covid chest x-ray detection CNN.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\users\\rahul\\appdata\\local\\programs\\python\\python37\\lib\\site-packages\\numpy\\_distributor_init.py:32: UserWarning: loaded more than 1 DLL from .libs:\n",
      "c:\\users\\rahul\\appdata\\local\\programs\\python\\python37\\lib\\site-packages\\numpy\\.libs\\libopenblas.GK7GX5KEQ4F6UYO3P26ULGBQYHGQO7J4.gfortran-win_amd64.dll\n",
      "c:\\users\\rahul\\appdata\\local\\programs\\python\\python37\\lib\\site-packages\\numpy\\.libs\\libopenblas.PYQHXLVVQ7VESDPUVUADXEVJOBGHJPAY.gfortran-win_amd64.dll\n",
      "c:\\users\\rahul\\appdata\\local\\programs\\python\\python37\\lib\\site-packages\\numpy\\.libs\\libopenblas.WCDJNK7YVMPZQ2ME2ZZHJJRJ3JIKNDB7.gfortran-win_amd64.dll\n",
      "  stacklevel=1)\n"
     ]
    }
   ],
   "source": [
    "# Importing the Keras libraries and packages\n",
    "import tensorflow as tf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d (Conv2D)              (None, 64, 64, 32)        896       \n",
      "_________________________________________________________________\n",
      "max_pooling2d (MaxPooling2D) (None, 32, 32, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_1 (Conv2D)            (None, 32, 32, 64)        18496     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 16, 16, 64)        0         \n",
      "_________________________________________________________________\n",
      "flatten (Flatten)            (None, 16384)             0         \n",
      "_________________________________________________________________\n",
      "dense (Dense)                (None, 128)               2097280   \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 1)                 129       \n",
      "=================================================================\n",
      "Total params: 2,116,801\n",
      "Trainable params: 2,116,801\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# Initialising the CNN\n",
    "classifier = tf.keras.models.Sequential()\n",
    "\n",
    "classifier.add(tf.keras.layers.Convolution2D(filters=32, kernel_size=3, padding=\"same\", input_shape=(64, 64, 3),\n",
    "                                             activation='relu'))\n",
    "\n",
    "classifier.add(tf.keras.layers.MaxPooling2D(pool_size=2, strides=2, padding='valid'))\n",
    "\n",
    "classifier.add(tf.keras.layers.Convolution2D(filters=64, kernel_size=3, padding=\"same\", activation=\"relu\"))\n",
    "classifier.add(tf.keras.layers.MaxPooling2D(pool_size=2, strides=2, padding='valid'))\n",
    "\n",
    "classifier.add(tf.keras.layers.Flatten())\n",
    "\n",
    "classifier.add(tf.keras.layers.Dense(units=128, activation='relu'))\n",
    "classifier.add(tf.keras.layers.Dense(units=1, activation='sigmoid'))\n",
    "\n",
    "classifier.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])\n",
    "classifier.summary()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 100 images belonging to 2 classes.\n",
      "Found 100 images belonging to 2 classes.\n"
     ]
    }
   ],
   "source": [
    "# Part 2 - Fitting the CNN to the images\n",
    "\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "\n",
    "train_datagen = ImageDataGenerator(rescale=1. / 255,\n",
    "                                   shear_range=0.2,\n",
    "                                   zoom_range=0.2,\n",
    "                                   horizontal_flip=True)\n",
    "\n",
    "test_datagen = ImageDataGenerator(rescale=1. / 255)\n",
    "\n",
    "training_set = train_datagen.flow_from_directory('G:/projects/corona x-ray images predication/Data/train',\n",
    "                                                 target_size=(64, 64),\n",
    "                                                 batch_size=32,\n",
    "                                                 class_mode='binary')\n",
    "\n",
    "test_set = test_datagen.flow_from_directory('G:/projects/corona x-ray images predication/Data/test',\n",
    "                                            target_size=(64, 64),\n",
    "                                            batch_size=32,\n",
    "                                            class_mode='binary')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "4/4 [==============================] - 23s 5s/step - loss: 0.7137 - accuracy: 0.4900 - val_loss: 0.6062 - val_accuracy: 0.5900\n",
      "Epoch 2/10\n",
      "4/4 [==============================] - 8s 2s/step - loss: 0.5019 - accuracy: 0.8400 - val_loss: 0.4093 - val_accuracy: 0.8300\n",
      "Epoch 3/10\n",
      "4/4 [==============================] - 7s 2s/step - loss: 0.3822 - accuracy: 0.8600 - val_loss: 0.2971 - val_accuracy: 0.9300\n",
      "Epoch 4/10\n",
      "4/4 [==============================] - 7s 2s/step - loss: 0.2365 - accuracy: 0.8800 - val_loss: 0.1829 - val_accuracy: 0.9500\n",
      "Epoch 5/10\n",
      "4/4 [==============================] - 7s 2s/step - loss: 0.2197 - accuracy: 0.9100 - val_loss: 0.2384 - val_accuracy: 0.9400\n",
      "Epoch 6/10\n",
      "4/4 [==============================] - 7s 2s/step - loss: 0.1762 - accuracy: 0.9400 - val_loss: 0.2341 - val_accuracy: 0.9400\n",
      "Epoch 7/10\n",
      "4/4 [==============================] - 7s 2s/step - loss: 0.1958 - accuracy: 0.9300 - val_loss: 0.1456 - val_accuracy: 0.9600\n",
      "Epoch 8/10\n",
      "4/4 [==============================] - 7s 2s/step - loss: 0.1741 - accuracy: 0.9300 - val_loss: 0.1918 - val_accuracy: 0.9500\n",
      "Epoch 9/10\n",
      "4/4 [==============================] - 7s 2s/step - loss: 0.0793 - accuracy: 0.9800 - val_loss: 0.1693 - val_accuracy: 0.9300\n",
      "Epoch 10/10\n",
      "4/4 [==============================] - 7s 2s/step - loss: 0.0768 - accuracy: 0.9700 - val_loss: 0.0883 - val_accuracy: 0.9700\n"
     ]
    }
   ],
   "source": [
    "#training \n",
    "history = classifier.fit(training_set,\n",
    "                         steps_per_epoch=4,\n",
    "                         epochs=10,\n",
    "                         validation_data=test_set,\n",
    "                         validation_steps=4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "classifier.save('G:/projects/corona x-ray images predication/my_model.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4/4 [==============================] - 4s 889ms/step - loss: 0.0883 - accuracy: 0.9700\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.08832349628210068, 0.9700000286102295]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# evaluation on test set\n",
    "loaded_model = tf.keras.models.load_model('G:/projects/corona x-ray images predication/my_model.h5')\n",
    "loaded_model.evaluate(test_set)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3deXxU9bn48c+TfSU7SxJCggKyKSCguFtX3EBrrVsXe1v0trba2/ZWe6+17e29t/fX5XavtV5rN1dUoBUVta4VTUCiCTuSQDYgMFkh+zy/P84EJhsMmMnJzDzv1yuvZOYs88zAnOec7/d7nq+oKsYYYyJXlNsBGGOMcZclAmOMiXCWCIwxJsJZIjDGmAhnicAYYyKcJQJjjIlwlghMRBGRR0Tk+wGuWykiFwc7JmPcZonAGGMinCUCY0KQiMS4HYMJH5YIzKjja5L5hoh8ICIHReT/RGSciDwvIi0i8rKIZPitf42IbBSRRhF5TUSm+y2bKyLv+bZ7Akjo91pXiUipb9u3ReTUAGO8UkQ2iEiziFSJyHf6LT/Ht79G3/LP+p5PFJEfi8guEWkSkbd8z10gItWDfA4X+/7+jogsF5E/i0gz8FkRWSgia32vUScivxSROL/tZ4rISyLiEZG9IvItERkvIodEJMtvvdNFpF5EYgN57yb8WCIwo9XHgUuAqcDVwPPAt4BsnP+3XwEQkanAY8DdQA6wGviriMT5DoorgD8BmcBTvv3i23Ye8DBwO5AF/BZYJSLxAcR3EPg0kA5cCfyziCz17bfAF+8vfDHNAUp92/0IOB04yxfTvwLeAD+TJcBy32v+BegBvur7TBYBFwFf9MWQCrwMvADkAicDr6jqHuA14Aa//d4KPK6qXQHGYcKMJQIzWv1CVfeqag3wJvCuqm5Q1Q7gWWCub71PAs+p6ku+A9mPgEScA+2ZQCzwU1XtUtXlQInfa3wB+K2qvquqPar6B6DDt91Rqeprqlqmql5V/QAnGZ3vW3wL8LKqPuZ73QOqWioiUcDngLtUtcb3mm/73lMg1qrqCt9rtqnqelV9R1W7VbUSJ5H1xnAVsEdVf6yq7araoqrv+pb9Aefgj4hEAzfhJEsToSwRmNFqr9/fbYM8TvH9nQvs6l2gql6gCsjzLavRvpUVd/n9PQn4mq9ppVFEGoGJvu2OSkTOEJFXfU0qTcAdOGfm+Pbx4SCbZeM0TQ22LBBV/WKYKiJ/E5E9vuai/wogBoCVwAwRmYxz1dWkqsUnGJMJA5YITKirxTmgAyAignMQrAHqgDzfc70K/P6uAv5TVdP9fpJU9bEAXvdRYBUwUVXTgAeA3tepAk4aZJv9QPsQyw4CSX7vIxqnWclf/1LBvwG2AFNUdQxO09mxYkBV24Enca5cPoVdDUQ8SwQm1D0JXCkiF/k6O7+G07zzNrAW6Aa+IiIxInIdsNBv298Bd/jO7kVEkn2dwKkBvG4q4FHVdhFZCNzst+wvwMUicoPvdbNEZI7vauVh4Ccikisi0SKyyNcnsQ1I8L1+LPDvwLH6KlKBZqBVRE4B/tlv2d+A8SJyt4jEi0iqiJzht/yPwGeBa4A/B/B+TRizRGBCmqpuxWnv/gXOGffVwNWq2qmqncB1OAe8Bpz+hGf8tl2H00/wS9/yHb51A/FF4Hsi0gJ8Gych9e53N3AFTlLy4HQUn+Zb/HWgDKevwgP8DxClqk2+fT6EczVzEOgzimgQX8dJQC04Se0JvxhacJp9rgb2ANuBC/2W/wOnk/o9X/+CiWBiE9MYE5lE5O/Ao6r6kNuxGHdZIjAmAonIAuAlnD6OFrfjMe6ypiFjIoyI/AHnHoO7LQkYsCsCY4yJeEG7IhCRh0Vkn4iUD7FcROTnIrJDnFIC84IVizHGmKEFs3DVIzijMf44xPLFwBTfzxk4Y6LPGGLdw7Kzs7WwsHB4IjTGmAixfv36/ara/94UIIiJQFXfEJHCo6yyBPij767Pd0QkXUQmqGrd0fZbWFjIunXrhjFSY4wJfyKya6hlbnYW59H3lvlq33MDiMgyEVknIuvq6+tHJDhjjIkUbiYCGeS5QXuuVfVBVZ2vqvNzcga9sjHGGHOC3EwE1Tg1YXrl49SNMcYYM4LcnOVoFXCniDyO00ncdKz+gaF0dXVRXV1Ne3v7sAY4GiUkJJCfn09srM0hYowZHkFLBCLyGHABkO2beel+nNrwqOoDOBOIXIFT3+UQcNuJvlZ1dTWpqakUFhbSt9BkeFFVDhw4QHV1NUVFRW6HY4wJE8EcNXTTMZYr8KXheK329vawTwIAIkJWVhbWYW6MGU5hU2Ii3JNAr0h5n8aYkeNmH4ExxoxqqsoL5XuobmgjNz2RvIxE8tITyU6JC6uTMksEw6CxsZFHH32UL37xi8e13RVXXMGjjz5Kenp6kCIzxpyoTbXN3L+qnJLKhgHL4mKiyEtPJDc9wffbSRB5vmQxPi2B+JhoF6I+MZYIhkFjYyO//vWvBySCnp4eoqOH/s+wevXqYIdmjDlOTW1d/O9L2/jj2krSEmP57+tms3jWeGob26ltbKOmsY3axjaqfb9f21rPvpaOPvsQgZyU+D5XEf0TxpjEmFFzVWGJYBjcc889fPjhh8yZM4fY2FhSUlKYMGECpaWlbNq0iaVLl1JVVUV7ezt33XUXy5YtA46Uy2htbWXx4sWcc845vP322+Tl5bFy5UoSExNdfmfGRA6vV1n+XjX/8/wWPIc6ueWMAr5+6TTSk+IASE+KY0bumEG37ejuYU9TOzWNbdQ0tFHb2E5N4yFqG9vZVNvMS5v20tnt7bNNSnxM3yuKjL5XFWNTE4iOGplEEXaJ4Lt/3cim2uZh3eeM3DHcf/XMIZf/4Ac/oLy8nNLSUl577TWuvPJKysvLDw/xfPjhh8nMzKStrY0FCxbw8Y9/nKysrD772L59O4899hi/+93vuOGGG3j66ae59dZbh/V9GGMGV17TxH0ry9mwu5F5Ben84XMLmZWXFvD28THRTMpKZlJW8qDLVZX9rZ19rygajvy9oaqRxkNdfbaJiRLGpyWQm55I0ZgoLu1Yw9hTFjH7zIs/0nsdTNglgtFg4cKFfcb5//znP+fZZ58FoKqqiu3btw9IBEVFRcyZMweA008/ncrKyhGL15hI1Xiokx++uJVHi3eTmRTHD68/lY/PyydqmM/ERYSc1HhyUuM5beLgfYIHO7oPJ4reBHFgfz2z655gcd0zZNJESeetYIng2I525j5SkpOPnBW89tprvPzyy6xdu5akpCQuuOCCQe+Ajo+PP/x3dHQ0bW1tIxKrMZHI61WeWFfF/3thC01tXXxmUSFfvWQqaYnu3bGfHB/DlHGpTBmXCgcPwLt/gt0PQkcTnHQR3Wd/lVn5ZwXltcMuEbghNTWVlpbBZ/xramoiIyODpKQktmzZwjvvvDPC0Rlj/JVWNXL/ynLer25iQWEG31syi+kTBm/7H3FNNbD2l7D+Eehqg+lXw7n/ArlziSF4B2xLBMMgKyuLs88+m1mzZpGYmMi4ceMOL7v88st54IEHOPXUU5k2bRpnnnmmi5EacwwdrVBdAp4PIXUCpE2EtHxIzHCGwoQwz8FO/t8LW3hiXRXZKfH89JNzWDInd3SM3DnwIfzjp1D6GKgXTr0Bzvkq5EwbkZcPuTmL58+fr/0nptm8eTPTp093KaKRF2nv1wTRwQOwe63zs+ttqHsftGfgenEpR5JCuu93WsGRxynjIXp0nlf2eJVH393Fj9Zso7Wjm9vOKuSui6eQmjAKCjfuKYe3/hc2PgNRsTDvU3DWVyBj0rC/lIisV9X5gy0bnf9yxpjgaNwNu9bC7red3/u3Os9Hx0Pe6XDO3VBwlnMmenAfNFZBUzU0+X437oaa9dDm6btfiYYxeX6JYuLAv+MGH1ETTOt3NfDtleVsrG1m0eQsvrtkJlPHpY54HANUlcCbP4ZtzztJ9qwvw5lfgtRxx942CCwRGBOuVKF+65GD/u61zgEdIH4MTDwDTvukc+DPnQuxCX23T5/oJIfBdLRCc40vUez2JQlfstj1NjTXDryySMz0Swz9E8VESM4etuan+pYOfvD8Fp5+r5rxYxL4xU1zuerUCe42A6lCxevwxo+g8k2nue3Cf4OFX3D+dpElAuOe5jqoetdpE02b6BwUksdCVNjUQhxZPV1Q90HfA3/vmXvyWJi0CBbd6fweNwuiPkIJhPgU56phqDbsnm5oqfO7mqg6kigO7IAPX4Wug323iUnwNTn1JokCyDoJChbBmNyAwuru8fKnd3bxk5e20d7Vw+3nT+YrH5tCcryLhzqv1znzf/PHztVU6gS47L9g3mecz3EUsERgRoaq0yF2+CD1NjRUDlwvOs6viaFg4IFhTN7AM9dI1XnI6djtbd+vXnfk4JpRBNMWOwfRSWdB5uSR7eyNjnESe/pEYNHA5arQ1uDX5FR1JGE0VcO2F52mqV7pk5z3Meks5wom66QB7+fdnQe4f9VGtuxp4dwp2dx/9UxOHuvigban22n7f/MnUL8ZMgrhqp/CnJshJv6Ym48kSwQmOLw9sLfcOejv+gfsfufIFzspyzlALfiCc3Yak3Ck/dm/PfrDV52zyv5TWSeP9euwnHjkaqL3cRiMcBnUIY/zOfYm07pS8HYD4pzhz73F+VwLFsGYCW5He3QikJTp/Ew4bfB1utph36YjiW77Gnj/MWdZcs7hJOfJOp3/WBfFs+/vJS89kQduncdlM8e71wzU3QGlf4F//Mw52cmZDtc9BDOvHbUd6qMzKhN6utqh9j3nC7t7LVQVQ4ev1EdaAZx04ZGz0+ypAw/U44a4EbCny68tul8zw95Nzpljd78b9GKT+yWKflcXqRNG7Reyj6aaIwfB3WudgyI4V02585wOxoKzYOJCSAzDCraxCZA3z/lZ9CXnKmL/9sOJUHf9A9m8ikzgPzSRr4w7jfw5FxGbEg/dGSN/5djR6oz/f/sX0LrH6V+57L9h6uWjvrkzBL4No9+JlqEG+OlPf8qyZctISkoKQmRB1N7kHOx7D1I166Gn01mWMx1mX+8cpCYtcg7AJyo61rmkzigcfLkqHDrQ92rCv5mhdoOz3J9EO23O/h2WY/Ig1t0if60d3VTW1TO26QMy9q8ntsXXsRuX4hzsZ17nJNK8ea7H6goRyJkKOVN5O/0q7q/cSGv7Lj6TX8vN42so2lsCr/8nvM6RZDlpEUw62/n8EgKvHXRcDnmg+Hfw7m+c5q6i8+C630LR+SFzZWr3EQyDyspKrrrqKsrLy497294KpNnZ2QFv48r7bdnbt31/70ankzcqxrm0L/B94QrOdC73R5POg87ZdZ/RLX5t0801g4+dd8l+HUOJdxol3lPYnjibg+nTmZCZMupLGY+EuqY2vv/cZp77oI6JmYl8+6qZXDx97JHP4GjNZ+NnHTk5KTjrow/VbNnr3AW87mHobIVpV8A5/wITF3zk9xkMdh9BkPmXob7kkksYO3YsTz75JB0dHVx77bV897vf5eDBg9xwww1UV1fT09PDfffdx969e6mtreXCCy8kOzubV1991e234lCFhgq/8eZvg2ensyw2CfLnw3n/6nyh8he4Mj78uMQlHz6THJS3B1r3HrmiCbIer/Le7kbWbNzD69vqOdTVw9jUeC6ePo5zT5lAa+xYDja1k9bQRq6vANnm2mZe3rSXjuMoZZybnsi4MSNXyjiYOru9/N9bFfzi79vp8Sp3XzyFO84/iYTYfiOfkjLhlCucH3BOAqrXHWli2/AnKP6tsyxzsl9iWBR4h3rDLnj75/Den8DbBbM+7twFPFTzZggIv0Tw/D2wp2x49zl+Niz+wZCL/ctQr1mzhuXLl1NcXIyqcs011/DGG29QX19Pbm4uzz33HODUIEpLS+MnP/kJr7766nFdEQw7b4/T/ux/o1HrHmdZYobzJTn9NqdZYsJpTnNNOImKDnh44olSVTbWNrNiQw2r3q9lX0sHqQkxXHHqHJbOzeOMosxjVrxUVQ4c7KSmoW8p496KlaVVjTT0K2UcHSWMH5MwcHKUjMTDM2wlxY3uw8Ab2+r5zqqN7Nx/kIunj+P+q2cwMTPAptS4ZJh8vvMDviG27x9p0tz6HJT+2VmWMv7I1cKkRTB2Rt8htvVbnbuAP3gSJMoZ/XP2Xc4IphA3uv8HhKA1a9awZs0a5s6dC0Brayvbt2/n3HPP5etf/zrf/OY3ueqqqzj33HPdDbRxN5Qt95UXeNepcAhOW3nhOUeG6mVPG/UdXaNZlecQq96v5dkNNezY10pstPCxU8aydE4eF54yduAZ7VGICNkp8WSnHL2UcV1T31r3vROlFFd42NPcTo+3b3NwZnLcgCkXc1LjXb+SUIXnPqjjhY17KMxK4vefXcCFp4z9aDuNjnWuaPPnw9lfccb47996JDHsWgsbnZLxxKdBwRlOc2ftBtj8N6dv5ow74Kw7g37yMJLCLxEc5cx9JKgq9957L7fffvuAZevXr2f16tXce++9XHrppXz72992IUKcuz7/71JnaGb2VJi51Dc+e5EzuiaC2pyDoeFgJ8+V1bGytObwfLcLizL5r2tnc8Xs8YdnvAqG5PgYTh6bysljBy+j0N3jZW9Lx+EE0Vv7vqahjZ31B3lz+34OdY6e/pKE2Ci+fulUPn/u5ONKmgGLioKx052fBf/kPNe/DMf2NU5H83nfcJJActbR9xmCwi8RuMC/DPVll13Gfffdxy233EJKSgo1NTXExsbS3d1NZmYmt956KykpKTzyyCN9th2xpqHOQ/DYTdDRAre/CRNOHZnXDXPtXT28snkfK0preG3rPrp6lCljU/jGZdNYMieX/IzRMSosJjrqcBPRgsKBy1WVprYu9rd2MBrGkeSkxgc1cQ4qvcD5Oe2TzuNDHudel7jR8W8YDJYIhoF/GerFixdz8803s2iRczdlSkoKf/7zn9mxYwff+MY3iIqKIjY2lt/85jcALFu2jMWLFzNhwoTgdxZ7vfDs7U4b6c1PWBL4iHq8yrs7D/DshhpeKN9DS0c348bE89mzClk6N48ZE8aE3IgeESE9KW7kD76j2WgbBRcENnw0BJ3w+/379+GNH8Kl/+m0cZrjpqpsrmthRWkNK0tr2NvcQUp8DItnjWfp3DzOnJzletu6MYOx4aPGGenwxg9h3qeduzTNcalpbGNlaQ0rNtSwbW8rMVHCBdPGct9VuVw8fVxw2q+NGSGWCCJBVTGsvBMmnQNX/Ng6gwPUdKiL1eV1PLuhhuIKp4rn/EkZfH/pLK6cPYGMZGs+MeEhbBKBqoZce+yJOO6mvMbd8PjNkJYHn/wTxNjB62jau3p4bes+nt1Qw6tb6uns8XJSTjJfv3QqS+bkBT5+3ZgQEhaJICEhgQMHDpCVlRXWyUBVOXDgAAkJARbT6miBR2+E7k747BMR0el1Irxe5d0KDytLa3iurI6W9m5yUuP51KJJLJ2Tx6y80Ov0NeZ4hEUiyM/Pp7q6mvr6erdDCbqEhATy8wMo4ubtgac/D/Vb4NblQ5dXiGBb9jSzYkMtq0prqG1qJzkumstmjefauXksmpxFTLTdSGciQ1ATgYhcDvwMiAYeUtUf9FueATwMnAS0A59T1eOu3BYbG0tRUdEwRBxGXr4ftr0AV/wITvqY29GMGnVNbawsrWXFhhq27GkhJko4b2oO31x8CpfOGE9inHX6msgTtEQgItHAr4BLgGqgRERWqeomv9W+BZSq6rUicopv/YuCFVPEeO+PTk30BV9w5kONcE1tXbxQXseKDbW8U3EAVZhbkM73lszkytkTyEoZXbNFGTPSgnlFsBDYoao7AUTkcWAJ4J8IZgD/DaCqW0SkUETGqereIMYV3irfgr/9i3MVcLm75Tbc1NHdw2tb61mxoYZXtuyjs9tLUXYyd180lSVzcinMHuUVU40ZQcFMBHlAld/jauCMfuu8D1wHvCUiC4FJQD5gieBEeHbCE7dCZhFc//vQmIVrGHm9yrpdDTy7oYbVZXU0tXWRlRzHzQsLuHZuHqfmp1mnrzGDCOaRYrBvXP+xjz8AfiYipUAZsAHoHrAjkWXAMoCCgoJhDjNMtDXCo77aKDc9Hp5TFw5h294WVmyoYWVpLTWNbSTGRnPZzHEsnZvHOSdnW6evMccQzERQDUz0e5wP1PqvoKrNwG0A4pyqVfh+6Lfeg8CD4JSYCFK8oaunG5bf5lwRfHplWNRHP5a9ze2sKnXKO2+qayY6Sjjn5Gy+cdk0LpkxjuT4yLoaMuajCOa3pQSYIiJFQA1wI3Cz/woikg4cUtVO4PPAG77kYI7Hi/fCh3+Ha37hzCUQplrau3ihfA8rSmt4+0On0/e0/DTuv3oGV52aS06qdfoacyKClghUtVtE7gRexBk++rCqbhSRO3zLHwCmA38UkR6cTuR/ClY8Yav4d1D8ICy606kjFGY6u728vq2eFaU1h6dqLMhM4ssfm8LSOblMzklxO0RjQl5Qr59VdTWwut9zD/j9vRaYEswYwtqHf4fnvwlTL4dLvud2NMNGVVm/q4EVpTX87YM6Gg91kZEUyycXTGTp3DzmTky3Tl9jhpE1pIaq+m3w5Gch5xT4+EN951YNUTv2tTqdvu/XUOVpIyE2iktmjOfaubmcOyWHWOv0NSYoLBGEokMeePQGp4DczY9D/ODTEoaCuqY2nvugjpWltZTVNBElcPbJ2dx90VQumzWeFOv0NSbo7FsWaro74clPQ3MNfOZvzpR6IUJVqdh/kJJKD8UVDZRUetjtOQTA7Lw0/v3K6VxzWi5jxwRYVM8YMywsEYQSVVj9Nah8E677HRT0vz9vdOnxKpvrmimu8FBS6aGksoH9rR0AZCTFsqAwk08vmsQF03KGnGzdGBN8lghCyTu/duoInft1OPUGt6MZoL2rh/erGp0z/soG3tvVQGuHc39gXnoi55ycxYKiTM4oyuSknBTr8DVmlLBEECq2vQgv/htMvwYu/De3owGgub2L9bsaKPGd8b9f1URnjxeAKWNTuGZOLgsLM1lQlEleeqLL0RpjhmKJIBTs3QTLPwcTToVrH4Aod0bP7Gtpp8TXtl9c4WHLnma8CtFRwqy8ND5z1iQWFGYyvzCTTJvG0ZiQYYlgtGutd2oIxaU4NYTiRqZqpqqy68Ahiis9h8/4Kw84HbsJsVHMK8jgyx+bwsKiTOYWpJMUZ/+VjAlV9u0dzbra4Ylb4GA93LYaxuQG7aV6vMrWPS2+9n3n4L+vxenYTUuMZUFhBjctLGBBUSazctOIi7Ex/caEC0sEo5Uq/PUuqHoXPvEI5M0b1t13dPdQVt10+KC/blcDLe1Ox+6EtATOnOx07C4szGTK2BSioqxj15hwZYlgtHrrJ/DB407H8MxrP/LuWju6D3fsFld6eL+qkY5up2P3pJxkrjp1AgsKM1lQmEl+RqKN6DEmglgiGI02rYJXvgezrofzvnFCu9jf2sE6vxu3NtY24VWIEpiZm8YtZ0xiYVEGCwozbapGYyKcJYLRprYUnr0d8ubDkl9CAGfmqkp1Q9vhG7eKKz3srD8IQHxMFHMmpvOlC09mQWEm8yZlWNkGY0wfdkQYTZrr4LGbIDETbnwUYgcfe+/1Ktv3tVLsG8ZZUuFhT3M7AGMSYphfmMknTp/IwqIMZuWlER8T+gXpjDHBY4lgtOg8BI/fBO1N8E8vQuq4I4u6vZTXNh0exllS2UBTWxcA48bEs6Awk4VFTvv+tHGp1rFrjDkulghGA68XVn7RaRa68VEOZU7nve37D4/o2VDVQHuX07FblJ3MZTPHHT74F2QmWceuMeYjsUQwCrS99J8kbnyWV/K/xM9fSWPjH9fQ7VWiBKZPGMONCwpYWJTJ/MIMxqZaZU5jzPCyROCCmsa2w8M4U7Y9y7fafsyT3efz75XnMCc/itvPn3y4Y3dMQqzb4RpjwpwlghH2p7WV3LdyIwBnxVfwB/k5tWlzmXzNg3xQkENCrHXsGmNGliWCEbZm014Ks5L47TXjmLrqbiQ2l9wvPE1ucpbboRljIpQVjBlB3T1e3tvVwEWTk5n292VIdzvc/CRYEjDGuMiuCEbQprpmDnV28fn9P4B9G+Hmp2DsKW6HZYyJcJYIRlBxhYfPRr/IhLpX4PL/gSkXux2SMcZY09BIKq7w8Km41yF/AZxxu9vhGGMMYIlgxKgqDZXvM1l3w+wbAqohZIwxI8ESwQj5sL6V8ztfxyvRMHOp2+EYY8xhlghGSPFOD9dEvU17/jmQMtbtcIwx5jDrLB4heze/SUFUPTrvk26HYowxfdgVwQjJr15Nl8Qi0692OxRjjOnDEsEIqD7QzAXdb1GTcz4kjHE7HGOM6cMSwQioXPciOdJE1KnXux2KMcYMYIlgBCRseYZWEslbsMTtUIwxZoCgJgIRuVxEtorIDhG5Z5DlaSLyVxF5X0Q2ishtwYzHFV3tnNLwOqXJ5xAdn+R2NMYYM0DQEoGIRAO/AhYDM4CbRGRGv9W+BGxS1dOAC4Afi0hcsGJyQ0v5alI4SMNJ17gdijHGDCqYVwQLgR2qulNVO4HHgf5tIwqkijPXYgrgAbqDGNOIa13/BPt1DLlzL3c7FGOMGVQwE0EeUOX3uNr3nL9fAtOBWqAMuEtVvf13JCLLRGSdiKyrr68PVrzDr72Z7Jq/84KeyeyJ2W5HY4wxgwpmIhismI72e3wZUArkAnOAX4rIgPGVqvqgqs5X1fk5OTnDH2mwbF1NrHayLecy4mKsX94YMzoF8+hUDUz0e5yPc+bv7zbgGXXsACqAsCnQ3/3+U1RrNmlTz3Y7FGOMGVIwE0EJMEVEinwdwDcCq/qtsxu4CEBExgHTgJ1BjGnkHNxPdMWr/LVnEQsnW7OQMWb0ClqtIVXtFpE7gReBaOBhVd0oInf4lj8A/AfwiIiU4TQlfVNV9wcrphG1aQWiPfxVz+bTBRluR2OMMUMKatE5VV0NrO733AN+f9cClwYzBteULacqpoCYrFkkx1ttP2PM6BVQ05CIPC0iV4qI9XgGorEKdq/lqY5FLCyyiemNMaNboAf23wA3A9tF5AciEjYdukFR/jQAK7rPZGIGP2AAABRnSURBVEFRpsvBGGPM0QWUCFT1ZVW9BZgHVAIvicjbInKbiMQGM8CQVL6cvakz2a3jWFBoicAYM7oF3NQjIlnAZ4HPAxuAn+EkhpeCElmoqt8Ke8p4OeY8poxNITM5rCpmGGPCUEC9mCLyDM74/j8BV6tqnW/REyKyLljBhaSy5ahE8ZBnDovm2NWAMWb0C3Q4yy9V9e+DLVDV+cMYT2hThfLlHJywiIqdqdxlzULGmBAQaNPQdBFJ730gIhki8sUgxRS6at8Dz05K0y8GYKF1FBtjQkCgieALqtrY+0BVG4AvBCekEFb2NETF8mz7PPLSE8lNT3Q7ImOMOaZAE0GUr1Q0cHiuAesF9eftgY3PoFMu4fXdXXY1YIwJGYEmgheBJ0XkIhH5GPAY8ELwwgpBu/4BLXXsm3QV+1s7LREYY0JGoJ3F3wRuB/4ZpybQGuChYAUVksqWQ2wyb0bNB3bY/QPGmJARUCLwTRbzG9+P6a+7EzathFOu5O3dbWQlx3FSTrLbURljTEACrTU0RUSWi8gmEdnZ+xPs4ELGh69AeyPM/gQllR4WFGbi16VijDGjWqB9BL/HuRroBi4E/ohzc5kBKHsKEjOpyz6TKk+b1RcyxoSUQBNBoqq+Aoiq7lLV7wAfC15YIaTzIGx9HmYsoXh3CwALrX/AGBNCAu0sbveVoN7um2ymBhgbvLBCyNbnoeuQ0yxU6iE5LprpE1LdjsoYYwIW6BXB3UAS8BXgdOBW4DPBCiqklD0FY/KgYBElFQ2cXphJTLRN22CMCR3HPGL5bh67QVVbVbVaVW9T1Y+r6jsjEN/odsgDO16GWdfR0NbN1r0tLCy0aSmNMaHlmIlAVXuA08WGwQy0aSV4u2HW9azb1QBg9w8YY0JOoH0EG4CVIvIUcLD3SVV9JihRhYrypyHrZJhwGiWlW4iLjuK0ienH3s4YY0aRQBNBJnCAviOFFIjcRNBcC5VvwQX3gAjFFR5Om5hGQmy025EZY8xxCfTO4tuCHUjIKX8GUJh1PYc6uymvaWLZeZPdjsoYY45boDOU/R7nCqAPVf3csEcUKsqXw4Q5kH0yG3bsp9urVmjOGBOSAm0a+pvf3wnAtUDt8IcTIg58CLUb4NLvA1Bc4SFK4PRJNmLIGBN6Am0aetr/sYg8BrwclIhCQdlyQGDmdYCTCKZPGENqQqy7cRljzAk40TufpgAFwxlIyFB1biKbdDak5dHZ7WVDVYM1CxljQlagfQQt9O0j2IMzR0Hk2fMBHNgOi5wpm8trm2jv8lp9IWNMyAq0aciK5/QqWw5RMTBjKeA0CwHMt0RgjAlRgc5HcK2IpPk9TheRpcELa5Tyep2byE66CJKcA39JhYfJOcnkpMa7HJwxxpyYQPsI7lfVpt4HqtoI3B+ckEaxqneguQZmfwIAr1cpqfRYs5AxJqQFmggGWy/Qoafho+wpiEmEaYsB2Lq3heb2bqsvZIwJaYEmgnUi8hMROUlEJovI/wLrj7WRiFwuIltFZIeI3DPI8m+ISKnvp1xEekRkdB5Ve7pg4wonCcSnAFBS6fQP2IghY0woCzQRfBnoBJ4AngTagC8dbQNf+epfAYuBGcBNIjLDfx1V/aGqzlHVOcC9wOuq6jm+tzBCPnwV2jyHm4XA6SiekJZAfkaii4EZY8xHE+iooYPAgDP6Y1gI7FDVnQAi8jiwBNg0xPo3AY8d52uMnPLlkJAGJ18EgKpSXOHhzMlZNlG9MSakBTpq6CURSfd7nCEiLx5jszygyu9xte+5wfafBFwOPD3E8mUisk5E1tXX1wcS8vDqPARbnoMZSyDGGR2023OIfS0d1ixkjAl5gTYNZftGCgGgqg0ce87iwU6TBxSu87ka+MdQzUKq+qCqzlfV+Tk5OQEFPKy2vQCdrTDr+sNP9d4/YInAGBPqAk0EXhE5XFJCRAoZ+qDeqxqY6Pc4n6EL1d3IqG4WehpSxkPhOYefKqn0kJ4Uy8k5KS4GZowxH12gQ0D/DXhLRF73PT4PWHaMbUqAKSJSBNTgHOxv7r+S70a184FbA4xlZLU1wvY1sODzEHVk0pniCg/zJ2USFWX9A8aY0BbQFYGqvgDMB7bijBz6Gs7IoaNt0w3cCbwIbAaeVNWNInKHiNzht+q1wBpfh/Tos/mv0NPZp1loX0s7lQcOcYY1CxljwkCgRec+D9yF07xTCpwJrKXv1JUDqOpqYHW/5x7o9/gR4JFAAx5xZU9BRhHkzTv8VEmFb6J6SwTGmDAQaB/BXcACYJeqXgjMBVwYvjPCWvZC5Zsw+3rwGyJaXHGAxNhoZuaOcTE4Y4wZHoEmgnZVbQcQkXhV3QJMC15Yo8TGZ0G9fW4iAyiubOD0SRnERp/odA7GGDN6BHokq/bdR7ACeElEVhIJU1WWPQXjZkPOkZzX1NbFlj3NVl/IGBM2Ar2z+Frfn98RkVeBNOCFoEU1GngqoGYdXPydPk+v3+VBFRYU2fzExpjwcNwVRFX19WOvFQbKlzu/Z328z9PFFQ3ERgtzJ1oiMMaEB2vkHkrZ0zDxTEjvOzVzSaWH2XlpJMZFD7GhMcaEFksEg9m7Eeo3O6OF/LR39fBBdaMNGzXGhBVLBIMpewokGmZe2+fpDbsb6epRu5HMGBNWLBH0p+o0C510ISRn91lUUulBBE6fZInAGBM+LBH0V1UMTbv7lJToVVzhYdq4VNISY10IzBhjgsMSQX/lyyEmAU65ss/T3T1e3tvdYM1CxpiwY4nAX0+3czfx1MsgoW/5iI21zRzq7LGOYmNM2LFE4K/idThYP2iz0OGJ6u2OYmNMmLFE4K/8aYgfA1MuHbDo3QoPhVlJjB2T4EJgxhgTPJYIenW1O3MPTL8aYvse7L1eZV2lx+oLGWPCkiWCXtvXQEfzgJISAB/Wt9JwqMv6B4wxYckSQa+ypyA5B4rOH7DoXd9E9TZiyBgTjiwRALQ3w7YXnTuJowfW4Sup9DA2NZ6CzCQXgjPGmOCyRACw5W/Q0zFgAhoAVaW4wsOCokxEbKJ6Y0z4sUQAULbcqTKav2DAouqGNuqa2m3YqDEmbFkiaK2Hna859w4McsZ/+P4B6x8wxoQpSwSbVoD2DCg53auk0sOYhBimjUsd4cCMMWZkWCIoWw4502HczEEXv1vhYX5hJlFR1j9gjAlPkZ0IGndD1TtDXg3sb+1gZ/1BaxYyxoS1yE4E5U87vwe5iQxgna9/wO4oNsaEs8hOBGVPOyOFMosGXfxuhYeE2Chm56WNcGDGGDNyIjcR7NsCe8sGrTTaq6TSw9yJGcTFRO7HZIwJf5F7hCtfDhI1YF7iXi3tXWyqbbb6QsaYsBeZiUDVqS1UdB6kjht0lfW7GvCqzT9gjAl/kZkIat6DhspjNgvFRAnzJqWPXFzGGOOCyEwEZU9BdJwz98AQSioamJmXRlLcwCJ0xhgTToKaCETkchHZKiI7ROSeIda5QERKRWSjiLwezHgA8PbAxmecWcgSBz/bb+/qobS6kYWFGUEPxxhj3Ba0010RiQZ+BVwCVAMlIrJKVTf5rZMO/Bq4XFV3i8jYYMVzWOVb0Lp3yJvIAD6obqKz28vCoqygh2OMMW4L5hXBQmCHqu5U1U7gcWBJv3VuBp5R1d0AqroviPE4yp6CuBSYevmQq/QWmps/ya4IjDHhL5iJIA+o8ntc7XvO31QgQ0ReE5H1IvLpIMYD3R2weRWcciXEJg65WnGFh6njUshIjgtqOMYYMxoEsyd0sCptOsjrnw5cBCQCa0XkHVXd1mdHIsuAZQAFBQUnHtGOl6G9adAJaHr1eJX1uxpYOjf3xF/HGGNCSDCvCKqBiX6P84HaQdZ5QVUPqup+4A3gtP47UtUHVXW+qs7Pyck58YjKlkNiJky+YMhVNtc109rRbfWFjDERI5iJoASYIiJFIhIH3Ais6rfOSuBcEYkRkSTgDGBzUKLpaIWtz/vmJY4dcrXiCpuIxhgTWYLWNKSq3SJyJ/AiEA08rKobReQO3/IHVHWziLwAfAB4gYdUtTwoAW1dDd1tRx0tBE4iyM9IZELa0H0IxhgTToJ6t5SqrgZW93vugX6Pfwj8MJhxAHDyxbDk1zDxzCFXUVVKKj2cP+0jND8ZY0yIiZzbZpMyYe4tR11l5/6DHDjYafWFjDERJTJLTAyht3/AKo4aYyKJJQI/JRUeslPimJyd7HYoxhgzYiwR+Cmu9LCgMBMRm6jeGBM5LBH41Da2Ud3QZvcPGGMijiUCn976Qnb/gDEm0lgi8Cmu8JAaH8P0CWPcDsUYY0aUJQKfkkoP8yZlEB1l/QPGmMhiiQBoONjJtr2t1ixkjIlIlgiw/gFjTGSzRICTCOJiojg1P83tUIwxZsRZIsDpKJ4zMZ34mGi3QzHGmBEX8YngYEc35bXNVl/IGBOxIj4RbNjdSI9Xrb6QMSZiRXwiKK44QJTA6TZRvTEmQlkiqPQwMzeNlPjIqchtjDH+IjoRdHZ72bC70eoLGWMiWkQngrKaRjq6vXb/gDEmokV0IiiuaABgQaH1DxhjIldEJ4KSSg8n5SSTlRLvdijGGOOaiE0EPV5nonprFjLGRLqITQRb97TQ0t5ticAYE/EiNhH0FpqzEUPGmEgXsYmguNJDbloC+RlJbodijDGuishEoKoUV1j/gDHGQIQmgl0HDlHf0mH1hYwxhghNBMW9E9FY/4AxxkRoIqjwkJkcx8ljU9wOxRhjXBeRiaCk0sP8SRmI2ET1xhgTcYlgb3M7uw4cso5iY4zxibhEUFxhE9UbY4y/iEsEJZUekuOimTFhjNuhGGPMqBDURCAil4vIVhHZISL3DLL8AhFpEpFS38+3gxkPOFcE8yZlEBMdcTnQGGMGFbRpuUQkGvgVcAlQDZSIyCpV3dRv1TdV9apgxeGv6VAXW/e2cOXsCSPxcsYYExKCeVq8ENihqjtVtRN4HFgSxNc7pnW7PKhiN5IZY4yfYCaCPKDK73G177n+FonI+yLyvIjMHGxHIrJMRNaJyLr6+voTDqi40kNstDBnYvoJ78MYY8JNMBPBYIP0td/j94BJqnoa8AtgxWA7UtUHVXW+qs7Pyck54YCKKzyclp9OQmz0Ce/DGGPCTTATQTUw0e9xPlDrv4KqNqtqq+/v1UCsiGQHI5i2zh7KqpusWcgYY/oJZiIoAaaISJGIxAE3Aqv8VxCR8eK7vVdEFvriORCMYDZUNdDtVasvZIwx/QRt1JCqdovIncCLQDTwsKpuFJE7fMsfAK4H/llEuoE24EZV7d98NCxio6O4cFoO8ybZRPXGGONPgnTcDZr58+frunXr3A7DGGNCioisV9X5gy2zu6qMMSbCWSIwxpgIZ4nAGGMinCUCY4yJcJYIjDEmwlkiMMaYCGeJwBhjIpwlAmOMiXAhd0OZiNQDu05w82xg/zCGE+rs8+jLPo8j7LPoKxw+j0mqOmjVzpBLBB+FiKwb6s66SGSfR1/2eRxhn0Vf4f55WNOQMcZEOEsExhgT4SItETzodgCjjH0efdnncYR9Fn2F9ecRUX0ExhhjBoq0KwJjjDH9WCIwxpgIFzGJQEQuF5GtIrJDRO5xOx43ichEEXlVRDaLyEYRucvtmNwmItEiskFE/uZ2LG4TkXQRWS4iW3z/Rxa5HZNbROSrvu9IuYg8JiIJbscUDBGRCEQkGvgVsBiYAdwkIjPcjcpV3cDXVHU6cCbwpQj/PADuAja7HcQo8TPgBVU9BTiNCP1cRCQP+AowX1Vn4Uy5e6O7UQVHRCQCYCGwQ1V3qmon8DiwxOWYXKOqdar6nu/vFpwvep67UblHRPKBK4GH3I7FbSIyBjgP+D8AVe1U1UZ3o3JVDJAoIjFAElDrcjxBESmJIA+o8ntcTQQf+PyJSCEwF3jX3Uhc9VPgXwGv24GMApOBeuD3vqayh0Qk2e2g3KCqNcCPgN1AHdCkqmvcjSo4IiURyCDPRfy4WRFJAZ4G7lbVZrfjcYOIXAXsU9X1bscySsQA84DfqOpc4CAQkX1qIpKB03JQBOQCySJyq7tRBUekJIJqYKLf43zC9BIvUCISi5ME/qKqz7gdj4vOBq4RkUqcJsOPicif3Q3JVdVAtar2XiEux0kMkehioEJV61W1C3gGOMvlmIIiUhJBCTBFRIpEJA6nw2eVyzG5RkQEpw14s6r+xO143KSq96pqvqoW4vy/+LuqhuVZXyBUdQ9QJSLTfE9dBGxyMSQ37QbOFJEk33fmIsK04zzG7QBGgqp2i8idwIs4Pf8Pq+pGl8Ny09nAp4AyESn1PfctVV3tYkxm9Pgy8BffSdNO4DaX43GFqr4rIsuB93BG2m0gTEtNWIkJY4yJcJHSNGSMMWYIlgiMMSbCWSIwxpgIZ4nAGGMinCUCY4yJcJYIjBlBInKBVTg1o40lAmOMiXCWCIwZhIjcKiLFIlIqIr/1zVfQKiI/FpH3ROQVEcnxrTtHRN4RkQ9E5FlfjRpE5GQReVlE3vdtc5Jv9yl+9f7/4rtr1RjXWCIwph8RmQ58EjhbVecAPcAtQDLwnqrOA14H7vdt8kfgm6p6KlDm9/xfgF+p6mk4NWrqfM/PBe7GmRtjMs6d3sa4JiJKTBhznC4CTgdKfCfricA+nDLVT/jW+TPwjIikAemq+rrv+T8AT4lIKpCnqs8CqGo7gG9/xapa7XtcChQCbwX/bRkzOEsExgwkwB9U9d4+T4rc12+9o9VnOVpzT4ff3z3Y99C4zJqGjBnoFeB6ERkLICKZIjIJ5/tyvW+dm4G3VLUJaBCRc33Pfwp43Te/Q7WILPXtI15Ekkb0XRgTIDsTMaYfVd0kIv8OrBGRKKAL+BLOJC0zRWQ90ITTjwDwGeAB34Hev1rnp4Dfisj3fPv4xAi+DWMCZtVHjQmQiLSqaorbcRgz3KxpyBhjIpxdERhjTISzKwJjjIlwlgiMMSbCWSIwxpgIZ4nAGGMinCUCY4yJcP8fM2reehpccoEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "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": [
    "import matplotlib.pyplot as plt\n",
    "plt.plot(history.history['accuracy'])\n",
    "plt.plot(history.history['val_accuracy'])\n",
    "plt.title('model accuracy')\n",
    "plt.ylabel('accuracy')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'test'], loc='upper left')\n",
    "plt.show()\n",
    "\n",
    "plt.plot(history.history['loss'])\n",
    "plt.plot(history.history['val_loss'])\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'test'], loc='upper left')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Covid\n"
     ]
    },
    {
     "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": [
    "# for only one prediction\n",
    "import numpy as np\n",
    "from tensorflow.keras.preprocessing import image\n",
    "\n",
    "test_image = image.load_img(r'G:\\projects\\corona x-ray images predication\\Data\\test\\Covid\\16672_1_1.jpg',target_size=(64, 64))\n",
    "plt.imshow(test_image)\n",
    "test_image = image.img_to_array(test_image)\n",
    "test_image = np.expand_dims(test_image, axis=0)\n",
    "result = classifier.predict(test_image)\n",
    "training_set.class_indices\n",
    "if result[0][0] == 1:\n",
    "    prediction = 'Normal'\n",
    "else:\n",
    "    prediction = 'Covid'\n",
    "print(prediction)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# plot confusion metrix\n",
    "y_pred = []\n",
    "y_test = []\n",
    "import os\n",
    "\n",
    "for i in os.listdir(\"G:/projects/corona x-ray images predication/Data/test/Normal\"):\n",
    "    img = image.load_img(\"G:/projects/corona x-ray images predication/Data/test/Normal/\" + i, target_size=(64, 64))\n",
    "    img = image.img_to_array(img)\n",
    "    img = np.expand_dims(img, axis=0)\n",
    "    p = classifier.predict_classes(img)\n",
    "    y_test.append(p[0, 0])\n",
    "    y_pred.append(1)\n",
    "\n",
    "for i in os.listdir(\"G:/projects/corona x-ray images predication/Data/test/Covid\"):\n",
    "    img = image.load_img(\"G:/projects/corona x-ray images predication/Data/test/Covid/\" + i, target_size=(64, 64))\n",
    "    img = image.img_to_array(img)\n",
    "    img = np.expand_dims(img, axis=0)\n",
    "    p = classifier.predict_classes(img)\n",
    "    y_test.append(p[0, 0])\n",
    "    y_pred.append(0)\n",
    "\n",
    "y_pred = np.array(y_pred)\n",
    "y_test = np.array(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      0.94      0.97        50\n",
      "           1       0.94      1.00      0.97        50\n",
      "\n",
      "    accuracy                           0.97       100\n",
      "   macro avg       0.97      0.97      0.97       100\n",
      "weighted avg       0.97      0.97      0.97       100\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import confusion_matrix\n",
    "import seaborn as sns\n",
    "from sklearn.metrics import classification_report\n",
    "\n",
    "cm = confusion_matrix(y_pred, y_test)\n",
    "sns.heatmap(cm, cmap=\"plasma\", annot=True)\n",
    "plt.show()\n",
    "\n",
    "print(classification_report(y_pred, y_test))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "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": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}