On Mon, 01 Mar 2010 12:25:20 -0000 (GMT) ted.hard...@manchester.ac.uk
<ted.hard...@manchester.ac.uk> wrote:
> > A similar type of overloading is used in the 'sp' class functions,
> > where you can basically treat a 'SpatialPointsDataFrame', a
> > 'SpatialLinesDataFrame' or a 'SpatialPolygonsDataFrame' as a data
> > frame,
> > with '$colname' indexing and '[' subsetting, even though the
> > *internals*
> > of the objects have a completely different (and very complex)
> > structure.
> > But as a convenience to the user, you don't need to bother with the
> > internals, and can handle the object *as if* it were a data frame. It's
> > a very comfortable way of working.
>
> I'm not sure that "SpatialPointsDataFrame" is a dataframe (despite
> the name)! Is it not simply a list? In which case, using "$" is
> what you have to do to get at its components.
That it's not a data frame is the point. :-)
And it not simply a list, it's a S4 object with the data (frame) stored
in a 'data' slot, and '$' overloaded so you can use it *as if* it was a
data frame. Example:
library(sp)
example("SpatialPolygonsDataFrame-class")
# Internal structure (warning: not pretty!)
str(ex_1.7$x)
# Extracting columns from the data frame
ex_1.7$z
# Both 'nrow' and '[' is overloaded, so you can use '['
# for normal subsetting. For example, to plot 10 random
# polygons, you can type
ex.sub=ex_1.7[sample(nrow(ex_1.7), 10), ]
plot(ex.sub)
In most cases you don't have to worry about how everything is stored
internally; things just work like you expect them to.
--
Karl Ove Hufthammer
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
