Transformations

to_edge_index(adjacency_matrix, *[, dtype])	Converts an adjacency matrix to an edge index representation.
to_sparse_adjacency_matrix(adjacency_matrix, *)	Converts an adjacency matrix to a sparse representation as a tuple of edge index and edge features.
to_adjacency_matrix(edge_index[, ...])	Converts an edge index representation to an adjacency matrix.
get_src_dst_features(edge_index, node_features)	Extracts source and destination node features based on the given edge indices.
get_unique_edge_indices(edge_index_1, ...)	Computes the indices of the edges in edge_index_1 that are NOT present in edge_index_2
add_self_loops(edge_index[, edge_features, ...])	Adds self-loops to the given graph represented by edge_index and edge_features.
remove_self_loops(edge_index[, edge_features])	Removes self-loops from the given graph represented by edge_index and edge_features.
to_undirected(edge_index[, edge_features])	Converts a graph given as edge_index and, optionally, edge_features to an undirected one.
remove_duplicate_directed_edges(edge_index)	Removes the duplicate directed edges in edge_index.
coalesce(edge_index)	Removes duplicate edges in edge_index and scatters the duplicate edges to the edge attributes following a given reduction operation (e.g. sum of duplicate edges).
get_isolated_nodes_mask(edge_index, num_nodes)	Returns a mask with isolated nodes set to False to filter them out if needed.
has_isolated_nodes(edge_index, num_nodes)	Function to check for isolated nodes.
has_self_loops(edge_index)	Function to check for self loops.