUq6+scDp0o6BRgKjASuAUZJai1b/QnAltqVaWbV1GtXPyIujYgJETEROBN4MCLO\nBpYBp5eHzQHuqlmVZlZVFQ3u7cUlwC2S/gFYCdxQnZLsvXZ9qb17+eJrfwLAlLZi4G532cnfsGsX\nAK/uHtJ97PRycefJnwZg2LJVxTlvvlnbgq3h7VPwI+LnwM/L5Q3AMdUvycxqrT8tvtVIy8iRAOz4\n/FQALrjq37v3/cmw35dL736V9qPtnwFg6bUzurf96jvfB+D+6/8VgGk//TYAh17yUPWLtqbiKbtm\nCbnFb0Cbf1xMiXj00wsrPue7Bz0KwL37faZ72znPfAmAxRMfAGDktJeqVaI1Obf4Zgm5xW8gXdNw\nbz6qmJwziPe/7+mcjTMB6HjgDwFYNbc4dtkbQwE4qOON7mPXby/GCNr+aVnxeKpF1daM3OKbJaSI\n+s2iHakxcaxm1u16zaJrGu7Vi68F3pmG2+XUtV/tXm45fQcAL3/5cABeOqJoxqcs3ARA56bN73v8\nu7esAGDr20Vv4C/mnFc8lt+886GzPJbyWrzca9/OLb5ZQn6NP4D0qU8C8NsLi5a4azbeivJNjA/+\nfhoAL91ySPc5B24v/gZ/wE8fLn6X2zvp3diWYirfS+e/DsBBy/pcujU5t/hmCTn4Zgm5q19ng4YP\n717u/OfXAHh46h0APN35FgAXXnYRAKN/8SwAB43Y1n3O21Wo4ZhxGwF4pgqPZc3JLb5ZQm7x6+yN\nEz7ZvXzf1Gvfte8v518AwP53FgN3lQzYmfWFW3yzhNzi19kf//3j3cuDyufdrmm4w+58pCbXbFML\nALvKuVot8kcfZucW3ywht/h18srXiw/I+NuxV3Rv212+CWfFfxUTdT7Or2ty7V3xdnm94mO67l1T\nXG8ynrKblVt8s4Tc4tdJ57Di9wGD3nmr7UNvFlNoD/3xc8UxVbhOz3kCa684olwq3qRz9oaTAZg6\n/2mgOnMCrDm5xTdLyME3S8hd/QH00tv7AdC54Zl+P1ZXF3/d9/6oe9va2cWn8/zn68V7+J5beBgA\n+29/uN/Xs+bmFt8sIbf4A+ivfvVnAEwpB9/6ouvTe7aV7+lf0/6D7n0zV30NgBGzNgCwP27preAW\n3ywht/j1Un4K2qAez7XXfPZmABYyZZ8fbuN3iwlBt3/jSuCdT+85+pE53ccc/NXVfSrVPvzc4psl\n5Ba/Xsr3xXRNmwU4YVjxzTbn/6j4PP1P3Fjsa3v+dwC8cMJHARjztXc+Offcjy8F4OThxbjAkh1j\nAfjGqlkAfOTfRtSkfPtwcYtvlpBb/AE0VMXtX3NS8W22v/xc8W04T+38GADnHPDMXs+d/9znALj3\n10cBMHm+R+ytcm7xzRLyN+nUScuUTwAw5eaN3dsu/9i7v6e+a8S/5zgAwMqd7zw/n/Xf84rHOafv\nf/u3Dy9/k46Z7VVFwZc0StJtktZKWiNphqQxku6X9FT5e3StizWz6qioqy9pMfCLiLhe0mBgOHAZ\n8HJEfE/SAmB0RFzyQY+TuavfpfXQid3La88tBvFWn/EvwPu7+lPv+SYAh1/7evc5sfLJepRpTapq\nXX1JBwCfB24AiIi3IuIVYDawuDxsMXBa38s1s3rqtcWXdBSwCFgNHEnxcS7zgS0RMao8RsD2rvW9\ncYtvVlvVHNxrBY4GrouI6cAOYEHPA6J49tjjM4ikeZI6JHXsYmcFlzOzWqsk+JuBzRGxvFy/jeKJ\n4AVJ4wDK39v2dHJELIqI9ohob2NINWo2s37qNfgR8TywSdLh5aaZFN3+JUDXW8HmAHfVpEIzq7pK\np+yeC9xUjuhvAM6heNL4maS5wEbgjNqUaGbVVlHwI+JxoH0PuzxSZ9aEPHPPLCEH3ywhB98sIQff\nLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98s\nIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywh\nB98sIQffLCEH3yyhioIv6QJJT0p6QtLNkoZKmiRpuaT1km6VNLjWxZpZdfQafEnjgfOA9og4AmgB\nzgQuB66KiMOA7cDcWhZqZtVTaVe/FRgmqRUYDmwFTgRuK/cvBk6rfnlmVgu9Bj8itgBXAM9SBP5V\nYAXwSkR0lodtBsbv6XxJ8yR1SOrYxc7qVG1m/VJJV380MBuYBBwMjABmVXqBiFgUEe0R0d7GkD4X\nambVU0lX/4vA0xHxYkTsAu4AjgdGlV1/gAnAlhrVaGZVVknwnwWOkzRckoCZwGpgGXB6ecwc4K7a\nlGhm1VbJa/zlFIN4jwGrynMWAZcAF0paDxwI3FDDOs2sihQRdbvYSI2JYzWzbtczy2Z5LOW1eFm9\nHeeZe2YJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhm\nCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJ\nOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJKSLqdzHpRWAH8Nu6XbR/PkLz1ArN\nVW8z1QrNU+8fRMRHezuorsEHkNQREe11vWgfNVOt0Fz1NlOt0Hz19sZdfbOEHHyzhAYi+IsG4Jp9\n1Uy1QnPV20y1QvPV+4Hq/hrfzAaeu/pmCdUt+JJmSVonab2kBfW6bqUkHSJpmaTVkp6UNL/cPkbS\n/ZKeKn+PHuhau0hqkbRS0t3l+iRJy8t7fKukwQNdYxdJoyTdJmmtpDWSZjTqvZV0Qfl/4AlJN0sa\n2sj3ti/qEnxJLcBC4GRgGnCWpGn1uPY+6AQuiohpwHHAt8oaFwBLI2IysLRcbxTzgTU91i8HroqI\nw4DtwNwBqWrPrgHujYipwJEUdTfcvZU0HjgPaI+II4AW4Ewa+97uu4io+Q8wA7ivx/qlwKX1uHY/\nar4LOAlYB4wrt40D1g10bWUtEyjCciJwNyCKCSate7rnA1zrAcDTlGNKPbY33L0FxgObgDFAa3lv\n/7RR721ff+rV1e+6mV02l9sakqSJwHRgOTA2IraWu54Hxg5QWe91NXAxsLtcPxB4JSI6y/VGuseT\ngBeBG8uXJtdLGkED3tuI2AJcATwLbAVeBVbQuPe2Tzy49x6S9gNuB86PiNd67ovi6X7A/wwi6SvA\ntohYMdC1VKgVOBq4LiKmU0zbfle3voHu7WhgNsWT1cHACGDWgBZVA/UK/hbgkB7rE8ptDUVSG0Xo\nb4qIO8rNL0gaV+4fB2wbqPp6OB44VdIzwC0U3f1rgFGSWstjGukebwY2R8Tycv02iieCRry3XwSe\njogXI2IXcAfF/W7Ue9sn9Qr+o8DkcmR0MMVgyZI6XbsikgTcAKyJiCt77FoCzCmX51C89h9QEXFp\nREyIiIkU9/LBiDgbWAacXh7WELUCRMTzwCZJh5ebZgKracB7S9HFP07S8PL/RFetDXlv+6yOgyan\nAL8B/g/4m4Ee3NhDfZ+l6Gr+L/B4+XMKxWvnpcBTwAPAmIGu9T11fwG4u1w+FHgEWA/8BzBkoOvr\nUedRQEd5f+8ERjfqvQX+DlgLPAH8BBjSyPe2Lz+euWeWkAf3zBJy8M0ScvDNEnLwzRJy8M0ScvDN\nEnLwzRJy8M0S+n8h38V1epy7OQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a042a358>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAADz1JREFUeJzt3X2QVfV9x/H3x+UZRcFEgmAKRpBS\nW8VsVGoSU4kpGkfMjDFaxzCWDn8kUXxofUj7R5o+TOw4PqRBW0ZjSGI1qTrKOFarK+k0iaKL2KI8\nCMEHQBCjqAlRZOHbP87ZdaGLe9m9T+v385rZufec8zv3fOcHn/s75+zv3lVEYGa5HNDoAsys/hx8\ns4QcfLOEHHyzhBx8s4QcfLOEHHyzhPoVfEmzJK2RtE7S1dUqysxqS32dwCOpBXgeOA3YCDwFnB8R\nK6tXnpnVwqB+7HsCsC4i1gNIuguYDewz+EM0NIYxsh+HNLMP8i7beS92qLd2/Qn+eGBDt+WNwIl7\nN5I0D5gHMIwRnKiZ/TikmX2QpdFWUbua39yLiIUR0RoRrYMZWuvDmVkF+hP8TcAR3ZYnlOvMrMn1\nJ/hPAZMlTZI0BDgPWFydssyslvp8jR8RHZK+ATwMtADfj4jnqlaZmdVMf27uEREPAg9WqRYzqxPP\n3DNLyME3S8jBN0vIwTdLyME3S8jBN0vIwTdLyME3S8jBN0vIwTdLyME3S8jBN0vIwTdLyME3S8jB\nN0vIwTdLyME3S8jBN0vIwTdLyME3S8jBN0vIwTdLyME3S8jBN0vIwTdLyME3S8jBN0vIwTdLyME3\nS8jBN0vIwTdLyME3S8jBN0vIwTdLqNfgSzpC0hJJKyU9J2l+uX6MpEckrS0fR9e+XDOrhkpG/A7g\nioiYBpwEfF3SNOBqoC0iJgNt5bKZDQC9Bj8iNkfE0+Xz3wCrgPHAbGBR2WwRcHatijSz6tqva3xJ\nE4HpwFJgbERsLjdtAcZWtTIzq5mKgy/pQOAe4NKIeLv7togIIPax3zxJ7ZLad7KjX8WaWXVUFHxJ\ngylCf0dE3FuuflXSuHL7OGBrT/tGxMKIaI2I1sEMrUbNZtZPldzVF3AbsCoiru+2aTEwp3w+B7i/\n+uWZWS0MqqDNycCFwApJz5Trvgl8B/ippLnAS8C5tSnRzKqt1+BHxM8B7WPzzOqWY2b14Jl7Zgk5\n+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4\nZgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhm\nCTn4ZgkNanQBVh0dp34SgM1f29G17n9mLALg2MfnAHD4giEAtCx5us7VWbPxiG+WkEf8AW73KdMB\n+O73vwfAUYPf/yfdXT4un3E7AGtadwHwVxNPql+B1pQ84pslVPGIL6kFaAc2RcSZkiYBdwGHAsuA\nCyPivdqUaXvb+YVWAK68+UcATBlcXL/v7hrnYf3OnQC8tXsoANOLB3ac/ikAhi9ZUezz7ru1L9ia\nyv6M+POBVd2WrwVuiIijgG3A3GoWZma1U9GIL2kC8EXgH4DLJQk4Ffizsski4FvALTWoMb2WUaO6\nnm//7FQALrvh3wD4k+G/Lbf8//fwH2z7YwDabp4BwC++9V0AHrn1XwCY9uNvAHDkVY9Xv2hrapWO\n+DcCV/L+/aJDgTcjoqNc3giM72lHSfMktUtq38mOnpqYWZ31OuJLOhPYGhHLJH1ufw8QEQuBhQCj\nNCb2u0Jj4w/ff0996lMLKt7v24c9BcBDBxYj/0UvfgGARRMfBWDUtNerVaINMJWc6p8MnCXpDGAY\nMAq4CThE0qBy1J8AbKpdmWZWTb2e6kfENRExISImAucBj0XEBcAS4Jyy2Rzg/ppVaWZV1Z8JPFcB\nd0n6e2A5cFt1SrJOndNw7zzue13rDmDIHm0uemkmAO2P/j4AK+a+33bJO8MAOKz9HQDWbStuDA7+\nxyXFa6kWVdtAsF/Bj4ifAT8rn68HTqh+SWZWa4qo3/22URoTJ2pm3Y43UHVOw71x0c3AntNwO521\n+ksAtJyzHYA3vng0AK8f8/4wPmXBBgA6NmzcY98HNi0DYPOu4kzgz+dc0rXNH+AZ2JZGG2/HG72e\ny3nKrllC/pBOE9En/wCAX19ejMSd03CXdZv+8NhvpwHw+l1HAHDotmLyzcE/fqJ47PZ6HXywsS3F\nHN7XL/1d17rDlvSpdBtgPOKbJeQRvwkcMGIEAB3/9DYAT0y9F4AXOorPPF3+zSu62o7+75cBOGzk\nVgB2VeH4J4x7qev5i1V4PWt+HvHNEvKI3wTeOaW4tn946s17rP+L+ZcBcNB9T3St6+263awSHvHN\nEvKI3wT+6O+eAeCA8n24czbe8PuerMnxBqsFgJ3lFI4W+bNT2XjEN0vIwTdLyKf6DfTmhcU34/zN\n2OsA2F1+AGfZfxaTdD7OL2ty3J2xqzxe8b0qD62a1rVtMp6ym4FHfLOEPOI3UMfw4vHgA4qR/vF3\niym0R/7wlWJ7lY7TOUFo9XXHlGuKD+lcsP50AKbOf6GrbTUmBFnz84hvlpBH/Cby+q4DAehY/2JV\nXq9zpF/znT8EYPXs4ks6/uN3xUd5XllwFAAHbXuih73tw8wjvllCHvGbyF/+4ssATCmvwfui80s8\nALaWH+9d1VqM9DNXfAWAkbPWA3AQHumz8ohvlpBH/EYqvyCpc6ruTZ++E4AFTNnvl3rp28WcgHu+\nen3Xus4v8jj+yTkAHP6llX0u1T5cPOKbJeTgmyXkU/1GKj8U1zl19pThxZ+0uvQHxffpf+L2Yv3g\nLb/p2uXVUz4KwJivFN+ce/HH2wA4fURxQ3Dx9rFdbb+6YhYAH/nXkTUp3wYuj/hmCXnEbyLDVPxz\nrDqt+DPWP/9M8Zdw1u74WFebiw5+scd957/yGQAe+uVxXesmz/ev66xnHvHNEvJf0mmglimfAGDK\nncW33F77scf32N75a77OewDdLd9RbDv/v+YVr3FR3yf92IeH/5KOme2Tr/EbaNfzvwJg7ZcnAjDt\n4osBWHnuP+9zn6kPfg2Ao28u/vrNlOUe6W3/ecQ3S8jX+GYfIr7GN7N9cvDNEnLwzRKqKPiSDpF0\nt6TVklZJmiFpjKRHJK0tH0fXulgzq45KR/ybgIciYipwLLAKuBpoi4jJQFu5bGYDQK/Bl3Qw8Fng\nNoCIeC8i3gRmA4vKZouAs2tVpJlVVyUj/iTgNeB2Scsl3SppJDA2IjaXbbYAY/f5CmbWVCoJ/iDg\neOCWiJgObGev0/ooJgP0OCFA0jxJ7ZLad7Kjv/WaWRVUEvyNwMaIWFou303xRvCqpHEA5ePWnnaO\niIUR0RoRrYMZWo2azayfeg1+RGwBNkg6ulw1E1gJLAbmlOvmAPfXpEIzq7pKP6RzMXCHpCHAeuAi\nijeNn0qaC7wEnFubEs2s2ioKfkQ8A7T2sMkT780GIM/cM0vIwTdLyME3S8jBN0vIwTdLyME3S8jB\nN0vIwTdLyME3S8jBN0vIwTdLyME3S8jBN0vIwTdLyME3S8jBN0vIwTdLyME3S8jBN0vIwTdLyME3\nS8jBN0vIwTdLyME3S8jBN0vIwTdLyME3S8jBN0vIwTdLyME3S8jBN0vIwTdLyME3S8jBN0vIwTdL\nqKLgS7pM0nOSnpV0p6RhkiZJWippnaSfSBpS62LNrDp6Db6k8cAlQGtEHAO0AOcB1wI3RMRRwDZg\nbi0LNbPqqfRUfxAwXNIgYASwGTgVuLvcvgg4u/rlmVkt9Br8iNgEXAe8TBH4t4BlwJsR0VE22wiM\n72l/SfMktUtq38mO6lRtZv1Syan+aGA2MAk4HBgJzKr0ABGxMCJaI6J1MEP7XKiZVU8lp/qfB16I\niNciYidwL3AycEh56g8wAdhUoxrNrMoqCf7LwEmSRkgSMBNYCSwBzinbzAHur02JZlZtlVzjL6W4\nifc0sKLcZyFwFXC5pHXAocBtNazTzKpIEVG3g43SmDhRM+t2PLNslkYbb8cb6q2dZ+6ZJeTgmyXk\n4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTg\nmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCb\nJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJaSIqN/BpNeA7cCv63bQ/vkIA6dWGFj1DqRaYeDU+3sR\n8dHeGtU1+ACS2iOita4H7aOBVCsMrHoHUq0w8OrtjU/1zRJy8M0SakTwFzbgmH01kGqFgVXvQKoV\nBl69H6ju1/hm1ng+1TdLqG7BlzRL0hpJ6yRdXa/jVkrSEZKWSFop6TlJ88v1YyQ9Imlt+Ti60bV2\nktQiabmkB8rlSZKWln38E0lDGl1jJ0mHSLpb0mpJqyTNaNa+lXRZ+X/gWUl3ShrWzH3bF3UJvqQW\nYAFwOjANOF/StHocez90AFdExDTgJODrZY1XA20RMRloK5ebxXxgVbfla4EbIuIoYBswtyFV9ewm\n4KGImAocS1F30/WtpPHAJUBrRBwDtADn0dx9u/8iouY/wAzg4W7L1wDX1OPY/aj5fuA0YA0wrlw3\nDljT6NrKWiZQhOVU4AFAFBNMBvXU5w2u9WDgBcp7St3WN13fAuOBDcAYYFDZt3/arH3b1596nep3\ndmanjeW6piRpIjAdWAqMjYjN5aYtwNgGlbW3G4Ergd3l8qHAmxHRUS43Ux9PAl4Dbi8vTW6VNJIm\n7NuI2ARcB7wMbAbeApbRvH3bJ765txdJBwL3AJdGxNvdt0Xxdt/wX4NIOhPYGhHLGl1LhQYBxwO3\nRMR0imnbe5zWN1HfjgZmU7xZHQ6MBGY1tKgaqFfwNwFHdFueUK5rKpIGU4T+joi4t1z9qqRx5fZx\nwNZG1dfNycBZkl4E7qI43b8JOETSoLJNM/XxRmBjRCwtl++meCNoxr79PPBCRLwWETuBeyn6u1n7\ntk/qFfyngMnlndEhFDdLFtfp2BWRJOA2YFVEXN9t02JgTvl8DsW1f0NFxDURMSEiJlL05WMRcQGw\nBDinbNYUtQJExBZgg6Sjy1UzgZU0Yd9SnOKfJGlE+X+is9am7Ns+q+NNkzOA54FfAX/d6JsbPdT3\naYpTzf8Fnil/zqC4dm4D1gKPAmMaXetedX8OeKB8fiTwJLAO+HdgaKPr61bncUB72b/3AaObtW+B\nvwVWA88CPwKGNnPf9uXHM/fMEvLNPbOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhP4PlMG9\nEPZw+JAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a06b15c0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ns4QcfLOE/h/HOb0QbV/MCQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a03c4358>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQg6+WUIOvllCDr5ZQv8PJD61RZyrokgAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a03ccf60>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ng2+WkINvltD/AUW/vRC8BP9NAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0aabc18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nWUIOvllC/weS870QydoBJQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0ab3e80>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nHb7QG4vNbwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0996c18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEnLwzRJy8M0ScvDNEnLwzRJy8M0S+j/E2NYS5+Lh7QAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0702860>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n3ywhB98sof8HtTy9EJys5eUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0e2bba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n4Jsl5OCbJeTgmyX0fzjzvRAi2RxUAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0e4b780>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n4Jsl9L/Vlb0QsjO8qAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0870c18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIQffLCEH3yyh/wNqF70Qrs2XnwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0e76160>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nOfhmCTn4Zgk5+GYJOfhmCTn4Zgn9PzMEw/5J42PVAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0e6fdd8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nkINvlpCDb5bQ/wMmRr0QokdXPQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0aa1f98>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nZgk5+GYJ/T/NFL0Ql40GFgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0927438>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQg6+WUL/B6tJvRCpQsgsAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0dcbe10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nnNm9EHwEZd8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a08bc748>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nZgk5+GYJOfhmCf0/StO9EP6WmmoAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a091bdd8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nszjAQes8+dAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a035cac8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEnLwzRJy8M0S+n9gaL0QlsvYTwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a035f710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ndeG9EAjg4f4AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a02a33c8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n+GYJ/T9vDLhgb9qxQgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a02b66d8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nN0vIwTdL6P8Ac2G7vWvxlGEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0d46eb8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n5OCbJfT/cuq4YMr8IHQAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a031da90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQXSedlcAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a030ee80>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nLCEH3yyh/weRwsRXtY6e4AAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a032b278>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nltD/A/envRD/O3VVAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a05b4198>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nhBx8s4QcfLOEHHyzhP4frYK7vVGZIg8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a08e9d30>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nwTdL6P8BUVXAQSkgjasAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a08fb9e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nyME3S+j/AOMEuGCYKZ80AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a08e3dd8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ncvDNEvp/qo3AQYpbpC0AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a09e1438>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ng2+WkINvltD/A1emuGCQy2B5AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a07da080>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSwHIz2cxCYwAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a078c208>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a07f9c88>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a07c86d8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nR+oykPYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a056f1d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0595048>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nPE4AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0578f98>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a081f978>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nyUZSQZq4AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0852630>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nhP4PkhXGR0i1tTsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a07552e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n0yyPWQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a083fe80>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nOkm2Ssy/JAsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a09b3c18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nuyTb8QOEAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a021d860>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n+GYJOfhmCTn4Zgn9HxyxutfKLVALAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0248518>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAAD0BJREFUeJzt3X+QVeV9x/H3h2X56S8wkSCYghGk\nxFYxG5WYxFZiqiYjZsYYrWMYSoY/kij+aI2mnUmadjqx4/gjDdoyGkMSq0nRUYZarSLpNImii9gS\nWRCCIiCIMagJUdyFb/84Z9cFwb3s3nv3br+f18zO3vOc59zznTP7uc+5Z597jyICM8tlUH8XYGb1\n5+CbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl1KfgSzpb0lpJ6yVdW62izKy21NsJPJKagOeAs4DNwFPA\nxRGxunrlmVktDO7DtqcA6yNiA4Cke4CZwAGDP0RDYxgj+7BLM3svb7GTt2OXeurXl+CPAzZ1W94M\nnLpvJ0lzgbkAwxjBqZrRh12a2XtZHksr6lfzi3sRsSAiWiKipZmhtd6dmVWgL8HfAhzTbXl82WZm\nDa4vwX8KmCRpoqQhwEXA4uqUZWa11Ov3+BHRIemrwMNAE/C9iHi2apWZWc305eIeEfEg8GCVajGz\nOvHMPbOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyz\nhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOE\nHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEegy+pGMkLZO0WtKzkuaV7aMlPSJpXfl7VO3LNbNqqGTE\n7wCujoipwGnAVyRNBa4FlkbEJGBpuWxmA0CPwY+IrRHxdPn4t0AbMA6YCSwsuy0Ezq9VkWZWXQf1\nHl/SBGAasBwYExFby1XbgDFVrczMaqbi4Es6BLgXuCIi3ui+LiICiANsN1dSq6TWdnb1qVgzq46K\ngi+pmSL0d0XEfWXzy5LGluvHAtv3t21ELIiIlohoaWZoNWo2sz6q5Kq+gDuAtoi4sduqxcCs8vEs\n4IHql2dmtTC4gj6nA5cCqyQ9U7Z9Hfg28BNJc4CNwIW1KdHMqq3H4EfEzwAdYPWM6pZjZvXgmXtm\nCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJ\nOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5\n+GYJOfhmCTn4Zgk5+GYJOfhmCTn4ZglVHHxJTZJWSlpSLk+UtFzSekk/ljSkdmWaWTUdzIg/D2jr\ntnw9cFNEHAfsAOZUszAzq52Kgi9pPPAZ4PZyWcCZwKKyy0Lg/FoUaGbVV+mIfzNwDbCnXD4SeC0i\nOsrlzcC4/W0oaa6kVkmt7ezqU7FmVh09Bl/SZ4HtEbGiNzuIiAUR0RIRLc0M7c1TmFmVDa6gz+nA\neZLOBYYBhwG3AEdIGlyO+uOBLbUr08yqqccRPyKui4jxETEBuAh4LCIuAZYBF5TdZgEP1KxKM6uq\nvvwf/2vAVZLWU7znv6M6JZlZrVVyqt8lIn4K/LR8vAE4pfolmVmteeaeWUIOvllCDr5ZQg6+WUIO\nvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+\nWUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5Z\nQg6+WUIVBV/SEZIWSVojqU3SdEmjJT0iaV35e1StizWz6qh0xL8FeCgipgAnAm3AtcDSiJgELC2X\nzWwA6DH4kg4HPgncARARb0fEa8BMYGHZbSFwfq2KNLPqqmTEnwi8AtwpaaWk2yWNBMZExNayzzZg\nTK2KNLPqqiT4g4GTgdsiYhqwk31O6yMigNjfxpLmSmqV1NrOrr7Wa2ZVUEnwNwObI2J5ubyI4oXg\nZUljAcrf2/e3cUQsiIiWiGhpZmg1ajazPuox+BGxDdgk6fiyaQawGlgMzCrbZgEP1KRCM6u6wRX2\nuwy4S9IQYAMwm+JF4yeS5gAbgQtrU6KZVVtFwY+IZ4CW/ayaUd1yzKwePHPPLCEH3ywhB98sIQff\nLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98s\nIQffLCEH3yyhSr9l16xHHWd+BICtXy5unPI/04s7rJ34+KyuPkfPHwJA07Kn61yddecR3ywhj/jW\nZ3vOmAbAd773XQCOay7+rPaU61dOv7Or79qW3QD81YTT6legvYtHfLOEPOJbr7V/urjHyjW3/hCA\nyc3F+/c95Vi/ob0dgNf3vHPPxGnlw13nfBSA4ctWFdu89VbtC7YuHvHNEvKIbxVpOuwwAHZ+ckpX\n25U3/SsAfzr8d2XL3uPI93d8DIClt07vavv5N78DwCO3/zMAU3/0VQCO/drj1S/aDsgjvllCDr5Z\nQj7Vt4ps/sE4AJ766PyKt/nWUU8B8NAhH+tqm/3CpwFYOOFRAA6b+mq1SrSD4BHfLCGP+PaeOqfh\n3n1SMTlnEEPe1Wf2xhkAtD76hwCsmlP0XfbmMACOan2zq+/6HcXFweZ/WFY8n2pRtfXEI75ZQoqI\nuu3sMI2OUzWjbvuz3uuchnvzwluBd6bhdnfems8B0HTBTgB+85njAXj1hGIYnzx/EwAdmza/a9sl\nW1YAsHV3cTbwF7MuL57LH97pk+WxlDfiNz2eR3nEN0uoovf4kq4EvgQEsAqYDYwF7gGOBFYAl0bE\n2zWq0+pEH/kwAL++qhiJO6fhrig+actjv5va1ffVe44B4MgdxeSbw3/0RPG7XN9Rwf7GNBVzeF+9\n4vcAHLWs16XbQehxxJc0DrgcaImIE4Am4CLgeuCmiDgO2AHMqWWhZlY9lV7VHwwMl9QOjAC2AmcC\nf16uXwh8E7it2gVa7Q0aMaLrccc/vgHAE1PuA+D5juIk7qqvXw3AqP9+savvUSO3A7C7CjWcMnYj\nAC9U4bmsZz2O+BGxBbgBeJEi8K9TnNq/FhGdZ3ObgXH7217SXEmtklrb2VWdqs2sTyo51R8FzAQm\nAkcDI4GzK91BRCyIiJaIaGlmaM8bmFnNVXKq/yng+Yh4BUDSfcDpwBGSBpej/nhgS+3KtFp684wP\ndz1+eMqte6370rwrATj0/uLCXSUX7KzxVfLvvBeB0ySNkCRgBrAaWAZcUPaZBTxQmxLNrNp6HPEj\nYrmkRcDTFC/4K4EFwL8D90j6+7LtjloWarXzx3/3TNfjQeVY0DkNd/j9T9Zkn81qAqC9nD/WpPpN\nJLMKr+pHxDeAb+zTvAE4peoVmVnN+UM6ib12afHNOH8z5oautj3lh3BW/GcxUeeD/KIm+26P3eX+\niu/ne6it2N8kPGW3Hjxl1ywhj/iJdQwvfh8+6J2P2j7+VvEv12N/8FLRpwr76T5BaM0NJ5SPig/p\nXLLhHACmzHseqM5kIOuZR3yzhDzi215e3X0IAB0bXujzc3WO9Gu//UddbWtmFl/S8R+/Lz7K89L8\n4wA4dMcTfd6fVc4jvllCHvFtL3/5888DMLl8D94bnV/isb38aG9by3e71s1Y9QUARp69AYBD8Ujf\nHzzimyXk4Jsl5FP9zMpvZhvU7fX/lo/fDcB8Jh/00238VjEh6N4v3gi88+09Jz85q6vP0Z9b3atS\nrbo84psl5BE/s/JzMZ3TZgHOGF7c2eaK7xffp/+hO4t1zdt+C8DLZ7y/q+/oLxTfnnvZB5cCcM6I\n4oLg4p1jAPjiquJrG973LyNrUr71nkd8s4Q84ttehqn4k2g7q7iN9c8+UdwNZ92uDwAw+/AXDrjt\nvJc+AcBDvzgJgEnz/K+6RuUR3ywh30knsabJHwJg8t0bu9qu/8Dje/XpvOLf/TpAp5W7inUX/9fc\n4nlm937Sj1WH76RjZgfk9/iJ7X7uVwCs+/yErrapl10GwOoL/2m/20x58Mtdj4+/tbj7zeSVHukH\nGo/4Zgk5+GYJ+eKe2f8jvrhnZgfk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk\n4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyVU1y/ikPQKsBP4dd122jfvY+DUCgOr3oFUKwycev8gIt7f\nU6e6Bh9AUmtEtNR1p700kGqFgVXvQKoVBl69PfGpvllCDr5ZQv0R/AX9sM/eGki1wsCqdyDVCgOv\n3vdU9/f4Ztb/fKpvllDdgi/pbElrJa2XdG299lspScdIWiZptaRnJc0r20dLekTSuvL3qP6utZOk\nJkkrJS0plydKWl4e4x9LGtLfNXaSdISkRZLWSGqTNL1Rj62kK8u/gV9KulvSsEY+tr1Rl+BLagLm\nA+cAU4GLJU2tx74PQgdwdURMBU4DvlLWeC2wNCImAUvL5UYxD2jrtnw9cFNEHAfsAOb0S1X7dwvw\nUERMAU6kqLvhjq2kccDlQEtEnAA0ARfR2Mf24EVEzX+A6cDD3ZavA66rx777UPMDwFnAWmBs2TYW\nWNvftZW1jKcIy5nAEkAUE0wG7++Y93OthwPPU15T6tbecMcWGAdsAkZT3FR2CfBnjXpse/tTr1P9\nzoPZaXPZ1pAkTQCmAcuBMRGxtVy1DRjTT2Xt62bgGui6cf2RwGsR0VEuN9Ixngi8AtxZvjW5XdJI\nGvDYRsQW4AbgRWAr8DqwgsY9tr3ii3v7kHQIcC9wRUS80X1dFC/3/f5vEEmfBbZHxEC5P/Vg4GTg\ntoiYRjFte6/T+gY6tqOAmRQvVkcDI4Gz+7WoGqhX8LcAx3RbHl+2NRRJzRShvysi7iubX5Y0tlw/\nFtjeX/V1czpwnqQXgHsoTvdvAY6QNLjs00jHeDOwOSKWl8uLKF4IGvHYfgp4PiJeiYh24D6K492o\nx7ZX6hX8p4BJ5ZXRIRQXSxbXad8VkSTgDqAtIm7stmoxMKt8PIvivX+/iojrImJ8REygOJaPRcQl\nwDLggrJbQ9QKEBHbgE2Sji+bZgCracBjS3GKf5qkEeXfRGetDXlse62OF03OBZ4DfgX8dX9f3NhP\nfR+nONX8X+CZ8udcivfOS4F1wKPA6P6udZ+6/wRYUj4+FngSWA/8GzC0v+vrVudJQGt5fO8HRjXq\nsQX+FlgD/BL4ITC0kY9tb348c88sIV/cM0vIwTdLyME3S8jBN0vIwTdLyME3S8jBN0vIwTdL6P8A\nNfu7JH4DyeoAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0224f60>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAAD0RJREFUeJzt3X2QVfV9x/H3h93l0ScwkSCYghGk\n1FYxG5WaxFZiqiYjZsYYrWMYSoY/kig+tEbTziRNO53YcXxIg7aMxpDEalJ0lKFWqyvpNImii9gS\neRCCD4AgxqAmRJGFb/84Z5eFAHvZvffu3X4/r5mde8/TPd/5zX7u79xzf/ccRQRmlsug/i7AzOrP\nwTdLyME3S8jBN0vIwTdLyME3S8jBN0uoT8GXdK6kNZLWSbq+WkWZWW2ptwN4JDUBLwDnABuBZ4BL\nI2Jl9cozs1po7sO2pwHrImI9gKT7gBnAAYM/WENiKCP6sEszO5h32c57sUM9rdeX4I8FNnSb3gic\nvu9KkuYAcwCGMpzTNb0PuzSzg1kabRWtV/OTexExPyJaI6K1hSG13p2ZVaAvwd8EHNdtelw5z8wa\nXF+C/wwwUdIESYOBS4BF1SnLzGqp15/xI6JD0peBR4Em4DsR8XzVKjOzmunLyT0i4mHg4SrVYmZ1\n4pF7Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJ\nOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5\n+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgn1GHxJx0laImmlpOclzS3nj5L0mKS15ePI2pdrZtVQSY/f\nAVwbEVOAM4AvSZoCXA+0RcREoK2cNrMBoMfgR8TmiHi2fP5rYBUwFpgBLChXWwBcWKsizay6Dukz\nvqTxwFRgKTA6IjaXi7YAo6tamZnVTMXBl3QYcD9wVUS83X1ZRAQQB9hujqR2Se072dGnYs2sOioK\nvqQWitDfExEPlLNfkzSmXD4G2Lq/bSNifkS0RkRrC0OqUbOZ9VElZ/UF3AWsioibuy1aBMwsn88E\nHqp+eWZWC80VrHMmcDmwQtJz5byvAt8EfiRpNvAycHFtSjSzausx+BHxE0AHWDy9uuWYWT145J5Z\nQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllC\nDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIO\nvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIVB19Sk6TlkhaX0xMkLZW0TtIPJQ2uXZlmVk2H0uPPBVZ1\nm74RuCUiTgC2AbOrWZiZ1U5FwZc0DvgUcGc5LeBsYGG5ygLgwloUaGbVV2mPfytwHbC7nD4aeDMi\nOsrpjcDY/W0oaY6kdkntO9nRp2LNrDp6DL6kTwNbI2JZb3YQEfMjojUiWlsY0puXMLMqa65gnTOB\nCySdDwwFjgBuA46S1Fz2+uOATbUr08yqqccePyJuiIhxETEeuAR4IiIuA5YAF5WrzQQeqlmVZlZV\nffke/yvANZLWUXzmv6s6JZlZrVVyqN8lIn4M/Lh8vh44rfolmVmteeSeWUIOvllCDr5ZQg6+WUIO\nvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+\nWUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5Z\nQg6+WUIVBV/SUZIWSlotaZWkaZJGSXpM0trycWStizWz6qi0x78NeCQiJgMnA6uA64G2iJgItJXT\nZjYA9Bh8SUcCHwfuAoiI9yLiTWAGsKBcbQFwYa2KNLPqqqTHnwC8DtwtabmkOyWNAEZHxOZynS3A\n6FoVaWbVVUnwm4FTgTsiYiqwnX0O6yMigNjfxpLmSGqX1L6THX2t18yqoJLgbwQ2RsTScnohxRvB\na5LGAJSPW/e3cUTMj4jWiGhtYUg1ajazPuox+BGxBdgg6cRy1nRgJbAImFnOmwk8VJMKzazqmitc\n7wrgHkmDgfXALIo3jR9Jmg28DFxcmxLNrNoqCn5EPAe07mfR9OqWY2b14JF7Zgk5+GYJOfhmCTn4\nZgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhm\nCTn4Zgk5+GYJVXqVXbND1nH2hwHY/MU9N1L5n2nFXddOfrK4Mvux8wYD0LTk2TpXl5t7fLOE3ONb\n1e0+ayoA3/rOtwE4oWXPv9nu8nH5tLsBWNO6C4C/Gn9G/Qo09/hmGbnHt6rZ+cninivX3f59ACa1\nFJ/fd3f187B+504A3tpd3Edxank7xR3nfQSAYUtWdK27+913a1twYu7xzRJyj2+91nTEEQBs//hk\nAK6+5V8B+NNhvynX+N1+5bvb/hiAttunAfDTr38LgMfu/GcApvzgy13rHv+VJ6tftAHu8c1ScvDN\nEvKhvvXaxu+NBeCZj8yreJtvHPMMAI8cVhzyz3rpkwAsGP84AEdMeaOaJdoBuMc3S8g9vh2SzmG4\nAPeeUgzQGcTgvdaZ9fJ0ANof/30AVsz+dteyJe8MBeCY9ncAWLetODHY8g9LitdSLaq2fbnHN0tI\nEVG3nR2hUXG6ptdtf1Y9ncNwb11we9e87kNxAS5Y/RkAmi7aDsCvPnUiAG+ctKcbnzRvAwAdGzbu\nte3iTcsA2Lzrna55fzHzyuL1/AOeii2NNt6OX/V43OQe3yyhij7jS7oa+AIQwApgFjAGuA84GlgG\nXB4R79WoTusn+vAfAPDLa4qeuHMYLsCy8te2T/xmCgBv3HccAEdvKwbeHPmDp4rHbq/X0cP+RjcN\n6Xr+xlW/BeCYJb0q3Q6ixx5f0ljgSqA1Ik4CmoBLgBuBWyLiBGAbMLuWhZpZ9VR6Vr8ZGCZpJzAc\n2AycDfx5uXwB8HXgjmoXaP1j0PDhAHT849sAPDX5AQBe7NhzUHfNV68FYOR/vwLAMSO2ArCrSjWc\nNuZlAF6q0uvZHj32+BGxCbgJeIUi8G9RHNq/GRGdR24bgbH7217SHEntktp3smN/q5hZnVVyqD8S\nmAFMAI4FRgDnVrqDiJgfEa0R0drCkJ43MLOaq+RQ/xPAixHxOoCkB4AzgaMkNZe9/jhgU+3KtHp7\n56zipN6jk2/fa/4X5l7d9fzwB4uTdz2dsLPGU8nXea8AZ0gaLknAdGAlsAS4qFxnJvBQbUo0s2rr\nscePiKWSFgLPUry5LwfmA/8O3Cfp78t5d9WyUKuvP/q75wAYVPYNncNwhz34dE3216ImAHZ2G0/W\npPoNLsumorP6EfE14Gv7zF4PnFb1isys5vwjHdvLm5cXV8b5m9E3AbC7/AHOsv8sBul8kJ/VZL87\nY1e5vz3X53tkVbHPiXjIbrV5yK5ZQu7xbS8dw4rHIwcVPf2T7xZfwR7/vVeL5VXaT+cAodU3nVTO\nKX6kc9n687rWmTz3RaB6A4JsD/f4Zgm5x7eDemPXYQB0rH+pKq/X2dOv+eYfArB6RnGRjv/4bfFT\nnlfnndC17uHbnqrKPu13ucc3S8g9vh3UX/70swBMKj+D90bnRTwAtpY/713VWvT001d8DoAR564H\n4HDcy9eDe3yzhBx8s4R8qG97K6/W1jlU97aP3gvAPCYd8ku9/I1iMND9n7+5a17nFXxOfXomAMd+\nZmWvS7Xec49vlpB7fNtb+buYzqGzZw0r7mxz1XeL6+l/6O49Q2pbtvwagNfOej8Aoz5XXDn3ig+2\nAXDe8OKE4KLto7u2+fyK4lIO7/uXETUp3yrjHt8sIff4dlBDVfyLrDqnuI31Tz42tGvZ2h0fAGDW\nkS/td9u5r34MgEd+dkrXvIlz/XVdI3CPb5aQ76Rje2ma9CEAJt1bXOH2xg88udfyQd36iu4/oQVY\nvqNYdul/zSleY1bvB/1Y7/hOOmZ2QP6Mb3vZ9cIvAFj72fEATLniCgBWXvxPB9xm8sNfBODE24s7\n30xa7p6+0bnHN0vIwTdLyCf3zP4f8ck9MzsgB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQff\nLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3yyhul6IQ9LrwHbgl3Xbad+8j4FTKwysegdSrTBw\n6v29iHh/TyvVNfgAktojorWuO+2lgVQrDKx6B1KtMPDq7YkP9c0ScvDNEuqP4M/vh3321kCqFQZW\nvQOpVhh49R5U3T/jm1n/86G+WUJ1C76kcyWtkbRO0vX12m+lJB0naYmklZKelzS3nD9K0mOS1paP\nI/u71k6SmiQtl7S4nJ4gaWnZxj+UNLi/a+wk6ShJCyWtlrRK0rRGbVtJV5f/Az+XdK+koY3ctr1R\nl+BLagLmAecBU4BLJU2px74PQQdwbURMAc4AvlTWeD3QFhETgbZyulHMBVZ1m74RuCUiTgC2AbP7\npar9uw14JCImAydT1N1wbStpLHAl0BoRJwFNwCU0dtseuoio+R8wDXi02/QNwA312Hcfan4IOAdY\nA4wp540B1vR3bWUt4yjCcjawGBDFAJPm/bV5P9d6JPAi5TmlbvMbrm2BscAGYBTFTWUXA3/WqG3b\n2796Hep3NmanjeW8hiRpPDAVWAqMjojN5aItwOh+KmtftwLXQddN6o8G3oyIjnK6kdp4AvA6cHf5\n0eROSSNowLaNiE3ATcArwGbgLWAZjdu2veKTe/uQdBhwP3BVRLzdfVkUb/f9/jWIpE8DWyNioNyP\nuhk4FbgjIqZSDNve67C+gdp2JDCD4s3qWGAEcG6/FlUD9Qr+JuC4btPjynkNRVILRejviYgHytmv\nSRpTLh8DbO2v+ro5E7hA0kvAfRSH+7cBR0lqLtdppDbeCGyMiKXl9EKKN4JGbNtPAC9GxOsRsRN4\ngKK9G7Vte6VewX8GmFieGR1McbJkUZ32XRFJAu4CVkXEzd0WLQJmls9nUnz271cRcUNEjIuI8RRt\n+UREXAYsAS4qV2uIWgEiYguwQdKJ5azpwEoasG0pDvHPkDS8/J/orLUh27bX6njS5HzgBeAXwF/3\n98mN/dT3UYpDzf8Fniv/zqf47NwGrAUeB0b1d6371P0nwOLy+fHA08A64N+AIf1dX7c6TwHay/Z9\nEBjZqG0L/C2wGvg58H1gSCO3bW/+PHLPLCGf3DNLyME3S8jBN0vIwTdLyME3S8jBN0vIwTdLyME3\nS+j/AEiouySYoZT+AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a020deb8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\njmEeAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a01b7b00>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nZgn9H/vnuSw9OcWMAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a01607b8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAADz5JREFUeJzt3X2QVfV9x/H3x93l0ScwkSCYghGk\n1FYxG5WaxFZiqiYjZsYYrWMYSoY/kig+tEbTziRNO53YcXxIg7aMxpDEalJ0lLFWq0g6TaLoIrZE\nHoTgAyCIMagJUWTh2z/O2etCF/eye+/du/l+XjM7ex5+557vnNnP/s4993fPUURgZrkcNNAFmFnj\nOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJ9Sv4ks6StFbSeknX1KooM6sv9XUAj6QW4DngTGAT8BRw\nUUSsql15ZlYPrf3Y9mRgfURsAJB0NzAT2G/wh2hoDGNkP3ZpZu/lbXbwTuxUb+36E/xxwMZu85uA\nU/ZtJGkuMBdgGCM4RTP6sUszey/LYklV7ep+cS8iFkREe0S0tzG03rszsyr0J/ibgaO7zY8vl5lZ\nk+tP8J8CJkmaKGkIcCGwuDZlmVk99fk9fkR0Svoy8DDQAnwnIp6tWWVmVjf9ubhHRDwIPFijWsys\nQTxyzywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98s\nIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywh\nB98sIQffLCEH3ywhB98sIQffLCEH3yyhXoMv6WhJSyWtkvSspHnl8tGSHpG0rvw9qv7lmlktVNPj\ndwJXRcRU4FTgS5KmAtcASyJiErCknDezQaDX4EfEloh4upz+NbAaGAfMBBaWzRYC59WrSDOrrQN6\njy9pAjANWAaMiYgt5aqtwJiaVmZmdVN18CUdDNwDXB4Rb3ZfFxEBxH62myupQ1LHLnb2q1gzq42q\ngi+pjSL0d0bEveXiVySNLdePBbb1tG1ELIiI9ohob2NoLWo2s36q5qq+gNuB1RFxQ7dVi4FZ5fQs\n4P7al2dm9dBaRZvTgEuAlZKeKZd9Ffgm8CNJc4AXgQvqU6KZ1VqvwY+InwDaz+oZtS3HzBrBI/fM\nEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0S\ncvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy\n8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRKqOviSWiStkPRAOT9R0jJJ6yX9UNKQ+pVpZrV0ID3+PGB1\nt/nrgBsj4lhgOzCnloWZWf1UFXxJ44FPAbeV8wLOABaVTRYC59WjQDOrvWp7/JuAq4E95fwRwOsR\n0VnObwLG9bShpLmSOiR17GJnv4o1s9roNfiSPg1si4jlfdlBRCyIiPaIaG9jaF9ewsxqrLWKNqcB\n50o6BxgGHArcDBwuqbXs9ccDm+tXppnVUq89fkRcGxHjI2ICcCHwWERcDCwFzi+bzQLur1uVZlZT\n/fkc/yvAlZLWU7znv702JZlZvVVzql8RET8GflxObwBOrn1JZlZvHrlnlpCDb5aQg2+WkINvlpCD\nb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINv\nlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+W\nkINvllBVwZd0uKRFktZIWi1puqTRkh6RtK78ParexZpZbVTb498MPBQRU4ATgNXANcCSiJgELCnn\nzWwQ6DX4kg4DPg7cDhAR70TE68BMYGHZbCFwXr2KNLPaqqbHnwi8CtwhaYWk2ySNBMZExJayzVZg\nTL2KNLPaqib4rcBJwK0RMQ3YwT6n9RERQPS0saS5kjokdexiZ3/rNbMaqCb4m4BNEbGsnF9E8Y/g\nFUljAcrf23raOCIWRER7RLS3MbQWNZtZP/Ua/IjYCmyUdFy5aAawClgMzCqXzQLur0uFZlZzrVW2\nuxS4U9IQYAMwm+Kfxo8kzQFeBC6oT4lmVmtVBT8ingHae1g1o7blmFkjeOSeWUIOvllCDr5ZQg6+\nWUIOvllCDr5ZQg6+WUIOvllCDr5ZQtUO2TWrWucZHwZgyxeLb2P+z/SFlXUnPF58veOo+UMAaFn6\ndIOrM3CPb5aSe3yrmT2nTwPgW9/5NgDHthV/Xnu6tVkx/Q4A1rbvBuCvJpzauAKtwj2+WULu8a3f\ndn2y+OLm1bd8H4DJbcX79z1lX79h165K2zf2FDdjmVbek2Xn2R8BYPjSlcU2b79d/4LNPb5ZRg6+\nWUI+1bcD0nLooZXpHR+fAsAVN/4rAH86/Dflmr37k+9u/+PK9JJbpgPw069/C4BHbvtnAKb+4MsA\nHPOVx2tftP0/7vHNEnKPbwdk0/fGVaaf+sj8qrb5xpFPVaYfOrjo/We/8EkAFk54FIBDp75WqxKt\nCu7xzRJyj29V6RqGe9eJ364sO4ghe7WZ/WJx79WOR38fgJVzirZL3xpWaXNkx1sArN9eXB9o+4el\nxWupHlXb/rjHN0tIxdOvGuNQjY5T5DtyDyZdw3BvWngL8O4w3O7OXfMZAFrO3wHArz5VPHvlteOL\nbnzy/I2Vtp0bN+217QOblwOwZXdxJvAXsy6rrPMXeA7csljCm/GrXs+f3OObJeT3+NYjffgPAPjl\nlUVP3DUMd3m3554+9pupALx299EAHLG9+Az+sB88Ufwu23VWsb8xLcUY3tcu/21l2ZFL+1S6VcE9\nvllC7vFtLweNGAFA5z++CcATU+4F4PnOdwC48qtXVdqO+u+XADhyZPGg5N012P/JY1+sTL9Qg9ez\nnrnHN0vIwTdLyKf6tpe3Ti8u6j085Za9ln9h3hUAHHLfE5Vl1Vy0s+bkHt8sIff4tpc/+rtnADio\n7BO6huEOv+/JuuyvTS0A7CrHkbWocQPKMnOPb5aQe3zj9UumV6b/Zsz1AOwpv4Cz/D+LQTof5Gd1\n2feu2F3ur7g/30Orp1bWTcJDduvFPb5ZQlX1+JKuAL4ABLASmA2MBe4GjgCWA5dExDt1qtPqqHP4\nu9OHHVT09I+/XQyhPeZ7LxdtarSvrgFCa64/vlxSfEnn4g1nAzBl3vOVtrUYEGQ967XHlzQOuAxo\nj4jjgRbgQuA64MaIOBbYDsypZ6FmVjvVvsdvBYZL2gWMALYAZwB/Xq5fCHwduLXWBdrAeG33wQB0\nbnih36/V1csDrP3mHwKwZmZxk47/+G3xVZ6X5x8LwCHbn8Dqr9cePyI2A9cDL1EE/g2K87PXI6Lr\nDHATMK6n7SXNldQhqWMXO3tqYmYNVs2p/ihgJjAROAoYCZxV7Q4iYkFEtEdEextD+1yomdVONaf6\nnwCej4hXASTdC5wGHC6ptez1xwOb61emNdpf/vSzAEwuL771Rdfde7aV3+kHWN1enOLPWPk5AEae\ntQGAQ/ApfiNV83HeS8CpkkZIEjADWAUsBc4v28wC7q9PiWZWa732+BGxTNIi4GmKT3VWAAuAfwfu\nlvT35bLb61mo1VG3O7R1DdW9+aN3ATCfyQf8ci9+oxgQdM/nbwDevXsPwElPzgLgqM+s6lOpVhtV\nXdWPiK8BX9tn8Qbg5JpXZGZ15yG7VgzLKnUNnT19ePFkm8u/W9xP/0N3FMvbtv660vaV098PwOjP\nFXfOvfSDSwA4e0RxXWDxjjEAfH7lu9eC3/cvI2tevh04D9k1S8g9vvVomIo/jdVnFk+z/cnHiqfh\nrNv5gUqb2Ye90OO2817+GAAP/exEACbN8xX7ZuMe3ywhP0nHaJn8ocr05LuKu9xe94G9n1PfdbW/\n6xpAdyt2Fusu+q+5xWvM7vtn/9Y/fpKOme2X3+Mbu5/7RWV63WcnADD10ksBWHXBP+13uykPfhGA\n424pnn4zeYV7+sHCPb5ZQg6+WUK+uGf2O8QX98xsvxx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8\ns4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyz\nhBx8s4QcfLOEHHyzhBx8s4QcfLOEGvrsPEmvAjuAXzZsp/3zPgZPrTC46h1MtcLgqff3IuL9vTVq\naPABJHVERHtDd9pHg6lWGFz1DqZaYfDV2xuf6psl5OCbJTQQwV8wAPvsq8FUKwyuegdTrTD46n1P\nDX+Pb2YDz6f6Zgk1LPiSzpK0VtJ6Sdc0ar/VknS0pKWSVkl6VtK8cvloSY9IWlf+HjXQtXaR1CJp\nhaQHyvmJkpaVx/iHkoYMdI1dJB0uaZGkNZJWS5rerMdW0hXl38DPJd0laVgzH9u+aEjwJbUA84Gz\nganARZKmNmLfB6ATuCoipgKnAl8qa7wGWBIRk4Al5XyzmAes7jZ/HXBjRBwLbAfmDEhVPbsZeCgi\npgAnUNTddMdW0jjgMqA9Io4HWoALae5je+Aiou4/wHTg4W7z1wLXNmLf/aj5fuBMYC0wtlw2Flg7\n0LWVtYynCMsZwAOAKAaYtPZ0zAe41sOA5ymvKXVb3nTHFhgHbARGA63lsf2zZj22ff1p1Kl+18Hs\nsqlc1pQkTQCmAcuAMRGxpVy1FRgzQGXt6ybgamBPOX8E8HpEdJbzzXSMJwKvAneUb01ukzSSJjy2\nEbEZuB54CdgCvAEsp3mPbZ/44t4+JB0M3ANcHhFvdl8Xxb/7Af8YRNKngW0RsXyga6lSK3AScGtE\nTKMYtr3XaX0THdtRwEyKf1ZHASOBswa0qDpoVPA3A0d3mx9fLmsqktooQn9nRNxbLn5F0thy/Vhg\n20DV181pwLmSXgDupjjdvxk4XFJr2aaZjvEmYFNELCvnF1H8I2jGY/sJ4PmIeDUidgH3UhzvZj22\nfdKo4D8FTCqvjA6huFiyuEH7rookAbcDqyPihm6rFgOzyulZFO/9B1REXBsR4yNiAsWxfCwiLgaW\nAueXzZqiVoCI2ApslHRcuWgGsIomPLYUp/inShpR/k101dqUx7bPGnjR5BzgOeAXwF8P9MWNHur7\nKMWp5v8Cz5Q/51C8d14CrAMeBUYPdK371P0nwAPl9DHAk8B64N+AoQNdX7c6TwQ6yuN7HzCqWY8t\n8LfAGuDnwPeBoc18bPvy45F7Zgn54p5ZQg6+WUIOvllCDr5ZQg6+WUIOvllCDr5ZQg6+WUL/B4us\nuyTb8QOEAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a01c0860>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/T8lG7skbqLqdQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a012b198>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nR+oykPYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a014fd30>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n5rskwUqJjQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a00f89e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a00d98d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJslGd4frDgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a00c3438>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAADy9JREFUeJzt3X2QVfV9x/H3x+UZn8BERDAFA0ip\nrWI2KjXGVGKKJiNmxhgdJ2EsHf5IovjQKtr+0aYzHe04PiTFtIzGkMSqqTrKOFariJ0mURTUFuVB\nCD4AghiDmhhFFr7945xdFwvsZffcu3f3+3nN7Ox5+J17vvODz/7OOfvbexURmFkuB/R2AWbWeA6+\nWUIOvllCDr5ZQg6+WUIOvllCDr5ZQj0KvqQZktZIWidpXlVFmVl9qbsTeCS1AC8BZwAbgWeACyJi\nZXXlmVk9DOjBsScC6yJiPYCku4CZwF6DP0iDYwjDe3BKM9uXD3iPD2O7umrXk+CPATZ0Wt8InPTx\nRpLmAHMAhjCMkzS9B6c0s31ZGotralf3h3sRsSAiWiOidSCD6306M6tBT4K/CTiq0/rYcpuZNbme\nBP8ZYKKk8ZIGAecDi6opy8zqqdv3+BHRJuk7wCNAC/DDiHixssrMrG568nCPiHgIeKiiWsysQTxz\nzywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQff\nLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98s\nIQffLCEH3ywhB98sIQffLCEH3yyhLoMv6ShJSyStlPSipLnl9pGSHpW0tvw+ov7lmlkVahnx24Ar\nImIKcDLwbUlTgHnA4oiYCCwu182sD+gy+BGxOSKeLZd/C6wCxgAzgYVls4XAOfUq0syqtV/3+JLG\nAVOBpcCoiNhc7toCjKq0MjOrm5qDL+lA4F7g0oh4t/O+iAgg9nLcHEnLJC3bwfYeFWtm1agp+JIG\nUoT+joi4r9z8hqTR5f7RwNY9HRsRCyKiNSJaBzK4iprNrIdqeaov4DZgVUTc0GnXImBWuTwLeKD6\n8sysHgbU0OYU4BvACknPl9uuAa4FfiZpNvAqcF59SjSzqnUZ/Ij4OaC97J5ebTlm1gieuWeWkINv\nlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+W\nkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQ\ng2+WkINvlpCDb5aQg2+WkINvllDNwZfUIuk5SQ+W6+MlLZW0TtLdkgbVr0wzq9L+jPhzgVWd1q8D\nboyICcA2YHaVhZlZ/dQUfEljgS8Dt5brAk4H7imbLATOqUeBZla9Wkf8m4ArgV3l+mHA2xHRVq5v\nBMbs6UBJcyQtk7RsB9t7VKyZVaPL4Ev6CrA1IpZ35wQRsSAiWiOidSCDu/MSZlaxATW0OQU4W9JZ\nwBDgYOBm4FBJA8pRfyywqX5lmlmVuhzxI+LqiBgbEeOA84HHI+JCYAlwbtlsFvBA3ao0s0r15Pf4\nVwGXS1pHcc9/WzUlmVm91XKp3yEingCeKJfXAydWX5KZ1Ztn7pkl5OCbJeTgmyXk4Jsl5OCbJeTg\nmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCb\nJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJbRfH6Fl+bSd/hkANn9rOwD/M21hx77jnpwFwJHzBwHQ\nsuTZBldn3eUR3ywhj/i2R7tOmwrA9374zwBMGFj8V9nVqc1z024HYE3rTgD+etzJjSvQesQjvllC\nHvFtNzu+1ArAlbf8BIBJA4v7913lWL9+x46Otu/sGgzA1OIb28/8LABDl6wojvngg/oXbN3iEd8s\nIY/4ibUcfDAA731+cse2y278NwD+bOjvyi27jw0/2vanHcuLb5kGwC/+7nsAPHrrvwAw5affAeDo\nq56svmirhEd8s4RqCr6kQyXdI2m1pFWSpkkaKelRSWvL7yPqXayZVaPWS/2bgYcj4lxJg4BhwDXA\n4oi4VtI8YB5wVZ3qtDrY+OMxADzz2fk1H/Pdw5/pWH74wOKy/6JXvgTAwnGPAXDwlLeqKtHqpMsR\nX9IhwOeB2wAi4sOIeBuYCbRP41oInFOvIs2sWrWM+OOBN4HbJR0HLAfmAqMiYnPZZgswqj4lWtXa\np+HeeXwxOecABv2/Nhe9Oh2AZY/9IQArZhdtl7w/pKPN4cveB2DdtuLh4MB/XFK8nupRtVWplnv8\nAcAJwA8iYirwHsVlfYeICCD2dLCkOZKWSVq2g+09rdfMKqAis/toIB0BPBUR48r1UymCPwH4QkRs\nljQaeCIijtnXax2skXGSpldSuO2/9mm4Ny28BfhoGm5nZ6/+KgAt574HwG++XPyTvnVsMYxPmr+h\no23bho27HfvgpuUAbN5ZXAn8xaxLitfyH+80zNJYzLvxmy6vuboc8SNiC7BBUnuopwMrgUXArHLb\nLOCBbtZqZg1W61P9i4E7yif664GLKH5o/EzSbOBV4Lz6lGg9pc/8EQC/vrwYidun4S4v77we/92U\njrZv3XUUAIdtKybfHPLTp4rv5f62Gs43qqWYw/vWpb8H4PAl3S7d6qSm4EfE80DrHnb5ut2sD/KU\n3X7sgGHDAGj7p3cBeGryfQC83PYhAJdfcwUAI/77tY5jDh++FYCdFZz/xNGvAvBKBa9l1fKUXbOE\nHHyzhHyp34+9f1rxUO+Rybfstv0v514GwEH3Fw/uanlgZ/2LR3yzhDzi92N/8g/PA3BA+fO9fRru\n0Pufrsv5BqoFgB3lnLAW7XtymPUej/hmCXnE72fe/sa0juW/HXU9ALvKP8JZ/p/FRJ1P8cu6nHtH\n7CzPV7w/38OrivNNxFN2m41HfLOEPOL3M21DP1o+5IBipH/yg2IK7dE/fr1oU9G52icIrb7+2HJL\n8Uc6F64/E4DJc18GqpkMZNXyiG+WkEf8BN7aeSAAbetf6fFrtY/yAGuu/WMAVs8s3qTjP35f/CnP\n6/MnAHDQtqd6fD6rD4/4Zgl5xE/gr37xNQAmlffg3dH+Jh5byz/tBVjVWoz001d8HYDhM9YDcBAe\n6ZudR3yzhBx8s4R8qd/fdHq3tfapujd/7k4A5jNpv1/u1e8WE4Lu/eYNwEfv3gNwwtPFO68d+dWV\n3SrVeo9HfLOEPOL3N53+LqZ96uxpQ4tPtrn0R8X76X/69mL7wC2/BeCN0z7ZcczIrxfvnHvxpxYD\ncOaw4oHgoveKj0345ooZHW0/8a/DKy/fGsMjvllCHvETGKLin3nVGcXHWP/81OLTcNZuPwKAiw55\nZa/Hzn39VAAe/uXxAEyc61/V9Qce8c0S6vKTdKrkT9Kpv5ZJn+5YnnRn8S631x3x5G5t2p/2tz8D\n6Oy57cW+C/5rTvEaF3V/0o81XmWfpGNm/Y/v8fuZnS/9qmN57dfGATDl4osBWHne9/d4zOSHvtWx\nfMwtxaffTHrOI31/5hHfLCHf45v1I77HN7O9cvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEqop\n+JIuk/SipBck3SlpiKTxkpZKWifpbkmDun4lM2sGXQZf0hjgEqA1Io4FWoDzgeuAGyNiArANmF3P\nQs2sOrVe6g8AhkoaAAwDNgOnA/eU+xcC51RfnpnVQ5fBj4hNwPXAaxSBf4fi0xHfjoj2z1/cCIzZ\n0/GS5khaJmnZDrZXU7WZ9Ugtl/ojgJnAeOBIYDgwY58HdRIRCyKiNSJaBzK424WaWXVqudT/IvBy\nRLwZETuA+4BTgEPLS3+AscCmOtVoZhWrJfivASdLGiZJwHRgJbAEOLdsMwt4oD4lmlnVarnHX0rx\nEO9ZYEV5zALgKuBySeuAw4Db6linmVXIb8Rh1o/4jTjMbK8cfLOEHHyzhBx8s4QcfLOEHHyzhBx8\ns4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyz\nhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOE\nHHyzhBx8s4QUEY07mfQm8B7w64adtGc+Qd+pFfpWvX2pVug79f5BRHyyq0YNDT6ApGUR0drQk3ZT\nX6oV+la9falW6Hv1dsWX+mYJOfhmCfVG8Bf0wjm7qy/VCn2r3r5UK/S9evep4ff4Ztb7fKlvllDD\ngi9phqQ1ktZJmteo89ZK0lGSlkhaKelFSXPL7SMlPSppbfl9RG/X2k5Si6TnJD1Yro+XtLTs47sl\nDertGttJOlTSPZJWS1olaVqz9q2ky8r/Ay9IulPSkGbu2+5oSPAltQDzgTOBKcAFkqY04tz7oQ24\nIiKmACcD3y5rnAcsjoiJwOJyvVnMBVZ1Wr8OuDEiJgDbgNm9UtWe3Qw8HBGTgeMo6m66vpU0BrgE\naI2IY4EW4Hyau2/3X0TU/QuYBjzSaf1q4OpGnLsHNT8AnAGsAUaX20YDa3q7trKWsRRhOR14EBDF\nBJMBe+rzXq71EOBlymdKnbY3Xd8CY4ANwEhgQNm3f96sfdvdr0Zd6rd3ZruN5bamJGkcMBVYCoyK\niM3lri3AqF4q6+NuAq4EdpXrhwFvR0Rbud5MfTweeBO4vbw1uVXScJqwbyNiE3A98BqwGXgHWE7z\n9m23+OHex0g6ELgXuDQi3u28L4of973+axBJXwG2RsTy3q6lRgOAE4AfRMRUimnbu13WN1HfjgBm\nUvywOhIYDszo1aLqoFHB3wQc1Wl9bLmtqUgaSBH6OyLivnLzG5JGl/tHA1t7q75OTgHOlvQKcBfF\n5f7NwKGSBpRtmqmPNwIbI2JpuX4PxQ+CZuzbLwIvR8SbEbEDuI+iv5u1b7ulUcF/BphYPhkdRPGw\nZFGDzl0TSQJuA1ZFxA2ddi0CZpXLsyju/XtVRFwdEWMjYhxFXz4eERcCS4Bzy2ZNUStARGwBNkg6\nptw0HVhJE/YtxSX+yZKGlf8n2mttyr7ttgY+NDkLeAn4FfA3vf1wYw/1fY7iUvN/gefLr7Mo7p0X\nA2uBx4CRvV3rx+r+AvBguXw08DSwDvh3YHBv19epzuOBZWX/3g+MaNa+Bf4eWA28APwEGNzMfdud\nL8/cM0vID/fMEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRL6P15vs25jioKgAAAAAElFTkSu\nQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a006d080>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0090cf8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAADzVJREFUeJzt3X+QVeV9x/H3x11+qwgmEgRTMIKU\n2iq6UalJbCWmaDJiZozRcRLG0uGPJIo/WkXbP9p0pqMdR2NaTMpoDEmsmqKjjGO1iqTTJIouaovy\nQwj+AAQxBjUhirvw7R/n7LIquJfde/fey/fzmtnZe855zj3feWY/+5xz7nPvVURgZrkcVO8CzGzg\nOfhmCTn4Zgk5+GYJOfhmCTn4Zgk5+GYJ9Sv4kmZKWitpvaT51SrKzGpLfZ3AI6kFeAE4E9gEPAVc\nGBGrqleemdVCaz/2PRlYHxEbACTdBcwC9hn8wRoSQxnRj0Oa2Ud5lx28FzvVW7v+BH8csLHH8ibg\nlA82kjQXmAswlOGcohn9OKSZfZTlsbSidjW/uRcRCyOiLSLaBjGk1oczswr0J/ibgaN6LI8v15lZ\ng+tP8J8CJkmaKGkwcAGwpDplmVkt9fkaPyI6JX0LeBhoAX4QEc9XrTIzq5n+3NwjIh4EHqxSLWY2\nQDxzzywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98s\nIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywh\nB98sIQffLCEH3ywhB98sIQffLCEH3yyhXoMv6ShJyyStkvS8pHnl+tGSHpG0rvw9qvblmlk1VDLi\ndwJXRsRU4FTgm5KmAvOBpRExCVhaLptZE+g1+BGxJSKeLh//FlgNjANmAYvKZouAc2tVpJlV135d\n40uaAEwDlgNjImJLuWkrMKaqlZlZzVQcfEkHA/cAl0XE2z23RUQAsY/95kpql9Tewc5+FWtm1VFR\n8CUNogj9HRFxb7n6NUljy+1jgW172zciFkZEW0S0DWJINWo2s36q5K6+gNuA1RFxY49NS4DZ5ePZ\nwP3VL8/MaqG1gjanAV8DVkp6tlx3LXAd8FNJc4CXgfNrU6KZVVuvwY+InwPax+YZ1S3HzAaCZ+6Z\nJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl\n5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk\n4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyVUcfAltUh6RtID5fJEScslrZd0t6TBtSvTzKppf0b8ecDq\nHsvXAzdFxDHAdmBONQszs9qpKPiSxgNfBG4tlwWcASwumywCzq1FgWZWfZWO+N8BrgJ2l8uHA29G\nRGe5vAkYt7cdJc2V1C6pvYOd/SrWzKqj1+BL+hKwLSJW9OUAEbEwItoiom0QQ/ryFGZWZa0VtDkN\nOEfS2cBQ4FDgZuAwSa3lqD8e2Fy7Ms2smnod8SPimogYHxETgAuAxyLiImAZcF7ZbDZwf82qNLOq\n6s/r+FcDV0haT3HNf1t1SjKzWqvkVL9bRPwM+Fn5eANwcvVLMrNa88w9s4QcfLOEHHyzhBx8s4Qc\nfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8s4QcfLOEHHyzhBx8\ns4T26zP3zBpR5xkndT/e8o3iS1v+d/oiAI5/fDYARy4ovtqxZdnTA1xdY/KIb5aQg2+WkE/1rWnt\nPn0aAN/9wb92rztmUPEn3fUlj89Mvx2AtW27APibCacOXIENzCO+WUIe8a3pdHyhDYCrbvkxAJMH\nDe7etrsc6zd0dADw1u7ii1qnld/XuvOsTwMwbNnKPfu8+25tC25AHvHNEvKIbw2v5dBDAdjxuSkA\nXH7TvwPw58N+V7b48Pj1w+1/CsDSW6YD8Iu//y4Aj9z6fQCm/uRb3W2Pvvrx6hfd4DzimyXkEd8a\n3qYfjQPgqU8vqHifbx/xFAAPHVyM/Be/9AUAFk14FIBDp75RzRKbjkd8s4Q84lvD6pqKe+cJxev0\nBzH4fdsvfnkGAO2P/mH3upVzirbL3hkKwBHt7wCwfntxf2DQPy0rnku1qro5eMQ3S0gRMWAHO1Sj\n4xTNGLDjWfPpmo0H8J1FtwB7ZuN1OWfNlwFoOW8HAL/54rHd2944rhjKJy/YCEDnxk3v2/eBzSsA\n2LLrne51fzn70uL5DoA38CyPpbwdv+n1fMYjvllCFQVf0mGSFktaI2m1pOmSRkt6RNK68veoWhdr\nZtVR6c29m4GHIuI8SYOB4cC1wNKIuE7SfGA+cHWN6rQDnE76IwB+fcWeU/CuqbgrirfY89jvpgLw\nxl1HAXD49mLizcifPNG9z8jyd2cvxxvTMqT78RuX/R6AI5b1qfSm1OuIL2kk8DngNoCIeC8i3gRm\nAYvKZouAc2tVpJlVVyUj/kTgdeB2SccDK4B5wJiI2FK22QqMqU2JdiA7aPhwADr/+W0Anphyb/e2\nFzvfA+CKa68EYNT/vALAESO2AbCrSjWcPPZlAF6q0vM1g0qu8VuBE4HvRcQ0YAfFaX23KF4a2OvL\nA5LmSmqX1N7Bzv7Wa2ZVUMmIvwnYFBHLy+XFFMF/TdLYiNgiaSywbW87R8RCYCEUL+dVoWY7gLxz\nenFt//CUWz607a/mXQ7AIfcV1/C9Xbdb5Xod8SNiK7BRUteLpTOAVcASYHa5bjZwf00qNLOqq/Su\n/iXAHeUd/Q3AxRT/NH4qaQ7wMnB+bUq0A9mf/OOzABxUjkFd03ABht33ZNWPN0gtAHT0OPdsUb4T\n0YqCHxHPAm172eRpeGZNyG/Ssbp482vFB2T83ZgbANhdvgFnxX9N7W7zSX5Z9eN2xK7yeLu71z20\nujjmJJp/ym6lPGXXLCEH3ywhn+pbXXQOK36PPKg4xX/83WIK7dE/enVPmyocp2uC0JobjivXFO/O\nu2jDWd1tpsx7EajehKBm4BHfLCGP+NYQ3th1MACdG16qyvN1jfRrr/tjANbMKj6Z5z9/X7yN59UF\nx3S3PWT7E2TjEd8sIY/41hD++hdfAWByeQ3eV12f4LOtfHvv6rZipJ+x8qsAjJi5AYBDyDfK9+QR\n3ywhj/hWH+WnwnVN1b35M3cCsIDJ+/1UL397evfje75+I7DnQzxOfLJ4O8mRX17V51IPRB7xzRLy\niG/1Ub4vpmvq7OnDim+2ueyHJ3U3+dTtxbZBW38LwGunfxyA0V8tPjn3kk8uBeCs4XvuCyzZUXwe\nzNdXzgTgY/82oiblNzuP+GYJecS3hjBUxZ/i6jO/373u558tvg1n3c5PAHDxyJf2uu+8Vz/b/fih\nX54AwKR5ue/a98YjvllCDr5ZQv4KLauLlsmfAmDyncUn3F7/icc/1Kbrpb6e750HeGZnsf7C/55b\nPMfF/Zv0cyDxV2iZ2T755p7Vxa4XfgXAuq9MAGDqJZcAsOr8f9nnPlMe/AYAx95SfPPN5Gc80veV\nR3yzhHyNb3YA8TW+me2Tg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINv\nlpCDb5aQg2+WUEXBl3S5pOclPSfpTklDJU2UtFzSekl3Sxpc62LNrDp6Db6kccClQFtEHAe0ABcA\n1wM3RcQxwHZgTi0LNbPqqfRUvxUYJqkVGA5sAc4AFpfbFwHnVr88M6uFXoMfEZuBG4BXKAL/FrAC\neDMiOstmm4Bxe9tf0lxJ7ZLaO9hZnarNrF8qOdUfBcwCJgJHAiOAmZUeICIWRkRbRLQNYkifCzWz\n6qnkVP/zwIsR8XpEdAD3AqcBh5Wn/gDjgc01qtHMqqyS4L8CnCppuCQBM4BVwDLgvLLNbOD+2pRo\nZtVWyTX+coqbeE8DK8t9FgJXA1dIWg8cDtxWwzrNrIr88dpmBxB/vLaZ7ZODb5aQg2+WkINvlpCD\nb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINv\nlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+W\nkINvlpCDb5aQg2+WkINvlpAiYuAOJr0O7AB+PWAH7Z+P0Ty1QnPV20y1QvPU+wcR8fHeGg1o8AEk\ntUdE24AetI+aqVZornqbqVZovnp741N9s4QcfLOE6hH8hXU4Zl81U63QXPU2U63QfPV+pAG/xjez\n+vOpvllCAxZ8STMlrZW0XtL8gTpupSQdJWmZpFWSnpc0r1w/WtIjktaVv0fVu9YuklokPSPpgXJ5\noqTlZR/fLWlwvWvsIukwSYslrZG0WtL0Ru1bSZeXfwPPSbpT0tBG7tu+GJDgS2oBFgBnAVOBCyVN\nHYhj74dO4MqImAqcCnyzrHE+sDQiJgFLy+VGMQ9Y3WP5euCmiDgG2A7MqUtVe3cz8FBETAGOp6i7\n4fpW0jjgUqAtIo4DWoALaOy+3X8RUfMfYDrwcI/la4BrBuLY/aj5fuBMYC0wtlw3Flhb79rKWsZT\nhOUM4AFAFBNMWvfW53WudSTwIuU9pR7rG65vgXHARmA00Fr27V80at/29WegTvW7OrPLpnJdQ5I0\nAZgGLAfGRMSWctNWYEydyvqg7wBXAbvL5cOBNyOis1xupD6eCLwO3F5emtwqaQQN2LcRsRm4AXgF\n2AK8Baygcfu2T3xz7wMkHQzcA1wWEW/33BbFv/u6vwwi6UvAtohYUe9aKtQKnAh8LyKmUUzbft9p\nfQP17ShgFsU/qyOBEcDMuhZVAwMV/M3AUT2Wx5frGoqkQRShvyMi7i1XvyZpbLl9LLCtXvX1cBpw\njqSXgLsoTvdvBg6T1Fq2aaQ+3gRsiojl5fJiin8Ejdi3nwdejIjXI6IDuJeivxu1b/tkoIL/FDCp\nvDM6mOJmyZIBOnZFJAm4DVgdETf22LQEmF0+nk1x7V9XEXFNRIyPiAkUfflYRFwELAPOK5s1RK0A\nEbEV2Cjp2HLVDGAVDdi3FKf4p0oaXv5NdNXakH3bZwN40+Rs4AXgV8Df1vvmxl7q+wzFqeb/Ac+W\nP2dTXDsvBdYBjwKj613rB+r+M+CB8vHRwJPAeuA/gCH1rq9HnScA7WX/3geMatS+Bf4BWAM8B/wY\nGNLIfduXH8/cM0vIN/fMEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRL6f7LRuGBs/ayIAAAA\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0074c88>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAADypJREFUeJzt3X2QVfV9x/H3h+VJ8AlMRARTMICU\n2ipmo1JjTCUmaDJiZozRcQxDyfBHEsWH1oe0f6TpTCd2HJWkmJbRGpJYTaqOMo7V6oqdJhF0UVsi\nD0LwARDEGNSEKO7Ct3+cs9vFAfaye+/de/1+XjM7e885v3PPd36zn/2dc+7v3quIwMxyGTTQBZhZ\n/Tn4Zgk5+GYJOfhmCTn4Zgk5+GYJOfhmCfUr+JJmSVonaYOk66tVlJnVlvo6gUdSC/AicA6wGXgG\nuCQiVlevPDOrhcH92PdUYENEbASQdA8wG9hv8IdqWAxnZD8OaWYH8h47eT92qbd2/Qn+OGBTj+XN\nwGkfbCRpPjAfYDgjOE0z+3FIMzuQFdFWUbua39yLiMUR0RoRrUMYVuvDmVkF+hP8LcBxPZbHl+vM\nrMH1J/jPAJMlTZQ0FLgYWFqdssyslvp8jR8RnZK+CTwKtAD/GhEvVK0yM6uZ/tzcIyIeBh6uUi1m\nVieeuWeWkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINv\nlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WkINvlpCDb5aQg2+W\nkINvlpCDb5aQg2+WkINvlpCDb5aQg2+WUK/Bl3ScpGWSVkt6QdKCcv1oSY9JWl/+HlX7cs2sGioZ\n8TuBayJiGnA68A1J04DrgbaImAy0lctm1gR6DX5EbI2IZ8vHvwPWAOOA2cCSstkS4IJaFWlm1XVQ\n1/iSJgDTgRXAmIjYWm7aBoypamVmVjMVB1/SocB9wJUR8U7PbRERQOxnv/mS2iW1d7CrX8WaWXVU\nFHxJQyhCf1dE3F+ufl3S2HL7WGD7vvaNiMUR0RoRrUMYVo2azayfKrmrL+AOYE1E3Nxj01JgTvl4\nDvBg9cszs1oYXEGbM4DLgFWSni/XfQv4LvAzSfOAV4CLalOimVVbr8GPiJ8D2s/mmdUtx8zqwTP3\nzBJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDN\nEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0S\ncvDNEnLwzRJy8M0ScvDNEnLwzRJy8M0Sqjj4klokPSfpoXJ5oqQVkjZI+qmkobUr08yq6WBG/AXA\nmh7LNwK3RMQkYAcwr5qFmVntVBR8SeOBLwC3l8sCzgbuLZssAS6oRYFmVn2Vjvi3AtcCe8rlo4C3\nIqKzXN4MjNvXjpLmS2qX1N7Brn4Va2bV0WvwJX0R2B4RK/tygIhYHBGtEdE6hGF9eQozq7LBFbQ5\nAzhf0nnAcOBwYCFwpKTB5ag/HthSuzLNrJp6HfEj4oaIGB8RE4CLgSci4lJgGXBh2WwO8GDNqjSz\nqurP6/jXAVdL2kBxzX9HdUoys1qr5FS/W0Q8CTxZPt4InFr9ksys1jxzzywhB98sIQffLCEH3ywh\nB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH\n3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQffLCEH3ywhB98sIQff\nLCEH3yyhioIv6UhJ90paK2mNpBmSRkt6TNL68veoWhdrZtVR6Yi/EHgkIqYCJwFrgOuBtoiYDLSV\ny2bWBHoNvqQjgE8DdwBExPsR8RYwG1hSNlsCXFCrIs2sugZX0GYi8AZwp6STgJXAAmBMRGwt22wD\nxtSmRLP+6zz7EwBs/fouAP5nxpLubSc9NQeAYxcNBaBl2bN1rq7+KjnVHwycAvwgIqYDO/nAaX1E\nBBD72lnSfEntkto72NXfes2sClRk9gANpGOA5RExoVw+kyL4k4DPRMRWSWOBJyPihAM91+EaHadp\nZlUKN6vEnrOmA3DrktsAmDRk/ye56zp2A/DXE06vfWE1siLaeCd+q97a9TriR8Q2YJOkrlDPBFYD\nS4E55bo5wIN9rNXM6qySa3yAy4G7JA0FNgJzKf5p/EzSPOAV4KLalGh28Do+1wrAtbf9GIApQ4rr\n9z3sAWBjR0d327f3DANgevGLXed+EoBDlq0q9nnvvdoXXGcVBT8ingda97HJ5+1mTajSEd+sYbUc\nfnj3452fngrAVbf8GwB/ccjvyy17X9X+cMefdz9uu20GAL/49vcAeOz2fwZg2k++CcDx1z1V/aIH\nmKfsmiXk4Jsl5FN9a3qbfzSu+/Ezn1xU0T7fOfqZ7sePHFqc9s99+XMALJnwOACHT3uzWiU2HI/4\nZgl5xLem1TUN9+6T/6l73SCG7tVm7ivFC0/tj/8xAKvmFW2XvTu8u83R7e8CsGFHcWNwyD8sK56r\n12kwzcsjvllCvU7ZrSZP2bVqqGQa7vlrvwRAy4U7AfjtF4qJp2+eWAzjUxZt6m7buWnzXvs+tGUl\nAFt3F2cCfznniu5tjf4GnqpN2TWzDx9f41vT0Cf+BIDfXF2MxF3TcFf2eNPnE7+fBsCb9xwHwFE7\nisk3R/xkefG7bNdZwfHGtBRzeN+88g/d645e1qfSG45HfLOEPOJbwxs0YgQAnf/4DgDLp94PwEud\n7wNw9beu6W476r9fBeDokdsB2F2F45869pXuxy9X4fkagUd8s4Q84lvDe/es4tr+0am37bX+awuu\nAuCwB5Z3r6vk2t084pul5OCbJeRTfWt4f/b3zwMwqBynuqbhHvLA0zU53hC1ANBRzm1rUf0mudWL\nR3yzhDziW8N667Lik3H+dsxNAOwp34Cz8j+LSTof45c1OW5H7C6PV3w+3yNrpnVvm0xjT9mtlEd8\ns4Q84lvD6jyk+H3EoGKkf+q9Ygrt8T96rdhepeN0TRBae9OJ5ZriTTqXbjwXgKkLXupuW40JQY3A\nI75ZQh7xrWm8uftQADo3vtzv5+oa5QHWffdPAVg7u/iQjv/4Q/FWntcWTQLgsB3L+bDxiG+WkEd8\naxp/9YsvAzClvAbvi64P8dhevrUXYE1rMdLPXPUVAEbO2gjAYXz4RvouHvHNEvKIb42r/ACprhl7\nCz91NwCLmHLQT/XKd4o5Afd99Wbg/z/EA+CUp4vvfj32S6v7XGqz8YhvlpCDb5aQT/WtcZXvjema\nOnvWIcU321z5w+Lz9D9+Z7F+yLbfde/y+lkfBWD0V4pPzr38Y20AnDuiuCG4dOcYAL66alb3Ph/5\nl5E1Kb+RecQ3S8gjvjWN4Sr+XNecU3yN9c/PLL4NZ/2uY7rbzD3i5X3uu+C1MwF45JcnAzB5wYf3\npbpKeMQ3S8jfpGMNq2XKxwGYcnfxKbc3HvPUXtu7XubrugfQ03O7im2X/Nf84jnm9n3STzPxN+mY\n2X5VdI0v6SrgaxT3WVcBc4GxwD3AURTvY7wsIt6vUZ2W0O4Xfw3A+i9PAGDa5ZcDsPqi7+93n6kP\nfx2AE24rvv1mynM5RvqD1euIL2kccAXQGhEnAi3AxcCNwC0RMQnYAcyrZaFmVj29XuOXwV8OnAS8\nAzwAfB+4CzgmIjolzQC+HRGfP9Bz+RrfrLaqdo0fEVuAm4BXga3A2xSn9m9FRNeHoGwGxu1rf0nz\nJbVLau9g176amFmdVXKqPwqYDUwEjgVGArMOuFMPEbE4IlojonUIw/pcqJlVTyV39T8LvBQRb0RE\nB3A/cAZwpKSum4PjgS01qtHMqqyS4L8KnC5phCQBM4HVwDLgwrLNHODB2pRoZtVWyTX+CuBe4FmK\nl/IGAYuB64CrJW2geEnvjhrWaWZV5Jl7Zh8inrlnZvvl4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk\n4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTg\nmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl5OCb\nJeTgmyWkiKjfwaQ3gJ3Ab+p20P75CM1TKzRXvc1UKzRPvX8UER/trVFdgw8gqT0iWut60D5qplqh\nueptplqh+ertjU/1zRJy8M0SGojgLx6AY/ZVM9UKzVVvM9UKzVfvAdX9Gt/MBp5P9c0SqlvwJc2S\ntE7SBknX1+u4lZJ0nKRlklZLekHSgnL9aEmPSVpf/h410LV2kdQi6TlJD5XLEyWtKPv4p5KGDnSN\nXSQdKeleSWslrZE0o1H7VtJV5d/AryTdLWl4I/dtX9Ql+JJagEXAucA04BJJ0+px7IPQCVwTEdOA\n04FvlDVeD7RFxGSgrVxuFAuANT2WbwRuiYhJwA5g3oBUtW8LgUciYipwEkXdDde3ksYBVwCtEXEi\n0AJcTGP37cGLiJr/ADOAR3ss3wDcUI9j96PmB4FzgHXA2HLdWGDdQNdW1jKeIixnAw8BophgMnhf\nfT7AtR4BvER5T6nH+obrW2AcsAkYDQwu+/bzjdq3ff2p16l+V2d22Vyua0iSJgDTgRXAmIjYWm7a\nBowZoLI+6FbgWmBPuXwU8FZEdJbLjdTHE4E3gDvLS5PbJY2kAfs2IrYANwGvAluBt4GVNG7f9olv\n7n2ApEOB+4ArI+Kdntui+Hc/4C+DSPoisD0iVg50LRUaDJwC/CAiplNM297rtL6B+nYUMJvin9Wx\nwEhg1oAWVQP1Cv4W4Lgey+PLdQ1F0hCK0N8VEfeXq1+XNLbcPhbYPlD19XAGcL6kl4F7KE73FwJH\nShpctmmkPt4MbI6IFeXyvRT/CBqxbz8LvBQRb0REB3A/RX83at/2Sb2C/wwwubwzOpTiZsnSOh27\nIpIE3AGsiYibe2xaCswpH8+huPYfUBFxQ0SMj4gJFH35RERcCiwDLiybNUStABGxDdgk6YRy1Uxg\nNQ3YtxSn+KdLGlH+TXTV2pB922d1vGlyHvAi8Gvgbwb65sY+6vsUxanm/wLPlz/nUVw7twHrgceB\n0QNd6wfq/gzwUPn4eOBpYAPw78Cwga6vR50nA+1l/z4AjGrUvgX+DlgL/Ar4MTCskfu2Lz+euWeW\nkG/umSXk4Jsl5OCbJeTgmyXk4Jsl5OCbJeTgmyXk4Jsl9H+Y9ch60LRo8wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe79ffdc6d8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n0P8B0He7JAGHDzEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe7a0004390>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "video_1 = []\n",
    "n_iters = 100\n",
    "x = 50\n",
    "y = 50\n",
    "\n",
    "for i in range(n_iters):\n",
    "    x_min = max(0, x - 10)\n",
    "    x_max = min(99-28, x + 10)\n",
    "    \n",
    "    y_min = max(0, y - 10)\n",
    "    y_max = min(99-28, y + 10)\n",
    "    \n",
    "    x = random.randint(x_min,x_max)\n",
    "    y = random.randint(y_min,y_max)\n",
    "    img = insert_num_background([x,y], x_train[1], [100,100])\n",
    "    plt.figure()\n",
    "    plt.imshow(img)\n",
    "\n",
    "    video_1.append(img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from __future__ import division\n",
    "import os\n",
    "import time\n",
    "import math\n",
    "from glob import glob\n",
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "from six.moves import xrange\n",
    "\n",
    "from ops import *\n",
    "from utils import *\n",
    "\n",
    "import pandas as pd\n",
    "\n",
    "# def conv_out_size_same(size, stride):\n",
    "# return int(math.ceil(float(size) / float(stride)))\n",
    "\n",
    "class DCGAN(object):\n",
    "    def __init__(self, sess, input_height=108, input_width=108, crop=True,\n",
    "         batch_size=64, sample_num = 64, output_height=64, output_width=64,\n",
    "         y_dim=10, z_dim=100, gf_dim=64, df_dim=64,\n",
    "         gfc_dim=1024, dfc_dim=1024, c_dim=3, dataset_name='faces',\n",
    "         input_fname_pattern='*.ppm', checkpoint_dir=None, sample_dir=None):\n",
    "\n",
    "    \"\"\"\"\n",
    "    Args:\n",
    "      sess: TensorFlow session\n",
    "      batch_size: The size of batch. Should be specified before training.\n",
    "      y_dim: (optional) Dimension of dim for y. [None]\n",
    "      z_dim: (optional) Dimension of dim for Z. [100]\n",
    "      gf_dim: (optional) Dimension of gen filters in first conv layer. [64]\n",
    "      df_dim: (optional) Dimension of discrim filters in first conv layer. [64]\n",
    "      gfc_dim: (optional) Dimension of gen units for for fully connected layer. [1024]\n",
    "      dfc_dim: (optional) Dimension of discrim units for fully connected layer. [1024]\n",
    "      c_dim: (optional) Dimension of image color. For grayscale input, set to 1. [3]\n",
    "    \"\"\"\n",
    "    self.sess = sess\n",
    "\n",
    "    self.batch_size = batch_size\n",
    "    self.sample_num = sample_num\n",
    "\n",
    "    self.gf_dim = gf_dim\n",
    "    self.df_dim = df_dim\n",
    "\n",
    "    self.gfc_dim = gfc_dim\n",
    "    self.dfc_dim = dfc_dim\n",
    "\n",
    "    # batch normalization : deals with poor initialization helps gradient flow\n",
    "    self.d_bn1 = batch_norm(name='d_bn1')\n",
    "    self.d_bn2 = batch_norm(name='d_bn2')\n",
    "\n",
    "    self.d_bn3 = batch_norm(name='d_bn3')\n",
    "\n",
    "    self.g_bn0 = batch_norm(name='g_bn0')\n",
    "    self.g_bn1 = batch_norm(name='g_bn1')\n",
    "    self.g_bn2 = batch_norm(name='g_bn2')\n",
    "\n",
    "    self.g_bn3 = batch_norm(name='g_bn3')\n",
    "\n",
    "    self.dataset_name = dataset_name\n",
    "    self.input_fname_pattern = input_fname_pattern\n",
    "    self.checkpoint_dir = checkpoint_dir\n",
    "\n",
    "    self.data = glob(os.path.join(\"/data\", self.dataset_name, self.input_fname_pattern))\n",
    "    self.labels = [get_condition(instance) for instance in self.data]\n",
    "    print (self.labels[:10])\n",
    "\n",
    "    imreadImg = imread(self.data[0])\n",
    "\n",
    "    if len(imreadImg.shape) >= 3: #check if image is a non-grayscale image by checking channel number\n",
    "          self.c_dim = imread(self.data[0]).shape[-1]\n",
    "    else:\n",
    "          self.c_dim = 1\n",
    "\n",
    "    self.grayscale = (self.c_dim == 1)\n",
    "\n",
    "    self.build_model()\n",
    "\n",
    "    def build_model(self):\n",
    "        if self.y_dim:\n",
    "            self.y = tf.placeholder(tf.float32, [self.batch_size, self.y_dim], name='y')\n",
    "        else:\n",
    "            self.y = None\n",
    "\n",
    "    if self.crop:\n",
    "      image_dims = [self.output_height, self.output_width, self.c_dim]\n",
    "    else:\n",
    "      image_dims = [self.input_height, self.input_width, self.c_dim]\n",
    "\n",
    "    self.inputs = tf.placeholder(\n",
    "      tf.float32, [self.batch_size] + image_dims, name='real_images')\n",
    "\n",
    "    inputs = self.inputs\n",
    "\n",
    "    self.z = tf.placeholder(\n",
    "      tf.float32, [None, self.z_dim], name='z')\n",
    "    self.z_sum = histogram_summary(\"z\", self.z)\n",
    "\n",
    "    self.G                  = self.generator(self.z, self.y)\n",
    "    self.D, self.D_logits   = self.discriminator(inputs, self.y, reuse=False)\n",
    "    self.sampler            = self.sampler(self.z, self.y)\n",
    "    self.D_, self.D_logits_ = self.discriminator(self.G, self.y, reuse=True)\n",
    "    \n",
    "    self.d_sum = histogram_summary(\"d\", self.D)\n",
    "    self.d__sum = histogram_summary(\"d_\", self.D_)\n",
    "    self.G_sum = image_summary(\"G\", self.G)\n",
    "\n",
    "    def sigmoid_cross_entropy_with_logits(x, y):\n",
    "      try:\n",
    "        return tf.nn.sigmoid_cross_entropy_with_logits(logits=x, labels=y)\n",
    "      except:\n",
    "        return tf.nn.sigmoid_cross_entropy_with_logits(logits=x, targets=y)\n",
    "\n",
    "    self.d_loss_real = tf.reduce_mean(\n",
    "      sigmoid_cross_entropy_with_logits(self.D_logits, tf.ones_like(self.D)))\n",
    "    self.d_loss_fake = tf.reduce_mean(\n",
    "      sigmoid_cross_entropy_with_logits(self.D_logits_, tf.zeros_like(self.D_)))\n",
    "    self.g_loss = tf.reduce_mean(\n",
    "      sigmoid_cross_entropy_with_logits(self.D_logits_, tf.ones_like(self.D_)))\n",
    "\n",
    "    self.d_loss_real_sum = scalar_summary(\"d_loss_real\", self.d_loss_real)\n",
    "    self.d_loss_fake_sum = scalar_summary(\"d_loss_fake\", self.d_loss_fake)\n",
    "                          \n",
    "    self.d_loss = self.d_loss_real + self.d_loss_fake\n",
    "\n",
    "    self.g_loss_sum = scalar_summary(\"g_loss\", self.g_loss)\n",
    "    self.d_loss_sum = scalar_summary(\"d_loss\", self.d_loss)\n",
    "\n",
    "    t_vars = tf.trainable_variables()\n",
    "\n",
    "    self.d_vars = [var for var in t_vars if 'd_' in var.name]\n",
    "    self.g_vars = [var for var in t_vars if 'g_' in var.name]\n",
    "\n",
    "    self.saver = tf.train.Saver()\n",
    "\n",
    "\n",
    "  def train(self, config):\n",
    "    d_optim = tf.train.AdamOptimizer(config.learning_rate, beta1=config.beta1) \\\n",
    "              .minimize(self.d_loss, var_list=self.d_vars)\n",
    "    g_optim = tf.train.AdamOptimizer(config.learning_rate, beta1=config.beta1) \\\n",
    "              .minimize(self.g_loss, var_list=self.g_vars)\n",
    "    try:\n",
    "      tf.global_variables_initializer().run()\n",
    "    except:\n",
    "      tf.initialize_all_variables().run()\n",
    "\n",
    "    self.g_sum = merge_summary([self.z_sum, self.d__sum,\n",
    "      self.G_sum, self.d_loss_fake_sum, self.g_loss_sum])\n",
    "    self.d_sum = merge_summary(\n",
    "        [self.z_sum, self.d_sum, self.d_loss_real_sum, self.d_loss_sum])\n",
    "    self.writer = SummaryWriter(\"./logs\", self.sess.graph)\n",
    "\n",
    "    sample_files = self.data[0:self.sample_num]\n",
    "\n",
    "    sample_z = np.random.uniform(-1, 1, size=(self.sample_num , self.z_dim))\n",
    "    sample_labels = self.labels[0:self.sample_num]\n",
    "\n",
    "    sample = [\n",
    "        get_image(sample_file,\n",
    "                  input_height=self.input_height,\n",
    "                  input_width=self.input_width,\n",
    "                  resize_height=self.output_height,\n",
    "                  resize_width=self.output_width,\n",
    "                  crop=self.crop,\n",
    "                  grayscale=self.grayscale) for sample_file in sample_files]\n",
    "    if (self.grayscale):\n",
    "      sample_inputs = np.array(sample).astype(np.float32)[:, :, :, None]\n",
    "    else:\n",
    "      sample_inputs = np.array(sample).astype(np.float32)\n",
    "\n",
    "    counter = 1\n",
    "    start_time = time.time()\n",
    "    could_load, checkpoint_counter = self.load(self.checkpoint_dir)\n",
    "    if could_load:\n",
    "      counter = checkpoint_counter\n",
    "      print(\" [*] Load SUCCESS\")\n",
    "    else:\n",
    "      print(\" [!] Load failed...\")\n",
    "\n",
    "    for epoch in xrange(config.epoch):\n",
    "      batch_idxs = min(len(self.data), config.train_size) // config.batch_size\n",
    "\n",
    "      for idx in xrange(0, batch_idxs):\n",
    "        batch_files = self.data[idx*config.batch_size:(idx+1)*config.batch_size]\n",
    "        batch = [\n",
    "            get_image(batch_file,\n",
    "                      input_height=self.input_height,\n",
    "                      input_width=self.input_width,\n",
    "                      resize_height=self.output_height,\n",
    "                      resize_width=self.output_width,\n",
    "                      crop=self.crop,\n",
    "                      grayscale=self.grayscale) for batch_file in batch_files]\n",
    "        if self.grayscale:\n",
    "          batch_images = np.array(batch).astype(np.float32)[:, :, :, None]\n",
    "        else:\n",
    "          batch_images = np.array(batch).astype(np.float32)\n",
    "\n",
    "        batch_z = np.random.uniform(-1, 1, [config.batch_size, self.z_dim]) \\\n",
    "              .astype(np.float32)\n",
    "\n",
    "        batch_labels = self.labels[idx*config.batch_size: (idx+1)*config.batch_size]\n",
    "\n",
    "        # Update D network\n",
    "        _, summary_str = self.sess.run([d_optim, self.d_sum],\n",
    "          feed_dict={ self.inputs: batch_images, self.z: batch_z, self.y: batch_labels })\n",
    "        self.writer.add_summary(summary_str, counter)\n",
    "\n",
    "        # Update G network\n",
    "        _, summary_str = self.sess.run([g_optim, self.g_sum],\n",
    "          feed_dict={ self.z: batch_z, self.y: batch_labels })\n",
    "        self.writer.add_summary(summary_str, counter)\n",
    "\n",
    "        # Run g_optim twice to make sure that d_loss does not go to zero (different from paper)\n",
    "        _, summary_str = self.sess.run([g_optim, self.g_sum],\n",
    "          feed_dict={ self.z: batch_z, self.y: batch_labels })\n",
    "        self.writer.add_summary(summary_str, counter)\n",
    "        \n",
    "        errD_fake = self.d_loss_fake.eval({ self.z: batch_z, self.y: batch_labels })\n",
    "        errD_real = self.d_loss_real.eval({ self.inputs: batch_images, self.y: batch_labels })\n",
    "        errG = self.g_loss.eval({self.z: batch_z, self.y: batch_labels})\n",
    "  \n",
    "        print(\"Epoch: [%2d] [%4d/%4d] time: %4.4f, d_loss: %.8f, g_loss: %.8f\" \\\n",
    "          % (epoch, idx, batch_idxs,\n",
    "            time.time() - start_time, errD_fake+errD_real, errG))\n",
    "\n",
    "      counter += 1\n",
    "      if np.mod(counter, 30) == 20:\n",
    "        try:\n",
    "          samples, d_loss, g_loss = self.sess.run(\n",
    "            [self.sampler, self.d_loss, self.g_loss],\n",
    "            feed_dict={\n",
    "                self.z: sample_z,\n",
    "                self.inputs: sample_inputs,\n",
    "                self.y: sample_labels,\n",
    "            },\n",
    "          )\n",
    "          save_images(samples, image_manifold_size(samples.shape[0]),\n",
    "                '/output/train_{:02d}_{:04d}.png'.format(epoch, idx))\n",
    "          print(\"[Sample] d_loss: %.8f, g_loss: %.8f\" % (d_loss, g_loss)) \n",
    "        except:\n",
    "          print(\"one pic error!...\")\n",
    "\n",
    "        self.save(\"/output\", counter)\n",
    "        print(\"After Save\")\n",
    "\n",
    "  def discriminator(self, image, y=None, reuse=False):\n",
    "    with tf.variable_scope(\"discriminator\") as scope:\n",
    "      if reuse:\n",
    "        scope.reuse_variables()\n",
    "\n",
    "      # yb = tf.reshape(y, [self.batch_size, 1, 1, self.y_dim])\n",
    "      # x = conv_cond_concat(image, yb)\n",
    "\n",
    "      h0 = conv2d(image, self.df_dim, name='d_h0_conv')\n",
    "      h0 = tf.layers.batch_normalization(h0, training=True)\n",
    "      h0 = lrelu(h0)\n",
    "\n",
    "      h1 = conv2d(h0, self.df_dim*2, name='d_h1_conv')\n",
    "      h1 = tf.layers.batch_normalization(h1, training=True)\n",
    "      h1 = lrelu(h1)\n",
    "\n",
    "      h2 = conv2d(h1, self.df_dim*4, name='d_h2_conv')\n",
    "      h2 = tf.layers.batch_normalization(h2, training=True)\n",
    "      h2 = lrelu(h2)\n",
    "\n",
    "      tags_vectors = tf.expand_dims(tf.expand_dims(y, 1), 2)\n",
    "      tags_vectors = tf.tile(tags_vectors, [1, 14, 14, 1])\n",
    "      condition_info = tf.concat([h2, tags_vectors], axis=-1)\n",
    "\n",
    "      h3 = conv2d(condition_info, self.df_dim*8, name='d_h3_conv')\n",
    "      h3 = tf.layers.batch_normalization(h3, training=True)\n",
    "      h3 = lrelu(h3)      \n",
    "\n",
    "      return tf.nn.sigmoid(h3), h3\n",
    "\n",
    "  def generator(self, z, y=None):\n",
    "    with tf.variable_scope(\"generator\") as scope:\n",
    "      s_h, s_w = self.output_height, self.output_width\n",
    "      s_h2, s_w2 = conv_out_size_same(s_h, 2), conv_out_size_same(s_w, 2)\n",
    "      s_h4, s_w4 = conv_out_size_same(s_h2, 2), conv_out_size_same(s_w2, 2)\n",
    "      s_h8, s_w8 = conv_out_size_same(s_h4, 2), conv_out_size_same(s_w4, 2)\n",
    "      s_h16, s_w16 = conv_out_size_same(s_h8, 2), conv_out_size_same(s_w8, 2)\n",
    "\n",
    "      noise_vector = concat([z, y], 1)\n",
    "\n",
    "      h0 = tf.contrib.layers.fully_connected(\n",
    "        noise_vector, s_h16*s_w16*self.gf_dim*8,\n",
    "        weights_initializer=tf.random_normal_initializer(stddev=0.02),\n",
    "        activation_fn=None\n",
    "        )\n",
    "\n",
    "      h0 = tf.layers.batch_normalization(h0, training=True)\n",
    "      h0 = tf.reshape(h0, [-1, s_h16, s_w16, self.gf_dim*8])\n",
    "      h0 = tf.nn.relu(h0)\n",
    "\n",
    "      self.h1, self.h1_w, self.h1_b = deconv2d(\n",
    "          h0, [self.batch_size, s_h8, s_w8, self.gf_dim*4], name='g_h1', with_w=True)\n",
    "\n",
    "      h1 = tf.nn.relu(self.g_bn1(self.h1))\n",
    "\n",
    "      h2, self.h2_w, self.h2_b = deconv2d(\n",
    "          h1, [self.batch_size, s_h4, s_w4, self.gf_dim*2], name='g_h2', with_w=True)\n",
    "      h2 = tf.nn.relu(self.g_bn2(h2))\n",
    "\n",
    "      h3, self.h3_w, self.h3_b = deconv2d(\n",
    "          h2, [self.batch_size, s_h2, s_w2, self.gf_dim*1], name='g_h3', with_w=True)\n",
    "      h3 = tf.nn.relu(self.g_bn3(h3))\n",
    "\n",
    "      h4, self.h4_w, self.h4_b = deconv2d(\n",
    "          h3, [self.batch_size, s_h, s_w, self.c_dim], name='g_h4', with_w=True)\n",
    "\n",
    "      return tf.nn.sigmoid(h4)\n",
    "\n",
    "  def sampler(self, z, y=None):\n",
    "    with tf.variable_scope(\"generator\") as scope:\n",
    "      scope.reuse_variables()\n",
    "\n",
    "      s_h, s_w = self.output_height, self.output_width\n",
    "      s_h2, s_w2 = conv_out_size_same(s_h, 2), conv_out_size_same(s_w, 2)\n",
    "      s_h4, s_w4 = conv_out_size_same(s_h2, 2), conv_out_size_same(s_w2, 2)\n",
    "      s_h8, s_w8 = conv_out_size_same(s_h4, 2), conv_out_size_same(s_w4, 2)\n",
    "      s_h16, s_w16 = conv_out_size_same(s_h8, 2), conv_out_size_same(s_w8, 2)\n",
    "\n",
    "      noise_vector = concat([z, y], 1)\n",
    "\n",
    "      h0 = tf.contrib.layers.fully_connected(\n",
    "        noise_vector, s_h16*s_w16*self.gf_dim*8,\n",
    "        weights_initializer=tf.random_normal_initializer(stddev=0.02),\n",
    "        activation_fn=None\n",
    "        )\n",
    "      \n",
    "      h0 = tf.layers.batch_normalization(h0, training=True)\n",
    "      h0 = tf.reshape(h0, [-1, s_h16, s_w16, self.gf_dim*8])\n",
    "      h0 = tf.nn.relu(h0)\n",
    "\n",
    "      self.h1, self.h1_w, self.h1_b = deconv2d(\n",
    "          h0, [self.batch_size, s_h8, s_w8, self.gf_dim*4], name='g_h1', with_w=True)\n",
    "\n",
    "      h1 = tf.nn.relu(self.g_bn1(self.h1, train = False))\n",
    "\n",
    "      h2, self.h2_w, self.h2_b = deconv2d(\n",
    "          h1, [self.batch_size, s_h4, s_w4, self.gf_dim*2], name='g_h2', with_w=True)\n",
    "      h2 = tf.nn.relu(self.g_bn2(h2, train = False))\n",
    "\n",
    "      h3, self.h3_w, self.h3_b = deconv2d(\n",
    "          h2, [self.batch_size, s_h2, s_w2, self.gf_dim*1], name='g_h3', with_w=True)\n",
    "      h3 = tf.nn.relu(self.g_bn3(h3, train = False))\n",
    "\n",
    "      h4, self.h4_w, self.h4_b = deconv2d(\n",
    "          h3, [self.batch_size, s_h, s_w, self.c_dim], name='g_h4', with_w=True)\n",
    "\n",
    "      return tf.nn.sigmoid(h4)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}