supure: Package for spectral unmixing with pure variables approach
In svkucheryavski/supure: Spectral Unmixing with Pure Variables

Description Details Author(s) References

supure is an R package implementing methods for spectral unmixing (curve resolution) based on the pure variables approach.

supure is an R package implementing methods for spectral unmixing (curve resolution) based on the pure variables approach. It includes variants of SIMPLSIMA [1,2] and a simple implementation of MCR-ALS with angle conctraint [3,4]. The package supplements a chapter in a book "Resolving spectral mixtures, Data handling in Science and Technology", edited by Cyril Ruckebusch which will be published by Elsevier in 2016. It also contains all datasets used in the chapter and a vignette with some of the examples used there as well.

There are two main classes: purity for unmixing pure variable approach and its variations and mcrals for the angle constrained MCR-ALS. Each class has several methods for visualisation and tuning the results of unmixing.

There are four datasets, which can be loaded using command data(puredata):

`carbs`	generated Raman spectra of ribose, fructose and lactose mixtures.
`simdata`	generated UV/Vis spectra of polyaromatic hydrocarbons mixtures.
`mir`	Mid infrared spectra of ethanol and glucose.
`hsi`	Raman hyperspectral image of oil-in-water emulsion.

Each dataset is a list with following fields:

`D`	data matrix with spectra of mixtures.
`S`	optional, matrix with spectra of pure components.
`C`	optional, matrix with concentrations of pure components.
`wavelength`	wavelength or wavenumbers.