Tables of a hypergraph — hg_get.net

One row per hyperedge: its identifier, size (number of distinct member states), the member states, and its weight (for hypergraphs built by window_hypergraph(), the number of windows collapsed into the hyperedge; NA for the other constructors, whose hyperedges are unweighted).

Usage

# S3 method for class 'net_hg'
hg_get(
  x,
  what = c("edges", "nodes", "memberships", "sets", "state_counts", "edge_data"),
  ...,
  sort_by = NULL,
  top = NULL
)

Arguments

x: A net_hg object.
what: "edges" (default) for one row per hyperedge, "nodes" for one row per node, or "memberships" for one row per node-in- hyperedge cell of the incidence matrix. A hypergraph of clustered sequences (group_hypergraph() on a clustering of sequences) also has "sets" and "state_counts". "edge_data" returns the per-hyperedge attribute table a constructor attached.
...: Additional arguments (ignored).
sort_by: NULL (construction order, default), "weight", or "size" - sort the table by that column, largest first (ties broken by hyperedge id so the order is deterministic).
top: Integer or NULL. Return only the first top rows, applied after any filter and after sort_by, so sort_by and top compose. Default NULL returns every row.

Value

A data.frame. For what = "edges", one row per hyperedge with columns hyperedge (character id), size (integer), members (the member nodes, comma separated), and, for a hypergraph of windows (window_hypergraph()), weight (the number of windows with that member set). sort_by is "size", or "weight" where it exists. For what = "nodes", one row per node with columns node and degree (the number of hyperedges it belongs to), plus block for a hypergraph with planted blocks (random_hypergraph() with type = "sbm"); sort_by = "degree" orders it. For what = "memberships", one row per non-zero incidence cell with columns node, hyperedge and weight (the incidence value: 1 for a binary hypergraph, the summed weight for a weighted one), in hyperedge order; sort_by = "weight" orders it. For what = "sets", one row per hyperedge with group, hyperedge, set (the states joined by " + "), size, count (sequences of the group with exactly that set) and share (count over the group's sequences), in group order and decreasing count. For what = "state_counts", one row per group and state with group, node, count (sequences of the group containing the state) and share. For what = "edge_data", one row per hyperedge with edge and its attributes; asking for it of a hypergraph without an attribute table raises hypergraphs_bad_input.

Examples

hg <- window_hypergraph(list(s1 = c("a", "b", "a", "c")), window = 2L)
hg_get(hg)
#>   hyperedge size members weight
#> 1        h1    2    a, b      2
#> 2        h2    2    a, c      1
hg_get(hg, sort_by = "weight")
#>   hyperedge size members weight
#> 1        h1    2    a, b      2
#> 2        h2    2    a, c      1
hg_get(hg, sort_by = "weight", top = 3)
#>   hyperedge size members weight
#> 1        h1    2    a, b      2
#> 2        h2    2    a, c      1
hg_get(hg, what = "nodes", sort_by = "degree")
#>   node degree
#> 1    a      2
#> 2    b      1
#> 3    c      1
hg_get(hg, what = "memberships")
#>   node hyperedge weight
#> 1    a        h1      2
#> 2    b        h1      2
#> 3    a        h2      1
#> 4    c        h2      1