Sorry, we seem to have misunderstood each other. Basically I have a
county-wide set of roads called "major roads". Then originally I had an
identically sized county-wide set of polygons of which I select (via R)
a smaller than county-wide subset of adjacent polygons and plot that
subset.
The "add = TRUE" in plot operations suggestion sounds interesting.
Your last paragraph is indeed all I am trying to achieve. I need a GIS
program to do this? I thought that R might be able to clip the major
roads file (via an intersect) to conform to the polygon shapes.
I don't know, maybe I am approaching this the wrong way (i.e. trying to
find the SQL style intersection of these two blobs). Perhaps another
approach may be better. I think I will do the following.
1. Plot the major roads and try to get the street names on them. No
polygon here.
2. Overlay what's going to be the large major roads over the smaller
polygon set. The "add = TRUE" suggestion. Thanks.
3. Figure out how to trim that major roads set.
I have your book "Applied Spatial Data Analysis with R (Use R)" on order
and it should arrive later today. Then also I am looking forward to a
good read of "Lattice Multivariate Visualization".
Thanks,
Jim Burke
Roger Bivand wrote:
On Mon, 16 Feb 2009, Jim Burke wrote:
(Just stick to plain text, but don't use fonts or bold or stuff - keep
it simple).
I have a nice solid SpatialPologonsDataFrame and a new line file
(wkbLineString) called "major roads.shp".
MY QUESTIONS.
*I would appreciate examples of*
*a)How to merge the intersection of a **SpatialPologonsDataFrame and
a wkbLineString **
Why, and what do you mean? You have some 30000 road lines, what do you
want to know? How much of which road is in which polygon? Again, you
need to think through your workflow. My inclination would be to
rasterise the lines in a GIS, and overlay the raster cell centre
points on the polygons, but maybe you want line length, or to retain
the line IDs (all 30000 of them).
b)How to plot the result
You can of course overplot lines on polygons, with add=TRUE in plot()
methods and sp.layout= in spplot() methods. But plotting the result of
the "intersection" means what?
c)How to display the street name on the plot (that's FNAME from the
major roads.dbf file).
30000 names? Do you have a very large screen? Lines objects have IDs,
but no label point, so there is no easy way of doing it, even label
alignment is hard.
I can't actually see the statistical question here, could you please
make it clearer?
Do you simply want to plot a subset of the road lines on your solid
polygons, adding names? Then you need to find out how to subset them,
perhaps by cookie-cutting using the union of your polygons, add label
points and rotation angles (probably by hand), and off you go, but
these are essentially GIS operations.
Roger
*
ADDITIONAL INFORMATION.
The SpatialPologonsDataFrame is a garden variety collection of
geographical polygons.
tx1_sp <- readShapePoly("precinct08.shp", IDvar="PCT",
proj4string=CRS("+proj=aea +ellps=GRS80 +datum=WGS84"))
The new file to intersect merge is major roads line map that should
overlay part of my SpatialPolygons. It does not behave like a Shapfile.
main_roads <-readOGR("main roads.shp", layer = "main roads")
OGR data source with driver: ESRI Shapefile
Source: "main roads.shp", layer: "main roads"
with 29161 rows and 33 columns
Feature type: wkbLineString with 2 dimensions
Thanks,
Jim Burke
_______________________________________________
R-sig-Geo mailing list
R-sig-Geo@stat.math.ethz.ch
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
_______________________________________________
R-sig-Geo mailing list
R-sig-Geo@stat.math.ethz.ch
https://stat.ethz.ch/mailman/listinfo/r-sig-geo