#run LDA
LDA<-lda(env,spa.group[,4])

#classification of the objects based on LDA
spe.class <- predict(LDA)$class

#posterior probabilities of the objects to belong to the groups
spe.post <- predict(LDA)$posterior

#table of prior versus predicted classifications
spe.table <- table(spa.group[,4], spe.class)

#proportion of correct classification
diag(prop.table(spe.table, 1))