Todo

New argument focus for Different Singular Value Partitionings, including GH, JK, SQRT, HJ.

New function ggbiplot() function using ggplot2 graphics to draw the biplot.

Average Environment Coordinate

Bootstrap testing for PCs (Forkman 2019 paper)

Bootstrap conf int

gge 1.9 (2024.10.28)

gge 1.8 (2023-08-20)

gge 1.7 (2021-10-31)

gge 1.6 (2020-12-16)

Brian Ripley wrote: “The future of OpenGL is uncertain (except on macOS, where it has no future). So it seems reasonable to require rgl only when essential to the package. These packages have it in Depends/Imports but seem not to actually call it in their checks (established using a fake install). It is possible that the sole purpose of the package might be to do interactive visualizations which are not checked, but that seems not to be the case here. We noticed calls to rgl functions in , but they would better be conditioned by if(interactive()) (see ‘Writing R Extensions’). Please move rgl to Suggests and use conditionally (see §1.1.3.1 of ‘Writing R Extensions’) at the next package update.””

gge 1.5 (2020-07-21)

gge 1.4 (2018-05-15)

gge 1.3 (2017-12-14)

gge 1.2 - (2017-05-26)

gge 1.1 (2016-10-08)

gge 1.0 (2015-12-14)

gge 0.1 - (2013-01-01)

gge 0.0 - (2004-01-01)

A history of NIPALS functions in gge

    1. Created nipals() based on pcaMethods::nipalsPca(). Modified the function for faster execution and submitted a patch back to pcaMethods.
    1. Henning Redestig created a C++ version of NIPALS for the pcaMethods package.
    1. The gge::nipals() R function is re-named rnipals(), and a new nipals() function is created, based on the C++ code in pcaMethods. Released gge version 1.2.
    1. Discovered that mixOmics::nipals() is a pure R function that is faster than the C++ version, so gge::nipals() was re-written into a pure R function. The C++ version was removed from the gge package.
    1. The gge::nipals function is moved to a new package, nipals::nipals. The function is optimized for performance, improved to better handle missing values and to orthogonalize the principal components.