Struct GraphPartitioner 

pub struct GraphPartitioner {
    pub num_partitions: usize,
    pub method: PartitionMethod,
    pub max_iterations: usize,
}

Configures and runs graph partitioning.

Fields

num_partitions: usize

Number of partitions k.

method: PartitionMethod

Algorithm selection.

max_iterations: usize

Maximum iterations (used by LabelPropagation and improvement loops).

Implementations

impl GraphPartitioner

pub fn new(num_partitions: usize) -> Self

Create a partitioner with the default method (Greedy) and 20 iterations.

pub fn with_method(self, method: PartitionMethod) -> Self

Select the partitioning method.

pub fn with_max_iterations(self, max_iter: usize) -> Self

Override the maximum iteration count.

pub fn partition( &self, nodes: &[String], edges: &[(String, String)], ) -> PartitionResult

Partition the graph and return a PartitionResult.

Arguments

• nodes – list of node identifiers (must be unique).
• edges – list of (from, to) string pairs; direction is ignored.

pub fn greedy_partition( nodes: &[String], edges: &[(String, String)], k: usize, ) -> Vec<usize>

Greedy round-robin assignment with a single local-improvement sweep.

Returns a per-node partition vector (indices into nodes).

pub fn label_propagation( nodes: &[String], edges: &[(String, String)], k: usize, max_iter: usize, ) -> Vec<usize>

Iterative label propagation.

Each node adopts the most frequent label among its neighbours. Labels are initialised in round-robin fashion and the propagation continues for at most max_iter rounds or until convergence.

pub fn bisection_partition( nodes: &[String], edges: &[(String, String)], k: usize, ) -> Vec<usize>

Recursive bisection: repeatedly split the largest partition by BFS.

pub fn count_cut_edges(assignments: &[usize], edges: &[(usize, usize)]) -> usize

Count edges whose endpoints are in different partitions.

pub fn balance_score(assignments: &[usize], k: usize) -> f64

Balance score in [0, 1]. 1.0 = perfectly balanced.

Defined as min_size / max_size where sizes are partition cardinalities. Returns 1.0 for empty assignments or k=1.

pub fn build_adjacency( nodes: &[String], edges: &[(String, String)], ) -> Vec<Vec<usize>>

Build an undirected adjacency list indexed by position in nodes.

Trait Implementations

impl Clone for GraphPartitioner

fn clone(&self) -> GraphPartitioner

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for GraphPartitioner

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.