Skip to content

module adabmDCA.graph

function update_mask_activation

update_mask_activation(Dkl: Tensor, mask: Tensor, nactivate: int) → Tensor

Updates the mask by removing the nactivate couplings with the smallest Dkl.

Args:

  • Dkl (torch.Tensor): Kullback-Leibler divergence matrix.
  • mask (torch.Tensor): Mask.
  • nactivate (int): Number of couplings to be activated at each graph update.

Returns:

  • torch.Tensor: Updated mask.

function update_mask_decimation

update_mask_decimation(mask: Tensor, Dkl: Tensor, drate: float) → Tensor

Updates the mask by removing the n_remove couplings with the smallest Dkl.

Args:

  • mask (torch.Tensor): Mask.
  • Dkl (torch.Tensor): Kullback-Leibler divergence matrix.
  • drate (float): Percentage of active couplings to be pruned at each decimation step.

Returns:

  • torch.Tensor: Updated mask.

function decimate_graph

decimate_graph(
    pij: Tensor,
    params: Dict[str, Tensor],
    mask: Tensor,
    drate: float
) → Tuple[Dict[str, Tensor], Tensor]

Performs one decimation step and updates the parameters and mask.

Args:

  • pij (torch.Tensor): Two-point marginal probability distribution.
  • params (Dict[str, torch.Tensor]): Parameters of the model.
  • mask (torch.Tensor): Mask.
  • drate (float): Percentage of active couplings to be pruned at each decimation step.

Returns:

  • Tuple[Dict[str, torch.Tensor], torch.Tensor]: Updated parameters and mask.

This file was automatically generated via lazydocs.