ABSTRACT
Logratios, i.e. log-transformed ratios, are the key to converting compositional data to scales that are additive and subcompositionally coherent. There are several variants of the simple logratio transformation that are worth considering for their practical and theoretical properties. In this chapter the additive logratio transformation is first considered, which gives a specific set of simple logratios. The set of centred logratios is mainly useful for facilitating several computations. The use of amalgamations in ratios is also considered, since amalgamations of parts are often used by practitioners of compositional data analysis. Finally, isometric logratios are treated, showing that they have interesting theoretical properties but are problematic in their substantive interpretation as statistical variables in practice.
