Edge Sampling

class RandomEdgeSampler(number_of_edges: int = 100, seed: int = 42)

An implementation of random edge sampling. Edges are sampled with the same uniform probability randomly. “For details about the algorithm see this paper.”

Parameters

• number_of_edges (int) – Number of edges. Default is 100.

• seed (int) – Random seed. Default is 42.

sample(graph: Union[networkx.classes.graph.Graph, networkit.graph.Graph]) → Union[networkx.classes.graph.Graph, networkit.graph.Graph]

Sampling edges randomly.

Arg types:

• graph (NetworkX or NetworKit graph) - The graph to be sampled from.

Return types:

• new_graph (NetworkX or NetworKit graph) - The graph of sampled edges.

class RandomNodeEdgeSampler(number_of_edges: int = 100, seed: int = 42)

An implementation of random node-edge sampling. The algorithm first randomly samples a node. From this node it samples an edge with a neighbor. “For details about the algorithm see this paper.”

Parameters

• number_of_edges (int) – Number of edges. Default is 100.

• seed (int) – Random seed. Default is 42.

sample(graph: Union[networkx.classes.graph.Graph, networkit.graph.Graph]) → Union[networkx.classes.graph.Graph, networkit.graph.Graph]

Sampling edges randomly from randomly sampled nodes.

Arg types:

• graph (NetworkX or NetworKit graph) - The graph to be sampled from.

Return types:

• new_graph (NetworkX or NetworKit graph) - The graph of sampled edges.

class HybridNodeEdgeSampler(number_of_edges: int = 100, seed: int = 42, p: float = 0.8)

An implementation of hybrid node-edge sampling. The algorithm alternates between two sampling methods. (A) Random uniform edge sampling. (B) The algorithm first randomly samples a node. From this node it samples an edge with a neighbor. “For details about the algorithm see this paper.”

Parameters

• number_of_edges (int) – Number of edges. Default is 100.

• seed (int) – Random seed. Default is 42.

• p (float) – Hybridization probability. Default is 0.8.

sample(graph: Union[networkx.classes.graph.Graph, networkit.graph.Graph]) → Union[networkx.classes.graph.Graph, networkit.graph.Graph]

Sampling edges randomly from randomly sampled nodes or sampling random edges.

Arg types:

• graph (NetworkX or NetworKit graph) - The graph to be sampled from.

Return types:

• new_graph (NetworkX or NetworKit graph) - The graph of sampled edges.

class RandomEdgeSamplerWithPartialInduction(p: float = 0.5, seed: int = 42)

An implementation of random edge sampling with partial edge set induction. The algorithm randomly samples edges in a streaming fashion with a fixed probability. Edges between nodes which are already in the sample are retained with an induction step. “For details about the algorithm see this paper.”

Parameters

• p (float) – Sampling probability. Default is 0.5.

• seed (int) – Random seed. Default is 42.

sample(graph: Union[networkx.classes.graph.Graph, networkit.graph.Graph]) → Union[networkx.classes.graph.Graph, networkit.graph.Graph]

Sampling edges randomly with partial induction.

Arg types:

• graph (NetworkX or NetworKit graph) - The graph to be sampled from.

Return types:

• new_graph (NetworkX or NetworKit graph) - The graph of sampled edges.

class RandomEdgeSamplerWithInduction(number_of_edges: int = 100, seed: int = 42)

An implementation of random edge sampling with edge set induction. The algorithm randomly samples edges with a fixed probability. Edges between nodes which are already in the sample are retained with an induction step. “For details about the algorithm see this paper.”

Parameters

• number_of_edges (int) – Number of edges. Default is 100.

• seed (int) – Random seed. Default is 42.

sample(graph: Union[networkx.classes.graph.Graph, networkit.graph.Graph]) → Union[networkx.classes.graph.Graph, networkit.graph.Graph]

Sampling edges randomly with induction.

Arg types:

• graph (NetworkX graph) - The graph to be sampled from.

Return types:

• new_graph (NetworkX graph) - The graph of sampled edges.