Skip to contents

Detect communities and leaders on a graph, then plot them

Source: R/plot_leaders.R
lcda_plot_communities.Rd

The one-call path from a bare graph to the paper's community-and-leader figure: run one of the LCDA algorithms, take the partition and the leader it designates for each community, and render them with [plot_partition()].

Usage

lcda_plot_communities(
  graph,
  method = c("gr", "grasp", "ecg"),
  args = list(),
  seed = NA_integer_,
  plot = TRUE,
  verbose = FALSE,
  ...
)

Arguments

graph

an [igraph::igraph] object (undirected, simple).

method

which algorithm to run: `"gr"` ([lcda_gr()], the reactive variant and the default), `"grasp"` ([lcda_grasp()]), or `"ecg"` ([lcda_ecg()], the ensemble consensus, which also yields node confidences).

args

a named list of extra arguments for the chosen algorithm, e.g. `list(B = 100, variant = 2)`.

seed

integer RNG seed for the run, or `NA` to leave the RNG untouched. Also seeds the layout, so the figure is reproducible.

plot

logical; set to `FALSE` to fit and return without drawing.

verbose

logical; passed to the algorithm.

...

further arguments passed to [plot_partition()] (`layout`, `legend`, `mark_communities`, ...).

Value

invisibly, the fitted result object (an `lcda_gr_result`, `lcda_grasp_result`, or `lcda_ecg_result`), with the layout used attached as the attribute `"lcda_layout"`. Feed it straight to [lcda_metrics()].

See also

[lcda_metrics()], [plot_partition()].

Examples

g <- igraph::make_graph("Zachary")
res <- lcda_plot_communities(g, method = "grasp", args = list(B = 20), seed = 1)

lcda_metrics(res, level = "leader")
#> # A tibble: 3 × 14
#>   algorithm  community leader leader_name community_size degree degree_within
#>   <chr>          <int>  <int> <chr>                <int>  <dbl>         <dbl>
#> 1 LCDA-GRASP         1     34 34                      17     17            14
#> 2 LCDA-GRASP         2      1 1                       12     16            10
#> 3 LCDA-GRASP         3      6 6                        5      4             3
#> # ℹ 7 more variables: degree_between <dbl>, eigen_centrality <dbl>,
#> #   nce_node <dbl>, participation <dbl>, degree_rank_in_community <int>,
#> #   degree_pctile_in_community <dbl>, source <chr>