Weights based on a function and 'listw' object

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Weights based on a function and 'listw' object

Virgilio Gómez Rubio
Hi,

I am using a 'listw' object (see package spdep) to store weights to
compute differentbstatistics, and I wonder whether it is possible to
define weights as, say, the inverse of the distance (or any other
function based on the centroids, for example).

Futhermore, is it efficient to use one of these objects when
all the areas are linked one another? Is it better to use a square nxn
matrix in this case?

Thank you for your help.

--
             Virgilio G?mez Rubio

Grup d'Estad?stica espacial i temporal
en Epidemiologia i medi ambient

Dpto. Estad?stica e I. O. - Facultat de Matem?tiques
Avda. Vicent A. Estell?s, 1 - 46100 Burjassot
Valencia - SPAIN

http://matheron.uv.es/~virgil

TLF: 00 34 96 354 43 62 - FAX: 00 34 96 354 47 35



Reply | Threaded
Open this post in threaded view
|

Weights based on a function and 'listw' object

Roger Bivand
Administrator
On Thu, 12 Feb 2004, Virgilio G?mez Rubio wrote:

> Hi,
>
> I am using a 'listw' object (see package spdep) to store weights to
> compute differentbstatistics, and I wonder whether it is possible to
> define weights as, say, the inverse of the distance (or any other
> function based on the centroids, for example).

See the example for nbdists(), and especially for nb2listw(), at the
bottom, where the output from nbdists() is fed into nb2listw() through the
glist= argument.

>
> Futhermore, is it efficient to use one of these objects when
> all the areas are linked one another? Is it better to use a square nxn
> matrix in this case?
>

Maybe, but you do lose the functions requiring a listw object, and you
need to define the weighting style yourself (for example row
standardised). In any case, weights at greater distances are very close to
zero, so either distance bands or k-nearest neighbours will give very
similar estimates (prove me wrong!!).

Roger

> Thank you for your help.
>
>

--
Roger Bivand
Econonic Geography Section, Department of Economics, Norwegian School of
Economics and Business Administration, Breiviksveien 40, N-5045 Bergen,
Norway, voice: +47-55959355, fax: +47-55959393; Roger.Bivand at nhh.no



Roger Bivand
Department of Economics
Norwegian School of Economics
Helleveien 30
N-5045 Bergen, Norway