Combine several feature geometries into one, without unioning or resolving internal boundaries
st_combine(x)st_union(x, y, ..., by_feature = FALSE)
object of class sf
, sfc
or sfg
object of class sf
, sfc
or sfg
(optional)
ignored
logical; if TRUE, union each feature, if FALSE return a single feature that is the geometric union of the set of features
st_combine
returns a single, combined geometry, with no resolved boundaries; returned geometries may well be invalid.
If y
is missing, st_union(x)
returns a single geometry with resolved boundaries, else the geometries for all unioned pairs of x[i] and y[j].
st_combine
combines geometries without resolving borders, using c.sfg (analogous to c for ordinary vectors).
If st_union
is called with a single argument, x
, (with y
missing) and by_feature
is FALSE
all geometries are unioned together and an sfg
or single-geometry sfc
object is returned. If by_feature
is TRUE
each feature geometry is unioned. This can for instance be used to resolve internal boundaries after polygons were combined using st_combine
. If y
is provided, all elements of x
and y
are unioned, pairwise (and by_feature
is ignored). The former corresponds to gUnaryUnion, the latter to gUnion.
Unioning a set of overlapping polygons has the effect of merging the areas (i.e. the same effect as iteratively unioning all individual polygons together). Unioning a set of LineStrings has the effect of fully noding and dissolving the input linework. In this context "fully noded" means that there will be a node or endpoint in the output for every endpoint or line segment crossing in the input. "Dissolved" means that any duplicate (e.g. coincident) line segments or portions of line segments will be reduced to a single line segment in the output. Unioning a set of Points has the effect of merging all identical points (producing a set with no duplicates).
# NOT RUN {
nc = st_read(system.file("shape/nc.shp", package="sf"))
st_combine(nc)
plot(st_union(nc))
# }
Run the code above in your browser using DataLab