logo       

Re: Regression analysis: how to convert shapefiles to listw-objects?: msg#00002

lang.r.geo

Subject: Re: Regression analysis: how to convert shapefiles to listw-objects?

On Tue, 2 May 2006, Anja Matatko wrote:

> Roger and list,
>
> thanks for your help with importing my shapefiles.
>
> As I don't want to use only the lm() command for regression analysis but
> also more complex ones as lm.morantest() and others, it seems that I
> need a listw-object instead of my shapefile.
>
> What I think I should do is:
> - create polylist or Spatial Polygons
> - use poly2nb()
> - use nb2listw ()
>
> My results:
> > abo_object_neu <- readShapePoly ("abo_poly",
> proj4string=CRS(as.character (NA)), verbose=FALSE)
> > summary(abo_object_neu)
> Object of class SpatialPolygonsDataFrame
> Coordinates:
> min max
> r1 2508089 2644341
> r2 5442170 5640160
> Is projected: NA
> proj4string : [NA]
> Data attributes: (several attributes, I deleted them here)
>
>
> I thought "SpatialPolygonsDataFrame" should be one of these
> "SpatialPolygon"-objects to serve as input for poly2nb, but trying this
> function, the answer was:
>
> > abo_nb <- poly2nb(abo_object_neu, row.names=NULL,
> snap=sqrt(.Machine$double.eps), queen=TRUE)
> Error in poly2nb(abo_object_neu, row.names = NULL, snap =
> sqrt(.Machine$double.eps), :
> Not a polygon list

I'll correct this in the next release, you need to cast the object
explicitly:

abo_nb <- poly2nb(as(abo_object_neu, "SpatialPolygons"))

should do the trick.

Roger

>
>
> My question: what is going wrong, are there other functions to do this
> conversion from shapefile to listw-object? Is there a direct shapefile
> to listw-conversion-function? I really tried to find it in help files,
> but had no success.
>
> Thanks, Anja
>
> > sessionInfo()
> Version 2.3.0 (2006-04-24)
> i386-pc-mingw32
>
> attached base packages:
> [1] "methods" "stats" "graphics" "grDevices" "utils"
> "datasets" "base"
>
> other attached packages:
> spdep boot SparseM maptools sp foreign tripack PK
> "0.3-22" "1.2-24" "0.68" "0.5-11" "0.8-14" "0.8-15" "1.2-10" "0.03"
>
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo@xxxxxxxxxxxxxxxxx
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>

--
Roger Bivand
Economic Geography Section, Department of Economics, Norwegian School of
Economics and Business Administration, Helleveien 30, N-5045 Bergen,
Norway. voice: +47 55 95 93 55; fax +47 55 95 95 43
e-mail: Roger.Bivand@xxxxxx


<Prev in Thread] Current Thread [Next in Thread>
Google Custom Search

News | FAQ | advertise