[R-sig-Geo] sp.correlogram

[Roger Bivand]

On Sun, 26 Aug 2007, Kitty Lee wrote:

> Hi. I have a question about sp.correlogram in SPDEP.
>
> I created a weight matrix of 5 nearest neighbors. To play around with 
> the sp.correlogram function, I specified an order of 10 (which is the 
> max. lag order). The plot came out with spatial autocorrelation with 10 
> lags.
>
> How to understand the 10 spatial lags? I mean the weight matrix only has 
> information of the nearest 5 neighbors (i.e. the matrix only provides 
> information for the first 5 lags). How does the program know which are 
> the 6th-10th lag?
>
> Sorry for the odd question. Perhaps it's because I misunderstood the 
> concept of lag in the program. Can someone clarify this for me?

Your k=5 k-nearest neighbours are first order neighbours. You asked for 10 
orders, so you "stepped out": the second order neighbours are the first 
order neighbours of your first order neighbours (omitting observations 
already included), and so on. Have a look (in practice) at what nblag() 
does if this is unclear, and note that for many lattices, there will be 
relatively few or no remaining neighbours at high lags.

Hope this helps,

Roger

>
> Thanks!
>
> K.
>
>
>
>
>
>
>
> ---------------------------------
>
> 	[[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at stat.math.ethz.ch
> 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 at nhh.no