Struct SparseTransition 

pub struct SparseTransition {
    pub nodes: Vec<NodeId>,
    pub row_ptr: Vec<usize>,
    pub col_idx: Vec<usize>,
    pub values: Vec<f32>,
    pub has_outgoing: Vec<bool>,
}

Row-stochastic sparse transition matrix in plain CSR form, plus the ordered NodeId -> matrix-row mapping. Public so tests / downstream code can inspect it.

Fields

nodes: Vec<NodeId>

Sorted NodeId sequence: nodes[i] is row/column i.

row_ptr: Vec<usize>

CSR row pointers. Length nodes.len() + 1.

col_idx: Vec<usize>

CSR column indices. Length nnz.

values: Vec<f32>

CSR values. Length nnz. Row-stochastic: each non-empty row sums to 1.0. Dead rows (no out-edges) have zero values in the CSR and are handled by teleporting their mass in ppr.

has_outgoing: Vec<bool>

True for rows that had at least one out-edge. Dead rows redistribute their mass to the personalization vector each iteration.

Trait Implementations

impl Clone for SparseTransition

fn clone(&self) -> SparseTransition

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for SparseTransition

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

Formats the value using the given formatter.