Struct MessagePassingNet 

pub struct MessagePassingNet {
    pub layers: Vec<GnnLayer>,
}

A multi-layer message passing neural network stacking GnnLayers.

Fields

layers: Vec<GnnLayer>

Ordered list of GNN layers.

Implementations

impl MessagePassingNet

pub fn new() -> Self

Create an empty MPNN.

pub fn add_layer(&mut self, layer: GnnLayer)

Add a GNN layer to the stack.

pub fn forward( &self, node_feats: &[f64], n_nodes: usize, adj: &[Vec<usize>], ) -> Vec<f64>

Run all layers in sequence over a fixed graph.

Returns the final node feature matrix [n_nodes × last_out_dim].

pub fn global_mean_pool( &self, node_feats: &[f64], n_nodes: usize, out_dim: usize, ) -> Vec<f64>

Aggregate node features to a single graph-level representation (mean pooling).

Trait Implementations

impl Clone for MessagePassingNet

fn clone(&self) -> MessagePassingNet

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for MessagePassingNet

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

Formats the value using the given formatter.

impl Default for MessagePassingNet

fn default() -> Self

Returns the “default value” for a type.