Struct graph_csr::Graph

pub struct Graph<'a, N> { /* private fields */ }

This structure holds a graph in the Compressed Sparse Row format for compression of data size. This graph is represented via Memory Mapping, allowing the graph to be loaded into memory as required. This makes it possible to load any-size graphs, even those that do not fit into memory!

Implementations

impl<'a, N> Graph<'a, N> where
    N: ValidGraphType,
    N: 'a, 

pub fn from_txt_adjacency_list<T>(
    stream: T,
    folder_name: &str
) -> Result<Graph<'a, N>, Error> where
    T: Read + Sized, 

Convenience method for reading an input stream in text format. Each line should contain two numbers, separated by a space. The graph will be converted to the underlying CSR representation, and stored in folder_name.

pub fn from_binary_adjancency<T>(
    stream: T,
    destination_folder_name: &str
) -> Result<Graph<'a, N>, Error> where
    T: Read + Sized, 

Same as from_txt_adjacency, except this time it assumes the edge list to be in binary representation.

pub fn from_adjacency_list<T>(
    stream: T,
    folder_name: &str
) -> Result<Graph<'a, N>, Error> where
    T: Iterator<Item = Result<(N, N)>> + Sized, 

Given a SORTED (by source) adjancency list file source_file_name, transforms this file into the underlying binary representation in CSR and returns a version of the Graph in this format. The graph will be stored in folder_name.

pub fn load_graph(graph_folder: &str) -> Result<Graph<'a, N>, Error>

Loads a graph from the underlying representation and returns it as a Graph struct.

pub fn iter(&'a self) -> GraphIterator<'a, N>

Returns an iterator over the edge list of each node.

pub fn n_nodes(&self) -> usize

Returns the number of nodes existing in the graph

pub fn n_edges(&self) -> usize

Returns the number of edges existing in the graph