Changes in version 1.0.5                        

    o   reduce reliance on bestglm as it appears to be no longer maintained

                 Changes in version 1.0.3 (2019-05-14)                  

    o   Fix for plot.bglmnet with min.prob in mplot

    o   Resolve warning messages around exporting objects

                 Changes in version 1.0.2 (2019-01-22)                  

    o   Compatibility update with a new version of dplyr.

                        Changes in version 1.0.0                        

    o   Version 1.0.0 release to coincide with appearance of
	Journal of Statistical Software article.

    o   Documentation updates.

    o   Examples simplified so that run time of each is <5 seconds

                 Changes in version 0.8.2 (2018-01-05)                  

    o   The bglmnet function has been been re-written

    o   Added boot_size plot option for bglmnet objects

    o   Vignettes removed, replaced with pkgdown website

                 Changes in version 0.8.1 (2017-11-18)                  

    o   More sensible nbest parameter behaviour in the vis function. 
	Now nbest = "all" reverts back to nbest = 1 for models with 
	more than 15 parameters.  This can be overridden by specifying 
	an integer, rather than specifying "all".  For large model sizes, 
	if the number specified is too large, it will lead to memory 
	overruns and failures with dependent packages.

    o   Fixed vignette/pkgdown issue where there were multiple 
	vignettes created with the same name.

                 Changes in version 0.8.0 (2017-11-06)                  

    o   Random seed parameter for reproducible results

    o   New website

    o   Small fixes for JSS article (e.g. googleVis axis padding)

                 Changes in version 0.7.8 (2016-08-16)                  

    o   reimplemented classic plots in ggplot2

    o   classic plots are now default (reviewer feedback)

    o   improved documentation
	  - more detail about procedures
	  - more detail about plotting methods

    o   fixed the passing of weights for weighted models

    o   can now obtain full loss v dimension which="lvk"
	plots for glms

    o   vis() default is now nbest="all".

    o   experimental support for glmulti as a backend
	instead of bestglm for glms for the vis function.
	This enables users to enforce marginality constraints.

                        Changes in version 0.7.6                        

    o   maintenance, larger update coming soon

                 Changes in version 0.7.5 (2015-11-16)                  

    o   added fev and wallabies dataset

                        Changes in version 0.7.4                        

    o   added tag option to googleVis plots to facilitate easier
	plotting in rmarkdown documents.

    o   citation updated to reference arXiv article

    o   bootstrapping glmnet (bglmnet) function refined and
	added to the mplot() shiny interface

                 Changes in version 0.7.1 (2015-09-09)                  

    o   Fix for undefined globals (CRAN submission)

                        Changes in version 0.7.0                        

    o   Release to coincide with JSS article

    o   Implements "all" for vis function

    o   Improved blmnet plots

    o   Consistency with interactive plotting methods

    o   Improved documentation

                        Changes in version 0.6.5                        

    o   Numerous refinements including consistency of 
	style between the classic plots

    o   ylim argument for classic boot and lvk plots

    o   legend.position argument for classic af plots

                        Changes in version 0.6.0                        

    o   First CRAN release

    o   Reimplemented vis function to 
	avoid massive memory use for moderate model
	sizes.  Now runs much faster and leaner.

                        Changes in version 0.5.5                        

    o   mplot interface now uses shinydashboard

    o   the scatterplot matrix from the mplot shiny
	interface has now been spun off into its 
	own package: parisD3

    o   issue with zooming on transparent reported to 
	GoogleCharts - the workaround is to not use
	backgroundColor = 'transparent' until it is fixed
	at the source

                        Changes in version 0.5.0                        

    o   Limited robustness via screening.

    o   Weights now get passed through in the adaptive fence.

                        Changes in version 0.4.9                        

    o   Changed parallel backend for af() from doMC
	to doParallel which should work for both
	unix-like systems and windows.

    o   Added redundant variable to vis().

    o   Fixed issue with deparse(model.formula)
	when the model.formula was too long for
	deparse to cope with.

                        Changes in version 0.4.7                        

    o   New data sets: diabetes and artificial example

                        Changes in version 0.4.6                        

    o   First public version