-
Notifications
You must be signed in to change notification settings - Fork 0
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
1 parent
9930562
commit 7b474c0
Showing
5 changed files
with
250 additions
and
11 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,155 @@ | ||
| import torch | ||
| from tqdm import tqdm, trange | ||
| import numpy as np | ||
| from sklearn.metrics import accuracy_score, mean_squared_error | ||
|
|
||
| class GeneralRNN(torch.nn.Module): | ||
| r"""A general RNN model for time-series prediction | ||
| """ | ||
|
|
||
| def __init__(self, args): | ||
| super(GeneralRNN, self).__init__() | ||
| self.model_type = args['model_type'] | ||
|
|
||
| self.input_size = args['in_dim'] | ||
| self.hidden_size = args['h_dim'] | ||
| self.output_size = args['out_dim'] | ||
| self.num_layers = args['n_layers'] | ||
| self.dropout = args['dropout'] | ||
| self.bidirectional = args['bidirectional'] | ||
|
|
||
| self.max_seq_len = args['max_seq_len'] | ||
|
|
||
| self.rnn_module = self._get_rnn_module(self.model_type) | ||
|
|
||
| self.rnn_layer = self.rnn_module( | ||
| input_size=self.input_size, | ||
| hidden_size=self.hidden_size, | ||
| num_layers=self.num_layers, | ||
| batch_first=True, | ||
| dropout=self.dropout, | ||
| bidirectional=self.bidirectional | ||
| ) | ||
|
|
||
| self.linear_layer = torch.nn.Linear( | ||
| in_features=self.hidden_size, | ||
| out_features=self.output_size | ||
| ) | ||
|
|
||
| def _get_rnn_module(self, model_type): | ||
| if model_type == "rnn": | ||
| return torch.nn.RNN | ||
| elif model_type == "lstm": | ||
| return torch.nn.LSTM | ||
| elif model_type == "gru": | ||
| return torch.nn.GRU | ||
|
|
||
| def forward(self, X): | ||
| # Dynamic RNN input for ignoring paddings | ||
| H_o, H_t = self.rnn_layer(X) | ||
| logits = self.linear_layer(H_o) | ||
|
|
||
| return logits | ||
|
|
||
| def rmse_error(y_true, y_pred): | ||
| """User defined root mean squared error. | ||
| Args: | ||
| - y_true: true labels | ||
| - y_pred: predictions | ||
| Returns: | ||
| - computed_rmse: computed rmse loss | ||
| """ | ||
| # Exclude masked labels | ||
| idx = (y_true >= 0) * 1 | ||
| # Mean squared loss excluding masked labels | ||
| computed_mse = np.sum(idx * ((y_true - y_pred) ** 2)) / np.sum(idx) | ||
| computed_rmse = np.sqrt(computed_mse) | ||
| return computed_rmse | ||
|
|
||
| def one_step_ahead_prediction(train_data, test_data): | ||
| """Use the previous time-series to predict one-step ahead feature values. | ||
| Args: | ||
| - train_data: training time-series | ||
| - test_data: testing time-series | ||
| Returns: | ||
| - perf: average performance of one-step ahead predictions (in terms of AUC or MSE) | ||
| """ | ||
| train_data = train_data | ||
| test_data = test_data | ||
|
|
||
| # Parameters | ||
| no, seq_len, dim = 256, 30, 92 | ||
|
|
||
| # Set model parameters | ||
| args = {} | ||
| args["device"] = "cuda" | ||
| args["task"] = "regression" | ||
| args["model_type"] = "gru" | ||
| args["bidirectional"] = False | ||
| args["epochs"] = 20 | ||
| args["batch_size"] = 256 | ||
| args["in_dim"] = dim | ||
| args["h_dim"] = dim | ||
| args["out_dim"] = dim | ||
| args["n_layers"] = 3 | ||
| args["dropout"] = 0.5 | ||
| args["max_seq_len"] = 30 # only 29 is used for prediction | ||
| args["learning_rate"] = 1e-3 | ||
| args["grad_clip_norm"] = 5.0 | ||
|
|
||
| # Set training features and labels | ||
| train_dataloader = torch.utils.data.DataLoader( | ||
| train_data, | ||
| batch_size=args["batch_size"], | ||
| shuffle=True | ||
| ) | ||
|
|
||
| # Set testing features and labels | ||
| test_dataloader = torch.utils.data.DataLoader( | ||
| test_data, | ||
| batch_size=no, | ||
| shuffle=True | ||
| ) | ||
| # Initialize model | ||
| model = GeneralRNN(args) | ||
| model.to(args["device"]) | ||
| criterion = torch.nn.MSELoss() | ||
| optimizer = torch.optim.Adam( | ||
| model.parameters(), | ||
| lr=args["learning_rate"] | ||
| ) | ||
|
|
||
| # Train the predictive model | ||
| logger = trange(args["epochs"], desc=f"Epoch: 0, Loss: 0") | ||
| for epoch in logger: | ||
| running_loss = 0.0 | ||
|
|
||
| for train_x in train_dataloader: | ||
| train_x = train_x.to(args["device"]) | ||
| # zero the parameter gradients | ||
| optimizer.zero_grad() | ||
| # forward | ||
| train_p = model(train_x[:,0:29,:]) | ||
| loss = criterion(train_p, train_x[:,1:30,:]) | ||
| # backward | ||
| loss.backward() | ||
| # optimize | ||
| optimizer.step() | ||
|
|
||
| running_loss += loss.item() | ||
|
|
||
| logger.set_description(f"Epoch: {epoch}, Loss: {running_loss:.4f}") | ||
|
|
||
| # Evaluate the trained model | ||
| with torch.no_grad(): | ||
| perf = 0 | ||
| for test_x in test_dataloader: | ||
| test_x = test_x.to(args["device"]) | ||
| test_p = model(test_x[:,0:29,:]).cpu() | ||
|
|
||
| test_p = np.reshape(test_p.numpy(), [-1]) | ||
| test_y = np.reshape(test_x[:,1:30,:].cpu().numpy(), [-1]) | ||
|
|
||
| perf += rmse_error(test_y, test_p) | ||
|
|
||
| return perf |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,83 @@ | ||
| from sklearn.manifold import TSNE | ||
| from sklearn.decomposition import PCA | ||
| import matplotlib.pyplot as plt | ||
| import numpy as np | ||
|
|
||
|
|
||
| def visualization(ori_data, generated_data, analysis): | ||
| """Using PCA or tSNE for generated and original data visualization. | ||
| Args: | ||
| - ori_data: original data | ||
| - generated_data: generated synthetic data | ||
| - analysis: tsne or pca | ||
| """ | ||
| # Analysis sample size (for faster computation) | ||
| anal_sample_no = min([1000, len(ori_data)]) | ||
| idx = np.random.permutation(len(ori_data))[:anal_sample_no] | ||
|
|
||
| # Data preprocessing | ||
| ori_data = np.asarray(ori_data) | ||
| generated_data = np.asarray(generated_data) | ||
|
|
||
| ori_data = ori_data[idx] | ||
| generated_data = generated_data[idx] | ||
|
|
||
| no, seq_len, dim = ori_data.shape | ||
|
|
||
| for i in range(anal_sample_no): | ||
| if (i == 0): | ||
| prep_data = np.reshape(np.mean(ori_data[0, :, :], 1), [1, seq_len]) | ||
| prep_data_hat = np.reshape(np.mean(generated_data[0, :, :], 1), [1, seq_len]) | ||
| else: | ||
| prep_data = np.concatenate((prep_data, | ||
| np.reshape(np.mean(ori_data[i, :, :], 1), [1, seq_len]))) | ||
| prep_data_hat = np.concatenate((prep_data_hat, | ||
| np.reshape(np.mean(generated_data[i, :, :], 1), [1, seq_len]))) | ||
|
|
||
| # Visualization parameter | ||
| colors = ["tab:blue" for i in range(anal_sample_no)] + ["tab:orange" for i in range(anal_sample_no)] | ||
|
|
||
| if analysis == 'pca': | ||
| # PCA Analysis | ||
| pca = PCA(n_components=2) | ||
| pca.fit(prep_data) | ||
| pca_results = pca.transform(prep_data) | ||
| pca_hat_results = pca.transform(prep_data_hat) | ||
|
|
||
| # Plotting | ||
| f, ax = plt.subplots(1) | ||
| plt.scatter(pca_results[:, 0], pca_results[:, 1], | ||
| c=colors[:anal_sample_no], alpha=0.2, label="Original") | ||
| plt.scatter(pca_hat_results[:, 0], pca_hat_results[:, 1], | ||
| c=colors[anal_sample_no:], alpha=0.2, label="Synthetic") | ||
|
|
||
| ax.legend() | ||
| plt.title('PCA plot') | ||
| plt.xlabel('x-pca') | ||
| plt.ylabel('y_pca') | ||
| plt.show() | ||
|
|
||
| elif analysis == 'tsne': | ||
|
|
||
| # Do t-SNE Analysis together | ||
| prep_data_final = np.concatenate((prep_data, prep_data_hat), axis=0) | ||
|
|
||
| # TSNE anlaysis | ||
| tsne = TSNE(n_components=2, verbose=1, perplexity=40, n_iter=300) | ||
| tsne_results = tsne.fit_transform(prep_data_final) | ||
|
|
||
| # Plotting | ||
| f, ax = plt.subplots(1) | ||
|
|
||
| plt.scatter(tsne_results[:anal_sample_no, 0], tsne_results[:anal_sample_no, 1], | ||
| c=colors[:anal_sample_no], alpha=0.2, label="Original") | ||
| plt.scatter(tsne_results[anal_sample_no:, 0], tsne_results[anal_sample_no:, 1], | ||
| c=colors[anal_sample_no:], alpha=0.2, label="Synthetic") | ||
|
|
||
| ax.legend() | ||
|
|
||
| plt.title('t-SNE plot') | ||
| plt.xlabel('x-tsne') | ||
| plt.ylabel('y_tsne') | ||
| plt.show() |