Computes mean L1Loss :return: mean L1Loss

Accuracy with masking. :param preds: :param labels: :param null_val: :return:

:param x: shape (batch_size, seq_len, num_sensor, input_dim) :param y: shape (batch_size, horizon, num_sensor, input_dim) :re

:param x: shape (seq_len, batch_size, num_sensor, input_dim) :param y: shape (horizon, batch_size, num_sensor, input_dim) :re

# L = D^-1/2 (D-A) D^-1/2 = I - D^-1/2 A D^-1/2 # D = diag(A 1) :param adj: :return:

Decoder forward pass :param encoder_hidden_state: (num_layers, batch_size, self.hidden_state_size) :param labels: (self.horiz

Encoder forward pass :param inputs: shape (seq_len, batch_size, num_sensor * input_dim) :return: encoder_hidden_state: (num_l

Generate samples from :param df: :param x_offsets: :param y_offsets: :param add_time_in_day: :param add_day_in_week:

Generate samples from :param df: :param x_offsets: :param y_offsets: :param add_time_in_day: :param add_day_in_week:

:param distance_df: data frame with three columns: [from, to, distance]. :param sensor_ids: list of sensor ids. :param normalized_k: ent

Sample from the Gumbel-Softmax distribution and optionally discretize. Args: logits: [batch_size, n_class] unnormalized log-probs temperatur

Calculates the historical average of sensor reading. :param df: :param period: default 1 week. :param test_ratio: :param nul

Accuracy with masking. :param preds: :param labels: :param null_val: :return:

Accuracy with masking. :param preds: :param labels: :param null_val: :return:

Assumes $x^{t+1} = x^{t}$ :param df: :param n_forward: :param test_ratio: :return:

Multivariate time series forecasting using Vector Auto-Regressive Model. :param df: pandas.DataFrame, index: time, columns: sensor id, cont

:param xs: :param ys: :param batch_size: :param pad_with_last_sample: pad with the last sample to make number of sam

:param num_units: :param adj_mx: :param max_diffusion_step: :param num_nodes: :param nonlinearity: :

Writes summary for a list of scalars. :param writer: :param names: :param values: :param global_step: :return:

Calculate the MAE, MAPE, RMSE :param df_pred: :param df_test: :param null_val: :return:

Encoder forward pass. :param inputs: shape (batch_size, self.num_nodes * self.input_dim) :param hidden_state: (num_layers, ba

:param inputs: shape (batch_size, self.num_nodes * self.output_dim) :param hidden_state: (num_layers, batch_size, self.hidden_state_s

:param inputs: shape (seq_len, batch_size, num_sensor * input_dim) :param labels: shape (horizon, batch_size, num_sensor * output)

Gated recurrent unit (GRU) with Graph Convolution. :param inputs: (B, num_nodes * input_dim) :param hx: (B, num_nodes * rnn_units)

Calculates the total number of trainable parameters in the current graph. :return: