{"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":"markdown","source":"

\n\n

Dogs vs Cats

\n\n\n\n

Table of content:

\n\n

1- Import relevant libraries

\n\n

2- Unzip data

\n\n

3- See some of the pictures

\n\n

4- Create a dataframe from train data

\n\n

5- Create a model

\n\n

\n5-1 Model Summary

\n\n

\n5-2 Split data to train and test section

\n\n

\n5-3 Genereate data by ImageDataGenereator

\n\n

\n5-4 Use Callbacks to have better contorl on training

\n\n

\n5-5 Fit model

\n\n

6- Plot results

\n\n

7- Evaluation on validation data with best model

\n\n

8- Create a dataframe from test data and prepare it

\n\n

9- Prediction on test data

\n\n

10- Submit the final result

\n","metadata":{}},{"cell_type":"markdown","source":"

\n\n

1- Import relevant libraries

\n
","metadata":{}},{"cell_type":"code","source":"import numpy as np\nimport pandas as pd\nimport matplotlib.pyplot as plt\nimport random\nimport tensorflow as tf\nfrom tensorflow import keras\nimport os\nfrom keras import layers\nfrom keras import models\nfrom keras.layers import Dropout\nfrom keras import optimizers\nfrom sklearn.model_selection import train_test_split\nfrom tensorflow.keras.preprocessing.image import ImageDataGenerator\nfrom tensorflow.keras import callbacks","metadata":{"execution":{"iopub.status.busy":"2022-10-13T11:18:11.627063Z","iopub.execute_input":"2022-10-13T11:18:11.627499Z","iopub.status.idle":"2022-10-13T11:18:18.674126Z","shell.execute_reply.started":"2022-10-13T11:18:11.627414Z","shell.execute_reply":"2022-10-13T11:18:18.672045Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"

\n\n

2- Unzip data

\n
","metadata":{}},{"cell_type":"code","source":"! unzip -qq \"../input/dogs-vs-cats/test1.zip\" \n! unzip -qq \"../input/dogs-vs-cats/train.zip\"","metadata":{"execution":{"iopub.status.busy":"2022-10-13T11:18:18.675990Z","iopub.execute_input":"2022-10-13T11:18:18.676694Z","iopub.status.idle":"2022-10-13T11:18:35.087384Z","shell.execute_reply.started":"2022-10-13T11:18:18.676652Z","shell.execute_reply":"2022-10-13T11:18:35.086007Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"

\n\n

3- See Some of pictures\n

","metadata":{}},{"cell_type":"code","source":"filenames = os.listdir('./train/')\nfor i in range(2):\n picture = np.array(plt.imread(os.path.join('./train',random.choice(filenames))))\n plt.imshow(picture)\n plt.show()\n plt.figure()","metadata":{"execution":{"iopub.status.busy":"2022-10-13T11:18:35.089581Z","iopub.execute_input":"2022-10-13T11:18:35.089971Z","iopub.status.idle":"2022-10-13T11:18:35.599285Z","shell.execute_reply.started":"2022-10-13T11:18:35.089930Z","shell.execute_reply":"2022-10-13T11:18:35.598253Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"

\n\n

4-Create a dataframe from train data\n

","metadata":{}},{"cell_type":"code","source":"categories = []\nfor filename in filenames:\n category= filename.split('.')[0]\n if category == 'cat':\n categories.append('0')\n else:\n categories.append('1')\n\ndf = pd.DataFrame({'filename':filenames,'category':categories})\nprint('df.shape:',df.shape)\ndf.head()","metadata":{"execution":{"iopub.status.busy":"2022-10-13T11:18:35.602204Z","iopub.execute_input":"2022-10-13T11:18:35.603089Z","iopub.status.idle":"2022-10-13T11:18:35.846774Z","shell.execute_reply.started":"2022-10-13T11:18:35.603045Z","shell.execute_reply":"2022-10-13T11:18:35.840903Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print('Number of dogs in the train dataset:',df.category[df.category=='1'].count())\nprint('Number of cats in the train dataset:',df.category[df.category=='0'].count())\ndf['category'].value_counts().plot.bar(color=['green','blue'])","metadata":{"execution":{"iopub.status.busy":"2022-10-13T11:18:35.851759Z","iopub.execute_input":"2022-10-13T11:18:35.852501Z","iopub.status.idle":"2022-10-13T11:18:36.164613Z","shell.execute_reply.started":"2022-10-13T11:18:35.852449Z","shell.execute_reply":"2022-10-13T11:18:36.163614Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"

\n\n

5- Create Model

\n
","metadata":{}},{"cell_type":"code","source":"Image_Size = (256,256)\nChannel_Size = 3\nBatch_Size = 64\n\n# Use CNN Artitechture\ndef create_model():\n model = models.Sequential()\n model.add(layers.Conv2D(32, (3, 3), activation='relu',\n input_shape=(256,256,3)))\n model.add(layers.MaxPooling2D((2, 2)))\n \n model.add(layers.Conv2D(64, (3, 3), activation='relu'))\n model.add(layers.MaxPooling2D((2, 2)))\n \n model.add(layers.Conv2D(128, (3, 3), activation='relu'))\n model.add(layers.MaxPooling2D((2, 2)))\n \n model.add(layers.Conv2D(256, (3, 3), activation='relu'))\n model.add(layers.MaxPooling2D((2, 2)))\n \n model.add(layers.Flatten())\n model.add(layers.Dense(128,activation='relu'))\n model.add(Dropout(0.2))\n \n model.add(layers.Dense(256,activation='relu'))\n model.add(Dropout(0.5))\n \n model.add(layers.Dense(2, activation='softmax'))\n \n model.compile(loss='categorical_crossentropy',\n optimizer='rmsprop',\n metrics=['accuracy'],)\n return model\n\nmy_model = create_model()","metadata":{"execution":{"iopub.status.busy":"2022-10-13T11:18:36.166136Z","iopub.execute_input":"2022-10-13T11:18:36.167194Z","iopub.status.idle":"2022-10-13T11:18:39.031478Z","shell.execute_reply.started":"2022-10-13T11:18:36.167156Z","shell.execute_reply":"2022-10-13T11:18:39.030446Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"

\n\n

5-1 Model Summary

\n
","metadata":{}},{"cell_type":"code","source":"my_model.summary()","metadata":{"execution":{"iopub.status.busy":"2022-10-13T11:18:39.032961Z","iopub.execute_input":"2022-10-13T11:18:39.033317Z","iopub.status.idle":"2022-10-13T11:18:39.041415Z","shell.execute_reply.started":"2022-10-13T11:18:39.033289Z","shell.execute_reply":"2022-10-13T11:18:39.040483Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"

\n\n

5-2 Split data to train and test section

\n
","metadata":{}},{"cell_type":"code","source":"train_df, validation_df = train_test_split(df, test_size=0.2, random_state=45)\ntrain_df = train_df.reset_index(drop=True)\nvalidation_df = validation_df.reset_index(drop=True)\nprint(f'train_df.shpae:{train_df.shape} & validation_df.shape:{validation_df.shape}')","metadata":{"execution":{"iopub.status.busy":"2022-10-13T11:18:39.043021Z","iopub.execute_input":"2022-10-13T11:18:39.043729Z","iopub.status.idle":"2022-10-13T11:18:39.062888Z","shell.execute_reply.started":"2022-10-13T11:18:39.043692Z","shell.execute_reply":"2022-10-13T11:18:39.062075Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"

\n\n

5-3 Genereate data by ImageDataGenereator

\n
","metadata":{}},{"cell_type":"code","source":"# We generate data and do data agumentation to avoid overfitting\n\n# Train_Section:\ntrain_datagen = ImageDataGenerator(\n rotation_range=15,\n rescale=1./255,\n shear_range=0.1,\n zoom_range=0.4,\n horizontal_flip=True)\n\ntrain_generator = train_datagen.flow_from_dataframe(\n train_df,\n './train',\n x_col='filename',\n y_col='category',\n target_size=Image_Size,\n class_mode='categorical',\n batch_size=Batch_Size\n)\n\n\n#Validation_Section:---> Note: You don not need to do data agumentation on validation data\nvalidation_datagen = ImageDataGenerator(\n rescale=1./255)\n\nvalidation_generator = train_datagen.flow_from_dataframe(\n validation_df,\n './train',\n x_col='filename',\n y_col='category',\n target_size=Image_Size,\n class_mode='categorical',\n batch_size=Batch_Size\n)\n","metadata":{"execution":{"iopub.status.busy":"2022-10-13T11:18:39.064155Z","iopub.execute_input":"2022-10-13T11:18:39.064569Z","iopub.status.idle":"2022-10-13T11:18:39.288282Z","shell.execute_reply.started":"2022-10-13T11:18:39.064531Z","shell.execute_reply":"2022-10-13T11:18:39.287241Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"

\n\n

5-4 Use Callbacks to have better contorl on training

\n
","metadata":{}},{"cell_type":"code","source":"# 1st_callbak : EarlyStoping\nMy_EarlyStopping_callback = callbacks.EarlyStopping(monitor=\"val_loss\",\n patience=3,\n verbose=1,\n mode=\"auto\")\n\n\n#2nd_callback : ModelCheckPoint\ncheckpoint_path = './Checkpint_Folder/cp.ckpt' # The path to save wights of the model after training\ncheckpoint_dir =os.path.dirname(checkpoint_path)\nMy_ModelCheckpoint_callback =callbacks.ModelCheckpoint(filepath=checkpoint_path,\n metrics='val_loss',\n verbose=1,\n save_weights_only = True,\n save_best_only=True,\n mode='auto',\n save_freq ='epoch')\n\nMy_callbacks = [My_EarlyStopping_callback, My_ModelCheckpoint_callback]","metadata":{"execution":{"iopub.status.busy":"2022-10-13T11:18:39.291934Z","iopub.execute_input":"2022-10-13T11:18:39.292219Z","iopub.status.idle":"2022-10-13T11:18:39.299223Z","shell.execute_reply.started":"2022-10-13T11:18:39.292193Z","shell.execute_reply":"2022-10-13T11:18:39.298068Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"

\n\n

5-5 Fit model

\n
","metadata":{}},{"cell_type":"code","source":"History = my_model.fit(\n train_generator,\n epochs=15,\n validation_data=(validation_generator),\n batch_size = Batch_Size,\n callbacks = My_callbacks\n)","metadata":{"execution":{"iopub.status.busy":"2022-10-13T11:18:39.300876Z","iopub.execute_input":"2022-10-13T11:18:39.301249Z","iopub.status.idle":"2022-10-13T11:19:09.277488Z","shell.execute_reply.started":"2022-10-13T11:18:39.301213Z","shell.execute_reply":"2022-10-13T11:19:09.274002Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"

\n\n

6- Plot results

\n
","metadata":{}},{"cell_type":"code","source":"accuracy = History.history['accuracy']\nval_accuracy = History.history['val_accuracy']\nloss = History.history['loss']\nval_loss = History.history['val_loss']\n\nepochs = range(len(accuracy))\n\nplt.plot(epochs,accuracy, label='accuracy',color='red')\nplt.plot(epochs,val_accuracy, label='val_accuracy',color='blue')\nplt.title('Training and validation loss')\nplt.legend()\nplt.figure()\n\nplt.plot(epochs,loss,label='loss',color='black')\nplt.plot(epochs,val_loss,label='val_loss',color='green')\nplt.title('Training and validation loss')\nplt.legend()\nplt.figure()","metadata":{"execution":{"iopub.status.busy":"2022-10-13T11:19:09.280308Z","iopub.status.idle":"2022-10-13T11:19:09.282887Z","shell.execute_reply.started":"2022-10-13T11:19:09.282606Z","shell.execute_reply":"2022-10-13T11:19:09.282633Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"

\n\n

7- Evaluation on validation data with best model

\n
","metadata":{}},{"cell_type":"code","source":"best_model = create_model()\nbest_model.load_weights(checkpoint_path)\nloss,accuracy = best_model.evaluate(validation_generator)\nprint('loss on validation data:',loss*100)\nprint('accuracy on validation data', accuracy*100)","metadata":{"execution":{"iopub.status.busy":"2022-10-13T11:19:09.286941Z","iopub.status.idle":"2022-10-13T11:19:09.289291Z","shell.execute_reply.started":"2022-10-13T11:19:09.289046Z","shell.execute_reply":"2022-10-13T11:19:09.289072Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"

\n\n

8- Create a dataframe from test data and prepare it

\n
\n","metadata":{}},{"cell_type":"code","source":"test_filenames = os.listdir('./test1/')\ntest_df = pd.DataFrame({'filename':test_filenames})\nprint('test_df.shape:',test_df.shape)\ntest_df.head()","metadata":{"execution":{"iopub.status.busy":"2022-10-13T11:19:09.292014Z","iopub.status.idle":"2022-10-13T11:19:09.292790Z","shell.execute_reply.started":"2022-10-13T11:19:09.292536Z","shell.execute_reply":"2022-10-13T11:19:09.292562Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"test_datagen = ImageDataGenerator(rescale=1./255)\ntest_generator = test_datagen.flow_from_dataframe(\n test_df, \n \"./test1/\", \n x_col='filename',\n y_col=None,\n class_mode=None,\n target_size=Image_Size,\n batch_size=Batch_Size,\n shuffle=False\n)","metadata":{"execution":{"iopub.status.busy":"2022-10-13T11:19:09.298941Z","iopub.status.idle":"2022-10-13T11:19:09.299734Z","shell.execute_reply.started":"2022-10-13T11:19:09.299471Z","shell.execute_reply":"2022-10-13T11:19:09.299497Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"

\n\n

9- Prediction on test data

\n
","metadata":{}},{"cell_type":"code","source":"Predict = best_model.predict(test_generator)\n# add the results to test_df as category column\ntest_df['category'] = np.argmax(Predict, axis=-1)","metadata":{"execution":{"iopub.status.busy":"2022-10-13T11:19:09.301160Z","iopub.status.idle":"2022-10-13T11:19:09.301905Z","shell.execute_reply.started":"2022-10-13T11:19:09.301659Z","shell.execute_reply":"2022-10-13T11:19:09.301684Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"

\n\n

10- Submit the final result

\n
","metadata":{}},{"cell_type":"code","source":"submission_df = test_df.copy()\nsubmission_df['id'] = submission_df['filename'].str.split('.').str[0]\nsubmission_df['label'] = submission_df['category']\nsubmission_df.drop(['filename', 'category'], axis=1, inplace=True)\nsubmission_df.to_csv('submission.csv', index=False)","metadata":{"execution":{"iopub.status.busy":"2022-10-13T11:19:09.303152Z","iopub.status.idle":"2022-10-13T11:19:09.303817Z","shell.execute_reply.started":"2022-10-13T11:19:09.303593Z","shell.execute_reply":"2022-10-13T11:19:09.303616Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}