Changes in version 0.2-9 (2020-10-19)                  

    o   editorial changes to Rd files

                 Changes in version 0.2-8 (2020-06-01)                  

    o   Warning: S4 exports gcv.enet specified in 'NAMESPACE' but not defined in package 'bujar' 

                 Changes in version 0.2-7 (2019-11-12)                  

    o   no change but resubmission

                        Changes in version 0.2-6                        

    o   no change but resubmission

                 Changes in version 0.2-5 (2019-02-27)                  

    o   changes are made in vignettes/dlbcl.Rnw since the current release of package gbm plot function ignores par(mfrow=()) option. Anyhow, grid.arrange simplies the codes when used with Sweave. 

                 Changes in version 0.2-4 (2019-01-15)                  

    o   drop bujar-package.Rd in man directory

                 Changes in version 0.2-3 (2017-04-27)                  

    o   minor changes to dlbcl.Rnw 

                 Changes in version 0.2-1 (2015-12-19)                  

    o   predict function is updated

    o   add data-raw for raw data processing R file

                       Changes in version 0.1-10                        

    o   mselect is an updated mstop in boosting for each BJ iteration. This was implemented for learner="tree" (with best.iter), now extends to learner="linear.regression" and "pspline".

    o   add rng for bujar function to be used for gbm internally.

    o   add name "Intercept" for twin=TRUE and learner="linear.regression"

                        Changes in version 0.1-9                        

    o   move boosting and penalized regression routine to a separate R file

    o   move prediction routine to a separate R file

                        Changes in version 0.1-8                        

    o   delete repeated computing pred.bj in bujar.R

    o   fix a bug with coefficients for learner="enet2", "mnet", "snet", due to the bugs in the mpath package 

                        Changes in version 0.1-7                        

    o   replace R/myglmboost.R with package bst for learner="linear.regression" and twin=TRUE

    o   update NAMESPACE

    o   add twin boosting for learner="pspline" and "tree"

    o   fix a bug of intercept estimate when center=TRUE and learner="linear.regression"

                        Changes in version 0.1-6                        

    o   in bujar function, intercept estimation fixed if argument center=TRUE

    o   twin boosting added using myglmboost.R and associated other functions

    o   fixed a bug with learner="pspline" importing bbs function from mboost library 

                 Changes in version 0.1-5 (2015-10-28)                  

    o   changes since 0.1-4: modify NAMESPACE and Imports to pass R CMD check

                        Changes in version 0.1-2                        

    o   learner = "acosso" is no longer supported. The function contain functions svd2 and La.svd2 which call entry points, which no longer exist in R-devel (3.0.0 to be) 

                        Changes in version 0.1-1                        

    o   added MCP and SCAD variable selection methods