malaria_cells.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np \n",
    "import matplotlib.pyplot as plt \n",
    "import seaborn as sns \n",
    "import os"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "from matplotlib.image import imread"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_dir = 'C:**\\\\Analysis Cell Images Using CCN\\\\cell_images'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['test', 'train']"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "os.listdir(data_dir)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_path = data_dir+'\\\\test\\\\'\n",
    "train_path = data_dir+'\\\\train\\\\'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'C:\\\\Users\\\\Mohamed Basher\\\\Desktop\\\\Analysis Cell Images Using CCN\\\\cell_images\\\\test\\\\'"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_path"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['parasitized', 'uninfected']"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "os.listdir(test_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['parasitized', 'uninfected']"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "os.listdir(train_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'C100P61ThinF_IMG_20150918_144104_cell_162.png'"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "os.listdir(train_path+'parasitized')[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "para_cell =train_path+'parasitized\\\\'+'C100P61ThinF_IMG_20150918_144104_cell_162.png'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'C:\\\\Users\\\\Mohamed Basher\\\\Desktop\\\\Analysis Cell Images Using CCN\\\\cell_images\\\\train\\\\parasitized\\\\C100P61ThinF_IMG_20150918_144104_cell_162.png'"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "para_cell"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(148, 142, 3)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "imread(para_cell).shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(imread(para_cell));"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'C100P61ThinF_IMG_20150918_144104_cell_128.png'"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "os.listdir(train_path+'uninfected')[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "uninfected_cell =train_path+'uninfected\\\\'+'C100P61ThinF_IMG_20150918_144104_cell_128.png'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(imread(uninfected_cell));"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "12480"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(os.listdir(train_path+'parasitized'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "12480"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(os.listdir(train_path+'uninfected'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Check Dimenstion of images \n",
    "\n",
    "dim1 = []\n",
    "dim2 = []\n",
    "\n",
    "for images_filename in os.listdir(test_path+'uninfected'):\n",
    "    \n",
    "    img =imread(test_path+'uninfected\\\\'+images_filename)\n",
    "    d1,d2,colors = img.shape\n",
    "    \n",
    "    dim1.append(d1)\n",
    "    dim2.append(d2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[118,\n",
       " 172,\n",
       " 130,\n",
       " 163,\n",
       " 172,\n",
       " 139,\n",
       " 151,\n",
       " 160,\n",
       " 145,\n",
       " 127,\n",
       " 121,\n",
       " 154,\n",
       " 145,\n",
       " 127,\n",
       " 154,\n",
       " 151,\n",
       " 148,\n",
       " 130,\n",
       " 160,\n",
       " 136,\n",
       " 139,\n",
       " 190,\n",
       " 127,\n",
       " 133,\n",
       " 136,\n",
       " 124,\n",
       " 151,\n",
       " 151,\n",
       " 151,\n",
       " 160,\n",
       " 118,\n",
       " 154,\n",
       " 154,\n",
       " 121,\n",
       " 127,\n",
       " 118,\n",
       " 130,\n",
       " 124,\n",
       " 163,\n",
       " 88,\n",
       " 139,\n",
       " 127,\n",
       " 121,\n",
       " 139,\n",
       " 166,\n",
       " 133,\n",
       " 124,\n",
       " 154,\n",
       " 139,\n",
       " 127,\n",
       " 124,\n",
       " 136,\n",
       " 166,\n",
       " 133,\n",
       " 145,\n",
       " 157,\n",
       " 181,\n",
       " 148,\n",
       " 112,\n",
       " 124,\n",
       " 112,\n",
       " 106,\n",
       " 121,\n",
       " 148,\n",
       " 148,\n",
       " 112,\n",
       " 121,\n",
       " 148,\n",
       " 97,\n",
       " 112,\n",
       " 142,\n",
       " 118,\n",
       " 130,\n",
       " 151,\n",
       " 133,\n",
       " 136,\n",
       " 115,\n",
       " 145,\n",
       " 115,\n",
       " 142,\n",
       " 136,\n",
       " 130,\n",
       " 142,\n",
       " 124,\n",
       " 142,\n",
       " 136,\n",
       " 166,\n",
       " 148,\n",
       " 136,\n",
       " 142,\n",
       " 133,\n",
       " 130,\n",
       " 139,\n",
       " 124,\n",
       " 139,\n",
       " 121,\n",
       " 130,\n",
       " 142,\n",
       " 133,\n",
       " 118,\n",
       " 157,\n",
       " 127,\n",
       " 109,\n",
       " 112,\n",
       " 139,\n",
       " 142,\n",
       " 142,\n",
       " 112,\n",
       " 118,\n",
       " 100,\n",
       " 127,\n",
       " 145,\n",
       " 136,\n",
       " 160,\n",
       " 187,\n",
       " 130,\n",
       " 130,\n",
       " 157,\n",
       " 130,\n",
       " 130,\n",
       " 151,\n",
       " 175,\n",
       " 145,\n",
       " 136,\n",
       " 157,\n",
       " 133,\n",
       " 127,\n",
       " 154,\n",
       " 133,\n",
       " 151,\n",
       " 130,\n",
       " 136,\n",
       " 160,\n",
       " 130,\n",
       " 139,\n",
       " 145,\n",
       " 145,\n",
       " 145,\n",
       " 154,\n",
       " 133,\n",
       " 160,\n",
       " 142,\n",
       " 157,\n",
       " 145,\n",
       " 136,\n",
       " 121,\n",
       " 115,\n",
       " 121,\n",
       " 121,\n",
       " 127,\n",
       " 145,\n",
       " 130,\n",
       " 133,\n",
       " 142,\n",
       " 127,\n",
       " 121,\n",
       " 139,\n",
       " 118,\n",
       " 118,\n",
       " 148,\n",
       " 121,\n",
       " 145,\n",
       " 130,\n",
       " 169,\n",
       " 142,\n",
       " 136,\n",
       " 148,\n",
       " 142,\n",
       " 169,\n",
       " 115,\n",
       " 109,\n",
       " 166,\n",
       " 145,\n",
       " 127,\n",
       " 124,\n",
       " 124,\n",
       " 130,\n",
       " 196,\n",
       " 97,\n",
       " 115,\n",
       " 133,\n",
       " 139,\n",
       " 148,\n",
       " 109,\n",
       " 127,\n",
       " 175,\n",
       " 148,\n",
       " 136,\n",
       " 175,\n",
       " 151,\n",
       " 154,\n",
       " 133,\n",
       " 145,\n",
       " 115,\n",
       " 130,\n",
       " 121,\n",
       " 133,\n",
       " 142,\n",
       " 121,\n",
       " 136,\n",
       " 142,\n",
       " 136,\n",
       " 145,\n",
       " 127,\n",
       " 151,\n",
       " 136,\n",
       " 145,\n",
       " 166,\n",
       " 139,\n",
       " 151,\n",
       " 127,\n",
       " 136,\n",
       " 142,\n",
       " 118,\n",
       " 145,\n",
       " 127,\n",
       " 127,\n",
       " 118,\n",
       " 121,\n",
       " 121,\n",
       " 121,\n",
       " 106,\n",
       " 88,\n",
       " 136,\n",
       " 151,\n",
       " 133,\n",
       " 88,\n",
       " 142,\n",
       " 103,\n",
       " 124,\n",
       " 148,\n",
       " 151,\n",
       " 139,\n",
       " 157,\n",
       " 148,\n",
       " 154,\n",
       " 115,\n",
       " 115,\n",
       " 136,\n",
       " 133,\n",
       " 130,\n",
       " 136,\n",
       " 112,\n",
       " 115,\n",
       " 136,\n",
       " 109,\n",
       " 145,\n",
       " 133,\n",
       " 136,\n",
       " 112,\n",
       " 124,\n",
       " 115,\n",
       " 100,\n",
       " 118,\n",
       " 148,\n",
       " 118,\n",
       " 94,\n",
       " 145,\n",
       " 115,\n",
       " 136,\n",
       " 103,\n",
       " 172,\n",
       " 154,\n",
       " 151,\n",
       " 172,\n",
       " 124,\n",
       " 133,\n",
       " 181,\n",
       " 142,\n",
       " 136,\n",
       " 148,\n",
       " 136,\n",
       " 115,\n",
       " 103,\n",
       " 118,\n",
       " 121,\n",
       " 136,\n",
       " 130,\n",
       " 154,\n",
       " 115,\n",
       " 127,\n",
       " 130,\n",
       " 133,\n",
       " 130,\n",
       " 133,\n",
       " 136,\n",
       " 172,\n",
       " 157,\n",
       " 145,\n",
       " 133,\n",
       " 109,\n",
       " 154,\n",
       " 130,\n",
       " 142,\n",
       " 121,\n",
       " 118,\n",
       " 130,\n",
       " 127,\n",
       " 106,\n",
       " 133,\n",
       " 112,\n",
       " 118,\n",
       " 76,\n",
       " 118,\n",
       " 115,\n",
       " 163,\n",
       " 145,\n",
       " 154,\n",
       " 121,\n",
       " 127,\n",
       " 130,\n",
       " 148,\n",
       " 118,\n",
       " 115,\n",
       " 127,\n",
       " 121,\n",
       " 109,\n",
       " 118,\n",
       " 124,\n",
       " 145,\n",
       " 166,\n",
       " 142,\n",
       " 136,\n",
       " 130,\n",
       " 145,\n",
       " 142,\n",
       " 148,\n",
       " 106,\n",
       " 124,\n",
       " 106,\n",
       " 115,\n",
       " 115,\n",
       " 127,\n",
       " 124,\n",
       " 163,\n",
       " 139,\n",
       " 124,\n",
       " 106,\n",
       " 130,\n",
       " 106,\n",
       " 103,\n",
       " 121,\n",
       " 130,\n",
       " 127,\n",
       " 127,\n",
       " 121,\n",
       " 121,\n",
       " 127,\n",
       " 124,\n",
       " 136,\n",
       " 118,\n",
       " 127,\n",
       " 115,\n",
       " 124,\n",
       " 112,\n",
       " 115,\n",
       " 124,\n",
       " 121,\n",
       " 121,\n",
       " 133,\n",
       " 133,\n",
       " 103,\n",
       " 118,\n",
       " 124,\n",
       " 154,\n",
       " 109,\n",
       " 133,\n",
       " 145,\n",
       " 124,\n",
       " 115,\n",
       " 127,\n",
       " 115,\n",
       " 115,\n",
       " 103,\n",
       " 136,\n",
       " 118,\n",
       " 109,\n",
       " 124,\n",
       " 118,\n",
       " 127,\n",
       " 124,\n",
       " 124,\n",
       " 112,\n",
       " 115,\n",
       " 124,\n",
       " 109,\n",
       " 130,\n",
       " 112,\n",
       " 127,\n",
       " 121,\n",
       " 112,\n",
       " 142,\n",
       " 130,\n",
       " 127,\n",
       " 127,\n",
       " 130,\n",
       " 148,\n",
       " 100,\n",
       " 127,\n",
       " 118,\n",
       " 112,\n",
       " 112,\n",
       " 121,\n",
       " 118,\n",
       " 121,\n",
       " 121,\n",
       " 112,\n",
       " 121,\n",
       " 121,\n",
       " 100,\n",
       " 130,\n",
       " 127,\n",
       " 130,\n",
       " 127,\n",
       " 124,\n",
       " 127,\n",
       " 118,\n",
       " 106,\n",
       " 127,\n",
       " 103,\n",
       " 106,\n",
       " 109,\n",
       " 118,\n",
       " 151,\n",
       " 121,\n",
       " 100,\n",
       " 139,\n",
       " 118,\n",
       " 133,\n",
       " 127,\n",
       " 145,\n",
       " 127,\n",
       " 127,\n",
       " 106,\n",
       " 109,\n",
       " 97,\n",
       " 175,\n",
       " 118,\n",
       " 127,\n",
       " 145,\n",
       " 121,\n",
       " 136,\n",
       " 124,\n",
       " 175,\n",
       " 172,\n",
       " 172,\n",
       " 130,\n",
       " 130,\n",
       " 130,\n",
       " 127,\n",
       " 142,\n",
       " 121,\n",
       " 112,\n",
       " 145,\n",
       " 115,\n",
       " 133,\n",
       " 121,\n",
       " 109,\n",
       " 106,\n",
       " 136,\n",
       " 103,\n",
       " 130,\n",
       " 124,\n",
       " 124,\n",
       " 124,\n",
       " 130,\n",
       " 139,\n",
       " 145,\n",
       " 97,\n",
       " 133,\n",
       " 103,\n",
       " 115,\n",
       " 130,\n",
       " 106,\n",
       " 133,\n",
       " 124,\n",
       " 118,\n",
       " 118,\n",
       " 109,\n",
       " 124,\n",
       " 118,\n",
       " 160,\n",
       " 91,\n",
       " 118,\n",
       " 130,\n",
       " 112,\n",
       " 124,\n",
       " 190,\n",
       " 121,\n",
       " 115,\n",
       " 136,\n",
       " 175,\n",
       " 127,\n",
       " 130,\n",
       " 118,\n",
       " 109,\n",
       " 94,\n",
       " 157,\n",
       " 100,\n",
       " 106,\n",
       " 148,\n",
       " 130,\n",
       " 133,\n",
       " 109,\n",
       " 91,\n",
       " 97,\n",
       " 85,\n",
       " 97,\n",
       " 97,\n",
       " 118,\n",
       " 115,\n",
       " 130,\n",
       " 124,\n",
       " 109,\n",
       " 97,\n",
       " 109,\n",
       " 136,\n",
       " 166,\n",
       " 139,\n",
       " 130,\n",
       " 163,\n",
       " 136,\n",
       " 136,\n",
       " 94,\n",
       " 160,\n",
       " 118,\n",
       " 112,\n",
       " 124,\n",
       " 94,\n",
       " 100,\n",
       " 91,\n",
       " 88,\n",
       " 112,\n",
       " 130,\n",
       " 133,\n",
       " 112,\n",
       " 109,\n",
       " 112,\n",
       " 118,\n",
       " 154,\n",
       " 136,\n",
       " 142,\n",
       " 130,\n",
       " 148,\n",
       " 166,\n",
       " 121,\n",
       " 115,\n",
       " 115,\n",
       " 136,\n",
       " 130,\n",
       " 127,\n",
       " 124,\n",
       " 145,\n",
       " 139,\n",
       " 136,\n",
       " 136,\n",
       " 112,\n",
       " 142,\n",
       " 130,\n",
       " 112,\n",
       " 118,\n",
       " 133,\n",
       " 121,\n",
       " 118,\n",
       " 154,\n",
       " 133,\n",
       " 124,\n",
       " 124,\n",
       " 148,\n",
       " 97,\n",
       " 109,\n",
       " 106,\n",
       " 112,\n",
       " 127,\n",
       " 88,\n",
       " 103,\n",
       " 94,\n",
       " 124,\n",
       " 85,\n",
       " 142,\n",
       " 136,\n",
       " 142,\n",
       " 127,\n",
       " 115,\n",
       " 130,\n",
       " 124,\n",
       " 142,\n",
       " 124,\n",
       " 124,\n",
       " 124,\n",
       " 148,\n",
       " 136,\n",
       " 130,\n",
       " 148,\n",
       " 115,\n",
       " 124,\n",
       " 136,\n",
       " 118,\n",
       " 142,\n",
       " 127,\n",
       " 154,\n",
       " 121,\n",
       " 178,\n",
       " 127,\n",
       " 130,\n",
       " 136,\n",
       " 145,\n",
       " 133,\n",
       " 106,\n",
       " 169,\n",
       " 133,\n",
       " 154,\n",
       " 118,\n",
       " 106,\n",
       " 133,\n",
       " 130,\n",
       " 142,\n",
       " 130,\n",
       " 148,\n",
       " 133,\n",
       " 145,\n",
       " 121,\n",
       " 121,\n",
       " 133,\n",
       " 154,\n",
       " 163,\n",
       " 130,\n",
       " 151,\n",
       " 139,\n",
       " 124,\n",
       " 118,\n",
       " 136,\n",
       " 115,\n",
       " 145,\n",
       " 127,\n",
       " 124,\n",
       " 139,\n",
       " 160,\n",
       " 148,\n",
       " 121,\n",
       " 154,\n",
       " 127,\n",
       " 124,\n",
       " 145,\n",
       " 124,\n",
       " 127,\n",
       " 136,\n",
       " 124,\n",
       " 115,\n",
       " 118,\n",
       " 133,\n",
       " 118,\n",
       " 106,\n",
       " 127,\n",
       " 121,\n",
       " 130,\n",
       " 103,\n",
       " 130,\n",
       " 172,\n",
       " 136,\n",
       " 139,\n",
       " 130,\n",
       " 148,\n",
       " 145,\n",
       " 115,\n",
       " 130,\n",
       " 118,\n",
       " 148,\n",
       " 127,\n",
       " 142,\n",
       " 121,\n",
       " 136,\n",
       " 130,\n",
       " 118,\n",
       " 133,\n",
       " 127,\n",
       " 136,\n",
       " 145,\n",
       " 136,\n",
       " 136,\n",
       " 118,\n",
       " 136,\n",
       " 124,\n",
       " 127,\n",
       " 121,\n",
       " 151,\n",
       " 142,\n",
       " 133,\n",
       " 133,\n",
       " 139,\n",
       " 136,\n",
       " 148,\n",
       " 142,\n",
       " 142,\n",
       " 109,\n",
       " 124,\n",
       " 121,\n",
       " 121,\n",
       " 130,\n",
       " 142,\n",
       " 139,\n",
       " 148,\n",
       " 136,\n",
       " 139,\n",
       " 118,\n",
       " 145,\n",
       " 121,\n",
       " 130,\n",
       " 124,\n",
       " 121,\n",
       " 109,\n",
       " 130,\n",
       " 136,\n",
       " 148,\n",
       " 127,\n",
       " 130,\n",
       " 142,\n",
       " 178,\n",
       " 121,\n",
       " 127,\n",
       " 136,\n",
       " 151,\n",
       " 121,\n",
       " 115,\n",
       " 115,\n",
       " 148,\n",
       " 124,\n",
       " 112,\n",
       " 118,\n",
       " 130,\n",
       " 124,\n",
       " 136,\n",
       " 160,\n",
       " 133,\n",
       " 145,\n",
       " 133,\n",
       " 133,\n",
       " 133,\n",
       " 136,\n",
       " 136,\n",
       " 139,\n",
       " 136,\n",
       " 160,\n",
       " 124,\n",
       " 103,\n",
       " 124,\n",
       " 130,\n",
       " 112,\n",
       " 154,\n",
       " 139,\n",
       " 139,\n",
       " 124,\n",
       " 142,\n",
       " 121,\n",
       " 145,\n",
       " 124,\n",
       " 145,\n",
       " 142,\n",
       " 133,\n",
       " 118,\n",
       " 151,\n",
       " 142,\n",
       " 133,\n",
       " 154,\n",
       " 124,\n",
       " 163,\n",
       " 139,\n",
       " 178,\n",
       " 115,\n",
       " 136,\n",
       " 142,\n",
       " 142,\n",
       " 145,\n",
       " 154,\n",
       " 115,\n",
       " 118,\n",
       " 124,\n",
       " 121,\n",
       " 136,\n",
       " 127,\n",
       " 124,\n",
       " 127,\n",
       " 106,\n",
       " 121,\n",
       " 127,\n",
       " 136,\n",
       " 109,\n",
       " 112,\n",
       " 124,\n",
       " 103,\n",
       " 139,\n",
       " 115,\n",
       " 136,\n",
       " 106,\n",
       " 127,\n",
       " 121,\n",
       " 130,\n",
       " 136,\n",
       " 136,\n",
       " 124,\n",
       " 115,\n",
       " 124,\n",
       " 121,\n",
       " 136,\n",
       " 121,\n",
       " 142,\n",
       " 124,\n",
       " 133,\n",
       " 124,\n",
       " 130,\n",
       " 127,\n",
       " 109,\n",
       " 157,\n",
       " 133,\n",
       " 124,\n",
       " 130,\n",
       " 145,\n",
       " 127,\n",
       " 118,\n",
       " 130,\n",
       " 157,\n",
       " 142,\n",
       " 139,\n",
       " 124,\n",
       " 139,\n",
       " 139,\n",
       " 136,\n",
       " 160,\n",
       " 97,\n",
       " 139,\n",
       " 133,\n",
       " 136,\n",
       " 136,\n",
       " 139,\n",
       " 142,\n",
       " 127,\n",
       " 112,\n",
       " 118,\n",
       " 118,\n",
       " 115,\n",
       " 106,\n",
       " 97,\n",
       " 130,\n",
       " 115,\n",
       " 136,\n",
       " 127,\n",
       " 103,\n",
       " 106,\n",
       " 118,\n",
       " 136,\n",
       " 133,\n",
       " 115,\n",
       " 133,\n",
       " 124,\n",
       " 133,\n",
       " 133,\n",
       " 118,\n",
       " 157,\n",
       " 112,\n",
       " 142,\n",
       " 151,\n",
       " 154,\n",
       " 169,\n",
       " 136,\n",
       " 130,\n",
       " 142,\n",
       " 121,\n",
       " 121,\n",
       " 148,\n",
       " 136,\n",
       " 142,\n",
       " 127,\n",
       " 142,\n",
       " 127,\n",
       " 118,\n",
       " 148,\n",
       " 154,\n",
       " 154,\n",
       " 166,\n",
       " 118,\n",
       " 142,\n",
       " 154,\n",
       " 139,\n",
       " 124,\n",
       " 127,\n",
       " 145,\n",
       " 115,\n",
       " 109,\n",
       " 121,\n",
       " 106,\n",
       " 130,\n",
       " 106,\n",
       " 148,\n",
       " 139,\n",
       " 127,\n",
       " 112,\n",
       " 136,\n",
       " 142,\n",
       " 112,\n",
       " 133,\n",
       " 130,\n",
       " 112,\n",
       " 118,\n",
       " 121,\n",
       " 112,\n",
       " 136,\n",
       " 121,\n",
       " 127,\n",
       " 124,\n",
       " 124,\n",
       " 124,\n",
       " 124,\n",
       " 118,\n",
       " 130,\n",
       " 112,\n",
       " 115,\n",
       " 118,\n",
       " 154,\n",
       " 109,\n",
       " 112,\n",
       " 136,\n",
       " 121,\n",
       " 121,\n",
       " 139,\n",
       " 115,\n",
       " 133,\n",
       " 160,\n",
       " 127,\n",
       " 133,\n",
       " 142,\n",
       " 109,\n",
       " 139,\n",
       " 133,\n",
       " 121,\n",
       " 100,\n",
       " 130,\n",
       " 145,\n",
       " 130,\n",
       " 103,\n",
       " 118,\n",
       " 115,\n",
       " 94,\n",
       " 124,\n",
       " 133,\n",
       " 136,\n",
       " 121,\n",
       " 85,\n",
       " 121,\n",
       " 103,\n",
       " 139,\n",
       " 124,\n",
       " 61,\n",
       " 106,\n",
       " 121,\n",
       " 124,\n",
       " 127,\n",
       " 118,\n",
       " 130,\n",
       " 154,\n",
       " 142,\n",
       " 130,\n",
       " 124,\n",
       " 157,\n",
       " 124,\n",
       " 127,\n",
       " 133,\n",
       " 133,\n",
       " 130,\n",
       " 199,\n",
       " 127,\n",
       " 121,\n",
       " 127,\n",
       " 139,\n",
       " 121,\n",
       " 118,\n",
       " 124,\n",
       " 112,\n",
       " 160,\n",
       " 127,\n",
       " 139,\n",
       " 124,\n",
       " 118,\n",
       " 115,\n",
       " 127,\n",
       " 118,\n",
       " 115,\n",
       " 130,\n",
       " 160,\n",
       " 103,\n",
       " 151,\n",
       " 118,\n",
       " 109,\n",
       " 118,\n",
       " 103,\n",
       " 121,\n",
       " 130,\n",
       " 127,\n",
       " 121,\n",
       " 103,\n",
       " 130,\n",
       " ...]"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dim1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "D:\\anaconda3\\envs\\mytfenv\\lib\\site-packages\\seaborn\\_decorators.py:43: FutureWarning: Pass the following variables as keyword args: x, y. From version 0.12, the only valid positional argument will be `data`, and passing other arguments without an explicit keyword will result in an error or misinterpretation.\n",
      "  FutureWarning\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.jointplot(dim1,dim2);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "130.92538461538462"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(dim1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "130.75"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(dim2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "image_shape = (130,130,3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Processing \n",
    "\n",
    "- ImageDataGenerator "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "50700"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "130*130*3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Help on class ImageDataGenerator in module tensorflow.python.keras.preprocessing.image:\n",
      "\n",
      "class ImageDataGenerator(keras_preprocessing.image.image_data_generator.ImageDataGenerator)\n",
      " |  ImageDataGenerator(featurewise_center=False, samplewise_center=False, featurewise_std_normalization=False, samplewise_std_normalization=False, zca_whitening=False, zca_epsilon=1e-06, rotation_range=0, width_shift_range=0.0, height_shift_range=0.0, brightness_range=None, shear_range=0.0, zoom_range=0.0, channel_shift_range=0.0, fill_mode='nearest', cval=0.0, horizontal_flip=False, vertical_flip=False, rescale=None, preprocessing_function=None, data_format=None, validation_split=0.0, dtype=None)\n",
      " |  \n",
      " |  Generate batches of tensor image data with real-time data augmentation.\n",
      " |  \n",
      " |   The data will be looped over (in batches).\n",
      " |  \n",
      " |  Arguments:\n",
      " |      featurewise_center: Boolean.\n",
      " |          Set input mean to 0 over the dataset, feature-wise.\n",
      " |      samplewise_center: Boolean. Set each sample mean to 0.\n",
      " |      featurewise_std_normalization: Boolean.\n",
      " |          Divide inputs by std of the dataset, feature-wise.\n",
      " |      samplewise_std_normalization: Boolean. Divide each input by its std.\n",
      " |      zca_epsilon: epsilon for ZCA whitening. Default is 1e-6.\n",
      " |      zca_whitening: Boolean. Apply ZCA whitening.\n",
      " |      rotation_range: Int. Degree range for random rotations.\n",
      " |      width_shift_range: Float, 1-D array-like or int\n",
      " |          - float: fraction of total width, if < 1, or pixels if >= 1.\n",
      " |          - 1-D array-like: random elements from the array.\n",
      " |          - int: integer number of pixels from interval\n",
      " |              `(-width_shift_range, +width_shift_range)`\n",
      " |          - With `width_shift_range=2` possible values\n",
      " |              are integers `[-1, 0, +1]`,\n",
      " |              same as with `width_shift_range=[-1, 0, +1]`,\n",
      " |              while with `width_shift_range=1.0` possible values are floats\n",
      " |              in the interval [-1.0, +1.0).\n",
      " |      height_shift_range: Float, 1-D array-like or int\n",
      " |          - float: fraction of total height, if < 1, or pixels if >= 1.\n",
      " |          - 1-D array-like: random elements from the array.\n",
      " |          - int: integer number of pixels from interval\n",
      " |              `(-height_shift_range, +height_shift_range)`\n",
      " |          - With `height_shift_range=2` possible values\n",
      " |              are integers `[-1, 0, +1]`,\n",
      " |              same as with `height_shift_range=[-1, 0, +1]`,\n",
      " |              while with `height_shift_range=1.0` possible values are floats\n",
      " |              in the interval [-1.0, +1.0).\n",
      " |      brightness_range: Tuple or list of two floats. Range for picking\n",
      " |          a brightness shift value from.\n",
      " |      shear_range: Float. Shear Intensity\n",
      " |          (Shear angle in counter-clockwise direction in degrees)\n",
      " |      zoom_range: Float or [lower, upper]. Range for random zoom.\n",
      " |          If a float, `[lower, upper] = [1-zoom_range, 1+zoom_range]`.\n",
      " |      channel_shift_range: Float. Range for random channel shifts.\n",
      " |      fill_mode: One of {\"constant\", \"nearest\", \"reflect\" or \"wrap\"}.\n",
      " |          Default is 'nearest'.\n",
      " |          Points outside the boundaries of the input are filled\n",
      " |          according to the given mode:\n",
      " |          - 'constant': kkkkkkkk|abcd|kkkkkkkk (cval=k)\n",
      " |          - 'nearest':  aaaaaaaa|abcd|dddddddd\n",
      " |          - 'reflect':  abcddcba|abcd|dcbaabcd\n",
      " |          - 'wrap':  abcdabcd|abcd|abcdabcd\n",
      " |      cval: Float or Int.\n",
      " |          Value used for points outside the boundaries\n",
      " |          when `fill_mode = \"constant\"`.\n",
      " |      horizontal_flip: Boolean. Randomly flip inputs horizontally.\n",
      " |      vertical_flip: Boolean. Randomly flip inputs vertically.\n",
      " |      rescale: rescaling factor. Defaults to None.\n",
      " |          If None or 0, no rescaling is applied,\n",
      " |          otherwise we multiply the data by the value provided\n",
      " |          (after applying all other transformations).\n",
      " |      preprocessing_function: function that will be applied on each input.\n",
      " |          The function will run after the image is resized and augmented.\n",
      " |          The function should take one argument:\n",
      " |          one image (Numpy tensor with rank 3),\n",
      " |          and should output a Numpy tensor with the same shape.\n",
      " |      data_format: Image data format,\n",
      " |          either \"channels_first\" or \"channels_last\".\n",
      " |          \"channels_last\" mode means that the images should have shape\n",
      " |          `(samples, height, width, channels)`,\n",
      " |          \"channels_first\" mode means that the images should have shape\n",
      " |          `(samples, channels, height, width)`.\n",
      " |          It defaults to the `image_data_format` value found in your\n",
      " |          Keras config file at `~/.keras/keras.json`.\n",
      " |          If you never set it, then it will be \"channels_last\".\n",
      " |      validation_split: Float. Fraction of images reserved for validation\n",
      " |          (strictly between 0 and 1).\n",
      " |      dtype: Dtype to use for the generated arrays.\n",
      " |  \n",
      " |  Examples:\n",
      " |  \n",
      " |  Example of using `.flow(x, y)`:\n",
      " |  \n",
      " |  ```python\n",
      " |  (x_train, y_train), (x_test, y_test) = cifar10.load_data()\n",
      " |  y_train = np_utils.to_categorical(y_train, num_classes)\n",
      " |  y_test = np_utils.to_categorical(y_test, num_classes)\n",
      " |  datagen = ImageDataGenerator(\n",
      " |      featurewise_center=True,\n",
      " |      featurewise_std_normalization=True,\n",
      " |      rotation_range=20,\n",
      " |      width_shift_range=0.2,\n",
      " |      height_shift_range=0.2,\n",
      " |      horizontal_flip=True)\n",
      " |  # compute quantities required for featurewise normalization\n",
      " |  # (std, mean, and principal components if ZCA whitening is applied)\n",
      " |  datagen.fit(x_train)\n",
      " |  # fits the model on batches with real-time data augmentation:\n",
      " |  model.fit(datagen.flow(x_train, y_train, batch_size=32),\n",
      " |            steps_per_epoch=len(x_train) / 32, epochs=epochs)\n",
      " |  # here's a more \"manual\" example\n",
      " |  for e in range(epochs):\n",
      " |      print('Epoch', e)\n",
      " |      batches = 0\n",
      " |      for x_batch, y_batch in datagen.flow(x_train, y_train, batch_size=32):\n",
      " |          model.fit(x_batch, y_batch)\n",
      " |          batches += 1\n",
      " |          if batches >= len(x_train) / 32:\n",
      " |              # we need to break the loop by hand because\n",
      " |              # the generator loops indefinitely\n",
      " |              break\n",
      " |  ```\n",
      " |  \n",
      " |  Example of using `.flow_from_directory(directory)`:\n",
      " |  \n",
      " |  ```python\n",
      " |  train_datagen = ImageDataGenerator(\n",
      " |          rescale=1./255,\n",
      " |          shear_range=0.2,\n",
      " |          zoom_range=0.2,\n",
      " |          horizontal_flip=True)\n",
      " |  test_datagen = ImageDataGenerator(rescale=1./255)\n",
      " |  train_generator = train_datagen.flow_from_directory(\n",
      " |          'data/train',\n",
      " |          target_size=(150, 150),\n",
      " |          batch_size=32,\n",
      " |          class_mode='binary')\n",
      " |  validation_generator = test_datagen.flow_from_directory(\n",
      " |          'data/validation',\n",
      " |          target_size=(150, 150),\n",
      " |          batch_size=32,\n",
      " |          class_mode='binary')\n",
      " |  model.fit(\n",
      " |          train_generator,\n",
      " |          steps_per_epoch=2000,\n",
      " |          epochs=50,\n",
      " |          validation_data=validation_generator,\n",
      " |          validation_steps=800)\n",
      " |  ```\n",
      " |  \n",
      " |  Example of transforming images and masks together.\n",
      " |  \n",
      " |  ```python\n",
      " |  # we create two instances with the same arguments\n",
      " |  data_gen_args = dict(featurewise_center=True,\n",
      " |                       featurewise_std_normalization=True,\n",
      " |                       rotation_range=90,\n",
      " |                       width_shift_range=0.1,\n",
      " |                       height_shift_range=0.1,\n",
      " |                       zoom_range=0.2)\n",
      " |  image_datagen = ImageDataGenerator(**data_gen_args)\n",
      " |  mask_datagen = ImageDataGenerator(**data_gen_args)\n",
      " |  # Provide the same seed and keyword arguments to the fit and flow methods\n",
      " |  seed = 1\n",
      " |  image_datagen.fit(images, augment=True, seed=seed)\n",
      " |  mask_datagen.fit(masks, augment=True, seed=seed)\n",
      " |  image_generator = image_datagen.flow_from_directory(\n",
      " |      'data/images',\n",
      " |      class_mode=None,\n",
      " |      seed=seed)\n",
      " |  mask_generator = mask_datagen.flow_from_directory(\n",
      " |      'data/masks',\n",
      " |      class_mode=None,\n",
      " |      seed=seed)\n",
      " |  # combine generators into one which yields image and masks\n",
      " |  train_generator = zip(image_generator, mask_generator)\n",
      " |  model.fit_generator(\n",
      " |      train_generator,\n",
      " |      steps_per_epoch=2000,\n",
      " |      epochs=50)\n",
      " |  ```\n",
      " |  \n",
      " |  Method resolution order:\n",
      " |      ImageDataGenerator\n",
      " |      keras_preprocessing.image.image_data_generator.ImageDataGenerator\n",
      " |      builtins.object\n",
      " |  \n",
      " |  Methods defined here:\n",
      " |  \n",
      " |  __init__(self, featurewise_center=False, samplewise_center=False, featurewise_std_normalization=False, samplewise_std_normalization=False, zca_whitening=False, zca_epsilon=1e-06, rotation_range=0, width_shift_range=0.0, height_shift_range=0.0, brightness_range=None, shear_range=0.0, zoom_range=0.0, channel_shift_range=0.0, fill_mode='nearest', cval=0.0, horizontal_flip=False, vertical_flip=False, rescale=None, preprocessing_function=None, data_format=None, validation_split=0.0, dtype=None)\n",
      " |      Initialize self.  See help(type(self)) for accurate signature.\n",
      " |  \n",
      " |  flow(self, x, y=None, batch_size=32, shuffle=True, sample_weight=None, seed=None, save_to_dir=None, save_prefix='', save_format='png', subset=None)\n",
      " |      Takes data & label arrays, generates batches of augmented data.\n",
      " |      \n",
      " |      Arguments:\n",
      " |          x: Input data. Numpy array of rank 4 or a tuple. If tuple, the first\n",
      " |            element should contain the images and the second element another numpy\n",
      " |            array or a list of numpy arrays that gets passed to the output without\n",
      " |            any modifications. Can be used to feed the model miscellaneous data\n",
      " |            along with the images. In case of grayscale data, the channels axis of\n",
      " |            the image array should have value 1, in case of RGB data, it should\n",
      " |            have value 3, and in case of RGBA data, it should have value 4.\n",
      " |          y: Labels.\n",
      " |          batch_size: Int (default: 32).\n",
      " |          shuffle: Boolean (default: True).\n",
      " |          sample_weight: Sample weights.\n",
      " |          seed: Int (default: None).\n",
      " |          save_to_dir: None or str (default: None). This allows you to optionally\n",
      " |            specify a directory to which to save the augmented pictures being\n",
      " |            generated (useful for visualizing what you are doing).\n",
      " |          save_prefix: Str (default: `''`). Prefix to use for filenames of saved\n",
      " |            pictures (only relevant if `save_to_dir` is set).\n",
      " |          save_format: one of \"png\", \"jpeg\"\n",
      " |              (only relevant if `save_to_dir` is set). Default: \"png\".\n",
      " |          subset: Subset of data (`\"training\"` or `\"validation\"`) if\n",
      " |            `validation_split` is set in `ImageDataGenerator`.\n",
      " |      \n",
      " |      Returns:\n",
      " |          An `Iterator` yielding tuples of `(x, y)`\n",
      " |              where `x` is a numpy array of image data\n",
      " |              (in the case of a single image input) or a list\n",
      " |              of numpy arrays (in the case with\n",
      " |              additional inputs) and `y` is a numpy array\n",
      " |              of corresponding labels. If 'sample_weight' is not None,\n",
      " |              the yielded tuples are of the form `(x, y, sample_weight)`.\n",
      " |              If `y` is None, only the numpy array `x` is returned.\n",
      " |  \n",
      " |  flow_from_dataframe(self, dataframe, directory=None, x_col='filename', y_col='class', weight_col=None, target_size=(256, 256), color_mode='rgb', classes=None, class_mode='categorical', batch_size=32, shuffle=True, seed=None, save_to_dir=None, save_prefix='', save_format='png', subset=None, interpolation='nearest', validate_filenames=True, **kwargs)\n",
      " |      Takes the dataframe and the path to a directory + generates batches.\n",
      " |      \n",
      " |       The generated batches contain augmented/normalized data.\n",
      " |      \n",
      " |      **A simple tutorial can be found **[here](\n",
      " |                                  http://bit.ly/keras_flow_from_dataframe).\n",
      " |      \n",
      " |      Arguments:\n",
      " |          dataframe: Pandas dataframe containing the filepaths relative to\n",
      " |            `directory` (or absolute paths if `directory` is None) of the images\n",
      " |            in a string column. It should include other column/s\n",
      " |              depending on the `class_mode`: - if `class_mode` is `\"categorical\"`\n",
      " |                (default value) it must include the `y_col` column with the\n",
      " |                class/es of each image. Values in column can be string/list/tuple\n",
      " |                if a single class or list/tuple if multiple classes. - if\n",
      " |                `class_mode` is `\"binary\"` or `\"sparse\"` it must include the given\n",
      " |                `y_col` column with class values as strings. - if `class_mode` is\n",
      " |                `\"raw\"` or `\"multi_output\"` it should contain the columns\n",
      " |                specified in `y_col`. - if `class_mode` is `\"input\"` or `None` no\n",
      " |                extra column is needed.\n",
      " |          directory: string, path to the directory to read images from. If `None`,\n",
      " |            data in `x_col` column should be absolute paths.\n",
      " |          x_col: string, column in `dataframe` that contains the filenames (or\n",
      " |            absolute paths if `directory` is `None`).\n",
      " |          y_col: string or list, column/s in `dataframe` that has the target data.\n",
      " |          weight_col: string, column in `dataframe` that contains the sample\n",
      " |              weights. Default: `None`.\n",
      " |          target_size: tuple of integers `(height, width)`, default: `(256, 256)`.\n",
      " |            The dimensions to which all images found will be resized.\n",
      " |          color_mode: one of \"grayscale\", \"rgb\", \"rgba\". Default: \"rgb\". Whether\n",
      " |            the images will be converted to have 1 or 3 color channels.\n",
      " |          classes: optional list of classes (e.g. `['dogs', 'cats']`). Default is\n",
      " |            None. If not provided, the list of classes will be automatically\n",
      " |            inferred from the `y_col`, which will map to the label indices, will\n",
      " |            be alphanumeric). The dictionary containing the mapping from class\n",
      " |            names to class indices can be obtained via the attribute\n",
      " |            `class_indices`.\n",
      " |          class_mode: one of \"binary\", \"categorical\", \"input\", \"multi_output\",\n",
      " |              \"raw\", sparse\" or None. Default: \"categorical\".\n",
      " |              Mode for yielding the targets:\n",
      " |              - `\"binary\"`: 1D numpy array of binary labels,\n",
      " |              - `\"categorical\"`: 2D numpy array of one-hot encoded labels.\n",
      " |                Supports multi-label output.\n",
      " |              - `\"input\"`: images identical to input images (mainly used to work\n",
      " |                with autoencoders),\n",
      " |              - `\"multi_output\"`: list with the values of the different columns,\n",
      " |              - `\"raw\"`: numpy array of values in `y_col` column(s),\n",
      " |              - `\"sparse\"`: 1D numpy array of integer labels, - `None`, no targets\n",
      " |                are returned (the generator will only yield batches of image data,\n",
      " |                which is useful to use in `model.predict_generator()`).\n",
      " |          batch_size: size of the batches of data (default: 32).\n",
      " |          shuffle: whether to shuffle the data (default: True)\n",
      " |          seed: optional random seed for shuffling and transformations.\n",
      " |          save_to_dir: None or str (default: None). This allows you to optionally\n",
      " |            specify a directory to which to save the augmented pictures being\n",
      " |            generated (useful for visualizing what you are doing).\n",
      " |          save_prefix: str. Prefix to use for filenames of saved pictures (only\n",
      " |            relevant if `save_to_dir` is set).\n",
      " |          save_format: one of \"png\", \"jpeg\"\n",
      " |              (only relevant if `save_to_dir` is set). Default: \"png\".\n",
      " |          subset: Subset of data (`\"training\"` or `\"validation\"`) if\n",
      " |            `validation_split` is set in `ImageDataGenerator`.\n",
      " |          interpolation: Interpolation method used to resample the image if the\n",
      " |            target size is different from that of the loaded image. Supported\n",
      " |            methods are `\"nearest\"`, `\"bilinear\"`, and `\"bicubic\"`. If PIL version\n",
      " |            1.1.3 or newer is installed, `\"lanczos\"` is also supported. If PIL\n",
      " |            version 3.4.0 or newer is installed, `\"box\"` and `\"hamming\"` are also\n",
      " |            supported. By default, `\"nearest\"` is used.\n",
      " |          validate_filenames: Boolean, whether to validate image filenames in\n",
      " |            `x_col`. If `True`, invalid images will be ignored. Disabling this\n",
      " |            option can lead to speed-up in the execution of this function.\n",
      " |            Defaults to `True`.\n",
      " |          **kwargs: legacy arguments for raising deprecation warnings.\n",
      " |      \n",
      " |      Returns:\n",
      " |          A `DataFrameIterator` yielding tuples of `(x, y)`\n",
      " |          where `x` is a numpy array containing a batch\n",
      " |          of images with shape `(batch_size, *target_size, channels)`\n",
      " |          and `y` is a numpy array of corresponding labels.\n",
      " |  \n",
      " |  flow_from_directory(self, directory, target_size=(256, 256), color_mode='rgb', classes=None, class_mode='categorical', batch_size=32, shuffle=True, seed=None, save_to_dir=None, save_prefix='', save_format='png', follow_links=False, subset=None, interpolation='nearest')\n",
      " |      Takes the path to a directory & generates batches of augmented data.\n",
      " |      \n",
      " |      Arguments:\n",
      " |          directory: string, path to the target directory. It should contain one\n",
      " |            subdirectory per class. Any PNG, JPG, BMP, PPM or TIF images inside\n",
      " |            each of the subdirectories directory tree will be included in the\n",
      " |            generator. See [this script](\n",
      " |              https://gist.github.com/fchollet/0830affa1f7f19fd47b06d4cf89ed44d)\n",
      " |                for more details.\n",
      " |          target_size: Tuple of integers `(height, width)`, defaults to `(256,\n",
      " |            256)`. The dimensions to which all images found will be resized.\n",
      " |          color_mode: One of \"grayscale\", \"rgb\", \"rgba\". Default: \"rgb\". Whether\n",
      " |            the images will be converted to have 1, 3, or 4 channels.\n",
      " |          classes: Optional list of class subdirectories\n",
      " |              (e.g. `['dogs', 'cats']`). Default: None. If not provided, the list\n",
      " |                of classes will be automatically inferred from the subdirectory\n",
      " |                names/structure under `directory`, where each subdirectory will be\n",
      " |                treated as a different class (and the order of the classes, which\n",
      " |                will map to the label indices, will be alphanumeric). The\n",
      " |                dictionary containing the mapping from class names to class\n",
      " |                indices can be obtained via the attribute `class_indices`.\n",
      " |          class_mode: One of \"categorical\", \"binary\", \"sparse\",\n",
      " |              \"input\", or None. Default: \"categorical\".\n",
      " |              Determines the type of label arrays that are returned: -\n",
      " |                \"categorical\" will be 2D one-hot encoded labels, - \"binary\" will\n",
      " |                be 1D binary labels, \"sparse\" will be 1D integer labels, - \"input\"\n",
      " |                will be images identical to input images (mainly used to work with\n",
      " |                autoencoders). - If None, no labels are returned (the generator\n",
      " |                will only yield batches of image data, which is useful to use with\n",
      " |                `model.predict_generator()`). Please note that in case of\n",
      " |                class_mode None, the data still needs to reside in a subdirectory\n",
      " |                of `directory` for it to work correctly.\n",
      " |          batch_size: Size of the batches of data (default: 32).\n",
      " |          shuffle: Whether to shuffle the data (default: True) If set to False,\n",
      " |            sorts the data in alphanumeric order.\n",
      " |          seed: Optional random seed for shuffling and transformations.\n",
      " |          save_to_dir: None or str (default: None). This allows you to optionally\n",
      " |            specify a directory to which to save the augmented pictures being\n",
      " |            generated (useful for visualizing what you are doing).\n",
      " |          save_prefix: Str. Prefix to use for filenames of saved pictures (only\n",
      " |            relevant if `save_to_dir` is set).\n",
      " |          save_format: One of \"png\", \"jpeg\"\n",
      " |              (only relevant if `save_to_dir` is set). Default: \"png\".\n",
      " |          follow_links: Whether to follow symlinks inside\n",
      " |              class subdirectories (default: False).\n",
      " |          subset: Subset of data (`\"training\"` or `\"validation\"`) if\n",
      " |            `validation_split` is set in `ImageDataGenerator`.\n",
      " |          interpolation: Interpolation method used to resample the image if the\n",
      " |            target size is different from that of the loaded image. Supported\n",
      " |            methods are `\"nearest\"`, `\"bilinear\"`, and `\"bicubic\"`. If PIL version\n",
      " |            1.1.3 or newer is installed, `\"lanczos\"` is also supported. If PIL\n",
      " |            version 3.4.0 or newer is installed, `\"box\"` and `\"hamming\"` are also\n",
      " |            supported. By default, `\"nearest\"` is used.\n",
      " |      \n",
      " |      Returns:\n",
      " |          A `DirectoryIterator` yielding tuples of `(x, y)`\n",
      " |              where `x` is a numpy array containing a batch\n",
      " |              of images with shape `(batch_size, *target_size, channels)`\n",
      " |              and `y` is a numpy array of corresponding labels.\n",
      " |  \n",
      " |  ----------------------------------------------------------------------\n",
      " |  Methods inherited from keras_preprocessing.image.image_data_generator.ImageDataGenerator:\n",
      " |  \n",
      " |  apply_transform(self, x, transform_parameters)\n",
      " |      Applies a transformation to an image according to given parameters.\n",
      " |      \n",
      " |      # Arguments\n",
      " |          x: 3D tensor, single image.\n",
      " |          transform_parameters: Dictionary with string - parameter pairs\n",
      " |              describing the transformation.\n",
      " |              Currently, the following parameters\n",
      " |              from the dictionary are used:\n",
      " |              - `'theta'`: Float. Rotation angle in degrees.\n",
      " |              - `'tx'`: Float. Shift in the x direction.\n",
      " |              - `'ty'`: Float. Shift in the y direction.\n",
      " |              - `'shear'`: Float. Shear angle in degrees.\n",
      " |              - `'zx'`: Float. Zoom in the x direction.\n",
      " |              - `'zy'`: Float. Zoom in the y direction.\n",
      " |              - `'flip_horizontal'`: Boolean. Horizontal flip.\n",
      " |              - `'flip_vertical'`: Boolean. Vertical flip.\n",
      " |              - `'channel_shift_intensity'`: Float. Channel shift intensity.\n",
      " |              - `'brightness'`: Float. Brightness shift intensity.\n",
      " |      \n",
      " |      # Returns\n",
      " |          A transformed version of the input (same shape).\n",
      " |  \n",
      " |  fit(self, x, augment=False, rounds=1, seed=None)\n",
      " |      Fits the data generator to some sample data.\n",
      " |      \n",
      " |      This computes the internal data stats related to the\n",
      " |      data-dependent transformations, based on an array of sample data.\n",
      " |      \n",
      " |      Only required if `featurewise_center` or\n",
      " |      `featurewise_std_normalization` or `zca_whitening` are set to True.\n",
      " |      \n",
      " |      When `rescale` is set to a value, rescaling is applied to\n",
      " |      sample data before computing the internal data stats.\n",
      " |      \n",
      " |      # Arguments\n",
      " |          x: Sample data. Should have rank 4.\n",
      " |           In case of grayscale data,\n",
      " |           the channels axis should have value 1, in case\n",
      " |           of RGB data, it should have value 3, and in case\n",
      " |           of RGBA data, it should have value 4.\n",
      " |          augment: Boolean (default: False).\n",
      " |              Whether to fit on randomly augmented samples.\n",
      " |          rounds: Int (default: 1).\n",
      " |              If using data augmentation (`augment=True`),\n",
      " |              this is how many augmentation passes over the data to use.\n",
      " |          seed: Int (default: None). Random seed.\n",
      " |  \n",
      " |  get_random_transform(self, img_shape, seed=None)\n",
      " |      Generates random parameters for a transformation.\n",
      " |      \n",
      " |      # Arguments\n",
      " |          seed: Random seed.\n",
      " |          img_shape: Tuple of integers.\n",
      " |              Shape of the image that is transformed.\n",
      " |      \n",
      " |      # Returns\n",
      " |          A dictionary containing randomly chosen parameters describing the\n",
      " |          transformation.\n",
      " |  \n",
      " |  random_transform(self, x, seed=None)\n",
      " |      Applies a random transformation to an image.\n",
      " |      \n",
      " |      # Arguments\n",
      " |          x: 3D tensor, single image.\n",
      " |          seed: Random seed.\n",
      " |      \n",
      " |      # Returns\n",
      " |          A randomly transformed version of the input (same shape).\n",
      " |  \n",
      " |  standardize(self, x)\n",
      " |      Applies the normalization configuration in-place to a batch of inputs.\n",
      " |      \n",
      " |      `x` is changed in-place since the function is mainly used internally\n",
      " |      to standardize images and feed them to your network. If a copy of `x`\n",
      " |      would be created instead it would have a significant performance cost.\n",
      " |      If you want to apply this method without changing the input in-place\n",
      " |      you can call the method creating a copy before:\n",
      " |      \n",
      " |      standardize(np.copy(x))\n",
      " |      \n",
      " |      # Arguments\n",
      " |          x: Batch of inputs to be normalized.\n",
      " |      \n",
      " |      # Returns\n",
      " |          The inputs, normalized.\n",
      " |  \n",
      " |  ----------------------------------------------------------------------\n",
      " |  Data descriptors inherited from keras_preprocessing.image.image_data_generator.ImageDataGenerator:\n",
      " |  \n",
      " |  __dict__\n",
      " |      dictionary for instance variables (if defined)\n",
      " |  \n",
      " |  __weakref__\n",
      " |      list of weak references to the object (if defined)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "help(ImageDataGenerator)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "image_gen = ImageDataGenerator(rotation_range=20,\n",
    "                              width_shift_range=0.1,\n",
    "                              height_shift_range=0.1,\n",
    "                              shear_range=0.1,\n",
    "                              zoom_range=0.1,\n",
    "                              horizontal_flip=True,\n",
    "                              fill_mode='nearest')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 24958 images belonging to 2 classes.\n"
     ]
    }
   ],
   "source": [
    "image_gen.flow_from_directory(train_path);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.models import Sequential"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.layers import Dense,MaxPool2D,Conv2D,Flatten,Dropout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "model =Sequential()\n",
    "\n",
    "#First Layer\n",
    "model.add(Conv2D(filters=32,\n",
    "                 kernel_size=(3,3),\n",
    "                 input_shape=image_shape,\n",
    "                 activation='relu'))\n",
    "\n",
    "model.add(MaxPool2D(pool_size=(2,2)))\n",
    "\n",
    "#Second layer\n",
    "model.add(Conv2D(filters=64,\n",
    "                 kernel_size=(3,3),\n",
    "                activation='relu'))\n",
    "\n",
    "model.add(MaxPool2D(pool_size=(2,2)))\n",
    "\n",
    "#Third layer\n",
    "model.add(Conv2D(filters=64,\n",
    "                 kernel_size=(3,3),\n",
    "                activation='relu'))\n",
    "\n",
    "model.add(MaxPool2D(pool_size=(2,2)))\n",
    "\n",
    "#flatten \n",
    "model.add(Flatten())\n",
    "\n",
    "#Dense Layer\n",
    "model.add(Dense(128,\n",
    "                activation='relu'))\n",
    "\n",
    "#prevent Overfitting\n",
    "model.add(Dropout(0.5))\n",
    "\n",
    "#Final Dense Layer\n",
    "model.add(Dense(1,\n",
    "                activation='sigmoid'))\n",
    "\n",
    "#Compile Model\n",
    "model.compile(loss='binary_crossentropy',\n",
    "              optimizer='adam',\n",
    "              metrics=['accuracy'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_5\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_9 (Conv2D)            (None, 128, 128, 32)      896       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_9 (MaxPooling2 (None, 64, 64, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_10 (Conv2D)           (None, 62, 62, 64)        18496     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_10 (MaxPooling (None, 31, 31, 64)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_11 (Conv2D)           (None, 29, 29, 64)        36928     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_11 (MaxPooling (None, 14, 14, 64)        0         \n",
      "_________________________________________________________________\n",
      "flatten_2 (Flatten)          (None, 12544)             0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 128)               1605760   \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 1)                 129       \n",
      "=================================================================\n",
      "Total params: 1,662,209\n",
      "Trainable params: 1,662,209\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Call Back\n",
    "from tensorflow.keras.callbacks import EarlyStopping"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "early_stop =EarlyStopping(monitor='val_loss',patience=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "#2**4\n",
    "batch_size=16 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 24958 images belonging to 2 classes.\n"
     ]
    }
   ],
   "source": [
    "train_image_gen = image_gen.flow_from_directory(train_path,\n",
    "                                               target_size=image_shape[:2],\n",
    "                                               color_mode='rgb',\n",
    "                                               batch_size=batch_size,\n",
    "                                               class_mode='binary')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 2600 images belonging to 2 classes.\n"
     ]
    }
   ],
   "source": [
    "test_image_gen = image_gen.flow_from_directory(test_path,\n",
    "                                               target_size=image_shape[:2],\n",
    "                                               color_mode='rgb',\n",
    "                                               batch_size=batch_size,\n",
    "                                               class_mode='binary',\n",
    "                                               shuffle=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'parasitized': 0, 'uninfected': 1}"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_image_gen.class_indices"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "1560/1560 [==============================] - 285s 183ms/step - loss: 0.2391 - accuracy: 0.9254 - val_loss: 0.2169 - val_accuracy: 0.9323\n",
      "Epoch 2/20\n",
      "1560/1560 [==============================] - 281s 180ms/step - loss: 0.1953 - accuracy: 0.9404 - val_loss: 0.1857 - val_accuracy: 0.9373\n",
      "Epoch 3/20\n",
      "1560/1560 [==============================] - 279s 179ms/step - loss: 0.1828 - accuracy: 0.9445 - val_loss: 0.1706 - val_accuracy: 0.9458\n",
      "Epoch 4/20\n",
      "1560/1560 [==============================] - 280s 180ms/step - loss: 0.1746 - accuracy: 0.9463 - val_loss: 0.1763 - val_accuracy: 0.9431\n",
      "Epoch 5/20\n",
      "1560/1560 [==============================] - 282s 181ms/step - loss: 0.1859 - accuracy: 0.9418 - val_loss: 0.1762 - val_accuracy: 0.9427\n"
     ]
    }
   ],
   "source": [
    "resul =model.fit_generator(train_image_gen,\n",
    "                            epochs=20,\n",
    "                            validation_data=test_image_gen,\n",
    "                            callbacks=[early_stop])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.models import load_model\n",
    "model.save('malaria_detector.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "losses = pd.DataFrame(model.history.history)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>loss</th>\n",
       "      <th>accuracy</th>\n",
       "      <th>val_loss</th>\n",
       "      <th>val_accuracy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.239130</td>\n",
       "      <td>0.925435</td>\n",
       "      <td>0.216925</td>\n",
       "      <td>0.932308</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.195330</td>\n",
       "      <td>0.940420</td>\n",
       "      <td>0.185721</td>\n",
       "      <td>0.937308</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.182850</td>\n",
       "      <td>0.944467</td>\n",
       "      <td>0.170634</td>\n",
       "      <td>0.945769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.174607</td>\n",
       "      <td>0.946270</td>\n",
       "      <td>0.176304</td>\n",
       "      <td>0.943077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.185863</td>\n",
       "      <td>0.941822</td>\n",
       "      <td>0.176211</td>\n",
       "      <td>0.942692</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       loss  accuracy  val_loss  val_accuracy\n",
       "0  0.239130  0.925435  0.216925      0.932308\n",
       "1  0.195330  0.940420  0.185721      0.937308\n",
       "2  0.182850  0.944467  0.170634      0.945769\n",
       "3  0.174607  0.946270  0.176304      0.943077\n",
       "4  0.185863  0.941822  0.176211      0.942692"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "losses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAD4CAYAAADiry33AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAAA1R0lEQVR4nO3dd3gVZfbA8e9JB0KoSYAkJPTeJFSlqDQb2AFBxV5Wd9XVXXZtiLq6srvu/lZd7OAuCogNRQSUqhQJnYSWhAAJkEILENLf3x9zwYAJuYF779zcnM/z8JA7M3fek4F7Mjln5h0xxqCUUsp3+dkdgFJKKffSRK+UUj5OE71SSvk4TfRKKeXjNNErpZSPC7A7gHM1btzYxMXF2R2GUkpVK+vWrcsxxoSXt87rEn1cXBwJCQl2h6GUUtWKiOypaJ2WbpRSysdpoldKKR+niV4ppXycUzV6ERkB/AvwB94zxrx6zvongHuBYiAbuNsYs6fM+jAgCfjSGPOIi2JXSvmQoqIi0tPTyc/PtzsUrxYSEkJ0dDSBgYFOv6fSRC8i/sCbwFAgHVgrInONMUllNtsAxBtj8kTkIeA1YHSZ9S8Cy52OSilV46Snp1O3bl3i4uIQEbvD8UrGGA4dOkR6ejotWrRw+n3OlG56A8nGmFRjTCEwExh1zuBLjDF5jpergejT60SkJxAJLHQ6KqVUjZOfn0+jRo00yZ+HiNCoUaMq/9bjTKKPAvaVeZ3uWFaRe4D5jqD8gL8DT55vABG5X0QSRCQhOzvbiZCUUr5Ik3zlLuQYubQZKyLjgXhgimPRw8C3xpj0873PGPOOMSbeGBMfHl7u9f6VOllQzF+/287eQ3mVb6yUUjWIM83YDCCmzOtox7KziMgQ4GlgkDGmwLG4HzBARB4GQoEgETlhjJl4cWH/2vH8YqavTGPnweO8P6GXq3evlKoBQkNDOXHihN1huJwzZ/RrgTYi0kJEgoAxwNyyG4hID+BtYKQxJuv0cmPMOGNMc2NMHFb55iN3JHmAJvVCeGxIG37YnsX3SZnuGEIppaqlShO9MaYYeARYAGwDZhtjEkVksoiMdGw2BeuM/VMR2SgicyvYnVvddWkL2kSEMunrRPKLSuwIQSnlA4wxPPXUU3Tu3JkuXbowa9YsAA4cOMDAgQPp3r07nTt3ZsWKFZSUlDBhwoQz277++us2R/9rTl1Hb4z5Fvj2nGXPlfl6iBP7mAZMq1p4VRPo78fkUZ0Z++5q3lqawhND27pzOKWUm7zwdSJJ+3Ndus+OzcJ4/rpOTm37+eefs3HjRjZt2kROTg69evVi4MCBfPzxxwwfPpynn36akpIS8vLy2LhxIxkZGWzduhWAo0ePujRuV/C5O2P7tWrEqO7NmLoshbSck3aHo5Sqhn788UfGjh2Lv78/kZGRDBo0iLVr19KrVy8+/PBDJk2axJYtW6hbty4tW7YkNTWVRx99lO+++46wsDC7w/8Vr5u90hX+fHUHftiWxaSvE/lwQi+9ZEupasbZM29PGzhwIMuXL2fevHlMmDCBJ554gjvuuINNmzaxYMECpk6dyuzZs/nggw/sDvUsPndGDxAZZjVml+7IZpE2ZpVSVTRgwABmzZpFSUkJ2dnZLF++nN69e7Nnzx4iIyO57777uPfee1m/fj05OTmUlpZy00038dJLL7F+/Xq7w/8VnzyjB7izfxyfJqTzwtdJDGgTTq0gf7tDUkpVEzfccAOrVq2iW7duiAivvfYaTZo0Yfr06UyZMoXAwEBCQ0P56KOPyMjI4K677qK0tBSAV155xebof02MMXbHcJb4+HjjqgePrEk9xOh3VvPoFa35/bB2LtmnUso9tm3bRocOHewOo1oo71iJyDpjTHx52/tk6ea0Pi0bcUOPKN5elspubcwqpWoon070AH+6uj3BAX4899VWvO23F6WU8gSfT/QRdUN4fGhbVuzKYUHiQbvDUUopj/P5RA9wR79Y2jepy+Svk8grLLY7HKWU8qgakegD/P148frO7D+WzxuLk+0ORymlPKpGJHqAXnENuemSaN5dkUpKtu/NTqeUUhWpMYkeYOJV7QkJ9GfS3ERtzCqlaowalejD6wbz5LB2rNiVw/yt2phVSl240NDQCtelpaXRuXNnD0ZzfjUq0QOM69Ocjk3DePGbJE4WaGNWKeX7fHYKhIpYjdlO3PSfVfx7cTITr2pvd0hKqXPNnwgHt7h2n026wFWvVrh64sSJxMTE8Jvf/AaASZMmERAQwJIlSzhy5AhFRUW89NJLjBo1qkrD5ufn89BDD5GQkEBAQAD/+Mc/uPzyy0lMTOSuu+6isLCQ0tJSPvvsM5o1a8att95Keno6JSUlPPvss4wePfqivm2ogYkeoGdsQ27pGc17K1K5uWcUrSPq2h2SUspmo0eP5rHHHjuT6GfPns2CBQv47W9/S1hYGDk5OfTt25eRI0dWaUbcN998ExFhy5YtbN++nWHDhrFz506mTp3K7373O8aNG0dhYSElJSV8++23NGvWjHnz5gFw7Ngxl3xvNTLRA/zxqvYsSDzIc18lMuPePjqVsVLe5Dxn3u7So0cPsrKy2L9/P9nZ2TRo0IAmTZrw+OOPs3z5cvz8/MjIyCAzM5MmTZo4vd8ff/yRRx99FID27dsTGxvLzp076devHy+//DLp6enceOONtGnThi5duvD73/+eP/7xj1x77bUMGDDAJd9bjavRn9Y4NJinhrdjZcohvtl8wO5wlFJe4JZbbmHOnDnMmjWL0aNHM2PGDLKzs1m3bh0bN24kMjKS/Px8l4x12223MXfuXGrVqsXVV1/N4sWLadu2LevXr6dLly4888wzTJ482SVj1dhED3Bbn1g6R4Xx0rwkTmhjVqkab/To0cycOZM5c+Zwyy23cOzYMSIiIggMDGTJkiXs2bOnyvscMGAAM2bMAGDnzp3s3buXdu3akZqaSsuWLfntb3/LqFGj2Lx5M/v376d27dqMHz+ep556ymVz2zuV6EVkhIjsEJFkEZlYzvonRCRJRDaLyA8iEutYHisi6x0PDE8UkQddErWL+PsJL47qTGZuAf/3wy67w1FK2axTp04cP36cqKgomjZtyrhx40hISKBLly589NFHtG9f9Ys3Hn74YUpLS+nSpQujR49m2rRpBAcHM3v2bDp37kz37t3ZunUrd9xxB1u2bKF37950796dF154gWeeecYl31el89GLiD+wExgKpANrgbHGmKQy21wOrDHG5InIQ8BgY8xoEQlyjFEgIqHAVqC/MWZ/ReO5cj56Z038bDNz1qXz7e8G0DZSG7NK2UHno3eeO+aj7w0kG2NSjTGFwEzgrOuLjDFLjDF5jpergWjH8kJjTIFjebCT43ncH0a0p05wgE5lrJTySc4k3ihgX5nX6Y5lFbkHmH/6hYjEiMhmxz7+er6zebs0rBPEH0a0Y3XqYeZu8rrwlFJeasuWLXTv3v2sP3369LE7rF9x6eWVIjIeiAcGnV5mjNkHdBWRZsCXIjLHGJN5zvvuB+4HaN68uStDctqYXs2ZtXYfL8/bxhXtI6gbEmhLHErVZMaYanWpc5cuXdi4caNHx7yQqoMzZ/QZQEyZ19GOZWcRkSHA08DIMuWassHtx6rR/+rCUGPMO8aYeGNMfHh4uLOxu9Tpxmz2iQL+9b02ZpXytJCQEA4dOqTl0/MwxnDo0CFCQkKq9D5nzujXAm1EpAVWgh8D3FZ2AxHpAbwNjDDGZJVZHg0cMsacEpEGwGXA61WK0IO6xdRnTK/mfLgyjZvjo2nfJMzukJSqMaKjo0lPTyc7O9vuULxaSEgI0dHRVXpPpYneGFMsIo8ACwB/4ANjTKKITAYSjDFzgSlAKPCp49euvcaYkUAH4O8iYgAB/maMcfEEFq71h+HtmL/1AM99mcisB/pWq18jlarOAgMDadGihd1h+KRKL6/0NDsurzzXJz/v5U+fb+H10d24oUfVfnIqpZQdLvbyyhpndHwM3WLq8/K87eTmF9kdjlJKXRRN9OXw8xNeGtWZQycLeH3RTrvDUUqpi6KJvgJdousxrk9zpq9MI2l/rt3hKKXUBdNEfx5PDmtH/dpBesesUqpa00R/HvVrBzFxRHsS9hzh8/W/unVAKaWqBU30lbi5ZzQ9mtfnlfnbOHZKG7NKqepHE30l/Bx3zB4+WaiNWaVUtaSJ3gmdo+oxvm8sH61KY2uGa57hqJRSnqKJ3km/H9qOBo7GbGmpNmaVUtWHJnon1asdyMSr2rN+71HmrE+3OxyllHKaJvoquOmSaOJjG/Dq/O0cy9PGrFKqetBEXwV+fsLkUZ05mlfI3xbusDscpZRyiib6KurYLIw7+sXxvzV72JKujVmllPfTRH8BHh/alkZ1gnlWG7NKqWpAE/0FqFcrkD9f3Z6N+47y6bp9lb9BKaVspIn+At3QI4pecVZj9mheod3hKKVUhTTRXyARqzGbm1/MlAXamFVKeS9N9BehQ9Mw7uwXx8c/72XTvqN2h6OUUuXSRH+RHhvahsahVmO2RBuzSikvpIn+IoWFBPL01R3YnH6MWWu1MauU8j5OJXoRGSEiO0QkWUQmlrP+CRFJEpHNIvKDiMQ6lncXkVUikuhYN9rV34A3GNW9GX1aNOS1Bds5fFIbs0op71JpohcRf+BN4CqgIzBWRDqes9kGIN4Y0xWYA7zmWJ4H3GGM6QSMAP4pIvVdFLvXEBFevL4zx/OLmbJgu93hKKXUWZw5o+8NJBtjUo0xhcBMYFTZDYwxS4wxeY6Xq4Fox/Kdxphdjq/3A1lAuKuC9yZtI+ty96VxzFy7jw17j9gdjlJKneFMoo8Cyhaf0x3LKnIPMP/chSLSGwgCUspZd7+IJIhIQnZ2thMheaffDWlLRN1gnvsqURuzSimv4dJmrIiMB+KBKecsbwr8F7jLGFN67vuMMe8YY+KNMfHh4dX3hD80OICnr+nIloxjfPLzXrvDUUopwLlEnwHElHkd7Vh2FhEZAjwNjDTGFJRZHgbMA542xqy+uHC933Vdm9KvZSOmLNjBoRMFlb9BKaXczJlEvxZoIyItRCQIGAPMLbuBiPQA3sZK8llllgcBXwAfGWPmuC5s72XdMduJkwXFvPad3jGrlLJfpYneGFMMPAIsALYBs40xiSIyWURGOjabAoQCn4rIRhE5/YPgVmAgMMGxfKOIdHf5d+Fl2kTW5Z7LWjArYR/r9mhjVillLzHGu5qG8fHxJiEhwe4wLtqJgmKG/H0ZDesE8fWjl+HvJ3aHpJTyYSKyzhgTX946vTPWTUKDA3jm2g4kHchlxpo9doejlKrBNNG70TVdmnJZ68ZMWbCDHG3MKqVs4juJvrQUfvoXnPCe6/BFhEkjO5FfVMKr8/WOWaWUPXwn0R9OgSV/gVnjoCjf7mjOaB0Ryr0DWjJnXToJaYftDkcpVQP5TqJv3AZumAr71sDcR8GLmsyPXtGaZvVCeParRIpLfnW/mFJKuZXvJHqATjfAFc/Altmw/G92R3NG7aAAnr22I9sO5PK/1dqYVUp5lm8leoABT0LX0bDkJdj6ud3RnDGicxMGtGnM3xfuJPu4NmaVUp7je4leBEb+G2L6wpcPQbp3XJMvIrwwshP5xSW8Mn+b3eEopWoQ30v0AAHBMGYGhEbCJ2PhqHc8+alleCj3D2zJ5+sz+Hm3NmaVUp7hm4keoE5juG02FOfDx6Oh4LjdEQHwm8tbE1W/Fs9+uZUibcwqpTzAdxM9QER7uGUaZG+HOfdAaYndEZ1pzO7IPM5Hq7Qxq5RyP99O9ACtr4SrX4NdC2Dhs3ZHA8DwTpEMbhfO64t2kpXrPdf8K6V8k+8neoBe90KfB2H1m5Dwgd3RWHfMXteJwuJS/vKtNmaVUu5VMxI9wLCXofVQmPckpCyxOxriGtfhwUEt+XLjflanHrI7HKWUD6s5id4/AG7+AMLbwew7IXun3RHx0ODWRDeoxXNfaWNWKeU+NSfRA4SEwdiZEBAEH98CJ+09k64V5M/z13ViZ+YJpq9MszUWpZTvqlmJHqBBLIz5GHIPwKzxUGzvXapDOkRwRfsIXl+0k0xtzCql3KDmJXqAmN5w/VuwdyV8/ZitE6CJCM9f15GiUsNL87Qxq5RyvZqZ6AG63AyD/wSbPoYfX7c1lNhGdXhoUCu+3rSflck5tsailPI9NTfRAwz6I3S+GX54AZK+sjWUhwa3IqZhLZ6bm0hhsTZmlVKu41SiF5ERIrJDRJJFZGI5658QkSQR2SwiP4hIbJl134nIURH5xpWBu4QIjHoTonvD5w9AxnrbQgkJ9GfSdZ1IzjrBhz/tti0OpZTvqTTRi4g/8CZwFdARGCsiHc/ZbAMQb4zpCswBXiuzbgpwu2vCdYPAEGsCtDrh1gRoxzJsC+XKDpEM6RDJv37YxYFjp2yLQynlW5w5o+8NJBtjUo0xhcBMYFTZDYwxS4wxeY6Xq4HoMut+ALxjRrGKhEbAbbOg8CR8MhoKTtgWyvPXdaREG7NKKRdyJtFHAWXn+U13LKvIPcD8qgQhIveLSIKIJGRn2/Rw78iOcMuHkJkIn99n2wRoMQ1r85vLWzNv8wF+3KWNWaXUxXNpM1ZExgPxWOUapxlj3jHGxBtj4sPDw10ZUtW0GQojXoUd38L3k2wL4/6BLYltVJvn5m7VxqxS6qI5k+gzgJgyr6Mdy84iIkOAp4GRxpjq+6y83vdbk6Ct/D9YN92WEEIC/Zk0shOp2Sd5/0dtzCqlLo4ziX4t0EZEWohIEDAGmFt2AxHpAbyNleSzXB+mB4nAiL9Cqytg3hOwe7ktYVzeLoJhHSP5vx92sf+oNmaVUheu0kRvjCkGHgEWANuA2caYRBGZLCIjHZtNAUKBT0Vko4ic+UEgIiuAT4ErRSRdRIa7/LtwNf8A64EljVrDrNshJ9mWMJ69tiMGw4vfJNkyvlLKN4ix8fb/8sTHx5uEBO94oDeHd8N7V0JIfbj3e6jd0OMhvLF4F39buJPpd/dmUFsb+xdKKa8mIuuMMfHlravZd8ZWpmELawK0Y/tg9h1QXOjxEO4b2JK4RrWZNDeRgmL7H4WolKp+NNFXpnlfGPkGpK2AeY97fAK04AB/XhjVmd05J3lvhTZmlVJVp4neGd1Gw8CnYMP/YOW/PT78oLbhXNW5Cf9evIv0I3mVv0EppcrQRO+swX+GTjfAoudgm+en7Xnm2o4Ioo1ZpVSVaaJ3lp8fXP8fiLrEunP2wCaPDh9VvxaPXtmaBYmZLNlRva9gVUp5lib6qgisBWM+gVoN4eMx1lOqPOjey1rSMrwOk+Ymkl+kjVmllHM00VdV3Ui4bSYU5MInY6yJ0DwkKMCPF0Z2Ys+hPN5dnuqxcZVS1Zsm+gvRpAvc9L5VvvniASj13Hw0A9qEc02XpryxJJl9h7Uxq5SqnCb6C9VuBAx/GbZ9DYtf9OjQz1zbAX8/4YWvtTGrlKqcJvqL0fdh6DkBfvwHbJjhsWGb1qvFb69sw/fbMvlhW6bHxlVKVU+a6C+GCFz9N2gxCL7+HaT95LGh7760Ba3C6zDpa23MKqXOTxP9xfIPhFunQ4M4mDUODqV4ZNigAD9eHNWZfYdPMXWZZ8ZUSlVPmuhdoVYD61GEAB+PhlNHPDJs/9aNua5bM95amsLeQ9qYVUqVTxO9qzRqBaNnwJE0mH0nlBR5ZNinr+5AoJ/wwteJHhlPKVX9aKJ3pbhL4bp/we5l8O2THpkArUm9EB4b0pYftmfxfZI2ZpVSv6aJ3tV6jIPLHod102D1fzwy5IRL42gTEaqNWaVUuTTRu8MVz0GH62DBn2HHd24fLtDfj8mjOpN+5BRvLdXGrFLqbJro3cHPD254G5p2gzl3w8Etbh+yX6tGjOrejKnLUkjL8dy0DEop76eJ3l2C6sDYmRBSz5oA7bj76+d/vroDQf5+TPo6EW97RKRSyj6a6N0prKk1AdqpwzBzLBSdcutwkWEhPDakDUt3ZLNQG7NKKQenEr2IjBCRHSKSLCITy1n/hIgkichmEflBRGLLrLtTRHY5/tzpyuCrhabd4MZ3IWM9fPGg2ydAu7N/HO0i6zL56yROFWpjVinlRKIXEX/gTeAqoCMwVkQ6nrPZBiDeGNMVmAO85nhvQ+B5oA/QG3heRBq4LvxqosO1MPQFSPoSlr7i1qGsxmwnMo6e4s0lyW4dSylVPThzRt8bSDbGpBpjCoGZwKiyGxhjlhhjTt+auRqIdnw9HFhkjDlsjDkCLAJGuCb0aqb/b6HH7bD8Ndg0y61D9WnZiBt7RPHO8lRSs0+4dSyllPdzJtFHAfvKvE53LKvIPcD8qrxXRO4XkQQRScjOznYipGpIBK75B8QNgLmPwN7Vbh1u4tXtCQ7w4/m52phVqqZzaTNWRMYD8cCUqrzPGPOOMSbeGBMfHh7uypC8S0AQ3PoR1IuBmbfB4d1uGyqibghPDGvLil05LEg86LZxlFLez5lEnwHElHkd7Vh2FhEZAjwNjDTGFFTlvTVK7YZw22woLbEeRZh/zG1D3d43lvZNrMZsXmGx28ZRSnk3ZxL9WqCNiLQQkSBgDDC37AYi0gN4GyvJZ5VZtQAYJiINHE3YYY5lNVvj1jD6v3AoGT6dACXuScIB/n68eH1n9h/L543F2phVqqaqNNEbY4qBR7AS9DZgtjEmUUQmi8hIx2ZTgFDgUxHZKCJzHe89DLyI9cNiLTDZsUy1GAjXvg4pi+G7X12x6jK94hpy0yXRvLsilRRtzCpVI4m3Neri4+NNQkKC3WF4zsJnYeX/wVVToM/9bhki+3gBV/x9Kd2i6/Pfe3ojIm4ZRyllHxFZZ4yJL2+d3hlrtyGToN018N0fYdcitwwRXjeYJ4e148fkHL7doo1ZpWoaTfR28/OHG9+ByE7w6V2QmeSWYcb1aU7HpmG8+E0SJwu0MatUTaKJ3hsEh8LYWdZEaB+PhhOuv5fgdGP2YG4+/7d4l8v3r5TyXprovUW9KBj7CZzMtq6xL8p3+RA9Yxtwa3w076/YTXLWcZfvXynlnTTRe5OoS6wyTvrP8NVv3PIowj+OaE/tIH+e+0rvmFWqptBE7206joQrn4etc2DZay7ffaPQYJ4a0Z6VKYf4ZvMBl+9fKeV9NNF7o8seh263wdK/wJY5Lt/9bb2b0zkqjJfmJXFCG7NK+TxN9N5IBK77JzTvD18+DPvWunT3/n7Ci6M6k5lbwA1v/sQXG9IpLnHvPPlKKftoovdWAcEw+n8Q1sx6OtXRvS7dfY/mDZg6vid+Ijw+axOD/7aU/65KI79IH1ailK/RO2O9XfZOeG+IdVXO3QsgJMyluy8tNSzensVbS5NZv/cojUODufuyOMb3jSUsJNClYyml3Od8d8Zqoq8OUpbA/26CVldYDxz3D3D5EMYY1uw+zFtLU1i+M5u6IQHc0S+Wuy5tQePQYJePp5RyLU30viDhQ/jmMejzEFz1qluH2pJ+jP8sS2b+1oME+fsxplcM9w1sSXSD2m4dVyl14c6X6F1/aqjcI/4uyNkFq9+0pjnuda/bhuoSXY+3xvUkJfsEby9LYcaavcxYs5eR3Zvx0KBWtIms67axlVKup2f01UlpiXXX7K5FMO5TaH2lR4bdf/QU765IZebP+zhVVMLwTpE8PLg13WLqe2R8pVTltHTjSwqOwwcjrKtw7lkEEe09NvThk4VM+2k301amkZtfzKWtG/Hw4Nb0b9VIpz5Wymaa6H3N0X3w7hUQWAvuWwx1Gnt0+BMFxXy8Zg/vrdhN1vECusXU5+HBrRjaIRI/P034StlBE70vSl8H066GZj3gjq+s6+49LL+ohM/Wp/P2slT2Hs6jTUQoDw5qxcjuzQj011s0lPIkTfS+auvnMOcu6DoGbphq3VFrg+KSUuZtOcB/lqaw/eBxourX4v6BLRndK4aQQH9bYlKqptFE78uWTYElL8EVz8LAJ20NxRjDkh1ZvLkkhXV7jtCoThB3X9aC8X1jqVdLb75Syp000fsyY+CLB2DzLLhlGnS6we6IMMbws+Pmq2U7s6kbHMD4frHcfWkLwuvqzVdKucNFPzNWREaIyA4RSRaRieWsHygi60WkWERuPmfdX0Vkq+PP6Av7FlSFRGDkvyGmL3zxIGSsszsiRIQ+LRsx/e7efPPoZQxsF87UZSlc9tfFPPvlVvYdzrM7RKVqlErP6EXEH9gJDAXSgbXAWGNMUplt4oAw4ElgrjFmjmP5NcBjwFVAMLAUuNIYk1vReHpGf4FO5lhX4hTnW1fi1Iu2O6KzpGaf4O1lqXy+IZ1SA6O6NePBwa1oqzdfKeUSF3tG3xtINsakGmMKgZnAqLIbGGPSjDGbgXPnuu0ILDfGFBtjTgKbgRFV/g5U5eo0httmQ9Ep+HgMFJywO6KztAwP5a83d2X5Hy5nQv845m89yLDXl3PfRwls2HvE7vCU8mnOJPooYF+Z1+mOZc7YBIwQkdoi0hi4HIg5dyMRuV9EEkQkITvb9Q/GrjEi2sMtH0JWEnx2r3UnrZdpWq8Wz17bkZUTr+B3V7bh592HueGtlYx9ZzUrdmXr4w2VcgO3XuxsjFkIfAusBD4BVgG/yj7GmHeMMfHGmPjw8HB3huT7Wg+Bq/4KO+fDoufsjqZCDeoE8fjQtqyceAXPXNOB1JwT3P7+z4x68ye+23qA0lJN+Eq5ijOJPoOzz8KjHcucYox52RjT3RgzFBCser9yp973Qe8HYNUbsG6a3dGcV53gAO4d0JLlf7icV27swrFTRTz4v/UMfX0Znybso0iffKXURXMm0a8F2ohICxEJAsYAc53ZuYj4i0gjx9ddga7AwgsNVlXB8L9A66Ew7/eQutTuaCoVHODP2N7NWfz7wfx7bA+CAvx5as5mBk9ZyrSfdnOq0PvKUEpVF05dRy8iVwP/BPyBD4wxL4vIZCDBGDNXRHoBXwANgHzgoDGmk4iEAOsdu8kFHjTGbDzfWHrVjQvl58L7w+D4frj3B2jcxu6InGaMYemObN5amszaNOvmq7sujeP2fnF685VS5dAbpmqyI3vgvSshuK6V7Gs3tDuiKlubdpi3liSzZEc2ocEBjOvbnHsua0FE3RC7Q1PKa2iir+n2/QzTroXoXnD7FxAQZHdEFyRpfy7/WZbCvM37CfD349b4aB4Y2IqYhvrkK6U00SvYMgc+uwe6j4dRb9g2AZorpOWc5O3lKXy2LoMSY7iua1MeGtyadk305itVc2miV5Ylr8CyV2HIC3DZY3ZHc9Eyc/N5b0UqM9bsJa+whCEdInj48tZc0ryB3aEp5XGa6JXFGOtGqq2fwej/Qofr7I7IJY7mFTJ95R4+XLmbo3lF9G3ZkIcHt2ZAm8b65CtVY2iiV78oyofp18LBrXD3fOvBJT7iZEExn/y8l/dW7OZgbj5dourx0OBWDO/UBH998pWqBowxF3xyoolene1EFrx7JZQWwd0LoEGs3RG5VEFxCV9uyGDqslR255ykZXgdHhzUiuu7RxEUoE++Ut5nc/pRpq1Mw1+EKbd0u6B9aKJXv5aZBB8Mh9JiuPQx6P8oBPnW1SslpYb5Ww/w1pIUkg7k0rReCPcNaMmY3jHUDgqwOzxVwxUWlzJ/6wGmr0xj/d6j1AnyZ0zv5jxzTYcLOqvXRK/Kd2SPNR9O0pcQFg1DX4DON1XrK3LKY4xh2c5s3lqaws+7D9OgdiB3XdqCO/vFUa+23nylPCvreD6frNnHjDV7yDpeQFyj2tzZP46bekYTFnLh/x810avzS/sJvpsIBzdDdG+46lWI6ml3VG6xbs9h3lqSwg/bs6gT5M/4vrHWzVdhevOVcq8Ne48wfWUa87YcoKjEMKhtOBMujWNQm3D8XNBD0kSvKldaAhs/hh8mw8ks6DYWrnwOwprZHZlbbDuQy3+WpvCN4+arm3tG8+DAVjRv5FvlK2WvguISvt1ygGkr97Bp31FCgwO4uWc0d/SLpWV4qEvH0kSvnFdwHFb8HVa9CX4BcNnjVv0+sJbdkbnFnkMneXt5KnMS0ikuLeXars14aHArOjQNszs0VY1l5uYzY81ePl6zl5wTBbQMr8Od/azyTGiwe/pDmuhV1R1Jg4XPwra5UC/Gqt93utHn6venZebm8/6Pu5mxeg8nC0u4sn0ED1/eip6x1W9uIGUPYwzr91pXz8zfcoASY7i8XQQT+sdxWevGLinPnI8menXh0n501O+3QEwfGPGKz9bvAY7lFTF9VRof/rSbI3lF9G7RkIcHt2JQ23C9+UqVK7+ohG82W1fPbMk4Rt3gAG7tFcMd/WKJbVTHY3FoolcXp7QENs5w1O+zodttjvp9U7sjc5u8wmJm/ryPd1ekcuBYPp2ahfHQ4FZc1bmp3nylADhw7BQzVu/lk5/3cuhkIa0jQrmzfxw39oiijpvKM+ejiV65Rn6uVb9f/Rb4BcKAx6HfIz5bvwfrWucvN2YwdWkKqTknaRwaxJAOkQzrFEn/Vo0JCfS3O0TlQcYYEvYcYdpPaXyXeJBSY7iyfSR3XRpH/1aNbP2tTxO9cq3Dqdb199u+hnrNHfX7G3y2fg/WzVffb8vkm80HWLI9ixMFxdQO8mdwu3CGdWzC5e0i9Jp8H5ZfVMLcjfuZtjKNpAO5hIUEMKZ3c27vG+s102RrolfusXsFfPcnyNwCzftZ9XsfmjunIgXFJaxOPczCxIMsSsok63gBAX5C35aNGNYpkqEdI2laz3d/y6lJMo6e4n+r9zDz570cySuiXWRd7uwfx/U9mnnd3dWa6JX7lJbAhv/B4hfhZA50d9Tv6zaxOzKPKC01bEo/ysKkTBYmHiQl+yQAXaPrMaxjJMM6NaFNRKg2cqsRYwxrdh9m+so0FiQeBGBox0gm9G9B35YNvfbfUhO9cr/8XFjxN1j9H0f9/glH/b5m3XGanHWCRUmZLEw6yIa9RwGIa1SbYZ2aMKxjJD2aN9Bmrpc6VVjClxszmL4yje0Hj1O/diBjejVnfN/mRDfwjvLM+WiiV55zONW6/n77N1b9fthk6Hi9T9fvK5KVm8+ibZksTMxkZUoORSVGm7leaN/hPKs8s3Yfx04V0aFpGBP6xzKqe1S1+ve56EQvIiOAfwH+wHvGmFfPWT8Q+CfQFRhjjJlTZt1rwDWAH7AI+J05z6Ca6H1E6jJY8GfI3ArN+zvq993tjso2x/OLWLojm4VJmdrM9QLGGFalHGLayjS+35aJiDC8k1We6RXXwGvLM+dzUYleRPyBncBQIB1YC4w1xiSV2SYOCAOeBOaeTvQi0h+YAgx0bPoj8CdjzNKKxtNE70NKS2D9R7D4Jcg7BD3GwRXPQd1IuyOzlTZz7ZNXWMwXG6zyzM7MEzSsE8TY3jGM6xNLs/rV+5hfbKLvB0wyxgx3vP4TgDHmlXK2nQZ8UybR9wPeAC4DBFgO3G6M2VbReJrofVD+MVg+BVZPhYBgq37f9zc1rn5fnrLN3AWJB0nVZq5b7D2Ux0er0pidsI/c/GI6NQtjQv84ruvWrFqVZ87nYhP9zcAIY8y9jte3A32MMY+Us+00yiR6x7K/AfdiJfo3jDFPl/O++4H7AZo3b95zz549Tn5rqlo5lGLV73fMg/qxMOxF6DCyRtbvK5KcdYKFSQdZmJjJxn1HAW3mXihjDD8m5zB9ZRo/bM/CX4QRnZswoX8cPWOrZ3nmfM6X6N16IaiItAY6ANGORYtEZIAxZkXZ7Ywx7wDvgHVG786YlI0atYKxH0PqUvjuzzD7Doi91KrfN72wx6f5mtYRobSOaM3Dg1uTmZvvuIInkw9/2s07y1O1meuEkwXFfL4+nWkr00jJtu5mfvTy1tzWJ5Ym9Wrmb5HOJPoMIKbM62jHMmfcAKw2xpwAEJH5QD9gxXnfpXxby8HwwHLY4Kjfvz0IeoyHK56t8fX7siLDQhjfN5bxfWPJPd3MTTzIN5sPMHPtvrObue0jqFerZjdz03JOMn1VGnMS0jleUEzX6Hr849ZuXNO1KcEBNfsHojOlmwCsZuyVWAl+LXCbMSaxnG2ncXaNfjRwHzACq3TzHfBPY8zXFY2nNfoa5tRRq36/5m0ICIGBv4c+D2n9/jwKiktYlXKIhUmZLErKJNvRzO3XqhHDOkYypAY1c0tLDct3ZTN9ZRpLdmQT6C9c3aUpd/aPo0dMfZ8rz5yPKy6vvBrr8kl/4ANjzMsiMhlIMMbMFZFewBdAAyAfOGiM6eS4YuctrKtuDPCdMeaJ842lib6GOpQCC5+BHd866vcvQYfrtH5fidJSw8b0oyxMtO7MTc2xmrndouudqeu39sFm7vH8Ij5bl85Hq/Y4JpsLZlyf5ozr07zGPhZSb5hS1UfKYljwNGQlQdwAGP4XaNrV7qiqjfKauS0a13FcwRNJ95jq3cxNyT7BRyvTmLMunZOFJXSPqc9dl8ZxVeemBAX42R2erTTRq+qlpBjWT4PFL8OpI3DJ7Vb9PjTC7siqlbLN3FVn7swNZmjHCIZ1bEK/Vo2qRTO3tNSwdGcW01buYfnObIL8/bi2q1We6RZT3+7wvIYmelU9nanfT4WAWjDwSej7kHUtvqqSss3cpTuyOVFQTJ0gfwa3i2BYp0gGt/O+Zm5ufhGfJqTz31VppB3KI6JuMOP7xjK2d3PC6+r/gXNpolfVW06yVb/fOR8axFn1+/bXav3+AlXWzB3asYmtlyEmZx1n2so0Pl+fQV5hCT1jGzChfxwjOjch0L9ml2fORxO98g3JP1j1++xtVv1+xCvQpIvdUVVrbmvmGgPH0q2pL8DxQ1kq/LvEwJrdh/hiw37W7T1KgL8fV7SP5MZLYmjbpO7Z73FifxVvc+5yzr+fyvbnRScbmuiV7ygphnUfwpK/OOr3dzjq9+F2R+YTKmvm9ohpgF95zdz8XNi/HtITIGOd9ffJLM8GbytX/OAAml0Cd3x5YRFoolc+59QRWPYa/PwOBNaGgU9Bnwe0fu9CB4+dnmb5IKtSDlFcajVzh3doyPXNcunul0LggfWQkQDZO7CuoAYatYaoeIiOh7Bm1jJjrPWOvzOO5LF0RxYJaYcpLC6ldXgdBrdrTNeoMPxFztr2zN/l7Kf8vzn764q2rfL+OP+2VdpfBdvWbw79fzW7jFM00SvflbPLKufsWgANWsDwl6Hd1V71K3W1ZgzkZnAydQ37E1cgGetolreD2lIAwAm/ME6Ed6d+m36ExPWGqJ5Qq0G5uyopNSxKymT6yjRWpR4iOMCP67tHcWf/ODo2C/Pkd+WTNNEr35f8vaN+vx1aDIThr0CTznZHVf0UHIf9G84uwZywHqeHfxA06Upxs0vYEdCObw9HMyvZn5yThedt5h7NK2Tm2n38d9UeMo6eIqp+Lcb3jWVMrxga1Amy6Rv1PZroVc1wpn7/sjU18iV3wuVPa/2+IqUl1g/G9LW/JPbs7WBKrfUNW/5SgomKt35wnlMaKy01bNh39Exdf3eZZu7QjpGkHznFFxsyKCgupW/Lhkzo34IhHSII0KtnXE4TvapZ8g5b9fu171r1+0F/gN4PQEANP3vMPWDV008n9f0boPCEtS6k/i8JPTreKsHUblil3RtjSMk+wYJE6yatTfuOEhLoxw09ormzfyztm2h5xp000auaKXsnLHwadi20zk6HvQztrqoZ9fvCk7B/49mJPdcx6axfoHV2HhUP0b2sxN6wpcuPS9bxfEIC/QkL8a4bsXyVbfPRK2Wr8LYw7lPY9b31/NqZY6HFIOv6+8hOdkfnOqWlkLPDkdATIH2dNVeQKbHW14+F5n1/OVtv0tUjs4NG1K2Zk4t5Iz2jVzVDSREkfGBdf1+QCz0nWPX7Oo3tjqzqjmeWOVNPgIwNUHjcWhdcD6Iu+aUME9VTexQ1hJZulDot7zAsfRXWvgdBoY76/f3eW78vOgUHNllJPX2tVYI5ts9a5xdg/WZStmHaqDX4aaOzJtJEr9S5sndY5Zzk76FhK+v6+7Yj7K3fl5bCoeSzz9YzE6G02FpfrzlE9/wlsTftBoE14wEjqnKa6JWqyK5FVsLP2QktL7fmv4/s6JmxT+aUqasnQMZ6KDhmrQuqC1E9fmmYRvXUxyyq89JmrFIVaTPUeobt2vdh6Ssw9VLoeZejft/IdeMU5cPBzWcn9qN7rHXiBxGdoPMNv5ytN24Lft4/V7yqHvSMXqnT8g5byX7t+xAcCoMmQq97q16/N8Z6NGLZEszBrVBaZK0Pi7LO0E/X1Zt1h6A6Lv92VM2ipRulqiJrm1XOSVlsNTeH/wXaDKu4fp93+JfpAk4n9/yj1rrAOtZVMGUTe1hTj30rqubQ0o1SVRHRAcZ/bt1oteDP8PGt0OoKK+E3bGmdnZc9Wz+car1P/CC8A3Qc+UsJJry9lmCU7ZxK9CIyAvgX4A+8Z4x59Zz1A4F/Al2BMcaYOY7llwOvl9m0vWP9lxcduVLuJAJth1sJfu17VknnP5daSbuk0NomtImVzHvcbv3drAcE17U3bqXKUWmiFxF/4E1gKJAOrBWRucaYpDKb7QUmAE+Wfa8xZgnQ3bGfhkAysNAVgSvlEf6B1nNqu9wKq96w6uxn5lqPqhnTKahqz5kz+t5AsjEmFUBEZgKjgDOJ3hiT5lhXep793AzMN8bkXXC0StmlTiMY8rzdUSh1QZy5hS4K2FfmdbpjWVWNAT4pb4WI3C8iCSKSkJ2dfQG7VkopVRGP3CstIk2BLsCC8tYbY94xxsQbY+LDw3VeDqWUciVnEn0GEFPmdbRjWVXcCnxhjCmq4vuUUkpdJGcS/VqgjYi0EJEgrBLM3CqOM5YKyjZKKaXcq9JEb4wpBh7BKrtsA2YbYxJFZLKIjAQQkV4ikg7cArwtIomn3y8icVi/ESxzQ/xKKaUqoXfGKqWUDzjfnbE6cbVSSvk4TfRKKeXjvK50IyLZwJ6L2EVjIMdF4biSxlU1GlfVaFxV44txxRpjyr0+3esS/cUSkYSK6lR20riqRuOqGo2rampaXFq6UUopH6eJXimlfJwvJvp37A6gAhpX1WhcVaNxVU2NisvnavRKKaXO5otn9EoppcrQRK+UUj6uWiZ6ERkhIjtEJFlEJpazPlhEZjnWr3HMt+MNcU0QkWwR2ej4c6+H4vpARLJEZGsF60VE/s8R92YRucRL4hosIsfKHK/nPBRXjIgsEZEkEUkUkd+Vs43Hj5mTcXn8mIlIiIj8LCKbHHG9UM42Hv9MOhmXLZ9Jx9j+IrJBRL4pZ51rj5cxplr9wXpubQrQEggCNgEdz9nmYWCq4+sxwCwviWsC8IYNx2wgcAmwtYL1VwPzAQH6Amu8JK7BwDc2HK+mwCWOr+sCO8v5t/T4MXMyLo8fM8cxCHV8HQisAfqes40dn0ln4rLlM+kY+wng4/L+vVx9vKrjGf2ZRxsaYwqB0482LGsUMN3x9RzgShG3P9zTmbhsYYxZDhw+zyajgI+MZTVQ3/GwGLvjsoUx5oAxZr3j6+NYs7ae+1Q1jx8zJ+PyOMcxOOF4Gej4c+5VHh7/TDoZly1EJBq4Bnivgk1ceryqY6J35tGGZ7Yx1jTLx4BGXhAXwE2OX/XniEhMOevt4KrHRbpDP8ev3vNFpJOnB3f8ytwD62ywLFuP2XniAhuOmaMMsRHIAhYZYyo8Xh78TDoTF9jzmfwn8Aegoudsu/R4VcdEX519DcQZY7oCi/jlJ7Yq33qs+Tu6Af8GvvTk4CISCnwGPGaMyfXk2OdTSVy2HDNjTIkxpjvWE+h6i0hnT4xbGSfi8vhnUkSuBbKMMevcPdZp1THRO/NowzPbiEgAUA84ZHdcxphDxpgCx8v3gJ5ujslZrnhcpMsZY3JP/+ptjPkWCBSRxp4YW0QCsZLpDGPM5+VsYssxqywuO4+ZY8yjwBJgxDmr7PhMVhqXTZ/JS4GRIpKGVeK9QkT+d842Lj1e1THRO/Now7nAnY6vbwYWG0dXw864zqnhjsSqsXqDucAdjitJ+gLHjDEH7A5KRJqcrkuKSG+s/69uTw6OMd8Hthlj/lHBZh4/Zs7EZccxE5FwEanv+LoWMBTYfs5mHv9MOhOXHZ9JY8yfjDHRxpg4rDyx2Bgz/pzNXHq8Ai70jXYxxhSLyOlHG/oDHxjHow2BBGPMXKwPw39FJBmr2TfGS+L6rViPXyx2xDXB3XEBiMgnWFdjNBbrkY/PYzWmMMZMBb7FuookGcgD7vKSuG4GHhKRYuAUMMYDP7DBOuO6HdjiqO8C/BloXiY2O46ZM3HZccyaAtNFxB/rB8tsY8w3dn8mnYzLls9kedx5vHQKBKWU8nHVsXSjlFKqCjTRK6WUj9NEr5RSPk4TvVJK+ThN9Eop5eM00SullI/TRK+UUj7u/wEFaZMEKHHeRAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "losses[['loss','val_loss']].plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAD4CAYAAADlwTGnAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAAA5CUlEQVR4nO3dd3gU1frA8e9JDxDSaUlIqAm9RYooVQQEqRfBgoDtehEUO5ar/hQ7XtFrRS5NQVQUpXcQEBRCb2nUJLQASUiAtN3z+2OWEEMgG7LJJtn38zz7MDvlzLsT9rw7Z2bOUVprhBBCOB4newcghBDCPiQBCCGEg5IEIIQQDkoSgBBCOChJAEII4aBc7B1AcQQEBOiwsDB7hyGEEBXK9u3bz2qtAwvOr1AJICwsjKioKHuHIYQQFYpS6lhh86UJSAghHJQkACGEcFCSAIQQwkFVqGsAhcnJySExMZHMzEx7hyIADw8PgoODcXV1tXcoQogiVPgEkJiYiJeXF2FhYSil7B2OQ9Nac+7cORITE6lXr569wxFCFKHCNwFlZmbi7+8vlX85oJTC399fzsaEqCAqfAIApPIvR+RvIUTFUeGbgIQQFYPWmqTUy+w4nkrqpWyGtA2mmrtUQfYkR18IUSqyck3sP3GBHcdS2H4shR3HUzh9IStv+adr4phwR2NG3BKCi3OlaIyocCQBVBC5ubm4uMifS5RfZy5ksuP4lco+lb2JaWSbzAAE+3rSsb4/bev60i7Ul6xcM+8vi+bVX/cxc/NRXuobQY+IGtKEWMakRrGBQYMGkZCQQGZmJk899RSPPfYYy5cv5+WXX8ZkMhEQEMCaNWvIyMhg/PjxREVFoZTi9ddfZ+jQoVSrVo2MjAwA5s+fz+LFi5k5cyajR4/Gw8ODnTt30rlzZ0aMGMFTTz1FZmYmnp6ezJgxg/DwcEwmEy+++CLLly/HycmJRx99lGbNmvHpp5/y66+/ArBq1Sq++OILFixYYMcjJSqLHJOZ6JPp+Sr8FBJTLgPg5uxEi2BvRncOo21dH9rW9aVGdY9ryvjhnx1ZeeA07y2L5uFZUXSq788r/ZrQPMi7rD+Ow6pUCeD/Fu3nwIkLNi2zaZ3qvH53sxuuM336dPz8/Lh8+TK33HILAwcO5NFHH2XDhg3Uq1eP8+fPA/DWW2/h7e3N3r17AUhJSSly/4mJiWzevBlnZ2cuXLjAxo0bcXFxYfXq1bz88sv8/PPPTJ06laNHj7Jr1y5cXFw4f/48vr6+jB07luTkZAIDA5kxYwYPPfRQyQ+IcEjnL2az01LZbz+Wwp7ENC7nmACoWd2ddqG+jL41jLahvjSrUx13F+ciy1RK0btZLXpE1GDuX8eZsjqW/v/dxJA2QTzXO5w6Pp6l/bEcXqVKAPby6aef5v2yTkhIYOrUqXTp0iXvXng/Pz8AVq9ezbx58/K28/X1LbLsYcOG4exsfJnS0tIYNWoUcXFxKKXIycnJK/fxxx/PayK6sr+RI0fy3XffMWbMGLZs2cLs2bNt9InF3+xfAPt+hsZ9IeIu8Cz671qemcyauDPp7DiWyvZjKew8nsLhsxcBcHFSNK1TneG3hNA21GjOqePtUaKmG1dnJ0bdGsbgtkF8se4Q0/84wpK9J3n4tnr8q1sDvDzkocLSUqkSQFG/1EvD+vXrWb16NVu2bKFKlSp069aN1q1bEx0dbXUZ+b88Be+hr1q1at70v//9b7p3786CBQs4evQo3bp1u2G5Y8aM4e6778bDw4Nhw4bJNYTScHAxzH8YnN3g4CJY5AoNekCzwUYy8Cj/zRkXMnPYdTw1ryln1/FU0rNyAfCr6kbbur78IzKYdnV9aRnsg6db0b/ub0Z1D1cm9o3ggY51mbwihi/WH+KHbQlMuKMRI9rXxVUuFNuc1AgllJaWhq+vL1WqVCE6Opo///yTzMxMNmzYwJEjR/KagPz8/OjVqxeff/45U6ZMAYwmIF9fX2rWrMnBgwcJDw9nwYIFeHl5XXdfQUFBAMycOTNvfq9evfj666/p3r17XhOQn58fderUoU6dOkyaNInVq1eX9qFwPIfWwvwxUKcNPPgrJMfC/l9g/68Qt8JICg16GskgvC94VLd3xGitOXz2Ijsslf2OY6nEnklHa1AKwmt6MaB1nbyLtaH+Vcr8wmywbxWmjGjDQ7fV4+0lB/n3b/uZsfkoL/Vtwh1N5EKxLUkCKKE+ffrw1Vdf0aRJE8LDw+nYsSOBgYFMnTqVIUOGYDabqVGjBqtWreLVV1/liSeeoHnz5jg7O/P6668zZMgQ3nvvPfr3709gYCCRkZF5F4QLeuGFFxg1ahSTJk2iX79+efMfeeQRYmNjadmyJa6urjz66KOMGzcOgPvvv5/k5GSaNGlSJsfDYRzbAt/fBwGN4YH54O4Fwe2M152TIDEKDvxqNA/FLgNnd2h4hyUZ9DHWLwOXsnPZlZDKzuNXm3NSLhlNh14eLrSt68tdLWrTLtSXViHe5aq5pWWwD/Me68jqg2d4d9lBHp0dRYd6frzSrwktg33sHV6loLTW9o7BapGRkbrggDAHDx6Uyu0Gxo0bR5s2bXj44YfLbJ+V/m9yYifMGgDVasCYZca/12M2Q1KUkQj2/wrpJ4xk0KiXkQwa9wH3ajYJS2tNYsrlv92Zc/BkOiaz8R1vEFiVdqG+eb/uGwRWw8mpYvyazjGZmbctgSmrYjl3MZtBrevwXO9wgn2r2Du0CkEptV1rHXnNfGsSgFKqD/AJ4AxM01q/V2B5KDAdCATOAw9orRPzLa8OHAB+1VqPs8xbD9QGLltWu1NrfeZGcUgCKJ527dpRtWpVVq1ahbu7e5ntt1L/Tc4chBl3gVs1eGgZeAdbv63ZDIlbryaDjFPg4gGN7rQkg97gVrXIYq7IzDGxLyntb/feJ6cbD1pVcXOmdYhPXoXfpq4PPlXcivlhy5/0zBy++v0Q0zYeQQMPda7H2O4NqF6OzlzKo5tOAEopZyAW6AUkAtuAe7XWB/Kt8xOwWGs9SynVAxijtR6Zb/knWJJDgQTwnNba6jEeJQFUDJX2b3L+MEzvC2jjl79/g5svy2yGhD+NZHDgN8g4DS6eRhJoNthICm5//3V7Ki0z75f99mMp7D+RRo7J+P7W9atiqex9aBvqS3hNr0r9dO2J1MtMXhnDLzuS8K3iyoQ7GnNfB7lQfD3XSwDWXANoD8RrrQ9bCpoHDMT4RX9FU+AZy/Q64Nd8O24H1ASWA9cEIESFkJYEswaCKavklT+AkxOE3mq8+rwHx7dcTQYHfkW7ViE1uCc7vbqx6FIztiZeJinV8qCVixOtgr15qHM92lp+4Qd6ld0ZXnlQx8eT/9zTmoc6GxeKX1+4n5mbjzKxbwR3Nq0pF4qtZE0CCAIS8r1PBDoUWGc3MASjmWgw4KWU8gdSgI+AB4A7Cil7hlLKBPwMTNKFnI4opR4DHgOoW7euFeEKYWMZyTB7IFxOgVELoYaNz26cnDkXcAs76jVih3qIzLiNND63il6Hf6eHWkRHPDjgdSup7fsT2KYfTUJq4OYiv3QBmgd5M/fRDqyLOcM7S6P557fbaR/mx8v9mtA6xMfe4ZV7troL6DngM6XUaGADkASYgLHAUq11YiEZ+X6tdZJSygsjAYwErnlSSWs9FZgKRhOQjeIVwjqXU+DbwZCWCCN/gaC2JS7SZNbEnEq33IZpNOkcPXcJMB60ahbUAnVLF6qHVKODcwwBR5cQeXAh7FkL0a8Zt5Q2G2zcYup6bRcLjkYpRY+ImnRpFMgPUQl8vCqWQZ//wYBWdXi+dzghfnKh+HqsSQBJQEi+98GWeXm01icwzgBQSlUDhmqtU5VSnYDblVJjgWqAm1IqQ2s9UWudZNk2XSk1F6OpSR5VFeVHVjrMGQZnY+DeeUZzzU1Iu5TDjoQUdh5LYbvlQauL2UY3CgHVjAetRrSvS7tQX1oEeePhmv9Bq7rQohfcNRmObjSaiQ4ugr0/gZuX8bBZs8HGw2cujtUMVJCLsxP3dwhlYOsgvv79EN9sPMzyfacY3TmMJ7o1xLuKXCguyJqLwC4YF4F7YlT824D7tNb7860TgHGB16yUehswaa1fK1DOaCBSaz3OUqaP1vqsUsoV+B5YrbX+6kaxyEXgiqFS/E1yLhuV/7HNcM8saHK3VZuZzZrDZzOMi7XHUtl+PIX4M8ZzHU4KImpVNy7WhvrQrq4fIX6exW+vNuXAkQ1Xk0FmKrhXh4h+RjKo3x1cKv4dPyV1Mu0yH62M5ecdiXh7uvJkj0Y80DHUIZvPbvoisNY6Vyk1DliBcRvodK31fqXUm0CU1noh0A14VymlMZqAniiiWHdghaXydwZWA98U5wNVVPl7/hTllCkHfhwFRzfB4K+LrPyPnL3I4t0n2H48hZ3HU0m7bDxo5e3pStu6PgxsVYd2ob60DPGxzQAozq7QsKfx6v8xHP7dSAbRi2D39+DuDU36G8mgXleHTQa1vT2ZPKwVYzqH8c7Sg7y5+ACztxzlxT4R9GleSy4UIw+ClbnykgBKc3yBivY3+RuzCX5+xOjSod9/4JbrP0B37NxFPl0Tz6+7kjCZNY1qVMu7775tqC/1A6qW7YNWudlweL0lGSyBrDTw8IEISzKo39VIHg5Ia8362GTeXXqQ2NMZRIb68kq/JrSpW7E77rNWSW4DrTiWTYRTe21bZq0W0Pe96y6eOHEiISEhPPGEcdLzxhtv4OLiwrp160hJSSEnJ4dJkyYxcODAIneVkZHBwIEDC91u9uzZTJ48GaUULVu25Ntvv+X06dM8/vjjHD58GIAvv/ySOnXq0L9/f/bt2wfA5MmTycjI4I033sjrqG7Tpk3ce++9NG7cmEmTJpGdnY2/vz9z5syhZs2ahY5bkJaWxp49e/L6Mfrmm284cOAAH3/8cUmObvliNsOiJ43Kv9eb1638j5+7xH/XxvHLziRcnBSjOoXxeNf6hfZ5X6Zc3KDxncYrN+tqMji4EHZ9Z/RSeiUZ1OviUMlAKUX38Brc3jCAn7Yn8tHKWAZ/sZn+LWvzQu8I6vo75oXiypUA7GD48OFMmDAhLwH8+OOPrFixgieffJLq1atz9uxZOnbsyIABA4o85fTw8GDBggXXbHfgwAEmTZrE5s2bCQgIyBtf4Mknn6Rr164sWLAAk8lERkZGkWMMZGdnc+UsKiUlhT///BOlFNOmTeODDz7go48+KnTcAldXV95++20+/PBDXF1dmTFjBl9//XVJD1/5oTWseAl2fgddXoDOT12zSsL5S3y+Lp752xNxclKM7BjK2G4N7F/xF8bF3XiorHFvIxkcWnv1CeSd34Knn9G01WwwhN0Ozo5RFbg4O3Fv+7oMaFWHrzcc5psNh1mx/xSjOoUxrkfDSvG0dHFUrr/6DX6pl5Y2bdpw5swZTpw4QXJyMr6+vtSqVYunn36aDRs24OTkRFJSEqdPn6ZWrVo3LEtrzcsvv3zNdmvXrmXYsGEEBAQAV/v7X7t2bV4f/87Oznh7exeZAIYPH543nZiYyPDhwzl58iTZ2dl54xdcb9yCHj16sHjxYpo0aUJOTg4tWrQo5tEqx9a9DX99BR3HQveX/7YoKfUyn62N56eoBJyU4v4OdflXt4bU8i6HFX9hXNyNW0fD+0JOJhxac3UMgx2zoIo/NBlgJIPQzg6RDKq6u/BMr8bc36Eu/1kZy//+OMJP2xMZ36MhIzuFWjWgTWVQ+f/SZWDYsGHMnz+fU6dOMXz4cObMmUNycjLbt2/H1dWVsLCwa/r5L8zNbpefi4sLZrM57/2NxhcYP348zzzzDAMGDGD9+vW88cYbNyz7kUce4Z133iEiIoIxY8YUK65ybdMU2PAhtBkJvd8x+kXG6G7g83Xx/BhlPAd5b/u6jO3egNreFXikKlcP426hiH7GnU7xq41ksOdH2D4DqgbmSwa3glPlrghrVvfg/X+0ZLTlQvGkJQeZveUYL/aJ4K4Wlf9CsePdD1UKhg8fzrx585g/fz7Dhg0jLS2NGjVq4Orqyrp16zh27JhV5Vxvux49evDTTz9x7tw5gLwmoJ49e/Lll18CYDKZSEtLo2bNmpw5c4Zz586RlZXF4sWLb7i/K+MLzJo1K2/+lXELrrhyVtGhQwcSEhKYO3cu9957r7WHp3zb+g2sfh2aD4W7PwGlOJWWyWu/7aPbh+v5MSqBYZEhrH++O28Nal6xK/+CXD2NZqB/TIfn4+Ge2RB2m3En0az+8FEELHnWuBvKbLJ3tKWqSe3qfPtwB2Y91J4qbs48MXcHQ7/czPZj5+0dWqmSBGADzZo1Iz09naCgIGrXrs39999PVFQULVq0YPbs2URERFhVzvW2a9asGa+88gpdu3alVatWPPOM0e3SJ598wrp162jRogXt2rXjwIEDuLq68tprr9G+fXt69ep1w32/8cYbDBs2jHbt2uU1LwG8+uqrpKSk0Lx5c1q1asW6devylt1zzz107tzZquEsy73d82Dpc8ZQjoO/5nRGDm8s3E+XD9cx96/jDG0XxLrnuvHO4BYEVfbxad2qQNOBMGymkQyGzTLOAHbOgZn94D9NYOnzxnMR+c4wK5uujQNZ8uTtfDC0JYkplxn65RbGztnOsXMX7R1aqZDbQEWx9O/fn6effpqePXted50K8Tc5sBB+GgVht3FmwLd8uSmJuX8dJ9es+UfbYMb1aChdCABkX4TYFUYzUdxKyM0Er9pGsmg2GILbGx3bVUKXsnP5ZsMRvt5wiByTmZEdwxjfoyG+VSveheISjQdQXkgCsJ/U1FTat29Pq1at+Omnn264brn/m8SvhrkjyKnZiv/Uep/p25LJNWsGtwlifI+GhPpb3ye/Q8nKgNjllmSwyugZ1asONBtkJIOgyEqZDM5cyOTj1bH8sC2Bau4ujO/RiAdvrVgXiiUBlCN79+5l5MiRf5vn7u7OX3/9ZaeIbKtc/02O/oH+biin3UK4O/1FzuV6MqhNEE/2aERYgFT8VstKhxhLMohfBaZsqB6cLxm0y7uYXlnEnErn3WUHWR+TTIifJy/0jqB/y9oV4kJxpU4AERERFeKP4Ai01kRHR5fLBJB2aCsecwaSaPJhRPa/ua11U8b3aEj9QNsMyeiwMtOuJoNDa4xk4B1yNRnUaVupksHGuGTeXnKQ6FPptA7x4ZV+TbglzM/eYd1QpU0AR44cwcvLC39/f0kCdqa15ty5c6Snp+c9U1AepFzMZsGKVQzZ/RjpePJNg895sE9nGtaQit/mMtMgZpnlzGANmHPAp66RCJoOgjptKkUyMJk1v+xIZPLKGE5fyKJPs1q82DeCeuX0LLLSJoCcnBwSExOLfb+8KB0eHh4EBwfj6mr/bgZSL2UzbeMR1m7ewkxex93FmfMjFlKvUXN7h+YYLqdCzFLLmcFaMOeCT+jV8Y+r+BtjILtVM14V8AG0y9kmpm08zFe/HyIr18wDHUN5smcj/MrZheJKmwCEKCjtcg7/23SEGZuOUC3rNIurvYW3Sy4uDy2DGtbdkits7NL5q8ng8HojGRTk7A7u1fIlhUL+da9WvGVl9CDbmfRMpqyOY97W41R1d2Fc94aMujWswNgO9iMJQFR6FzJzmL7pCP/bdIT0zFyGhbsxKfUF3DPPwqhFUKe1vUMUYCSDxCjIumDcZpr3Sr86nZV+/WW5xTjbd/EsJDnkO+vIW1YwoRSWVKqCa9Ub3ukUdzqdd5dFszb6DEE+nrzQJ5y7W9Yp215hCyEJQFRa6Zk5zPjjKNM2HuZCZi53Nq3JM7cFErH8Xkg5AiMXQN2O9g5T2IopF7IzCk8O1ySPjL+ve73EYsq2fv+uVa8mhGvOOozpxItOrD58kSMXFP6+vvRt25BGIbUL3861SqlfF3GM7qCFQ8nIymXW5qN8s/EwqZdyuKNJDSbc0ZjmAU7GIO7n4uC+H6Xyr2ycXcDTx3jZSm425Fw0nnUoNLFk5FtWyL+XzkNqQt774OwMRptzwRXIwBgm67rU388yrne2ctsE8Lpxh5LFJQlAVDgXs3KZteUo32w4TMqlHHpE1GDCHY1oGexzdSjHE7tg+LfQoLu9wxUVgYub8fK0URcnWhtnFdkXybx0gQV/RrNoWxwuuZfo29iL/hFeeDllWc5KrpNYLiZDytGry9s/apvY8pEmIFFhXMrOZfaWY0zdcJjzF7PpFh7IhDsa0zrEx1ghNxvm3Wc86TvkG2g5zK7xCpHf2YwsPlkdx9ytx6ni6szY7g0Z07lsLhTLNQBRYV3ONvHdn8f4esMhzmZkc3ujAJ7u1Zi2+YfzM+XCzw/Bgd+g/xSIrETdVYtKJf5MBu8ti2b1wdME+XjyfO9wBrQq3QvFkgBEhZOZY1T8X/1+mLMZWdzWMICnezWiXWiBpy7NZlg4DnbNgTvfhlvH2SdgIYphy6FzvLP0IHuT0mgeVJ1X7mpKpwb+pbIvSQCiwsjMMfH91uN8sf4QyelZ3NrAn6d7NS78cXutYdkLsHUqdJ0I3V8q+4CFuElms2bh7hN8uCKGpNTL3NGkBhP7NrH5U+qSAES5l5lj4odtCXyxPp7TF7LoUM+Pp3s1pmP9G/wqWvMmbPwIOo2DOydVim4GhOPJzDEx44+jfLEunks5Ju5tH8KEOxoTUM3dJuVLAhDlVlauiR+3JfD5ukOcupBJ+zA/JvRqxK0NAm684cb/wJr/g7aj8kbzEqIiO5eRxadr4pjz13E8XJ35V7cGPHxbvRJfKJYEIMqd7FwzP21P4PO18ZxIyyQy1JenezXm1gZWdOy39RtjNK8Ww2Dw15V+7FrhWA4lZ/D+smhWHjhNbW8PnrsznMFtgm76QrE8CCbKjRyTmfnbE/lsbTxJqZdpU9eH94a25PZGAdb16LprrlH5h/eDQV9K5S8qnQaB1Zj6YCRbj5zn7SUHePan3TSsUY1WV255thFJAKLM5JjM/LIjkf+ujScx5TKtQnx4e3BzujYOtL4r7/2/wm9PQP1uxmDmzvbvdVSI0tK+nh8LxnZm29HzNq/8QRKAKAO5JjMLdibx37XxHD9/iZbB3rw1sDndwotR8YMxDOHPj0DwLTBiLrh6lF7QQpQTTk6KDje6EaIEJAGIUpNrMvPbrhP8d20cR89dolmd6kx7MJKeTWoUf/Ceo5vghwegRhOjfx+38jnwhhAViSQAYXMms2bh7iQ+XRPPkbMXaVq7OlNHtqNX05o3N2pb4naYO9wYTGTkAtt2AiaEA5MEIGzGZNYs3nOCT9fEcSj5IhG1vPjqgXbc2bTmzT/mfno/fDfEGD3qwV+hahG3hgohrCYJQJSY2axZsvckn6yJI/5MBuE1vfji/rb0aVarZP2bnI2H2YOM/tJHLYTqdWwWsxBCEoAoAbNZs2zfKT5ZE0vs6Qwa1ajGZ/e14a7mtUvesVXqcaNPf22CB5eAb5hNYhZCXCUJQBSb2axZeeAUU1bHEX0qnQaBVfn03jb0a1EbZ1v0aJh+2qj8s9Jh9CIIbFzyMoUQ15AEIKymtWblgdNMWR3HwZMXqB9QlSnDW3N3qzq2qfjBGFnp20FGEnjwV6jdyjblCiGuIQlAFElrzeqDZ5iyOpb9Jy4Q5l+F/9zTigGt6uDifP0Bsost84JxwffcIbj/Rwhpb7uyhRDXkAQgrktrzdroM0xZHcfepDTq+lVh8rBWDGpt44ofIPsSfD8CTu2F4d8ZT/oKIUqVJABxDa0162OTmbIqlt2JaYT4efLB0JYMbhuEq60rfoDcLPhxJBzbDEOnQXhf2+9DCHENSQDiGu8sPcg3G48Q5OPJe0NaMLRdcOlU/GAZyvFhYxzfuz+FFv8onf0IIa4hCUD8zZGzF5n+x1GGtA3ivSEtcXMppYofrg7leHAR9H4X2o0qvX0JIa5h1bdbKdVHKRWjlIpXSk0sZHmoUmqNUmqPUmq9Uiq4wPLqSqlEpdRn+ea1U0rttZT5qbqpPgKErU1eGYObsxMT+0aUbuWvNSx7HnZ/D91fgU5jS29fQohCFfkNV0o5A58DfYGmwL1KqaYFVpsMzNZatwTeBN4tsPwtYEOBeV8CjwKNLK8+xY5e2NTexDSW7DnJI7fXo4ZXKfe0ueb/YNs0uHU8dHm+dPclhCiUNT/x2gPxWuvDWutsYB4wsMA6TYG1lul1+ZcrpdoBNYGV+ebVBqprrf/UxpBks4FBN/shhG28vzwa3yquPNalfunuaMNk2PQxRD4Evd6SoRyFsBNrEkAQkJDvfaJlXn67gSGW6cGAl1LKXynlBHwEPFdImYlFlAmAUuoxpVSUUioqOTnZinDFzdgUd5ZN8Wd5ontDvDxKcZCVP7+CtW9By+Fw10dS+QthR7Zq5H0O6KqU2gl0BZIAEzAWWKq1TrzRxjeitZ6qtY7UWkcGBgbaJlrxN2az5v3l0QT5ePJAx9DS29HO72D5ixDRHwZ+AU6leI1BCFEka+4CSgJC8r0PtszLo7U+geUMQClVDRiqtU5VSnUCbldKjQWqAW5KqQzgE0s51y1TlJ2l+06yNymNycNa4eFaSuPr7l8AC8dDgx6WoRzlBjQh7M2ab+E2oJFSqh5GJT0CuC//CkqpAOC81toMvARMB9Ba359vndFApNZ6ouX9BaVUR+Av4EHgvyX+NKLYckxmJq+IIbymF4PbFNoKV3KxK4yhHEM6GE/5uriXzn6EEMVS5Dm41joXGAesAA4CP2qt9yul3lRKDbCs1g2IUUrFYlzwfduKfY8FpgHxwCFgWfHDFyX1w7YEjp67xPO9w23XoVt+RzbADyOhZnO47wcZylGIckQZN+FUDJGRkToqKsreYVQal7Jz6frhekL9qvDT451ubrjGG0mMMrp19g6G0UuhaukMbC2EuDGl1HatdWTB+XIVzoHN+OMoyelZTOwbYfvK/9Reo2fPqoHw4G9S+QtRDkkCcFApF7P5av0h7mhSg8gwP9sWfjYOvh0MbtWMyt+rlm3LF0LYhCQAB/XF+ngysnN5vneEbQtOOWY0+4BR+fuW4m2lQogSkXvxHFBS6mVmbTnGkDbBhNfysl3B6aeMyj87A0YvgYBGtitbCGFzkgAc0JRVsaDh6V42rKAvnjMq/4wzxi//Wi1sV7YQolRIE5CDiTudzs87EhnZKZRg3yq2KTQzzbjgm3IU7psHIbfYplwhRKmSMwAH88GKGKq6ufBE94a2KTD7IswdDqf3wYi5UK+LbcoVQpQ6OQNwINuPnWfVgdM81qU+flXdSl5gbhb88AAk/AVDvoHGvUtephCizMgZgIPQWvP+shgCqrnz8O31Sl6gKRfmPwSH1sKAz6D5kKK3EUKUK3IG4CDWxZxh69HzPNWzIVXcSpj3zWb4bSxEL4Y+70PbkbYJUghRpiQBOACTWfPB8hhC/aswon3dkhWmNSx9Fvb8AD1ehY6P2yZIIUSZkwTgAH7blUT0qXSeuzMcV+cS/Mm1hlWvQdR06DwBbi84zo8QoiKRBFDJZeWa+GhlLM2DqtOvRe2SFbZhMmz+FG55BO54Q0bzEqKCkwRQyc358zhJqZd5sU8ETiXp7nnLF7BuErS6F/p+KJW/EJWAJIBKLD0zh8/WxdO5oT+3NyrBcJo7ZsOKl6DJ3cYdPzKUoxCVgnyTK7FvNh7h/MVsXuxTgg7f9v0MC5+EhnfA0P/JUI5CVCKSACqp5PQspm08TL8WtWkZ7HNzhcQsg18eg9Bb4Z5vZShHISoZSQCV1Gdr48jKNfPsnY1vroDD6+HHUUanbvfOAzcb9RskhCg3JAFUQsfPXWLu1uMMvyWE+oHVil/AkQ3w/X3g3wAe+AU8qts+SCGE3UmDbiX00aoYnJ0UT/UsZnfPl1Ng9RuwfSb4N4SRC6CKjUcLE0KUG5IAKpn9J9L4bdcJxnZrQM3qHtZtpDXsnW/c6XPpPHQaB91eAvebOHsQQlQYkgAqmQ+Wx+Dt6co/uzawboNzh2DJM0abf1A7o8mndstSjVEIUT5IAqhEthw6x++xybx8VwTenq43Xjk3GzZ/Ar9/aNzdc9dkiHwInJzLJlghhN1JAqgktNa8tzya2t4ePNgp7MYrH9sMiybA2RhoOgj6vAfVS9hNhBCiwpEEUEms2H+K3QmpfDC0JR6u1/kVf+m80Znbzm/Buy7c9xM0vrNsAxVClBuSACqBXJOZD1bE0LBGNYa0Dbp2Ba1h9zxY+QpcToXOT0HXF8GtapnHKoQoPyQBVALztydyOPkiX49sh0vB7p7PxsOSp417+4Nvgf5ToFZzu8QphChfJAFUcJezTXy8Opa2dX24s2nNqwtys2DTx7DxI3DxhH7/gXZjpCM3IUQeSQAV3MzNRzl9IYtPR7RBXemi+chGWPw0nIuD5kOh97vgVfPGBQkhHI4kgAos7VIOX66Pp3t4IB3q+8PFc7DyVdg9F3xC4YGfjV48hRCiEJIAKrAvfo8nPSuXF3qHw87vYOW/IesC3PYMdHleOnATQtyQJIAK6mTaZWb+cZTHmuTSZMV9cGwThHSEu6dAjSb2Dk8IUQFIAqigPl+5j/HqR544usj4pX/3J9DmQbnIK4SwmiSACihpxzIe3juBes6noNk90PttqFbD3mEJISoYSQAVSUYyrHyFoD0/cEzVIm3YT3g3kyd5hRA3RxJARWA2G903rHoNc/ZFPssdhLr9WcY3k147hRA3TxJAeXcmGhZPgONb0HU78dzlMfx+3o/fuzW1d2RCiApOrhiWVzmXYc2b8NVtkBwNAz7j91tn8ktCNcb3aEg1d8ndQoiSsSoBKKX6KKVilFLxSqmJhSwPVUqtUUrtUUqtV0oF55u/Qym1Sym1Xyn1eL5t1lvK3GV5yVXMK+LXwBcdjW4cWgyDcVGYWz/A+yviCPHz5L4OofaOUAhRCRT5M1Ip5Qx8DvQCEoFtSqmFWusD+VabDMzWWs9SSvUA3gVGAieBTlrrLKVUNWCfZdsTlu3u11pH2fIDVWjpp2HFy7BvvjEm76hFUK8LAIt2JXHw5AWmDG+Nm4ucuAkhSs6adoT2QLzW+jCAUmoeMBDInwCaAs9YptcBvwJorbPzreOONDkVzmyGHTONAdlzLhvj8XaeAK7GmL7ZuWY+WhlLk9rVGdCqjj0jFUJUItZUyEFAQr73iZZ5+e0GhlimBwNeSil/AKVUiFJqj6WM9/P9+geYYWn++bfK68ns75RSjymlopRSUcnJyVaEW8Gc3g/Texudt9VqCf/aDN0m5lX+AN9vPc7x85d4oU84Tk6FHiYhhCg2W/0ifw7oqpTaCXQFkgATgNY6QWvdEmgIjFJKXemW8n6tdQvgdstrZGEFa62naq0jtdaRgYGBNgq3HMi+BKteh6+7wLl4GPSV0eQT0Ohvq13MyuW/a+PoUM+Pbo0r0ecXQtidNU1ASUBIvvfBlnl5LL/qhwBY2vqHaq1TC66jlNqHUdnP11onWeanK6XmYjQ1zb7Jz1GxxK2CJc9A6nFo8wD0eguq+BW66rSNRzibkc3UByO4zkmSEELcFGvOALYBjZRS9ZRSbsAIYGH+FZRSAUqpK2W9BEy3zA9WSnlapn2B24AYpZSLUirAMt8V6A/ss8UHKtfST8FPo2HOP8DFA0YvgYGfX7fyP5eRxdQNh+jdrCZt6/qWbaxCiEqvyDMArXWuUmocsAJwBqZrrfcrpd4EorTWC4FuwLtKKQ1sAJ6wbN4E+MgyXwGTtdZ7lVJVgRWWyt8ZWA18Y+PPVn6YTRA13bivPzcLur8KnZ8EF/cbbvbZungu55h4vnd4GQUqhHAkVj1NpLVeCiwtMO+1fNPzgfmFbLcKuKa/Aq31RaBdcYOtkE7thUUTICkK6nczhmb0b1DkZgnnLzHnz+MMaxdCwxpepR6mEMLxyOOkpSX7Iqx/F7Z8AZ6+MOQb46EuK9vxP14Vi1IwoVejolcWQoibIAmgNMQsh6XPQVoCtB0Fd7xx3Xb+whw8eYEFu5J47Pb61Pb2LL04hRAOTRKALV04ActehIMLITACxiyH0E7FLubDFTF4ubvwr25FNxUJIcTNkgRgC2YTbJsGa94Ccw70fA06jQcXt2IXtfXIedZGn+GFPuH4VCn+9kIIYS1JACV1YpfRXfOJndCgJ/SbDH71b6oorTXvLTtIzerujLm1nk3DFEKIgiQB3KysDFj3Dvz1JVQJgKH/g+ZDrb7IW5hVB06z43gq7wxugaebsw2DFUKIa0kCuBnRS2Dp83AhCSIfgp6vg6dPiYo0mTUfroihfkBV7okMtk2cQghxA5IAiiMt0bjIG70YajSDYTMhpL1Niv55RyJxZzL44v62uDhLp6lCiNInCcAaplzYOhXWvW1c8L3j/6DTE+DsapPiM3NMTFkVS6tgb/o2r2WTMoUQoiiSAIqStAMWPQWn9kCjO+GuD8E3zKa7+HbLMU6kZTJ5WCvp8E0IUWYkAVxP5gXjF//WqVC1BgybBU0Hlugib2EuZObw+fp4bm8UwK0NA2xathBC3IgkgIK0hoOLYNkLRu+dtzwCPf8NHt6lsruvfz9E6qUcXuwTUSrlCyHE9UgCyC/1uHF3T+xyqNkChs+B4NLrs+7MhUz+t+kId7eqQ/Og0kkwQghxPZIAwLjI+9eXxn39AHdOgg7/AufSPTyfrIkj16R5tlfjUt2PEEIURhJAYpTRXfPpvdC4L9z1AfjULfXdHjl7kXnbErivfV3CAqqW+v6EEKIgx00AmWnGAC3b/gdetWH4dxDR3+YXea9n8soY3JydGN+zYZnsTwghCnK8BKA1HPgVlk2Ei2egwz+h+yvgUb3MQtibmMaSPScZ36MhNbw8ymy/QgiRn2MlgJSjsOQ5iF8FtVvBvd9DUNsyD+P95dH4VnHlsS4312mcEELYgmMkAFMObPkc1r8HTs7Q5z245dFSv8hbmE1xZ9kUf5ZX+zXBy8M2TxILIcTNcIwEMGcYHF5ntPH3fR+87dPZmtmseX95NEE+njzQMdQuMQghxBWOkQDaP2q8IvrZNYyl+06yNymNycNa4eEq3T0LIezLMRKAnSt+gByTmckrYgiv6cXgNkH2DkcIIZB+h8vID9sSOHruEs/3DsfZSTp8E0LYnySAMnApO5dP1sQRGepLzyY17B2OEEIAkgDKxIw/jpKcnsXEvhHS3bMQotyQBFDKUi5m89X6Q9zRpAaRYX72DkcIIfJIAihlX6yPJyM7l+d7S3fPQojyRRJAKUpKvcysLccY0iaY8Fpe9g5HCCH+RhJAKZqyKhY0PN2rkb1DEUKIa0gCKCVxp9P5eUciIzuFEuxbxd7hCCHENSQBlJIPVsRQ1c2FJ7pLd89CiPJJEkAp2H7sPKsOnOaxLvXxq+pm73CEEKJQkgBsTGvN+8tiCKjmzsO317N3OEIIcV2SAGxsXcwZth49z1M9G1LFzTG6WhJCVEySAGzIZNZ8sDyGUP8qjGhf+uMKCyFESUgCsKHfdiURfSqdZ+8Mx9VZDq0QonyTWspGsnJNfLQylmZ1qtO/RW17hyOEEEWSBGAjc/48TlLqZV7sE4GTdPcshKgArEoASqk+SqkYpVS8UmpiIctDlVJrlFJ7lFLrlVLB+ebvUErtUkrtV0o9nm+bdkqpvZYyP1UVuJvM9MwcPlsXT+eG/tzeKMDe4QghhFWKTABKKWfgc6Av0BS4VynVtMBqk4HZWuuWwJvAu5b5J4FOWuvWQAdgolKqjmXZl8CjQCPLq0/JPor9fLPxCOcvZvNiH+nuWQhRcVhzBtAeiNdaH9ZaZwPzgIEF1mkKrLVMr7uyXGudrbXOssx3v7I/pVRtoLrW+k+ttQZmA4NK8kHsJTk9i2kbD9OvRW1aBvvYOxwhhLCaNQkgCEjI9z7RMi+/3cAQy/RgwEsp5Q+glApRSu2xlPG+1vqEZfvEIsqsED5bG0dWrpln72xs71CEEKJYbHUR+Dmgq1JqJ9AVSAJMAFrrBEvTUENglFKqZnEKVko9ppSKUkpFJScn2yhc2zh+7hJztx5n+C0h1A+sZu9whBCiWKxJAElASL73wZZ5ebTWJ7TWQ7TWbYBXLPNSC64D7ANut2wffKMy8203VWsdqbWODAwMtCLcsvPRqhicnRRP9ZTunoUQFY81CWAb0EgpVU8p5QaMABbmX0EpFaCUulLWS8B0y/xgpZSnZdoXuA2I0VqfBC4opTpa7v55EPjNJp+ojOw/kcZvu07wUOd61KzuYe9whBCi2IpMAFrrXGAcsAI4CPyotd6vlHpTKTXAslo3IEYpFQvUBN62zG8C/KWU2g38DkzWWu+1LBsLTAPigUPAMtt8pLLxwfIYvD1d+WfXBvYORQghbopVvZVprZcCSwvMey3f9HxgfiHbrQJaXqfMKKB5cYItL7YcOsfvscm8fFcE3p6u9g5HCCFuijwJXExaa95bHk1tbw8e7BRm73CEEOKmSQIophX7T7E7IZWn72iMh6uzvcMRQoibJgmgGHJNZj5YEUPDGtUY0rZCPrYghBB5JAEUw/ztiRxOvsjzvcNxke6ehRAVnNRiVrqcbeLj1bG0revDnU2L9SybEEKUS5IArDRz81FOX8iSDt+EEJWGJAArpF3K4cv18XQPD6RDfX97hyOEEDYhCcAKX/weT3pWLi/0ibB3KEIIYTOSAIpwMu0yM/84yqDWQTSpXd3e4QghhM1IAijCJ6vjMGvNM72ku2chROUiCeAG4s9k8GNUAvd3CCXEr4q9wxFCCJuSBHADk1fE4OnqzLgeDe0dihBC2JwkgOvYeTyF5ftP8WiX+gRUc7d3OEIIYXOSAAqhteb95dH4V3Xjkdvr2zscIYQoFZIACvF7bDJ/Hj7P+B4NqeZuVY/ZQghR4UgCKMBs1ry/PIYQP0/u6xBq73CEEKLUSAIoYNGeExw8eYFne4Xj5iKHRwhReUkNl092rpmPVsbSpHZ1BrSqY+9whBCiVEkCyOf7rcc5fv4SL/QJx8lJOnwTQlRukgAsLmbl8t+1cXSo50e3xoH2DkcIIUqdJACLaRuPcDYjmxf7SnfPQgjHIAkAOJeRxdQNh+jdrCZt6/raOxwhhCgTkgCAz9bFcznHxPO9w+0dihBClBmHTwAJ5y8x58/jDGsXQsMaXvYORwghyozDJ4CPV8WiFEzo1cjeoQghRJly6ARw8OQFFuxKYvStYdT29rR3OEIIUaYcOgF8uCIGL3cX/tWtgb1DEUKIMuewCWDrkfOsjT7D490a4FPFzd7hCCFEmXPIBKC15r1lB6lZ3Z0xt9azdzhCCGEXDpkAVh04zY7jqTzVszGebs72DkcIIezC4RKAyaz5cEUM9QOqck9ksL3DEUIIu3G4BPDzjkTizmTwXO9wXJwd7uMLIUQeh6oBM3NMTFkVS6tgb/o2r2XvcIQQwq4cKgF8u+UYJ9IyebGPdPgmhBAOkwAuZObw+fp4bm8UwK0NA+wdjhBC2J3DJICvfz9E6qUcXuwTYe9QhBCiXHCIBHDmQib/23SEu1vVoXmQt73DEUKIcsEhEsAna+LINWme7dXY3qEIIUS54RAJoK5fFR7tUp+wgKr2DkUIIcoNqxKAUqqPUipGKRWvlJpYyPJQpdQapdQepdR6pVSwZX5rpdQWpdR+y7Lh+baZqZQ6opTaZXm1ttmnKuCfXRtI278QQhRQZAJQSjkDnwN9gabAvUqppgVWmwzM1lq3BN4E3rXMvwQ8qLVuBvQBpiilfPJt97zWurXltatEn0QIIUSxWHMG0B6I11of1lpnA/OAgQXWaQqstUyvu7Jcax2rtY6zTJ8AzgCBtghcCCFEyViTAIKAhHzvEy3z8tsNDLFMDwa8lFL++VdQSrUH3IBD+Wa/bWka+lgp5V7YzpVSjymlopRSUcnJyVaEK4QQwhq2ugj8HNBVKbUT6AokAaYrC5VStYFvgTFaa7Nl9ktABHAL4Ae8WFjBWuupWutIrXVkYKCcPAghhK24WLFOEhCS732wZV4eS/POEAClVDVgqNY61fK+OrAEeEVr/We+bU5aJrOUUjMwkogQQogyYs0ZwDagkVKqnlLKDRgBLMy/glIqQCl1payXgOmW+W7AAowLxPMLbFPb8q8CBgH7SvA5hBBCFFORCUBrnQuMA1YAB4Eftdb7lVJvKqUGWFbrBsQopWKBmsDblvn3AF2A0YXc7jlHKbUX2AsEAJNs9JmEEEJYQWmt7R2D1SIjI3VUVJS9wxBCiApFKbVdax15zfyKlACUUsnAsZvcPAA4a8NwbEXiKh6Jq3gkruKprHGFaq2vuYumQiWAklBKRRWWAe1N4ioeiat4JK7icbS4HKIvICGEENeSBCCEEA7KkRLAVHsHcB0SV/FIXMUjcRWPQ8XlMNcAhBBC/J0jnQEIIYTIRxKAEEI4qEqXAKwYvMZdKfWDZflfSqmwchLXaKVUcr4nph8pg5imK6XOKKUK7YZDGT61xLxHKdW2tGOyMq5uSqm0fMfqtTKKK0QptU4pdcAyyNFThaxT5sfMyrjK/JgppTyUUluVUrstcf1fIeuU+ffRyrjK/PuYb9/OSqmdSqnFhSyz7fHSWleaF+CM0d10fYyup3cDTQusMxb4yjI9AvihnMQ1GvisjI9XF6AtsO86y+8ClgEK6Aj8VU7i6gYstsP/r9pAW8u0FxBbyN+xzI+ZlXGV+TGzHINqlmlX4C+gY4F17PF9tCauMv8+5tv3M8Dcwv5etj5ele0MwJrBawYCsyzT84Gelg7p7B1XmdNabwDO32CVgRgd+Wlt9OTqc6UTPzvHZRda65Na6x2W6XSMvrEKjo1R5sfMyrjKnOUYZFjeulpeBe86KfPvo5Vx2YUyhtPtB0y7zio2PV6VLQFYM3hN3jra6OguDfCndFkTF8BQS7PBfKVUSCHLy5q1cdtDJ8sp/DKlVLOy3rnl1LsNxq/H/Ox6zG4QF9jhmFmaM3ZhjAa4Smt93eNVht9Ha+IC+3wfpwAvAObrLLfp8apsCaAiWwSEaWNc5VVczfLiWjsw+jZpBfwX+LUsd66MMS9+BiZorS+U5b5vpIi47HLMtNYmrXVrjHFE2iulmpfFfotiRVxl/n1USvUHzmitt5f2vq6obAmgyMFr8q+jlHIBvIFz9o5La31Oa51leTsNaFfKMVnDmuNZ5rTWF66cwmutlwKuSqmAsti3UsoVo5Kdo7X+pZBV7HLMiorLnsfMss9UjPHC+xRYZI/vY5Fx2en72BkYoJQ6itFM3EMp9V2BdWx6vCpbAihy8BrL+1GW6X8Aa7Xlioo94yrQTjwAox3X3hYCD1rubOkIpOmrI7nZjVKq1pV2T2WMNe1EGVQaln3+Dziotf7PdVYr82NmTVz2OGZKqUCllI9l2hPoBUQXWK3Mv4/WxGWP76PW+iWtdbDWOgyjjlirtX6gwGo2PV7WDAlZYWitc5VSVwavcQama8vgNUCU1nohxhflW6VUPMaFxhHlJK4nlTHATq4lrtGlHZdS6nuMu0MClFKJwOsYF8TQWn8FLMW4qyUeuASMKe2YrIzrH8C/lFK5wGVgRBkkcTB+oY0E9lrajwFeBurmi80ex8yauOxxzGoDs5RSzhgJ50et9WJ7fx+tjKvMv4/XU5rHS7qCEEIIB1XZmoCEEEJYSRKAEEI4KEkAQgjhoCQBCCGEg5IEIIQQDkoSgBBCOChJAEII4aD+H6YXABb1NvEpAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "losses[['accuracy','val_accuracy']].plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From <ipython-input-69-f4d82b300d09>:1: Model.predict_generator (from tensorflow.python.keras.engine.training) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Please use Model.predict, which supports generators.\n"
     ]
    }
   ],
   "source": [
    "preds = model.predict_generator(test_image_gen)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "predaction = preds >0.5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[False],\n",
       "       [False],\n",
       "       [False],\n",
       "       ...,\n",
       "       [ True],\n",
       "       [ True],\n",
       "       [ True]])"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predaction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 0, ..., 1, 1, 1])"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_image_gen.classes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import classification_report ,confusion_matrix "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.96      0.94      0.95      1300\n",
      "           1       0.94      0.96      0.95      1300\n",
      "\n",
      "    accuracy                           0.95      2600\n",
      "   macro avg       0.95      0.95      0.95      2600\n",
      "weighted avg       0.95      0.95      0.95      2600\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(test_image_gen.classes,predaction))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1218,   82],\n",
       "       [  56, 1244]], dtype=int64)"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "confusion_matrix(test_image_gen.classes,predaction)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "con_mat = confusion_matrix(test_image_gen.classes,predaction)\n",
    "sns.heatmap(con_mat,annot=True,fmt='.0f',cmap='Greens');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'C:\\\\Users\\\\Mohamed Basher\\\\Desktop\\\\Analysis Cell Images Using CCN\\\\cell_images\\\\train\\\\parasitized\\\\C100P61ThinF_IMG_20150918_144104_cell_162.png'"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "para_cell"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.preprocessing import image "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [],
   "source": [
    "my_image = image.load_img(para_cell,target_size=image_shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<PIL.Image.Image image mode=RGB size=130x130 at 0x25840F06F88>"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "my_image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [],
   "source": [
    "#convert image to array \n",
    "\n",
    "my_img_arr = image.img_to_array(my_image)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(130, 130, 3)"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "my_img_arr.shape #(1,130,130,3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [],
   "source": [
    "my_img_arr = np.expand_dims(my_img_arr,axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, 130, 130, 3)"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "my_img_arr.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.00081311]], dtype=float32)"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.predict(my_img_arr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'parasitized': 0, 'uninfected': 1}"
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_image_gen.class_indices"
   ]
  },
  {
   "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": "3.7.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}