Skip to contents

Graph visualization for clustered terms

Source: R/VizGraph.R
viz_graph.Rd

Graph visualization for clustered terms

Usage

viz_graph(
  similarity_matrix,
  clusters,
  plot_threshold = 0.5,
  plot_name = "ClusterGraph",
  max_nodes = NULL,
  min_degree = NULL,
  node_sizes = NULL,
  show_density = FALSE,
  seed = NULL,
  save_plot = "svg",
  print_plot = TRUE,
  path = NULL,
  plot_width = 3000,
  plot_height = 2000,
  plot_unit = "px"
)

Arguments

similarity_matrix

Square numeric matrix of term similarity values. Row/column names must match clusters names.

clusters

Named vector of cluster labels for each term (e.g., "cluster1").

plot_threshold

Similarity threshold used to define edges. Values below the threshold are removed. Default = 0.5.

plot_name

Optional: String added to output files of the plot. Default = "ClusterGraph".

max_nodes

Optional: Maximum nodes for plotting. If set, keeps nodes from the largest component up to this limit (by degree).

min_degree

Optional: Minimum degree filter for graph plotting.

node_sizes

Optional: Named numeric vector of node sizes, with names matching term IDs. Values are scaled for plotting.

show_density

Optional: If TRUE, add a hull background per cluster. Default = FALSE.

seed

Optional: Random seed for graph layout reproducibility. Default = NULL.

save_plot

Optional: Select the file type of output plots. Options are svg, pdf, png or NULL. Default = "svg"

print_plot

Optional: If TRUE prints an overview of resulting plots. Default = TRUE

path

Optional: String which is added to the resulting folder name. default: NULL

Value

Graph plot as a ggplot object.

Examples


# Load example data
d <- metsigdb_kegg()

# Run clustering with graph plotting
r <- cluster_pk(
    d,
    metadata_info = c(
        metabolite_column = "MetaboliteID",
        pathway_column = "term"
    ),
    input_format = "long",
    similarity = "jaccard",
    threshold = 0.2,
    clust = "community",
    min = 2,
    plot_name = "GraphExample_long_format",
    save_plot = "png",
    min_degree = 1,
    seed = 123,
    show_density = FALSE,
    max_nodes = 1000
) 
#> Warning: No shared levels found between `names(values)` of the manual scale and the
#> data's fill values.
#> Warning: ggrepel: 388 unlabeled data points (too many overlaps). Consider increasing max.overlaps

# run graph plotting separately from clustering output but without node_sizes
viz_graph(
    r$similarity_matrix,
    r$clusters,
    plot_threshold = 0.5,
    plot_name = "ClusterGraph",
    max_nodes = NULL,
    min_degree = NULL,
    node_sizes = NULL,
    show_density = TRUE,
    seed = 123,
    save_plot = NULL,
    print_plot = FALSE,
    path = NULL
)
#> Warning: ggrepel: 451 unlabeled data points (too many overlaps). Consider increasing max.overlaps