{"cells":[{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"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 in \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 \"../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# Any results you write to the current directory are saved as output.","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","trusted":true},"cell_type":"code","source":"import numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\nfrom tqdm.auto import tqdm\nfrom glob import glob\nimport time, gc\nimport cv2\n\nimport matplotlib.image as mpimg\nfrom keras.preprocessing.image import ImageDataGenerator\nfrom keras.models import Sequential\nfrom keras.models import clone_model\nfrom keras.layers import Dense,Conv2D,Flatten,MaxPool2D,Dropout,BatchNormalization\nfrom keras.optimizers import Adam\nfrom keras.callbacks import ReduceLROnPlateau\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.metrics import confusion_matrix\nimport PIL.Image as Image, PIL.ImageDraw as ImageDraw, PIL.ImageFont as ImageFont\nfrom matplotlib import pyplot as plt\nimport seaborn as sns","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train_df_ = pd.read_csv('/kaggle/input/bengaliai-cv19/train.csv')\ntest_df_ = pd.read_csv('/kaggle/input/bengaliai-cv19/test.csv')\nclass_map_df = pd.read_csv('/kaggle/input/bengaliai-cv19/class_map.csv')\nsample_sub_df = pd.read_csv('/kaggle/input/bengaliai-cv19/sample_submission.csv')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"test_df_.head()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"sample_sub_df.head(10)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"class_map_df.head(10)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"print(f'Size of training data: {train_df_.shape}')\nprint(f'Size of test data: {test_df_.shape}')\nprint(f'Size of class map: {class_map_df.shape}')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"HEIGHT = 236\nWIDTH = 236\n\ndef get_n(df, field, n, top=True):\n    top_graphemes = df.groupby([field]).size().reset_index(name='counts')['counts'].sort_values(ascending=not top)[:n]\n    top_grapheme_roots = top_graphemes.index\n    top_grapheme_counts = top_graphemes.values\n    top_graphemes = class_map_df.iloc[top_grapheme_roots]\n    top_graphemes.drop(['component_type', 'label'], axis=1, inplace=True)\n    top_graphemes.loc[:, 'count'] = top_grapheme_counts\n    return top_graphemes\n\ndef image_from_char(char):\n    image = Image.new('RGB', (WIDTH, HEIGHT))\n    draw = ImageDraw.Draw(image)\n    myfont = ImageFont.truetype('/kaggle/input/bengali-fonts/hind_siliguri_normal_500.ttf', 120)\n    w, h = draw.textsize(char, font=myfont)\n    draw.text(((WIDTH - w) / 2,(HEIGHT - h) / 3), char, font=myfont)\n\n    return image","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"print(f'Number of unique grapheme roots: {train_df_[\"grapheme_root\"].nunique()}')\nprint(f'Number of unique vowel diacritic: {train_df_[\"vowel_diacritic\"].nunique()}')\nprint(f'Number of unique consonant diacritic: {train_df_[\"consonant_diacritic\"].nunique()}')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"top_10_roots = get_n(train_df_, 'grapheme_root', 10)\ntop_10_roots","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"f, ax = plt.subplots(2, 5, figsize=(16, 8))\nax = ax.flatten()\n\nfor i in range(10):\n    ax[i].imshow(image_from_char(top_10_roots['component'].iloc[i]), cmap='Greys')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"bottom_10_roots = get_n(train_df_, 'grapheme_root', 10, False)\nbottom_10_roots","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"f, ax = plt.subplots(2, 5, figsize=(16, 8))\nax = ax.flatten()\n\nfor i in range(10):\n    ax[i].imshow(image_from_char(bottom_10_roots['component'].iloc[i]), cmap='Greys')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"top_5_vowels = get_n(train_df_, 'vowel_diacritic', 5)\ntop_5_vowels","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"f, ax = plt.subplots(1, 5, figsize=(16, 8))\nax = ax.flatten()\n\nfor i in range(5):\n    ax[i].imshow(image_from_char(top_5_vowels['component'].iloc[i]), cmap='Greys')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"top_5_consonants = get_n(train_df_, 'consonant_diacritic', 5)\ntop_5_consonants","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"f, ax = plt.subplots(1, 5, figsize=(16, 8))\nax = ax.flatten()\n\nfor i in range(5):\n    ax[i].imshow(image_from_char(top_5_consonants['component'].iloc[i]), cmap='Greys')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train_df_ = train_df_.drop(['grapheme'], axis=1, inplace=False)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train_df_[['grapheme_root', 'vowel_diacritic', 'consonant_diacritic']] = train_df_[['grapheme_root', 'vowel_diacritic', 'consonant_diacritic']].astype('uint8')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"IMG_SIZE=64\nN_CHANNELS=1","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def resize(df, size=64, need_progress_bar=True):\n    resized = {}\n    if need_progress_bar:\n        for i in tqdm(range(df.shape[0])):\n            image = cv2.resize(df.loc[df.index[i]].values.reshape(137,236),(size,size))\n            resized[df.index[i]] = image.reshape(-1)\n    else:\n        for i in range(df.shape[0]):\n            image = cv2.resize(df.loc[df.index[i]].values.reshape(137,236),(size,size))\n            resized[df.index[i]] = image.reshape(-1)\n    resized = pd.DataFrame(resized).T\n    return resized","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def get_dummies(df):\n    cols = []\n    for col in df:\n        cols.append(pd.get_dummies(df[col].astype(str)))\n    return pd.concat(cols, axis=1)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model = Sequential()\n\nmodel.add(Conv2D(filters=32, kernel_size=(3, 3), padding='SAME', activation='relu', input_shape=(64, 64, 1)))\nmodel.add(Conv2D(filters=32, kernel_size=(3, 3), padding='SAME', activation='relu'))\nmodel.add(BatchNormalization(momentum=0.15))\nmodel.add(MaxPool2D(pool_size=(2, 2)))\nmodel.add(Conv2D(filters=32, kernel_size=(5, 5), padding='SAME', activation='relu'))\nmodel.add(Dropout(rate=0.3))\n\nmodel.add(Conv2D(filters=64, kernel_size=(3, 3), padding='SAME', activation='relu'))\nmodel.add(Conv2D(filters=64, kernel_size=(3, 3), padding='SAME', activation='relu'))\nmodel.add(BatchNormalization(momentum=0.15))\nmodel.add(MaxPool2D(pool_size=(2, 2)))\nmodel.add(Conv2D(filters=64, kernel_size=(5, 5), padding='SAME', activation='relu'))\nmodel.add(Dropout(rate=0.3))\n\nmodel.add(Flatten())\nmodel.add(Dense(256, activation = \"relu\"))\nmodel.add(Dropout(0.40))\nmodel.add(Dense(192, activation = \"relu\"))\nmodel.add(Dropout(0.40))\n# model.add(Dense(186, activation = \"softmax\"))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model_root = clone_model(model)\nmodel_vowel = clone_model(model)\nmodel_consonant = clone_model(model)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model_root.add(Dense(168, activation = 'softmax'))\nmodel_vowel.add(Dense(11, activation = 'softmax'))\nmodel_consonant.add(Dense(7, activation = 'softmax'))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model_root.compile(optimizer=\"adam\", loss=['categorical_crossentropy'], metrics=['accuracy'])\nmodel_vowel.compile(optimizer=\"adam\", loss=['categorical_crossentropy'], metrics=['accuracy'])\nmodel_consonant.compile(optimizer=\"adam\", loss=['categorical_crossentropy'], metrics=['accuracy'])","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model_root.summary()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model_vowel.summary()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model_consonant.summary()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Set a learning rate annealer. Learning rate will be half after 3 epochs if accuracy is not increased\nlearning_rate_reduction = ReduceLROnPlateau(monitor='val_accuracy', \n                                            patience=3, \n                                            verbose=1,\n                                            factor=0.5, \n                                            min_lr=0.00001)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"batch_size = 64\nepochs = 12","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model_dict = {\n    'grapheme_root': model_root,\n    'vowel_diacritic': model_vowel,\n    'consonant_diacritic': model_consonant\n}","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"histories = []\nfor i in range(4):\n    train_df = pd.merge(pd.read_parquet(f'/kaggle/input/bengaliai-cv19/train_image_data_{i}.parquet'), train_df_, on='image_id').drop(['image_id'], axis=1)\n    \n    # Visualize few samples of current training dataset\n    fig, ax = plt.subplots(nrows=3, ncols=4, figsize=(16, 8))\n    count=0\n    for row in ax:\n        for col in row:\n            col.imshow(resize(train_df.drop(['grapheme_root', 'vowel_diacritic', 'consonant_diacritic'], axis=1).iloc[[count]], need_progress_bar=False).values.reshape(64, 64))\n            count += 1\n    plt.show()\n    \n    X_train = train_df.drop(['grapheme_root', 'vowel_diacritic', 'consonant_diacritic'], axis=1)\n    X_train = resize(X_train)/255\n    # CNN takes images in shape `(batch_size, h, w, channels)`, so reshape the images\n    X_train = X_train.values.reshape(-1, IMG_SIZE, IMG_SIZE, N_CHANNELS)\n    \n    for target in ['grapheme_root', 'vowel_diacritic', 'consonant_diacritic']:\n        Y_train = train_df[target]\n        Y_train = pd.get_dummies(Y_train).values\n\n        print(f'Training images: {X_train.shape}')\n        print(f'Training labels: {Y_train.shape}')\n\n        # Divide the data into training and validation set\n        x_train, x_test, y_train, y_test = train_test_split(X_train, Y_train, test_size=0.15, random_state=666)\n        del Y_train\n        \n        # Data augmentation for creating more training data\n        datagen = ImageDataGenerator(\n            featurewise_center=False,  # set input mean to 0 over the dataset\n            samplewise_center=False,  # set each sample mean to 0\n            featurewise_std_normalization=False,  # divide inputs by std of the dataset\n            samplewise_std_normalization=False,  # divide each input by its std\n            zca_whitening=False,  # apply ZCA whitening\n            rotation_range=8,  # randomly rotate images in the range (degrees, 0 to 180)\n            zoom_range = 0.15, # Randomly zoom image \n            width_shift_range=0.15,  # randomly shift images horizontally (fraction of total width)\n            height_shift_range=0.15,  # randomly shift images vertically (fraction of total height)\n            horizontal_flip=False,  # randomly flip images\n            vertical_flip=False)  # randomly flip images\n\n        # This will just calculate parameters required to augment the given data. This won't perform any augmentations\n        datagen.fit(x_train)\n        \n         # Fit the model\n        history = model_dict[target].fit_generator(datagen.flow(x_train, y_train, batch_size=batch_size),\n                                      epochs = epochs, validation_data = (x_test,y_test),\n                                      steps_per_epoch=x_train.shape[0] // batch_size, \n                                      callbacks=[learning_rate_reduction])\n    \n        histories.append(history)\n        del x_train\n        del x_test\n        del y_train\n        del y_test\n        gc.collect()\n    # Delete to reduce memory usage\n    del X_train\n    del train_df\n    gc.collect()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"%matplotlib inline\ndef plot_loss(his, epoch, title):\n    plt.style.use('ggplot')\n    plt.figure()\n    plt.plot(np.arange(0, epoch), his.history['loss'], label='train_loss')\n    plt.plot(np.arange(0, epoch), his.history['val_loss'], label='val_loss')\n    plt.title(title)\n    plt.xlabel('Epoch #')\n    plt.ylabel('Loss')\n    plt.legend(loc='upper right')\n    plt.show()\n\ndef plot_acc(his, epoch, title):\n    plt.style.use('ggplot')\n    plt.figure()\n    plt.plot(np.arange(0, epoch), his.history['accuracy'], label='train_acc')\n    plt.plot(np.arange(0, epoch), his.history['val_accuracy'], label='val_accuracy')\n    plt.title(title)\n    plt.xlabel('Epoch #')\n    plt.ylabel('Accuracy')\n    plt.legend(loc='upper right')\n    plt.show()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"for dataset in range(4):\n    for target in ['grapheme_root', 'vowel_diacritic', 'consonant_diacritic']:\n        plot_loss(histories[0], epochs, f'Dataset: {dataset}, Training on: {target}')\n        plot_acc(histories[0], epochs, f'Dataset: {dataset}, Training on: {target}')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"del histories\ndel model\ngc.collect()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"preds_dict = {\n    'grapheme_root': [],\n    'vowel_diacritic': [],\n    'consonant_diacritic': []\n}","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"components = ['consonant_diacritic', 'grapheme_root', 'vowel_diacritic']\ntarget=[] # model predictions placeholder\nrow_id=[] # row_id place holder\nfor i in range(4):\n    df_test_img = pd.read_parquet('/kaggle/input/bengaliai-cv19/test_image_data_{}.parquet'.format(i)) \n    df_test_img.set_index('image_id', inplace=True)\n\n    X_test = resize(df_test_img, need_progress_bar=False)/255\n    X_test = X_test.values.reshape(-1, IMG_SIZE, IMG_SIZE, N_CHANNELS)\n\n    for pred in preds_dict:\n        preds_dict[pred]=np.argmax(model_dict[pred].predict(X_test), axis=1)\n\n    for k,id in enumerate(df_test_img.index.values):  \n        for i,comp in enumerate(components):\n            id_sample=id+'_'+comp\n            row_id.append(id_sample)\n            target.append(preds_dict[comp][k])\n    del df_test_img\n    del X_test\n    gc.collect()\n\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"df_sample = pd.DataFrame(\n    {\n        'row_id': row_id,\n        'target':target\n    },\n    columns = ['row_id','target'] \n)\ndf_sample.to_csv('submission.csv',index=False)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# components = ['consonant_diacritic', 'grapheme_root', 'vowel_diacritic']\n# target=[] # model predictions placeholder\n# row_id=[] # row_id place holder\n# n_cls = [7,168,11] # number of classes in each of the 3 targets\n# for i in range(4):\n#     df_test_img = pd.read_parquet('/kaggle/input/bengaliai-cv19/test_image_data_{}.parquet'.format(i)) \n#     df_test_img.set_index('image_id', inplace=True)\n\n#     X_test = resize(df_test_img)/255\n#     X_test = X_test.values.reshape(-1, IMG_SIZE, IMG_SIZE, N_CHANNELS)\n\n#     for pred in preds_dict:\n#         preds_dict[pred]=np.argmax(model_dict[pred].predict(X_test), axis=1)\n\n#     for k,id in enumerate(df_test_img.index.values):  \n#         for i,comp in enumerate(components):\n#             id_sample=id+'_'+comp\n#             row_id.append(id_sample)\n#             target.append(preds_dict[comp][k])\n\n# df_sample = pd.DataFrame(\n#     {'row_id': row_id,\n#     'target':target\n#     },\n#     columns =['row_id','target'] \n# )\n# df_sample.to_csv('submission.csv',index=False)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"","execution_count":null,"outputs":[]}],"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":4,"nbformat_minor":1}