core

Core classes for (temporal) graphs, paths, and higher-order De Bruijn graphs.

The classes in the core module can be used to implement integrated pipelines to preprocess time-stamped network data, do inference and model selection of higher-order De Bruijn graph models and address temporal graph learning tasks based on time-aware graph neural networks.

Example

import pathpyG as pp
pp.config['torch']['device'] = 'cuda'

# Generate toy example temporal graph
g = pp.TemporalGraph.from_edge_list([
    ('b', 'c', 2),
    ('a', 'b', 1),
    ('c', 'd', 3),
    ('d', 'a', 4),
    ('b', 'd', 2),
    ('d', 'a', 6),
    ('a', 'b', 7)])

# Create Multi-Order model that models time-respecting paths
m = pp.MultiOrderModel.from_temporal_graph(g, delta=1, max_order=3)
print(m.layers[1])
print(m.layers[2])
print(m.layers[3])