pymatgen.analysis.diffusion.neb.periodic_dijkstra module¶

Dijkstra’s path search on a graph where the nodes are on a periodic graph.

get_optimal_pathway_rev(path_parent: dict, leaf_node: tuple)[source]¶: Follow a leaf node all the way up to source.

periodic_dijkstra(G: Graph, sources: set, weight: str = 'weight', max_image: int = 2, target_reached: Callable = <function <lambda>>)[source]¶

Find the lowest cost pathway from a source point in the periodic graph. Since the search can move many cells away without finding the target we have to limit how many cells away from (0,0,0) to search.

Parameters:

G (Graph) – The graph object with additional “to_jimage” fields to indicate edges across periodic images.
sources (set) – the index of the source node
weight – the weight of the edges.
max_image (int, optional) – Defaults to 3.
target_reached (callable, optional) – A function of (site_index, jimage) used to check for stop iteration. This function is always called on the top of heap so it might miss the optimal path but often can find a “good enough” path very quickly.

Returns:

a dictionary of the best cost found to periodic node keyed by (site_index, jimage) path_parent: dictionary of optimal path parent for each node given in index-image pairs.

Return type:

best_ans

periodic_dijkstra_on_sgraph(sgraph: StructureGraph, sources: set, weight: str = 'weight', max_image: int = 1, target_reached: Callable = <function <lambda>>)[source]¶