pointLabel(x, y = NULL, labels = seq(along = x), cex = 1, method = c("SANN", "GA"), allowSmallOverlap = FALSE, trace = FALSE, doPlot = TRUE, ...)
plot.default
, these provide the x and y coordinates for
the point labels. Any reasonable way of defining the coordinates is
acceptable. See the function xy.coords
for details. text
, a character vector or expression specifying the text to be
written. An attempt is made to coerce other language objects
(names and calls) to expressions, and vectors and other
classed objects to character vectors by as.character
. text
. TRUE
, labels are allowed
a small overlap. The overlap allowed is 2% of the diagonal
distance of the plot area.TRUE
, status updates are given as the optimization algorithms
progress.TRUE
, the labels are plotted on the
existing graph with text
.text
to specify
labeling parameters such as col
. xy
list giving the x
and y
positions of the
label as would be placed by text(xy, labels)
.
With the default settings, simulating annealing solves faster than the genetic algorithm. It is an open question as to which settles into a global optimum the best (both algorithms have parameters that may be tweaked).
The label positioning problem is NP-hard (nondeterministic
polynomial-time hard). Placement becomes difficult and slows
considerably with large numbers of points. This function places all
labels, whether overlaps occur or not. Some placement algorithms
remove labels that overlap.
Note that only cex
is used to calculate string width and
height (using strwidth
and strheight
), so passing a
different font may corrupt the label dimensions. You could get around
this by adjusting the font parameters with par
prior to running
this function.
http://en.wikipedia.org/wiki/Automatic_label_placement
http://i11www.iti.uni-karlsruhe.de/map-labeling/bibliography/
http://www.eecs.harvard.edu/~shieber/Projects/Carto/carto.html
http://www.szoraster.com/Cartography/PracticalExperience.htm
The genetic algorithm code was adapted from the python code at
http://meta.wikimedia.org/wiki/Map_generator.
The simulated annealing code follows the algorithm and guidelines in:
Jon Christensen, Joe Marks, and Stuart Shieber. Placing text labels on maps and diagrams. In Paul Heckbert, editor, Graphics Gems IV, pages 497-504. Academic Press, Boston, MA, 1994. http://www.eecs.harvard.edu/~shieber/Biblio/Papers/jc.label.pdf
text
, thigmophobe.labels
in package plotrix n <- 50
x <- rnorm(n)*10
y <- rnorm(n)*10
plot(x, y, col = "red", pch = 20)
pointLabel(x, y, as.character(round(x,5)), offset = 0, cex = .7)
plot(x, y, col = "red", pch = 20)
pointLabel(x, y, expression(over(alpha, beta[123])), offset = 0, cex = .8)
Run the code above in your browser using DataCamp Workspace