marmap (version 1.0.10)

space.pies: Automatic placement of piecharts on maps

Description

Attemps to automatically place piecharts on maps, avoiding overlap. Work in progress...

Usage

space.pies(x, y, pie.slices, pie.colors=NULL, pie.radius=1, pie.space=5,
              link=TRUE, seg.lwd=1, seg.col=1, seg.lty=1, coord=NULL)

Value

Piechart(s) added to a plot.

Arguments

x

the longitude of the anchor point for the piechart

y

the latitude of the anchor point for the piechart

pie.slices

a table with the counts to draw pies (col: pie categories, or slices; rows: sites on the map)

pie.colors

a table with the colors to draw pies (col: pie categories, or slices; rows: sites on the map)

pie.radius

size of the piechart

pie.space

factor of spacing between the anchor and the pie (the larger, the farther the pie from the anchor)

link

logical; whether to add a segment to link pie and anchor

seg.lwd

the line width of the link

seg.col

the line color of the link

seg.lty

the line type of the link

coord

when coord = NULL (default), placement is automatic. Otherwise, a 2-col table of lon/lat for pies.

Author

Eric Pante, using functions plotrix::floating.pie and maptools::pointLabel.

Details

space.pies tries to position piecharts on a map while avoiding overlap between them. The function heavily relies on two other functions. floating.pie from package plotrix is used to draw individual piecharts. floating.pie treats one pie at a time; space.pies can handle one or multiple pies by looping floating.pie. pointLabels from package maptools was modified to find the best placement for the pies, given their size and distance from their anchor point. pointLabels was originally meant to automatically place text labels, not objects; the modified version contained in space.pies uses the coordinates chosen by pointLabels for text. The algorithm used is simulating annealing (SANN). You can get a different result each time you run space.pies, because pointLabel finds one good solution out of many. If you are not satisfied by the solution, you can try running the function again.

The argument coord allows to choose between the automatic placement outlined above, and a user-defined list of longitudes and latitudes (in a two-column table format) for plotting the piecharts.

Anchor point: spatial location of the data corresponding to the piechart (e.g. a sampling point).

References

Bivand, R. and Lewin-Koh, N. (2013) maptools: Tools for reading and handling spatial objects. R package version 0.8-25. http://CRAN.R-project.org/package=maptools

Lemon, J. (2006) Plotrix: a package in the red light district of R. R-News, 6(4): 8-12.

SANN code implemented in pointLabel based on: 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.

See Also

plot.bathy, plotrix::floating.pie, maptools::pointLabel

Examples

# fake frequencies to feed to space.pies()
	sample(seq(10,90,5), 11)-> freq.a
	100-freq.a -> freq.b
	rep("lightblue",11) -> col.a
	rep("white",11) -> col.b

# some coordinates on the NW Atlantic coast, and on seamounts
	x = c(-74.28487,-73.92323,-73.80753,-72.51728,-71.12418,
		  -69.81176,-69.90715,-70.43201,-70.17135,-69.43912,-65.49608)
	y = c(39.36714,39.98515,40.40316,40.79654,41.49872,41.62076,
		  41.99805,42.68061,43.40714,43.81499,43.36471)
	pts.coast = data.frame(x,y, freq.a, freq.b, col.a, col.b)

	x = c(-66.01404,-65.47260,-63.75456,-63.26082,-62.12838,
	      -60.46885,-59.96952,-56.90925,-52.20397,-51.32288,-50.72461)
	y = c(39.70769,39.39064,38.83020,38.56479,38.01881,38.95405,
	      37.55675,34.62617,36.15592,36.38992,35.91779)
	pts.smt = data.frame(x,y, freq.a, freq.b, col.a, col.b)

# prepare the plot
	data(nw.atlantic) ; atl <- as.bathy(nw.atlantic)
	plot(atl, deep=-8000, shallow=0, step=1000,col="grey")
	points(pts.coast,pch=19,col="blue", cex=0.5)
	points(pts.smt,pch=19,col="blue", cex=0.5)

# automatic placement of piecharts with space.pies
	space.pies(pts.coast[,1], pts.coast[,2],
	           pie.slices=pts.coast[,3:4], pie.colors=pts.coast[,5:6], pie.radius=0.5)
	space.pies(pts.smt[,1], pts.smt[,2],
	           pie.slices=pts.smt[,3:4], pie.colors=pts.coast[,5:6], pie.radius=0.5)