RELEASE HISTORY OF THE "whitening" PACKAGE
1.4.0
- depends now R >= 3.5.0
- new data set pitprops14
- new function explainedVariation()
- reverted back to a single Cholesky whitening procedure (as Chol-cov and Chol-prec are really
the same, just reversed ordering). Default is now lower triangular loadings.
- whiteningCrossCov() depricated.
- Instead whiteningLoadings() returns both loadings Phi and cross-correlation loadings Psi
with the columns corresponding to the latent variables
- simOrtho() has positiveDiag option
- whiten() and other functions now label the latent variables (L1, L2, etc)
- whitening matrix and cross-cov matrix now also labelled (both rows and columns)
- whitening method used now attached as method attribute
Version 1.3.0:
- update URLs
- new function simOrtho() to simulate random orthogonal matrices.
Version 1.2.0:
- new whitening options "Chol.cov" and "Chol.prec" added.
- new function whiteningCrossCov() to compute the cross-covariance matrix
for a given whitening transformation.
Version 1.1.1:
- References to Kessy et al (2018) and Jendoubi and Strimmer (2019) updated.
- whiten() now has "center" option.
- New data set "forina1986" added (extended version of UCI wine data set)
Version 1.1.0:
- The functions scca() and cca() now also compute the
loadings for X and Y.
- The data sets "lusc" and "nutrimouse" are now included.
- Updated nutrimouse data analysis script, and added R code
for example analysis of lusc data.
- Two functions loadplot() and corplot() have been added.
Version 1.0.0:
- Public release on 3 March 2018.
This package contains functions for computing whitening matrices,
for whitening data, and for (shrinkage) estimation of canonical
correlations and directions.