{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"# This Python 3 environment comes with many helpful analytics libraries installed\n# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python\n# For example, here's several helpful packages to load\n\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n\n# Input data files are available in the read-only \"../input/\" directory\n# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n\nimport os\nfor dirname, _, filenames in os.walk('/kaggle/input'):\n    for filename in filenames:\n        print(os.path.join(dirname, filename))\n\n# You can write up to 20GB to the current directory (/kaggle/working/) that gets preserved as output when you create a version using \"Save & Run All\" \n# You can also write temporary files to /kaggle/temp/, but they won't be saved outside of the current session","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"![image.png](attachment:67044e36-1a5d-409c-b513-2ddffbcd19a9.png)","metadata":{},"attachments":{"67044e36-1a5d-409c-b513-2ddffbcd19a9.png":{"image/png":"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"}}},{"cell_type":"markdown","source":"*Great notebook and resource on how to set up the directory for using Tensorflow's ImageDataGenerator object:*\n\nhttps://www.kaggle.com/code/ggsri123/p1-dogsvscats-using-tensorflow  \nhttps://vijayabhaskar96.medium.com/tutorial-image-classification-with-keras-flow-from-directory-and-generators-95f75ebe5720","metadata":{"execution":{"iopub.status.busy":"2022-07-29T16:42:51.522821Z","iopub.execute_input":"2022-07-29T16:42:51.523544Z","iopub.status.idle":"2022-07-29T16:42:51.534222Z","shell.execute_reply.started":"2022-07-29T16:42:51.523498Z","shell.execute_reply":"2022-07-29T16:42:51.533062Z"}}},{"cell_type":"code","source":"import os\nimport zipfile\nimport tensorflow as tf\nfrom tensorflow.keras.optimizers import RMSprop\nfrom tensorflow.keras.preprocessing.image import ImageDataGenerator\nfrom sklearn.model_selection import train_test_split\nimport tqdm\nimport shutil","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# extract test file into tmp folder\n\nzip_ref = zipfile.ZipFile('../input/dogs-vs-cats/test1.zip', 'r')\nzip_ref.extractall(\"../tmp\")\nzip_ref.close()","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# extract train file into tmp folder\n\nzip_ref = zipfile.ZipFile('../input/dogs-vs-cats/train.zip', 'r')\nzip_ref.extractall(\"../tmp\")\nzip_ref.close()","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# set up training and testing directories \n\nbase_dir = '../tmp'\ntrain_dir = os.path.join(base_dir, 'train')\ntest_dir = os.path.join(base_dir, 'test1')","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# visualize the cat images\n\nimport matplotlib.pyplot as plt\nfrom matplotlib.image import imread\n\nfolder  = train_dir + '/'\nfor i in range(9):\n    \n    # 3x3 subplot\n    plt.subplot(330 + 1 + i)\n    \n    # get files labeld as cat\n    filename = folder + 'cat.' + str(i) +  '.jpg'\n    \n    image = imread(filename)\n    plt.imshow(image)\nplt.show()","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# visualize the dog images\n\nfolder  = train_dir + '/'\nfor i in range(9):\n    \n    # 3x3 subplot\n    plt.subplot(330 + 1 + i)\n    \n    # get files labeld as dog\n    filename = folder + 'dog.' + str(i) +  '.jpg'\n    \n    image = imread(filename)\n    plt.imshow(image)\nplt.show()","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# visualize the test images\n\nfolder  = test_dir + '/'\nfor i in range(1,10):\n    \n    # 3x3 subplot\n    plt.subplot(330 + 1 + i-1)\n    \n    # get files labeld as dog\n    filename = folder + str(i) +  '.jpg'\n    \n    image = imread(filename)\n    plt.imshow(image)\nplt.show()","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# number of images in the training file\ntrain_file_list = os.listdir('../tmp/train')\ntest_file_list = os.listdir('../tmp/test1')\nprint('Number of train images:', len(train_file_list))\n\n# get lists of dog and cat file names to move them later\ndog_filenames = [fn for fn in train_file_list if fn.startswith('dog')]\ncat_filenames = [fn for fn in train_file_list if fn.startswith('cat')]\n\n# shuffle into train and validation sets (90/10 split); dataset_filenames is a tuple of (dog train, dog val, cat train, cat val)\ndataset_filenames = train_test_split(dog_filenames, cat_filenames, test_size=0.1, shuffle=True, random_state=42)\n\ntrain_dog, valid_dog, train_cat, valid_cat = [fns for fns in dataset_filenames]\n\n# check totals\ntrain_dog_total, valid_dog_total, train_cat_total, valid_cat_total = [len(fns) for fns in dataset_filenames]\ntrain_total = train_dog_total + train_cat_total\nvalid_total = valid_dog_total + valid_cat_total\n\nprint('Train: {}, val: {}, test: {}'.format(train_total, valid_total, len(test_file_list)))","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# create the directories for data preprocessing\n\nTRAIN_DIR_DOG = train_dir + '/dog'\nTRAIN_DIR_CAT = train_dir + '/cat'\n\nvalid_dir = base_dir + '/valid'\nVALID_DIR_DOG = valid_dir + '/dog'\nVALID_DIR_CAT = valid_dir + '/cat'\n\nTEST_DIR = test_dir + '/test'","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# make training and validation directories, move files to appropriate directories\n\nmake_dirs = [TRAIN_DIR_DOG, VALID_DIR_DOG, TRAIN_DIR_CAT, VALID_DIR_CAT]\n\nfor dir, fns in zip(make_dirs, dataset_filenames):\n    os.makedirs(dir, exist_ok=True)\n    for fn in tqdm.tqdm(fns):\n        shutil.move(os.path.join(train_dir, fn), dir)\n    print('elements in {}: {}'.format(dir, len(os.listdir(dir))))\n    \nos.makedirs(TEST_DIR, exist_ok=True)\nfor fn in tqdm.tqdm(os.listdir(test_dir)):\n    shutil.move(os.path.join(test_dir, fn), TEST_DIR)\nprint('elements in {}: {}'.format(TEST_DIR, len(os.listdir(TEST_DIR))))    ","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from tensorflow.keras.applications.inception_v3 import InceptionV3\n\n\n# Initialize the base model.\n# Set the input shape and remove the dense layers.\npre_trained_model = InceptionV3(input_shape = (150, 150, 3), \n                                include_top = False, \n                                weights = 'imagenet')\n\n\n# Freeze the weights of the layers.\nfor layer in pre_trained_model.layers:\n  layer.trainable = False","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"last_layer = pre_trained_model.get_layer('mixed7')\nlast_output = last_layer.output\n\nx = tf.keras.layers.Flatten()(last_output)\nx = tf.keras.layers.Dense(1024, activation='relu')(x)\nx = tf.keras.layers.Dropout(0.2)(x)\nx = tf.keras.layers.Dense(1, activation='sigmoid')(x)\n\nmodel = tf.keras.Model(pre_trained_model.input, x)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# lr_schedule = tf.keras.optimizers.schedules.InverseTimeDecay(0.001, decay_steps = 100 * 1000, # steps_per_epoch * 1000\n#                                                             decay_rate=1,\n#                                                             staircase=False)\n\n# model.compile(optimizer=tf.keras.optimizers.Adam(lr_schedule),\n#              loss='binary_crossentropy',\n#              metrics=['accuracy'])\n\ncallbacks = tf.keras.callbacks.EarlyStopping(monitor='val_loss',\n                                            patience=200,\n                                            verbose=1,\n                                            mode='auto',\n                                            restore_best_weights=False)\n\nmodel.compile(optimizer=tf.keras.optimizers.RMSprop(learning_rate=1e-3),\n              loss='binary_crossentropy',\n              metrics=['accuracy'])","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Data Preprocessing","metadata":{}},{"cell_type":"code","source":"train_datagen = ImageDataGenerator(rotation_range=40,\n                                   width_shift_range=0.2,\n                                   height_shift_range=0.2,\n                                   rescale = 1.0/255.,\n                                   shear_range=0.2,\n                                   zoom_range=0.2,\n                                   horizontal_flip=True,\n                                   fill_mode='nearest')\n\ntest_datagen = ImageDataGenerator(rescale = 1.0/255.)\n\n\ntrain_generator = train_datagen.flow_from_directory(train_dir,\n                                                   batch_size=32,\n                                                   class_mode='binary',\n                                                    shuffle=True,\n                                                   target_size=(150, 150))\n\nvalidation_generator = test_datagen.flow_from_directory(valid_dir,\n                                                       batch_size=32,\n                                                       class_mode='binary',\n                                                        shuffle=True,\n                                                       target_size=(150, 150))\n\ntest_generator = test_datagen.flow_from_directory(test_dir,\n                                                 batch_size=1,\n                                                 class_mode=None,\n                                                  shuffle=False,\n                                                  target_size=(150, 150))\n\nSTEP_SIZE_TRAIN = train_generator.n // train_generator.batch_size\nSTEP_SIZE_VALID = validation_generator.n // validation_generator.batch_size","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"history = model.fit(train_generator,\n                    steps_per_epoch = 100,\n                   epochs=20,\n                   validation_data = validation_generator,\n                    validation_steps = 50,\n                   verbose=2,\n                    callbacks=[callbacks]\n                   ) ","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"acc = history.history['accuracy']\nval_acc = history.history['val_accuracy']\nloss = history.history['loss']\nval_loss = history.history['val_loss']\n\nepochs = range(len(acc))\n\nplt.plot(epochs, acc, 'b', label='Training Accuracy')\nplt.plot(epochs, val_acc, 'r', label='Validation Accuracy')\nplt.title('Training and Validation Accuracy')\nplt.legend()\n\nplt.figure()\n\nplt.plot(epochs, loss, 'b', label='Training Loss')\nplt.plot(epochs, val_loss, 'r', label='Validation Loss')\nplt.title('Training and Validation Loss')\nplt.legend()\n\nplt.show()","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"predictions = model.predict(test_generator,\n                           verbose=1)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}