apply_edges()

void MayaFlux::Kinesis::Stochastic::Weights::apply_edges ( std::span< const std::pair< size_t, size_t > >  edges, double  weight = 1.0, bool  symmetric = true, bool  normalize = true  )

inlinenoexcept

Write an adjacency edge list into this table.

Parameters

edges	Index pairs, the same shape Kinesis::EdgeList is (Spatial::ProximityGraphs' sequential_chain(), k_nearest_neighbors(), minimum_spanning_tree(), and so on all return one)
weight	Value written at each edge
symmetric	When true (the default), also writes the reverse direction for every edge, matching an undirected graph. Some EdgeList producers (k_nearest_neighbors, nearest_neighbor_graph) are already directed; pass false for those.
normalize	When true (the default), every row touched by an edge is normalized afterward, so the result is directly usable as a transition table.

Note

An index outside this table's bounds is skipped rather than trusted, since edges typically comes from an external computation over a possibly different point count.

Definition at line 224 of file Weights.hpp.

    {
        for (const auto& [a, b] : edges) {
            if (a < m_rows && b < m_cols) {
                (*this)(a, b) = weight;
            }
            if (symmetric && b < m_rows && a < m_cols) {
                (*this)(b, a) = weight;
            }
        }
        if (normalize) {
            for (size_t i = 0; i < m_rows; ++i) {
                normalize_row(i);
            }
        }
    }

References a, b, m_cols, m_rows, normalize_row(), and weight.

Here is the call graph for this function: