Local Probability Transfer

This example compares global propagation with radius-limited local propagation. The global two-step run reaches node 2; the local run with radius=1 keeps the transfer inside the one-hop active neighborhood.

Python

import json
from tonggraph import Graph

graph = Graph()
a = graph.add_node()
b = graph.add_node()
c = graph.add_node()

graph.add_edge(a, b, "P", properties={"probability": "0.5"})
graph.add_edge(b, c, "P", properties={"probability": "0.25"})

result = {
    "global_two_step": {
        str(k): v for k, v in sorted(graph.propagate({a: 1.0}, 2).items())
    },
    "local_radius_1": {
        str(k): v
        for k, v in sorted(
            graph.local_propagate({a: 1.0}, radius=1, edge_type="P").items()
        )
    },
}

print(json.dumps(result, indent=2, sort_keys=True))

Output:

{
  "global_two_step": {
    "0": 1.0,
    "1": 0.5,
    "2": 0.125
  },
  "local_radius_1": {
    "0": 1.0,
    "1": 0.5
  }
}