Allocate redundant shapes or patterns for confusable colors — sc_redundant

Finds label pairs whose worst CIE2000 distance across normal and requested CVD simulations is below min_cie2000, then assigns auxiliary encoding groups with a deterministic DSATUR heuristic so every such pair differs. Colors are returned unchanged. Shapes are ggplot2-compatible integers. Default pattern names are compatible with ggpattern; texture_group can also be used as a grouping variable for other texture systems such as scatterHatch.

Usage

sc_redundant_encoding(
  x,
  channel = c("shape", "pattern", "both"),
  min_cie2000 = 8,
  cvd = c("deutan", "protan", "tritan"),
  shapes = c(16L, 17L, 15L, 18L, 8L, 3L),
  patterns = c("none", "stripe", "crosshatch", "circle", "wave", "weave")
)

Arguments

x: An sc_color_map or fully named color vector.
channel: Allocate "shape", "pattern", or joint shape-pattern combinations with "both". A conflicting pair differs in at least one enabled aesthetic.
min_cie2000: Pairs below this worst-vision CIE2000 distance must receive different auxiliary encodings.
cvd: Vision simulations used with normal vision.
shapes: Unique integer plotting shapes.
patterns: Unique non-empty pattern names.

Value

A data frame containing label, unchanged color, encoding group, texture group, shape, pattern, and number of confusable neighbors, with sc_encoding settings and conflicts records as attributes.

Details

Allocation is deterministic for one full label set, but graph priorities can change when labels are added or removed. Allocate once for the full label universe and reuse or subset the returned table across figures. The sc_encoding and conflicts attributes record the method, settings, pools, and conflicting pairs; save the object as RDS if those attributes must persist, because ordinary data-frame CSV export does not retain them.

Examples

colors <- c(A = "#0072B2", B = "#0072B2", C = "#D55E00")
sc_redundant_encoding(colors, channel = "shape")
#>   label   color encoding_group texture_group shape pattern conflict_count
#> 1     A #0072B2              1             1    16    none              1
#> 2     B #0072B2              2             2    17    none              1
#> 3     C #D55E00              1             1    16    none              0