Thank you Roger, your professional knowledge has rebuilt my confidence in
spdep.
As to the example, I have a new understanding: when a 'cold spots' was put
in the center, the surrounding grids became 'hot spots' relatively, it
explained the result when distance was set to 1 unit. The idea of 'high' and
'low' in global G is comparative, not measured by mean value.
I will take the example to CrimeStat.
Merry Chrismas everyone!
--
View this message in context: http://r-sig-geo.2731867.n2.nabble.com/about-global-getis-g-tp7587571p7587579.html
Sent from the R-sig-geo mailing list archive at Nabble.com.