acompmargin {compositions}

Marginal compositions in Aitchison Compositions

Description

Compute marginal compositions of selected parts, by computing the rest as the geometric mean of the non-selected parts.

Usage

          acompmargin(X,d=c(1,2),name="*",pos=length(d)+1,what="data")
          

Arguments

X

composition or dataset of compositions

d

vector containing the indices xor names of the columns selected

name

The new name of the amalgamation column

pos

The position where the new amalgamation column should be stored. This defaults to the last column.

what

The role of X either "data" for data (or means) to be transformed or "var" for (acomp-clr)-variances to be transformed.

Details

The amalgamation column is simply computed by taking the geometric mean of the non-selected components. This is consistent with the acomp approach and gives clear ternary diagrams. However, this geometric mean is difficult to interpret.

Value

A closed compositions with class "acomp" containing the variables given by d and the the amalgamation column.

Missing Policy

MNAR has the highest priority, MAR afterwards, and WZERO (BDL,SZ) values are considered as 0 and finally reported as BDL.

Author(s)

Raimon Tolosana-Delgado, K.Gerald v.d. Boogaart http://www.stat.boogaart.de

References

Vera Pawlowsky-Glahn (2003) personal communication. Universitat de Girona.

van den Boogaart, K.G. and R. Tolosana-Delgado (2008) "compositions": a unified R package to analyze Compositional Data, Computers & Geosciences, 34 (4), pages 320-338, doi: 10.1016/j.cageo.2006.11.017.

See Also

rcompmargin, acomp

Examples

data(SimulatedAmounts)
plot.acomp(sa.lognormals5,margin="acomp")
plot.acomp(acompmargin(sa.lognormals5,c("Pb","Zn")))
plot.acomp(acompmargin(sa.lognormals5,c(1,2)))