This package provides a number of useful functions that facilitate
the analysis of spatially autocorrelated data based on the eigenfunction
decomposition of an exogenously given connectivity matrix
* W*. The main function

`getEVs()`

specifies a projection matrix `getEVs()`

also provides the Moran coefficient associated
with each eigenvector.Subsequently, these eigenvectors can be used to perform
semiparametric spatial filtering in a regression framework. The
functions `lmFilter()`

and `glmFilter()`

in this
package implement unsupervised eigenvector selection based on a stepwise
regression procedure and different objective functions, including i) the
maximization of model fit, ii) minimization of residual autocorrelation,
iii) the statistical significance of residual autocorrelation, and iv)
the statistical significance of candidate eigenvectors. If other
selection criteria are required or a model needs to be fitted that is
currently not supported by these two functions, supervised eigenvector
selection can be performed as well using the basic `stats::lm()`

and `stats::glm()`

commands in conjunction with the output generated by
`getEVs()`

. This option provides additional flexibility.