gnn_tracking.metrics.graph_construction

gnn_tracking.metrics.graph_construction#

Functions#

get_good_node_mask(→ torch.Tensor)

Get a mask for nodes that are included in metrics and more.

get_efficiency_purity_edges(→ dict[str, float])

Calculate efficiency and purity for edges based on data.true_edge_index.

Module Contents#

gnn_tracking.metrics.graph_construction.get_good_node_mask(data: torch_geometric.data.Data, *, pt_thld: float = 0.9, max_eta: float = 4.0) torch.Tensor#

Get a mask for nodes that are included in metrics and more. This includes lower limit on pt, not noise, reconstructable, cut on eta.

gnn_tracking.metrics.graph_construction.get_efficiency_purity_edges(data: torch_geometric.data.Data, pt_thld: float = 0.9, max_eta: float = 4.0) dict[str, float]#

Calculate efficiency and purity for edges based on data.true_edge_index.

Only edges where at least one of the two nodes is accepted by the pt threshold (and reconstructable etc.) are considered.