plot graph with legend right of it
plot multivariate data into colour channels using geom_tile
normalize.colrange normalizes the range of each column to [0, 1]
normalize.range normalizes the range of all columns to [0, 1]
normalize.null does not touch the values
normalize.minmax normalizes the range of each column j to [min_j, max_j]
legends for mixed colour plots
legendright(p, l, legend.width = 8, legend.unit = "lines")qmixtile(object, purecol = stop("pure component colors needed."),
mapping = aes_string(x = "x", y = "y", fill = "spc"), ...,
map.tileonly = FALSE)
normalize.colrange(x, na.rm = TRUE, legend = FALSE, n = 100, ...)
normalize.range(x, na.rm = TRUE, legend = FALSE, n = 100, ...)
normalize.null(x, na.rm = TRUE, legend = FALSE, n = 100, ...)
normalize.minmax(x, min = 0, max = 1, legend = FALSE, n = 100, ...)
qmixlegend(x, purecol, dx = 0.33, ny = 100, labels = names(purecol),
normalize = normalize.colrange, ...)
colmix.rgb(x, purecol, against = 1, sub = TRUE,
normalize = normalize.colrange, ...)
plot object
legend object
size of legend part
matrix to be plotted with mixed colour channels
pure component colours, names determine legend labels
see geom_tile
qmixtile: handed to colmix.rgb
qmixlegend and colmix.rgb hand further arguments to the normalize function
matrix with component intensities in columns
see link[base]{min}
should a legend be produced instead of normalized values?
of colours to produce in legend
numeric with value corresponding to "lowest" colour for each column
numeric with value corresponding to "hightest" colour for each column
width of label bar
number of colours in legend
component names
function to normalize the values.
value to mix against (for sub = TRUE only, 1 = white, 0 = black)
subtractive color mixing?
invisible NULL
list with components ymin, max and fill to specify value and fill colour value (still numeric!) for the legend, otherwise the normalized values
ggplot object with legend
character with colours