Up | Previous | Table of Contents | Next | Version Française
![[R 3.0 icon]](../../../../../images/r3-icon-t.gif){kind=icon}
What does AUTOCORRELATION do ?
Program AUTOCORRELATION analyzes the spatial autocorrelation
structure of a variable by producing structure functions, called
correlograms, following various schemes of connections or distances
among the data locations. This is a strictly univariate method; program
MANTEL
makes it possible to produce a correlogram from multivariate
data. For quantitative data, autocorrelation is measured using Moran's
I and Geary's c indices. For ordinal or nominal data,
standard normal deviates (S.N.D.'s) are computed for each
distance class. Each value is accompanied by the probability of it not
being significantly different from zero (one-tailed test). How to
interpret correlograms is discussed by
Legendre & Fortin (1989).
In the Macintosh version, the production of that list of link edges
has been put in a different program called
LINKS. Finally,
AUTOCORRELATION may also produce a file containing an upper triangular
matrix of distance classes among localities; that file, whose default
name is CLASSEF, is required by program
MANTEL to compute a
multivariate correlogram.
Up | Previous | Table of Contents | Next | Version Française