Convert Node Features to Tensor

Usage

nodes_to_tensor(nodes, adj = NULL, node_id = NULL)

Arguments

nodes: Dataframe with node features
adj: Adjacency matrix with id_map attribute
node_id: Column name for node IDs

Value

A dense torch_tensor of dimension n_nodes x n_features. When adj and node_id are both supplied, rows are reordered to match the node ordering of adj and the identifier column is dropped.

Examples

if (FALSE) { # torch::torch_is_installed()
adj <- adj_from_edgelist(from = c("a", "b", "c"), to = c("b", "c", "a"))
nodes <- data.frame(id = c("c", "a", "b"), deg = c(2, 2, 2), pop = c(10, 20, 30))

nodes_to_tensor(nodes, adj, node_id = "id")
}