TrendBasis¶

class TrendBasis(degree: int, backcast_size: int, forecast_size: int)[source]¶

Bases: torch.nn.modules.module.Module

Polynomial trend basis function.

Initialize trend basis function.

Parameters

degree (int) – Degree of polynomial for trend modeling.
backcast_size (int) – Number of backcast values.
forecast_size (int) – Number of forecast values.

Methods

forward(theta)

Forward pass.

Attributes

forward(theta: torch.Tensor) → Tuple[torch.Tensor, torch.Tensor][source]¶

Forward pass.

Parameters: theta (torch.Tensor) – Basis function parameters.
Returns: Tuple with backcast and forecast.
Return type: Tuple[torch.Tensor, torch.Tensor]