{"cells":[{"metadata":{"trusted":true,"_uuid":"8cf9948884f182c9fbe8ba317443c0ab3f78caa7","scrolled":true},"cell_type":"code","source":"import os\nimport numpy as np\nimport matplotlib.pyplot as plt\n\nfrom random import shuffle\nfrom tqdm import tqdm\nfrom keras.preprocessing.image import ImageDataGenerator, img_to_array, array_to_img, load_img\nfrom keras import optimizers\nfrom keras.models import Sequential\nfrom keras.layers import Dropout, Flatten, Dense, Activation\nfrom keras.layers.convolutional import Conv2D, MaxPooling2D\nfrom keras.utils import to_categorical\n\nTRAIN_DIR = '../input/train'\nTEST_DIR = '../input/test'\n\nimage_width = 160\nimage_height = 160\npool_size = 2\nlearning_rate = 1e-4\nbatch_size = 32\nepochs = 20\n\nclasses = {'dog': 1, 'cat': 0}\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"cb32a3d3d449880d49d6fbe34c711b346f750b4b"},"cell_type":"code","source":"import random \nrandom.seed(1)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"8dec60abf5ffa08746a995fdf8f454cde3596297"},"cell_type":"code","source":"def read_image(dir_path, file_name):\n    image_path = os.path.join(dir_path, file_name)\n    raw_image = load_img(image_path, target_size=(image_width, image_height))\n    if raw_image is not None:\n        img_array = img_to_array(raw_image)\n        return img_array\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"d8c7a401bd8e00018ad0fc1189a6576d534447c2"},"cell_type":"code","source":"def display_image(image, prediction=None):\n    label_name = None\n    accuracy = 0.0\n    if prediction >= 0.5:\n        label_name = 'Dog'\n        accuracy = prediction\n    else:\n        label_name = 'Cat'\n        accuracy = 1-prediction\n    \n    plt.title('label: {}, accuracy: {:.2%}'.format(label_name, accuracy))\n    plt.imshow(array_to_img(image))\n    plt.show()\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"f0159c8d8e3c195c5f5f542102e1ea4f89022e04"},"cell_type":"code","source":"def load_train_data(is_train=True):\n    file_list = os.listdir(TRAIN_DIR)\n    shuffle(file_list)\n    \n    file_list = file_list[:20000]\n    bucket_size = int(len(file_list)*0.9)\n    if is_train:\n        file_list = file_list[:bucket_size]\n    else:\n        file_list = file_list[bucket_size:]\n\n    features = []\n    labels = []\n    for file_name in tqdm(file_list):\n        # 1 = dog, 0 = cat\n        label = classes[file_name.split('.')[0]]\n        image = read_image(TRAIN_DIR, file_name)\n        if image is not None and label is not None:\n            features.append(image)\n            labels.append(label)\n    return np.array(features), labels\n\ndef load_test_data():\n    test_data = []\n    names = []\n    for file_name in tqdm(os.listdir(TEST_DIR)):\n        image = read_image(TEST_DIR, file_name)\n        if image is not None:\n            test_data.append(image)\n            names.append(file_name.split('.')[0])\n    return np.array(test_data), names\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"d6587771c12b80611ab0b27551baf2f09d4d5b29","scrolled":false},"cell_type":"code","source":"X_train, y_train = load_train_data()\nX_train /= 255\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"62e135d2257a062dae9044365a05ba8e227b21e0"},"cell_type":"code","source":"# for i in range(5):\n#     display_image(X_train[i], y_train[i])\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"3c20ec1ea738762438a3be6fd54c5906492c03b8","scrolled":true},"cell_type":"code","source":"model = Sequential()\n\nmodel.add(Conv2D(32, (3,3), input_shape=(image_width, image_height, 3), padding='same', activation='relu'))\nmodel.add(Conv2D(32, (3,3), padding='same', activation='relu'))\nmodel.add(MaxPooling2D(pool_size=(pool_size, pool_size)))\nmodel.add(Dropout(0.25))\n\nmodel.add(Conv2D(64, (3,3), padding='same', activation='relu'))\nmodel.add(Conv2D(64, (3,3), padding='same', activation='relu'))\nmodel.add(MaxPooling2D(pool_size=(pool_size, pool_size)))\nmodel.add(Dropout(0.25))\n\nmodel.add(Conv2D(128, (3,3), padding='same', activation='relu'))\nmodel.add(Conv2D(128, (3,3), padding='same', activation='relu'))\n# model.add(Conv2D(128, (3,3), padding='same', activation='relu'))\nmodel.add(MaxPooling2D(pool_size=(pool_size, pool_size)))\nmodel.add(Dropout(0.25))\n\nmodel.add(Flatten())\nmodel.add(Dense(1024, activation='relu'))\nmodel.add(Dropout(0.5))\nmodel.add(Dense(2, activation='softmax'))\n\nmodel.summary()\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"d9b5b1e23489f1d52b382f9c79fa53bf47a78a85","scrolled":true},"cell_type":"code","source":"model.compile(loss='categorical_crossentropy', optimizer=optimizers.RMSprop(lr=learning_rate), metrics=['accuracy'])\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"c6f41f3c3403cb00d3261904faa5ec023a530dde"},"cell_type":"code","source":"# datagen = ImageDataGenerator(rescale=1./255, validation_split=0.2)\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"scrolled":true,"_uuid":"b2cb06454161fd5699d75c96166eef575123b21f"},"cell_type":"code","source":"y_train = to_categorical(y_train)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"b378a3ef3f5f329e6eb2e131c7265d634038b9f2","scrolled":false},"cell_type":"code","source":"history = model.fit(X_train, y_train, batch_size=batch_size, epochs=epochs, validation_split=0.2, verbose=1)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"9cb39709217c6b5b6536c61c6b7a3c10265c513b","scrolled":false},"cell_type":"code","source":"# train_generator = datagen.flow(X_train, y_train, batch_size=batch_size, subset='training')\n# validation_generator = datagen.flow(X_train, y_train, batch_size=batch_size, subset='validation')\n\n# history = model.fit_generator(\n#     train_generator,\n#     validation_data=validation_generator,\n#     steps_per_epoch=train_generator.n/batch_size,\n#     validation_steps=validation_generator.n/batch_size,\n#     epochs=epochs,\n#     verbose=1)\n","execution_count":null,"outputs":[]},{"metadata":{"scrolled":false,"trusted":true,"_uuid":"b0c79048fa0efaf1013ee700dc484fb669dc2fca"},"cell_type":"code","source":"plt.plot(history.history['acc'])\nplt.plot(history.history['val_acc'])\nplt.title('Model accuracy')\nplt.ylabel('accuracy')\nplt.xlabel('epoch')\nplt.legend(['train', 'validate'])\nplt.show()\n\nplt.plot(history.history['loss'])\nplt.plot(history.history['val_loss'])\nplt.title('Model loss')\nplt.ylabel('loss')\nplt.xlabel('epoch')\nplt.legend(['train', 'validate'])\nplt.show()\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"5e747197cd382ad0d1c0a7673c0ccecb166ae7c3"},"cell_type":"code","source":"X_evaluate, y_evaluate = load_train_data(False)\nX_evaluate /= 255\n\ny_evaluate = to_categorical(y_evaluate)\n\nevaluation = model.evaluate(X_evaluate, y_evaluate)\nprint('Model loss: %s\\nModel accuracy: %s' % (evaluation[0], evaluation[1]))\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"8d3e4a22323845e31101b827af6843de917cc95d","scrolled":true},"cell_type":"code","source":"# X_test, names = load_test_data()\n# X_test /= 255\n\n# predictions = model.predict(X_test)\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"604b69214a473ea152ae82dac548a15f26ed6c40"},"cell_type":"code","source":"# for i in range(5):\n#     display_image(X_test[i], predictions[i][0])\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"8e9f647e622d20109f0c0eed8b3a64eaf6f8e887"},"cell_type":"code","source":"# with open('dogs-vs-cats-results.csv', 'w') as f:\n#     f.write('id,label\\n')\n#     for i in tqdm(range(predictions.shape[0])):\n#         f.write('{},{}\\n'.format(names[i], predictions[i][0]))\n","execution_count":null,"outputs":[]}],"metadata":{"kernelspec":{"display_name":"Python 3","language":"python","name":"python3"},"language_info":{"name":"python","version":"3.6.6","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat":4,"nbformat_minor":1}