easy_tpp.model.torch_model.torch_sahp

Classes

SAHP(model_config)

Torch implementation of Self-Attentive Hawkes Process, ICML 2020.

class easy_tpp.model.torch_model.torch_sahp.SAHP(model_config)

Torch implementation of Self-Attentive Hawkes Process, ICML 2020. Part of the code is collected from https://github.com/yangalan123/anhp-andtt/blob/master/sahp

I slightly modify the original code because it is not stable.

__init__(model_config)

Initialize the model

Parameters:

model_config (EasyTPP.ModelConfig) – config of model specs.

state_decay(encode_state, mu, eta, gamma, duration_t)

Equation (15), which computes the pre-intensity states

Parameters:

• encode_state (tensor) – [batch_size, seq_len, hidden_size].

• mu (tensor) – [batch_size, seq_len, hidden_size].

• eta (tensor) – [batch_size, seq_len, hidden_size].

• gamma (tensor) – [batch_size, seq_len, hidden_size].

• duration_t (tensor) – [batch_size, seq_len, num_sample].

Returns:

hidden states at event times.

Return type:

tensor

forward(time_seqs, time_delta_seqs, event_seqs, attention_mask)

Call the model

Parameters:

• time_seqs (tensor) – [batch_size, seq_len], timestamp seqs.

• time_delta_seqs (tensor) – [batch_size, seq_len], inter-event time seqs.

• event_seqs (tensor) – [batch_size, seq_len], event type seqs.

• attention_mask (tensor) – [batch_size, seq_len, hidden_size], attention masks.

Returns:

hidden states at event times.

Return type:

tensor

loglike_loss(batch)

Compute the loglike loss.

Parameters:

batch (tuple, list) – batch input.

Returns:

loglike loss, num events.

Return type:

list

compute_states_at_sample_times(encode_state, sample_dtimes)

Compute the hidden states at sampled times.

Parameters:

• encode_state (tensor) – three tensors with each shape [batch_size, seq_len, hidden_size].

• sample_dtimes (tensor) – [batch_size, seq_len, num_samples].

Returns:

[batch_size, seq_len, num_samples, hidden_size]， hidden state at each sampled time.

Return type:

tensor

compute_intensities_at_sample_times(time_seqs, time_delta_seqs, type_seqs, sample_dtimes, **kwargs)

Compute hidden states at sampled times.

Parameters:

• time_seqs (tensor) – [batch_size, seq_len], times seqs.

• time_delta_seqs (tensor) – [batch_size, seq_len], time delta seqs.

• type_seqs (tensor) – [batch_size, seq_len], event type seqs.

• sample_dtimes (tensor) – [batch_size, seq_len, num_samples], sampled inter-event timestamps.

Returns:

[batch_size, seq_len, num_samples, num_event_types], intensity at all sampled times.

Return type:

tensor