Create Sparse Adjacency Matrix from Edge List
Usage
adj_from_edgelist(from, to, weight = NULL, n = NULL, symmetric = TRUE)
Arguments
- from
Integer or character vector of source nodes
- to
Integer or character vector of target nodes
- weight
Numeric vector of edge weights
- n
Number of nodes
- symmetric
Make adjacency symmetric
Value
A sparse COO torch_tensor of dimension n x n, carrying the
attributes node_ids (the original node identifiers, or NULL when
from and to are already integers) and id_map (a named integer
vector mapping identifiers to row and column positions, or NULL).
Examples
if (FALSE) { # torch::torch_is_installed()
adj_from_edgelist(from = c(1, 2, 3), to = c(2, 3, 1))
# Character node identifiers are mapped to positions
adj <- adj_from_edgelist(
from = c("a", "b", "c"),
to = c("b", "c", "a"),
weight = c(0.5, 1, 2)
)
attr(adj, "id_map")
}