Extract Specific Motif Instances (Subgraphs)

Convenience wrapper for motifs(x, named_nodes = TRUE, ...). Returns one row per concrete node-triple and MAN type. At individual level, observed counts sessions/units exhibiting that combination, so one triple can occupy multiple rows when its type differs across units. The same MAN type can also appear in many rows, each with its own z / p. For per-triple significance use plot(., type = "significance") or plot(., type = "triads"); the per-type plots ("types", "patterns") deliberately drop the significance decoration here, because aggregating per type requires a rule (median? max-|z|?) that isn't pinned and would be misleading by default.

Usage

subgraphs(...)

Arguments

...

Arguments forwarded to motifs(). See ?motifs for the full parameter list (x, actor, window, window_type, pattern, include, exclude, significance, n_perm, cores, min_count, edge_method, edge_threshold, min_transitions, top, seed). named_nodes is fixed to TRUE and must not be supplied.

Value

A cograph_motif_result object with named_nodes = TRUE. Contains $results (data frame with columns triad, node1, node2, node3, observed, type, and when significance = TRUE also expected, z, p, sig), $type_summary, $level, $n_units, and $params. At individual level, each result row is a node-triple and MAN-type combination, and observed counts sessions/units exhibiting it. In instance mode, $type_summary is built via table(results$type) so it counts how many node-triples fall under each MAN type.

Details

The "triads" diagram uses a canonical representative of the row's MAN isomorphism class. Concrete labels identify the participating nodes; their positions in that representative diagram do not encode the nodes' observed source/sink roles.

See also

motifs()

Other motifs: extract_motifs(), extract_triads(), get_edge_list(), motif_census(), motifs(), plot.cograph_motif_analysis(), plot.cograph_motifs(), triad_census()

Examples

mat <- matrix(c(0,3,2,0, 0,0,5,1, 0,0,0,4, 2,0,0,0), 4, 4, byrow = TRUE)
rownames(mat) <- colnames(mat) <- c("Plan","Execute","Monitor","Adapt")
subgraphs(mat, significance = FALSE)
#> Showing triangle patterns (count >= 5). For all MAN types use pattern = 'all'.
#> Motif Subgraphs 
#> Level: aggregate | States: 4 | Pattern: triangle 
#> Min count: >= 5 
#> 
#> Type distribution:
#> 
#> 030C 030T 
#>    2    2 
#> 
#> Top 4 results:
#>                      triad   node1   node2   node3 type observed
#>   Plan - Execute - Monitor    Plan Execute Monitor 030T       10
#>  Execute - Monitor - Adapt Execute Monitor   Adapt 030T       10
#>     Plan - Monitor - Adapt    Plan Monitor   Adapt 030C        8
#>     Plan - Execute - Adapt    Plan Execute   Adapt 030C        6