predict.genDiag: Predict method for generalised diagonalisation objects

In gmGeostats: Geostatistics for Compositional Analysis

Predict method for generalised diagonalisation objects

Description

Predict method for generalised diagonalisation objects

Usage

## S3 method for class 'genDiag'
predict(object, newdata = NULL, ...)

Arguments

object	a generalized diagonalisation object, as obtained from a call to Maf, and on the same page, information on the other diagonalisation methods UWEDGE or RJD
newdata	a matrix or data.frame of factor scores to convert back to the original scale (default: the scores element from object)
...	not used, kept for generic compatibility

Value

A data set or compositional object of the nature of the original data used for creating the genDiag object.

See Also

Other generalised Diagonalisations: Maf(), coloredBiplot.genDiag()

Examples

data("jura", package="gstat")
juracomp = compositions::acomp(jura.pred[, -(1:6)]) 
lrvg = logratioVariogram(data=juracomp, loc=jura.pred[,1:2])
mf = Maf(juracomp, vg=lrvg)
mf
biplot(mf)
predict(mf) 
unclass(predict(mf)) - unclass(juracomp) # predict recovers the original composition

gmGeostats documentation built on Aug. 21, 2025, 5:43 p.m.