EMMAgeo-package {EMMAgeo}

End-member modelling algorithm and supporting functions for unmixing grain-size distributions and further compositional data.

Description

EMMAgeo provides a set of functions for end-member modelling analysis (EMMA) of grain-size data and other cases of compositional data. EMMA describes a multivariate data set of m samples, each comprising n parameters (e.g. grain-size classes), as a linear combination of end-member loadings (the underlying distributions) and end-member scores (the contribution of each loading to each sample).
EMMA can be run in two principal ways, a deterministic and a robust, including modelling the uncertainties. The deterministic way can be accessed simply with the function EMMA(). For the robust way there are two protocols that need to be respected. There is a compact protocol, which is mainly automated but needs adjustments by the user, and there is an extended protocol, which allows access to all parameterisation steps of robust EMMA.
The package contains further auxiliary functions to check and prepare input data, test parameters and use a graphic user interface for deterministic EMMA. The package also contains an example data set, comprising meaured grain-size distributions of real world sediment end-members.

Details

Package:	EMMAgeo
Type:	Package
Version:	0.9.7
Date:	2019-05-10
License:	GPL-3

Author(s)

Michael Dietze, Elisabeth Dietze