[R-sig-Geo] Spatial Lags Excluding Neighbors' Missing Attribute Values

[Roger Bivand]

On Mon, 5 Jan 2009, Christopher Moore wrote:

> Dear Roger,
>
> I am glad to hear that the NAOK argument will be implemented in the next 
> release. Will it be implemented in functions besides lag.listw, such as 
> moran.plot?

The draft release 0.4-32 is on R-Forge, you can try it by:

install.packages("spdep",repos="http://R-Forge.R-project.org")

with the usual precautions. I ought to revisit several functions - some 
take an NAOK or similar argument, others do not. In fact, for analysis 
they should do something more like what moran.test() does, using 
na.action= as an argument, but usage isn't consistent. Probably this needs 
to be factored out as a helper function that can be plugged into the test 
and model functions. I'd be interested in some (off-list) indications of 
whether NA handling matters and why, with some example data sets to add to 
spdep if possible.

For na.omit, moran.test() subsets the variable and the weights object 
internally to remove NA observations.

Roger

>
> I agree that "missing data is missing, really," which motivated me to develop 
> an ad hoc alternative to treating NA as zero, one that uses all available 
> information in a manner similar to pairwise deletion. I can not make a case 
> for including that approach in the spdep package (which is great, by the 
> way). However, I do think the approach is worthy of further methodological 
> inquiry, perhaps as a complement to multiple imputation on the attribute 
> side.
>
> Thanks,
> Chris
>
> P.S. My apologies for double posting to the list. I mistakenly used an 
> uninformative subject line in the first posting.
>
>

-- 
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