Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
st: sppack [was: SSPACK]
From
[email protected]
To
[email protected]
Subject
st: sppack [was: SSPACK]
Date
Thu, 31 Mar 2011 13:21:55 -0500
Ashley Arnio <[email protected]> asked about creating k-nearest-neighbor (KNN)
spatial-weighting matrices:
> Are there any plans to include a k-nn matrix in the near future?
Ashley can use -spmat idistance- with the undocumented option -knn()- to create
a KNN spatial-weighting matrix. I will illustrate with the ancillary
pollute.dta dataset that comes with the sppack suite.
. use pollute, clear
. spmat idistance dobj longitude latitude, id(id) dfunction(dhaversine) knn(5)
-spmat idistance ..., knn(5)- creates a spatial-weighting matrix in which the
five nearest neighbors of each spatial unit are weighted by their inverse
distances and all others recieive zero weights. We can use -spmat summarize-
with the option -links- to verify that each spatial unit has indeed five
neighbors.
. spmat summarize dobj
We did not document the -knn()- option because the command is slow for large
problems. We plan to address this issue in the future.
--Rafal
[email protected]
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/