Provides (weighted) Partial least squares Regression for generalized linear models and repeated k-fold cross-validation of such models using various criteria. It allows for missing data in the explanatory variables. Bootstrap confidence intervals constructions are also available.
This file documents software changes.
1.2.3 Complete S3 methods to pass CRAN checks. 1.2.2 Sped up examples to pass CRAN checks. 1.2.1 Updated code to pass CRAN checks. 1.2.0 Added naive AIC and BIC output for data with missing values. 1.1.1 donttest update requested by CRAN. 1.1.0 Verbose option (request from Max Kuhn, caret package). 1.0.1 Fix for sPLS like option. 1.0.0 Major upgrade to enhance generics support. Package vignette (130 p.) release. Help of package completed. 0.9.0 Tilted bootstrap and demos fixed. 0.8.3 Code upgrade requested by R Core Team due to change in namespace policies. 0.8.2 Improvement of access to T-np or model based Bootstrap. 0.8.1 Graphical assessment of variable selection. Selection of bootstap techniques for CI plots. 0.8.0 Improved bootstrap stability. 0.7.9 polr demo fixed. 0.7.8 Minor code enhancements. 0.7.7 Minor code enhancements. 0.7.6 Demos fixed. 0.7.5 Minor code enhancements. 0.7.4 Minor code enhancements. 0.7.3 Minor code fixes. 0.7.2 Cross-validation fully compatible with weighted plsR or plsRglm models. 0.7.1 Minor code changes. 0.7.0 Estimation of Degrees of Freedom for plsR models. 0.6.7 Weighted plsR models. 0.6.5 Sparse option to set to 0 the coefficients of non-significant predictors and to select the number of components. 0.6.3 Minor fixes to cope with upcoming changes in R 2.13. 0.6.2 Support for logistic, probit, cloglog and cauchit links for cumulative link models. 0.6.1 Support for various glm or polr options. 0.6.0 Support for formulas. 0.5.1 Minor code changes. 0.5.0 Support for R or user defined "family" definition of glms added. 0.4.0 Support for Gamma(link = "inverse"), inverse.gaussian(link = "1/mu^2") and poisson(link = "log") added. 0.3.4 Minor code fixes. 0.3.3 Boostrap functionnalities improvement. Bootstrap no longer stops if model fitting fails for some subsample. 0.2.0-0.3.2 Support for polr, crossvalidations and bootstraps. 0.1.2-0.1.5 Completing help files.