FederatedGraphPartitioner<T>

Models & Types ยท AiDotNet.FederatedLearning.Graph

Partitions a graph across federated clients using various strategies.

Inheritance: FederatedLearningComponentBase<T> โ†’ FederatedGraphPartitioner<T>

For Beginners

Before federated graph learning can begin, the graph must be divided among clients. The partitioner splits nodes into groups, trying to minimize the number of edges that cross partition boundaries (cross-client edges) while keeping partitions balanced.

How It Works

Strategies:

Constructors

Constructor Summary
FederatedGraphPartitioner(FederatedGraphOptions options) Initializes a new instance of FederatedGraphPartitioner.

Methods

Method Summary
ValueTuple<Tensor<T>, Tensor<T>> ExtractSubgraph(Tensor<T> adjacency, Tensor<T> nodeFeatures, int[] assignments, int partitionId) Extracts subgraph data for a specific partition.
int[] Partition(Tensor<T> adjacency, Tensor<T> nodeFeatures, int numPartitions) Partitions a graph into the specified number of parts.