aggregate: aggregation of spatial objects

aggregateR Documentation

aggregation of spatial objects


spatial aggregation of thematic information in spatial objects


## S3 method for class 'Spatial'
aggregate(x, by = list(ID = rep(1, length(x))),
	FUN, ..., dissolve = TRUE, areaWeighted = FALSE)



object deriving from Spatial, with attributes


aggregation predicate; if by is a Spatial object, the geometry by which attributes in x are aggregated; if by is a list, aggregation by attribute(s), see


aggregation function, e.g. mean; see details


arguments passed on to function FUN, unless minDimension is specified, which is passed on to function over


logical; should, when aggregating based on attributes, the resulting geometries be dissolved? Note that if x has class SpatialPointsDataFrame, this returns an object of class SpatialMultiPointsDataFrame; deprecated


logical; should the aggregation of x be weighted by the areas it intersects with each feature of by? See value; deprecated.


For as far as these functions use package rgeos, (lines, polygons, dissolve = TRUE), they are deprecated as rgeos will retire; try using sf::aggregate instead.

FUN should be a function that takes as first argument a vector, and that returns a single number. The canonical examples are mean and sum. Counting features is obtained when summing an attribute variable that has the value 1 everywhere.


The aggregation of attribute values of x either over the geometry of by by using over for spatial matching, or by attribute values, using aggregation function FUN.

If areaWeighted is TRUE, FUN is ignored and the area weighted mean is computed for numerical variables, or if all attributes are factors, the area dominant factor level (area mode) is returned. This computes the intersection of x and by; see examples below. As this uses code from package rgeos, it is deprecated as package rgeos will retire.

If by is missing, aggregates over all features.


uses over to find spatial match if by is a Spatial object


Edzer Pebesma,


coordinates(meuse) <- ~x+y
coordinates(meuse.grid) <- ~x+y
gridded(meuse.grid) <- TRUE
i = cut(meuse.grid$dist, c(0,.25,.5,.75,1), include.lowest = TRUE)
j = sample(1:2, 3103,replace=TRUE)
## Not run: 
if (require(rgeos)) {
	# aggregation by spatial object:
	ab = gUnaryUnion(as(meuse.grid, "SpatialPolygons"), meuse.grid$part.a)
	x = aggregate(meuse["zinc"], ab, mean)
	# aggregation of multiple variables
	x = aggregate(meuse[c("zinc", "copper")], ab, mean)
	# aggregation by attribute, then dissolve to polygon:
	x = aggregate(meuse.grid["dist"], list(i=i), mean)
	x = aggregate(meuse.grid["dist"], list(i=i,j=j), mean)
	spplot(x["dist"], col.regions=bpy.colors())
	spplot(x["i"], col.regions=bpy.colors(4))
	spplot(x["j"], col.regions=bpy.colors())

## End(Not run)

x = aggregate(meuse.grid["dist"], list(i=i,j=j), mean, dissolve = FALSE)
spplot(x["j"], col.regions=bpy.colors())

if (require(gstat) && require(rgeos)) {
	x = idw(log(zinc)~1, meuse, meuse.grid, debug.level=0)[1]
	i = cut(x$var1.pred, seq(4, 7.5, by=.5), 
		include.lowest = TRUE)
	# xa = aggregate(x["var1.pred"], list(i=i), mean)
	# spplot(xa[1],col.regions=bpy.colors(8))

if (require(rgeos)) {
# Area-weighted example, using two partly overlapping grids:

  gt1 = SpatialGrid(GridTopology(c(0,0), c(1,1), c(4,4)))
  gt2 = SpatialGrid(GridTopology(c(-1.25,-1.25), c(1,1), c(4,4)))

  # convert both to polygons; give p1 attributes to aggregate
  p1 = SpatialPolygonsDataFrame(as(gt1, "SpatialPolygons"), 
		  data.frame(v = 1:16, w=5:20, x=factor(1:16)), match.ID = FALSE)
  p2 = as(gt2, "SpatialPolygons")

  # plot the scene:
  plot(p1, xlim = c(-2,4), ylim = c(-2,4))
  plot(p2, add = TRUE, border = 'red')
  i = gIntersection(p1, p2, byid = TRUE)
  plot(i, add=TRUE, density = 5, col = 'blue')
  # plot IDs p2:
  ids.p2 = sapply(p2@polygons, function(x) slot(x, name = "ID"))
  text(coordinates(p2), ids.p2)
  # plot IDs i:
  ids.i = sapply(i@polygons, function(x) slot(x, name = "ID"))
  text(coordinates(i), ids.i, cex = .8, col = 'blue')

  # compute & plot area-weighted average; will warn for the factor
  #ret = aggregate(p1, p2, areaWeighted = TRUE)

  # all-factor attributes: compute area-dominant factor level:
  #ret = aggregate(p1["x"], p2, areaWeighted = TRUE) 

sp documentation built on July 9, 2023, 7:39 p.m.