nsprcomp: Non-Negative and Sparse PCA

Two methods for performing a constrained principal component analysis (PCA), where non-negativity and/or sparsity constraints are enforced on the principal axes (PAs). The function 'nsprcomp' computes one principal component (PC) after the other. Each PA is optimized such that the corresponding PC has maximum additional variance not explained by the previous components. In contrast, the function 'nscumcomp' jointly computes all PCs such that the cumulative variance is maximal. Both functions have the same interface as the 'prcomp' function from the 'stats' package (plus some extra parameters), and both return the result of the analysis as an object of class 'nsprcomp', which inherits from 'prcomp'. See <https://sigg-iten.ch/learningbits/2013/05/27/nsprcomp-is-on-cran/> and Sigg et al. (2008) <doi:10.1145/1390156.1390277> for more details.

Getting started

Package details

AuthorChristian Sigg [aut, cre] (<https://orcid.org/0000-0003-1067-9224>), R Core team [ctb] (prcomp interface, formula implementation and documentation)
MaintainerChristian Sigg <christian@sigg-iten.ch>
LicenseGPL (>= 2)
URL https://sigg-iten.ch/research/
Package repositoryView on CRAN
Installation Install the latest version of this package by entering the following in R:

Try the nsprcomp package in your browser

Any scripts or data that you put into this service are public.

nsprcomp documentation built on May 1, 2019, 7:56 p.m.