read.gwt2nb {spdep} | R Documentation |
The "gwt" functions read and write GeoDa GWT files (the example file baltk4.GWT was downloaded from the site given in the reference), and the "dat" functions read and write Matlab sparse matrix files as used by James LeSage's Spatial Econometrics Toolbox (the example file wmat.dat was downloaded from the site given in the reference).
read.gwt2nb(file, region.id=NULL) write.sn2gwt(sn, file, shpfile=NULL, ind=NULL) read.dat2listw(file) write.sn2dat(sn, file)
file |
name of file with weights data |
region.id |
region IDs |
sn |
a spatial.neighbour object |
shpfile |
Shapefile name taken from GWT file for this dataset |
ind |
region id indicator variable name |
Now attempts to honour the region.id argument given when reading GWT files.
read.gwt2nb
returns a neighbour "nb" object with the generalised weights stored as a list element called "dlist" of the "GeoDa" attribute.
Roger Bivand Roger.Bivand@nhh.no
Luc Anselin (2003) GeoDa 0.9 User's Guide, pp. 80–81, Spatial Analysis Laboratory, Department of Agricultural and Consumer Economics, University of Illinois, Urbana-Champaign, http://agec221.agecon.uiuc.edu/csiss/pdf/geoda093.pdf; also http://spatial-econometrics.com/data/contents.html
data(baltimore) STATION <- baltimore$STATION gwt1 <- read.gwt2nb(system.file("etc/weights/baltk4.GWT", package="spdep")[1], STATION) cat(paste("Neighbours list symmetry;", is.symmetric.nb(gwt1, FALSE, TRUE), "\n")) listw1 <- nb2listw(gwt1, style="B", glist=attr(gwt1, "GeoDa")$dist) tmpGWT <- tempfile() write.sn2gwt(listw2sn(listw1), tmpGWT) gwt2 <- read.gwt2nb(tmpGWT, STATION) cat(paste("Neighbours list symmetry;", is.symmetric.nb(gwt2, FALSE, TRUE), "\n")) data(oldcol) diffnb(gwt1, gwt2) tmpMAT <- tempfile() COL.W <- nb2listw(COL.nb) write.sn2dat(listw2sn(COL.W), tmpMAT) listwmat1 <- read.dat2listw(tmpMAT) diffnb(listwmat1$neighbours, COL.nb, verbose=TRUE) listwmat2 <- read.dat2listw(system.file("etc/weights/wmat.dat", package="spdep")[1]) diffnb(listwmat1$neighbours, listwmat2$neighbours, verbose=TRUE)