{"cells":[{"metadata":{},"cell_type":"markdown","source":"### TabNet, a novel deep learning architecture for tabular learning. TabNet uses a sequential attention mechanism to choose a subset of semantically meaningful features to process at each decision step.TabNet uses canonical DNNs to act as decision trees .\n![image.png](attachment:image.png)\n\n\n### Above represented encoder decoder architecture for TabNet. \n#### (a) TabNet encoder for classification or regression, composed of a feature transformer, an attentive transformer and feature masking at each decision step. A split block divides the processed representation into two, to be used by the attentive transformer of the subsequent step as well as for constructing the overall output. At each decision step, the feature selection mask can provide interpretable information about the model’s functionality, and the masks can be aggregated to obtain global feature important attribution. (b) TabNet decoder, composed of a feature transformer block at each step. (c) A feature transformer block example – 4-layer network is shown, where 2 of the blocks are shared across all decision steps and 2 are decision step-dependent. Each layer is composed of a fully-connected (FC) layer, BN and GLU nonlinearity. (d) An attentive transformer block example – a single layer mapping is modulated with a prior scale information which aggregates how much each feature has been used before the current decision step. Normalization of the coefficients is done using sparsemax for sparse selection of the most salient features at each decision step. (Collected from Original Paper)\n","attachments":{"image.png":{"image/png":"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"}}},{"metadata":{},"cell_type":"markdown","source":"## Installation"},{"metadata":{"trusted":true},"cell_type":"code","source":"!pip install pytorch-tabnet","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Library"},{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"cell_type":"code","source":"#===========================================================\n# Library\n#===========================================================\nimport os\nimport gc\nfrom logging import getLogger, INFO, StreamHandler, FileHandler, Formatter\nfrom contextlib import contextmanager\nimport time\nimport glob\n\nimport numpy as np\nimport pandas as pd\nimport scipy as sp\nimport random\n\nimport matplotlib.pyplot as plt\nimport seaborn as sns\n\nfrom functools import partial\n\nfrom sklearn.model_selection import StratifiedKFold, GroupKFold, KFold\nfrom sklearn import preprocessing\nimport category_encoders as ce\nfrom sklearn.metrics import mean_squared_error\n\nimport torch\nfrom sklearn.preprocessing import LabelEncoder\n\nimport pandas as pd\nimport numpy as np\nnp.random.seed(0)\n\nfrom pytorch_tabnet.tab_model import TabNetRegressor ##Import Tabnet \n\n\n\nfrom pathlib import Path\n\nimport lightgbm as lgb\n\nimport warnings\nwarnings.filterwarnings(\"ignore\")","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","trusted":true},"cell_type":"code","source":"os.listdir('../input/indoor-location-navigation/')","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Utils"},{"metadata":{"trusted":true},"cell_type":"code","source":"#===========================================================\n# Utils\n#===========================================================\ndef get_logger(filename='log'):\n    logger = getLogger(__name__)\n    logger.setLevel(INFO)\n    handler1 = StreamHandler()\n    handler1.setFormatter(Formatter(\"%(message)s\"))\n    handler2 = FileHandler(filename=f\"{filename}.log\")\n    handler2.setFormatter(Formatter(\"%(message)s\"))\n    logger.addHandler(handler1)\n    logger.addHandler(handler2)\n    return logger\n\nlogger = get_logger()\n\n\n@contextmanager\ndef timer(name):\n    t0 = time.time()\n    yield\n    logger.info(f'[{name}] done in {time.time() - t0:.0f} s')\n\n\ndef seed_everything(seed=777):\n    random.seed(seed)\n    os.environ['PYTHONHASHSEED'] = str(seed)\n    np.random.seed(seed)\n    torch.manual_seed(seed)\n    torch.cuda.manual_seed(seed)\n    torch.backends.cudnn.deterministic = True\n\n    \ndef load_df(path, df_name, debug=False):\n    if path.split('.')[-1]=='csv':\n        df = pd.read_csv(path)\n        if debug:\n            df = pd.read_csv(path, nrows=1000)\n    elif path.split('.')[-1]=='pkl':\n        df = pd.read_pickle(path)\n    if logger==None:\n        print(f\"{df_name} shape / {df.shape} \")\n    else:\n        logger.info(f\"{df_name} shape / {df.shape} \")\n    return df","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Config"},{"metadata":{"trusted":true},"cell_type":"code","source":"#===========================================================\n# Config\n#===========================================================\nOUTPUT_DICT = ''\n\nID = 'Id'\nTARGET_COLS = ['x', 'y', 'f']\nSEED = 42\nseed_everything(seed=SEED)\n\nN_FOLD = 3","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Data Loading"},{"metadata":{"trusted":true,"_kg_hide-input":true},"cell_type":"code","source":"sample_submission = pd.read_csv('../input/indoor-location-navigation/sample_submission.csv')","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## MODEL"},{"metadata":{},"cell_type":"markdown","source":"### As per TabNet Paper , TabNet has three specialities : \n    a) Unlike tree-based methods, TabNet inputs raw tabular data\n    b)TabNet uses sequential aention \n    c) Trained using Gradient -Descent Based Optimizations\n![image.png](attachment:image.png)","attachments":{"image.png":{"image/png":"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"}}},{"metadata":{"trusted":true},"cell_type":"code","source":"# the metric used in this competition\ndef comp_metric(y_pred, y_true):\n    xhat = y_pred[:, 0]\n    yhat = y_pred[:, 1]\n    fhat = y_pred[:, 2]\n    x = y_true[:, 0]\n    y = y_true[:, 1]\n    f = y_true[:, 2]\n    intermediate = np.sqrt(np.power(xhat - x,2) + np.power(yhat-y,2)) + 15 * np.abs(fhat-f)\n    return intermediate.sum()/xhat.shape[0]","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def my_loss_fn(y_pred, y_true):\n    xhat = y_pred[:, 0]\n    yhat = y_pred[:, 1]\n    fhat = y_pred[:, 2]\n    x = y_true[:, 0]\n    y = y_true[:, 1]\n    f = y_true[:, 2]\n    intermediate = torch.sqrt(torch.pow(xhat - x,2) + torch.pow(yhat-y,2)) + 15 * torch.absolute(fhat-f)\n    return torch.mean(intermediate)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"#===========================================================\n# model\n#===========================================================\ndef run_single_tabnet(clf,train_df, test_df, folds, features, target, fold_num=0, categorical=[]):\n    \n    trn_idx = folds[folds.fold != fold_num].index\n    val_idx = folds[folds.fold == fold_num].index\n    logger.info(f'len(trn_idx) : {len(trn_idx)}')\n    logger.info(f'len(val_idx) : {len(val_idx)}')\n    X_train= train_df.iloc[trn_idx][features].values ###Converted this into Numpy array because TabNet will give error otherwise .\n    y_train=target.iloc[trn_idx].values\n    X_valid = train_df.iloc[val_idx][features].values\n    y_valid= target.iloc[val_idx].values\n\n    oof = np.zeros((len(train_df), target.shape[-1]))\n    predictions = np.zeros((len(test_df), target.shape[-1]))\n    \n\n    clf.fit(\n                X_train=X_train, y_train=y_train, ##Train features and train targets\n                eval_set=[(X_valid, y_valid)],\n                loss_fn = my_loss_fn,\n                weights =0,#0 for no balancing,1 for automated balancing,dict for custom weights per class\n                max_epochs=1000,##Maximum number of epochs during training , Default 1000. I used 10\n                patience=70, ##Number of consecutive non improving epoch before early stopping , Default 50\n                batch_size=1024, ##Training batch size\n                virtual_batch_size=128 )##Batch size for Ghost Batch Normalization (virtual_batch_size < batch_size)\n\n    oof[val_idx.tolist()] = clf.predict(train_df.iloc[val_idx][features].values)\n\n    fold_importance_df = pd.DataFrame()\n    fold_importance_df[\"Feature\"] = features\n    fold_importance_df[\"importance\"] = clf.feature_importances_\n    fold_importance_df[\"fold\"] = fold_num\n\n    predictions += clf.predict(test_df[features].values)\n    \n    # RMSE\n    logger.info(\"fold{} RMSE score: {:<8.5f}\".format(fold_num, np.sqrt(mean_squared_error(target.loc[val_idx], oof[val_idx.tolist()]))))\n    \n    return oof, predictions, fold_importance_df\n\n\ndef run_kfold_tabnet(clf,train, test, folds, features, target, n_fold=5, categorical=[]):\n    \n    logger.info(f\"================================= {n_fold}fold TabNet =================================\")\n    \n    oof = np.zeros((len(train), target.shape[-1]))\n    predictions = np.zeros((len(test), target.shape[-1]))\n    feature_importance_df = pd.DataFrame()\n\n    for fold_ in range(n_fold):\n        print(\"Fold {}\".format(fold_))\n        _oof, _predictions, fold_importance_df = run_single_tabnet(clf,train,\n                                                                     test,\n                                                                     folds,\n                                                                     features,\n                                                                     target,\n                                                                     fold_num=fold_,\n                                                                     categorical=categorical)\n        feature_importance_df = pd.concat([feature_importance_df, fold_importance_df], axis=0)\n        oof += _oof\n        predictions += _predictions / n_fold\n\n    # RMSE\n    logger.info(\"CV RMSE score: {:<8.5f}\".format(np.sqrt(mean_squared_error(target, oof))))\n\n    logger.info(f\"=========================================================================================\")\n    \n    return feature_importance_df, predictions, oof\n\n    \ndef show_feature_importance(feature_importance_df, sitename):\n    cols = (feature_importance_df[[\"Feature\", \"importance\"]]\n            .groupby(\"Feature\")\n            .mean()\n            .sort_values(by=\"importance\", ascending=False)[:50].index)\n    best_features = feature_importance_df.loc[feature_importance_df.Feature.isin(cols)]\n\n    plt.figure(figsize=(8, 16))\n    sns.barplot(x=\"importance\", y=\"Feature\", data=best_features.sort_values(by=\"importance\", ascending=False))\n    plt.title('Features importance (averaged/folds)')\n    plt.tight_layout()\n    plt.savefig(OUTPUT_DICT+f'feature_importance_{sitename}.png')\n    plt.show()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_kg_hide-input":true},"cell_type":"code","source":"# prediction_dict = {}\n# oof_dict = {}\n\n# # for TARGET in TARGET_COLS: ## I think this model will work for multiple targets altogether , let me try that later\n\n# logger.info(f'### TABNET for {TARGET_COLS} ###')\n\n# target = train[TARGET_COLS]\n# test[TARGET] = np.nan\n\n# # features\n# cat_features = []\n# num_features = [c for c in test.columns if (test.dtypes[c] != 'object') & (c not in cat_features)]\n# features = num_features + cat_features\n# drop_features = [ID] + TARGET_COLS\n# features = [c for c in features if c not in drop_features]\n\n# if cat_features:\n#     ce_oe = ce.OrdinalEncoder(cols=cat_features, handle_unknown='impute')\n#     ce_oe.fit(train)\n#     train = ce_oe.transform(train)\n#     test = ce_oe.transform(test)\n\n# cat_idxs = [ i for i, f in enumerate(features) if f in cat_features]\n\n# cat_dims = [ categorical_dims[f] for i, f in enumerate(features) if f in cat_features]    \n\n# clf = TabNetRegressor(\n#                     n_d = 16,##Width of the decision prediction layer. Bigger values gives more capacity to the model with the risk of overfitting. Values typically range from 8 to 64.\n#                     n_a = 16,##Width of the attention embedding for each mask. According to the paper n_d=n_a is usually a good choice. (default=8)\n#                     n_steps = 3,##Number of steps in the architecture (usually between 3 and 10)\n#                     gamma =1.3,##This is the coefficient for feature reusage in the masks. A value close to 1 will make mask selection least correlated between layers. Values range from 1.0 to 2.0.\n#                     cat_idxs=cat_idxs, ##List of categorical features indices.\n#                     cat_dims=cat_dims,\n#                     cat_emb_dim =1, ##List of embeddings size for each categorical features. (default =1)\n#                     n_independent =2,##Number of independent Gated Linear Units layers at each step. Usual values range from 1 to 5.\n#                     n_shared =2,##Number of shared Gated Linear Units at each step Usual values range from 1 to 5\n#                     epsilon  = 1e-15,##Should be left untouched.\n#                     seed  =0,##Random seed for reproducibility\n#                     momentum = 0.02, ##Momentum for batch normalization, typically ranges from 0.01 to 0.4 (default=0.02)\n# #                         lr = 0.01, ##Initial learning rate used for training. As mentionned in the original paper, a large initial learning of 0.02 with decay is a good option.\n#                     clip_value =None,\n#                     lambda_sparse =1e-3,##This is the extra sparsity loss coefficient as proposed in the original paper. The bigger this coefficient is, the sparser your model will be in terms of feature selection. Depending on the difficulty of your problem, reducing this value could help.\n#                     optimizer_fn =torch.optim.Adam, ## Optimizer\n#                     scheduler_fn = None, #torch.optim.lr_scheduler.ReduceLROnPlateau, ## LR scheduler \n#                     scheduler_params = None,#{\"mode\":'min', \"factor\":0.1, \"patience\":10, \"verbose\":\"False\"}, ## LR scheduler parameters dictionary\n#                     verbose =1,\n#                     device_name = 'auto' ## Auto or 'gpu' ## I have no GPU\n\n#                     )    \n\n# feature_importance_df, predictions, oof = run_kfold_tabnet(clf,train, test, folds, features, target, \n#                                                              n_fold=N_FOLD, categorical=cat_features)\n\n# prediction_dict = predictions\n# oof_dict = oof\n\n# show_feature_importance(feature_importance_df, TARGET)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Training"},{"metadata":{"trusted":true},"cell_type":"code","source":"# get our train and test files\nfeature_dir = '../input/generate-wifi-features-5-times-faster/'\n\ntrain_files = sorted(glob.glob(os.path.join(feature_dir, '*_train.csv')))\ntest_files = sorted(glob.glob(os.path.join(feature_dir, '*_test.csv')))\nssubm = pd.read_csv('../input/indoor-location-navigation/sample_submission.csv', index_col=0)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_kg_hide-input":true},"cell_type":"code","source":"subm_predictions = list()\n\nfor e, train_file in enumerate(train_files):\n    train = pd.read_csv(train_file).reset_index().rename(columns={'index':'Id'})\n    train = train.sample(frac=1, random_state=10)\n    train['Id'] = range(0,len(train))\n    train = train.set_index('Id')\n    train = train.reset_index()\n    \n    test = pd.read_csv(test_files[e]).reset_index().rename(columns={'index':'Id'})\n    \n    \n#     # Comment this to train using all data\n#     train_size = int(len(train) * 0.8)\n#     # --- Data Validation ---\n#     # Valid features + targets\n#     local_valid = train.iloc[train_size:, :]\n#     # Train features + targets\n#     train = train.iloc[:train_size, :]\n\n     \n    \n    folds = train[[ID]+TARGET_COLS].copy()\n    Fold = KFold(n_splits=N_FOLD, shuffle=True, random_state=SEED)\n    for n, (train_index, val_index) in enumerate(Fold.split(folds, folds[TARGET_COLS])):\n        folds.loc[val_index, 'fold'] = int(n)\n    folds['fold'] = folds['fold'].astype(int)\n    folds.head()\n\n    prediction_dict = {}\n    oof_dict = {}\n\n    # for TARGET in TARGET_COLS: ## I think this model will work for multiple targets altogether , let me try that later\n    site = train_file.split('/')[-1].split('_')[0]\n    logger.info(f'### TABNET for {TARGET_COLS} of site {e}: {site} ###')\n\n    target = train[TARGET_COLS]\n    test[TARGET_COLS] = np.nan\n\n    # features\n    cat_features = []\n    num_features = [c for c in test.columns if (test.dtypes[c] != 'object') & (c not in cat_features)]\n    features = num_features + cat_features\n    drop_features = [ID] + TARGET_COLS\n    features = [c for c in features if c not in drop_features]\n\n    if cat_features:\n        ce_oe = ce.OrdinalEncoder(cols=cat_features, handle_unknown='impute')\n        ce_oe.fit(train)\n        train = ce_oe.transform(train)\n        test = ce_oe.transform(test)\n\n    cat_idxs = [ i for i, f in enumerate(features) if f in cat_features]\n\n    cat_dims = [ categorical_dims[f] for i, f in enumerate(features) if f in cat_features]    \n\n    clf = TabNetRegressor(\n                        n_d = 32,##Width of the decision prediction layer. Bigger values gives more capacity to the model with the risk of overfitting. Values typically range from 8 to 64.\n                        n_a = 32,##Width of the attention embedding for each mask. According to the paper n_d=n_a is usually a good choice. (default=8)\n                        n_steps = 3,##Number of steps in the architecture (usually between 3 and 10)\n                        gamma =1.3,##This is the coefficient for feature reusage in the masks. A value close to 1 will make mask selection least correlated between layers. Values range from 1.0 to 2.0.\n                        cat_idxs=cat_idxs, ##List of categorical features indices.\n                        cat_dims=cat_dims,\n                        cat_emb_dim =1, ##List of embeddings size for each categorical features. (default =1)\n                        n_independent =2,##Number of independent Gated Linear Units layers at each step. Usual values range from 1 to 5.\n                        n_shared =2,##Number of shared Gated Linear Units at each step Usual values range from 1 to 5\n                        epsilon  = 1e-15,##Should be left untouched.\n                        seed  =42,##Random seed for reproducibility\n                        momentum = 0.02, ##Momentum for batch normalization, typically ranges from 0.01 to 0.4 (default=0.02)\n                        optimizer_params = dict(lr=0.01),##Initial learning rate used for training. As mentionned in the original paper, a large initial learning of 0.02 with decay is a good option.\n                        clip_value =None,\n                        lambda_sparse =1e-5,##This is the extra sparsity loss coefficient as proposed in the original paper. The bigger this coefficient is, the sparser your model will be in terms of feature selection. Depending on the difficulty of your problem, reducing this value could help.\n                        optimizer_fn =torch.optim.Adam, ## Optimizer\n                        scheduler_fn = None, #torch.optim.lr_scheduler.ReduceLROnPlateau, ## LR scheduler \n                        scheduler_params = None,#{\"mode\":'min', \"factor\":0.1, \"patience\":10, \"verbose\":\"False\"}, ## LR scheduler parameters dictionary\n                        verbose =1,\n                        device_name = 'cuda' ## Auto or 'gpu' ## I have no GPU\n\n                        )    \n\n    feature_importance_df, prediction_dict, oof_dict = run_kfold_tabnet(clf,train, test, folds, features, target, \n                                                                 n_fold=N_FOLD, categorical=cat_features)\n\n    \n    \n    test_predsx = prediction_dict[:, 0]\n    test_predsy = prediction_dict[:, 1]\n    test_predsf = prediction_dict[:, 2]\n    \n    test_preds = pd.DataFrame(np.stack((test_predsf, test_predsx, test_predsy))).T\n    test_preds.columns = ssubm.columns\n    test_preds.index = test[\"site_path_timestamp\"]\n    test_preds[\"floor\"] = test_preds[\"floor\"].astype(int)\n    subm_predictions.append(test_preds)\n    \n    # save tabnet model\n    saving_path_name = f\"tabnet_model_test_{e}\"\n    saved_filepath = clf.save_model(saving_path_name)\n    \n    score = comp_metric(oof_dict, train[['x', 'y', 'f']].to_numpy())\n    logger.info(f'Local OOF Score site {e}: {score}')\n    \n#     local_valid_score = comp_metric(clf.predict(local_valid.drop(columns = [ID] + TARGET_COLS + ['path']).to_numpy()), local_valid[['x', 'y', 'f']].to_numpy())\n#     logger.info(f'Local Valid Score site {e}: {local_valid_score}')\n    \n#     show_feature_importance(feature_importance_df, site)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Submission"},{"metadata":{"trusted":true},"cell_type":"code","source":"# generate prediction file \nall_preds = pd.concat(subm_predictions)\nall_preds = all_preds.reindex(ssubm.index)\nall_preds.to_csv('submission.csv')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"last_subm = pd.read_csv('../input/lightgdm-baseline-wifi-features/submission.csv')\nlast_subm.head(30)","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":4}