Re: st: Good spatial manual

[Gordon Hughes <G.A.Hughes@ed.ac.uk>]

The presentations listed in previous responses to this query are 
useful but they are certainly not manuals and they are far from comprehensive.

You can find a list of SSC spatial routines and related material via 
--net search spatial--.  In most cases the help files provide 
relevant references, but I do not think that there is anything 
approaching a manual or text on how carry out spatial analysis in 
Stata.  In part this is because the area is still developing quite 
rapidly, while many researchers have used other statistical 
environments for their work - Matlab, Gauss and R are heavily 
favoured.  You could consider using --spwmatrix-- and --splagvar-- 
followed by --ivreg2-- to estimate a variety of spatial models via 
GMM/IV, but there are alternative routines available - e.g. Maurizio 
Pisati's routines.

Two additional remarks:

1.  In most cases the routines are unable to cope with missing data, 
especially unbalanced panel data.  As this is a characteristic 
feature of many spatial datasets, you need to think how to deal with 
it and choose routines that are designed to cope with it.

2.  The number of spatial units (10s, 100s, 1000s, ...) is very 
important for the route that you take.

Finally, a general question for all readers.  The Drukker 
presentation refer to forthcoming stata user modules --gs2sls-- 
--spreg-- and --spivreg--.  Can anyone provide information on whether 
these are actually available and, if so, how one can obtain 
them?  There is a completely different routine (really in testing 
form) from William MacMillan at the University of Michigan under the 
name --spreg-- .

Gordon Hughes


===================

Professor Gordon Hughes
Department of Economics
University of Edinburgh

E-mail : G.A.Hughes@ed.ac.uk

*
*   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/