Skip to contents

Edge-color formulas in netplot

Source: R/coloring.R
netplot-formulae.Rd

Edge colors in both nplot() and set_edge_gpar() can be specified using a formula based on ego() and alter() (source and target). This way the user can set various types of combination varying the mixing of the colors, the alpha levels, and the actual mixing colors to create edge colors.

Usage

color_formula(x, col, alpha, env, type, mix = 1, postfix = NULL)

ego(...)

alter(...)

Arguments

x

An object of class netplot.

col

Any valid color. Can be a single color or a vector.

alpha

Number. Alpha levels

env, type, postfix

For internal use only.

mix

Number. For mixing colors between ego and alter

...

Passed to color_formula.

Value

Nothing. These functions are called internally when using formulas. color_formula modifies the environment env.

Details

The formula is one-sided and combines the special terms ego() (the source vertex) and alter() (the target vertex) with +. Each term draws its color from the corresponding endpoint's vertex.color unless a col is supplied, and accepts:

  • col: the base color to use for that endpoint.

  • alpha: transparency, from 0 (fully transparent) to 1 (opaque).

  • mix: how much weight that endpoint's color receives when the two are blended along the edge.

For instance, ~ ego(alpha = .1, col = "gray") + alter (the default) fades each edge from a faint gray at the source to the target's color, producing the characteristic netplot look.

This grammar is specific to edge colors. To map other aesthetics (vertex color, shape, size, or edge width) from a graph attribute, pass a formula naming the attribute directly to nplot() (e.g. vertex.color = ~ group); see the "Mapping attributes with formulas" section of nplot().

See also

nplot() for the attribute-mapping formulas, and set_edge_gpar().

Examples

if (require(gridExtra) & require(magrittr)) {
  library(igraph)
  net <- make_ring(4)

  set.seed(1)
  np <- nplot(net, vertex.color = grDevices::hcl.colors(4), vertex.size.range=c(.1, .1))
  np %<>% set_edge_gpar(lwd = 4)

  grid.arrange(
    np,
    np %>% set_edge_gpar(col =~ego + alter),
    np %>% set_edge_gpar(col =~ego(alpha=0) + alter),
    np %>% set_edge_gpar(col =~ego + alter(alpha=0)),
    np %>% set_edge_gpar(col =~ego(mix=0) + alter(mix=1)),
    np %>% set_edge_gpar(col =~ego(mix=1) + alter(mix=0))
  )
}
#> Loading required package: gridExtra