Linear color interpolation function with adjustable range, baseline, and floor

Source: R/colorjam-gradient.R

Linear color interpolation function with adjustable range, baseline, and floor

col_linear_xf(
  x = 1,
  floor = 0,
  baseline = 0,
  lens = 0,
  n = 6,
  colramp = "Purples",
  open_floor = FALSE,
  debug = FALSE,
  ...
)

Arguments

x

numeric value used as a threshold, where numeric values at or above this value x are assigned the last color in the color gradient. Negative values at or below this negative value -x are assigned the first color in the color gradient.

floor

numeric optional value where numeric values between -x and x are assigned the middle color in the color gradient. Note that values at exactly x or -x are assigned the next respective color away from the middle color. When floor=0 or floor=NULL no floor is applied, and colors are assigned using a continuous range of numeric values from -x to x with length n.

baseline

numeric value to define the baseline value, used when zero is not the initial value. Note that baseline can be either higher or lower than x, and colors from colramp will be applied starting at baseline through x.

lens

numeric value indicating a color lens applied to the color gradient, passed to jamba::getColorRamp(). Lens values lens > 0 will condense the color gradient, making smaller changes more visually distinct; lens < 0 expands the color gradient, making smaller changes less visually distinct.

n

integer number of colors used for the initial color gradient. This value is forced to be an odd number, so the "middle color" will always be represented as one strict color. Note that when using a floor, the first non-middle color is used for the floor assignment which means a smaller n value will assign a more visibly distinct color than using a larger n. See examples.

colramp

character passed to jamba::getColorRamp() which recognizes one of several forms of input:

  • character string matching the name of a color ramp from RColorBrewer (see divergent palettes with RColorBrewer::display.brewer.all(type="div")). Note that adding "_r" will reverse the color gradient, so the default "BuRd_r" will create a color gradient with "blue-white-red" - with red for high values consistent with "heat" in "heatmaps" - where heat is red.

  • character vector of R colors, which define a specific color ramp. This vector will be expanded to n length.

open_floor

logical indicating whether colors below the assigned floor will still receive non-middle color. Setting open_floor=TRUE is the best method to compare the effect of assigning the strict middle-color to values below the floor, versus using gradient colors below the floor, while all remaining numeric-color assignments are held constant.

debug

logical indicating whether to produce a plot that shows the resulting color gradient.

...

additional arguments are ignored.

Details

This function is the linear equivalent of col_div_xf(), in that it takes linear/sequential color gradient instead of a divergent color gradient.

See also

Other colorjam gradients: col_div_xf(), make_jam_divergent(), twostep_gradient()

Other colorjam assignment: col_div_xf(), group2colors(), matrix2heatColors(), rainbowJamMulti(), vals2colorLevels(), vibrant_color_by_hue()

Examples

col_fn1 <- col_linear_xf(x=3, baseline=0, floor=0)
col_fn2 <- col_linear_xf(x=3, baseline=0, floor=1)

col_fn2o <- col_linear_xf(x=3, baseline=0, floor=1, open_floor=TRUE)

test_seq <- seq(from=0, to=3, by=0.05);
names(test_seq) <- round(test_seq, digits=2);

opar <- par("mfrow"=c(1, 1));
bp0 <- barplot(abs(test_seq),
   las=2, yaxt="n",
   main="floor=0",
   col=col_fn1(test_seq),
   border="#22222222")
abline(v=bp0[abs(test_seq) == 1,], lty="dashed")

bp1 <- barplot(abs(test_seq),
   las=2, yaxt="n",
   main="floor=1",
   col=col_fn2(test_seq),
   border="#22222222")
abline(v=bp1[abs(test_seq) == 1,], lty="dashed")

bp2 <- barplot(abs(test_seq),
   las=2, yaxt="n",
   main="floor=1\nopen_floor=TRUE",
   col=col_fn2o(test_seq),
   border="#22222222")
abline(v=bp2[abs(test_seq) == 1,], lty="dashed")

par(opar)

col_fn3 <- col_linear_xf(x=3, baseline=6, floor=5)
test_seq <- seq(from=0, to=7, by=0.1);
names(test_seq) <- round(test_seq, digits=2);
bp3 <- barplot(abs(test_seq),
   las=2, yaxt="n",
   main="baseline=6, x=3, floor=5",
   col=col_fn3(test_seq),
   border="#22222222")
abline(v=bp3[abs(test_seq) == 5,], lty="dashed")
abline(v=bp3[abs(test_seq) == 3,], lty="dashed")