R help - Apr 2013 - Spatial Ananlysis: zero.policy=TRUE doesn't work for no neighbour regions??

Hello,

I'm new to R and to Spatial Analysis and got a problem trying to create a
Spatial Weights Matrix.

*I us the following code to create the Neighbourslist:*
>library(maptools)
>library(spdep)
>library(rgdal)
>location_County<- readShapePoly("....")
>proj4string(location_County)<- CRS("+proj=longlat ellps=WGS84")
>location_nbq<- poly2nb(location_County)
>summary(location_nbq)
*And get this Output:*

Neighbour list object:
Number of regions: 3109 
Number of nonzero links: 18246 
Percentage nonzero weights: 0.1887671 
Average number of links: 5.868768 
4 regions with no links:
35 689 709 881
Link number distribution:

   0    1    2    3    4    5    6    7    8    9   10   11   13   14 
   4   29   40   94  283  616 1045  703  228   51   12    2    1    1 
29 least connected regions:
45 49 587 645 844 853 1206 1286 1391 1416 1456 1478 1485 1545 1546 1548 1558
1612 1621 1663 1672 1675 1760 1794 1795 2924 2925 2952 3107 with 1 link
1 most connected region:
1385 with 14 links

*As there are some regions without neighbours in my data I use the following
code to create the Weights Matrix:*
> W_Matrix<- nb2listw(location_nbq, style="W", zero.policy=TRUE)
>W_Matrix
*And get this Output:*

Fehler in print.listw(list(style = "W", neighbours = list(c(23L, 31L,
42L :
  regions with no neighbours found, use zero.policy=TRUE

/("Error in print.listw(list(style = "W", neighbours =
list(c(23L, 31L, 42L
: 
  regions with no neighbours found, use zero.policy=TRUE")/

As I use "zero.policy=TRUE" I just don't understand what I'm
doing wrong...
My question would be: How could I create a Weights Matrix allowing for
no-neighbour areas? 

Thanks
Michael






--
View this message in context:
http://r.789695.n4.nabble.com/Spatial-Ananlysis-zero-policy-TRUE-doesn-t-work-for-no-neighbour-regions-tp4664367.html
Sent from the R help mailing list archive at Nabble.com.

Molo <kurz_m <at> uni-hohenheim.de> writes:
> 
...
> *As there are some regions without neighbours in my data I use the
following
> code to create the Weights Matrix:*
> 
> > W_Matrix<- nb2listw(location_nbq, style="W",
zero.policy=TRUE)
> >W_Matrix
> 
> *And get this Output:*
> 
...
> /("Error in print.listw(list(style = "W", neighbours =
list(c(23L, 31L, 42L
> : 
>   regions with no neighbours found, use zero.policy=TRUE")/
> 
> As I use "zero.policy=TRUE" I just don't understand what
I'm doing wrong...
> My question would be: How could I create a Weights Matrix allowing for
> no-neighbour areas? 
You have not grasped the fact that your object W_Matrix has been created
correctly, but that spdep:::print.listw also needs a zero.policy=TRUE, so:

print(W_Matrix, zero.policy=TRUE)

will work. If you want to set this globally for all subsequent function
calls in your current session, use set.ZeroPolicyOption(TRUE).

Hope this clarifies,

Roger

PS. Consider posting questions of this kind to R-sig-geo
> 
> Thanks
> Michael