Changes in cvTools version 0.3.2

    + Bugfix in cvTool(): matrix of estimated prediction errors is now returned 
      correctly for more than one replication (this bug was introduced in the 
      bugfix for one replication from version 0.3.1).
    
    + Added tests for cvTool(), cvFit() and cvTuning() using package 'testthat'.


Changes in cvTools version 0.3.1

    + Bugfix in cvTool(): matrix of estimated prediction errors is now returned 
      correctly for one replication.
    
    + Bugfix in cvTuning(): "function" method now correctly stores matched call.


Changes in cvTools version 0.3.0

    + Changed some argument and component names regarding standard errors.



Changes in cvTools version 0.2.0

    + New generic function repCV() for (repeated) K-fold cross-validation.  The 
      example functions cvLm(), cvLmrob() and cvLts() are now aliases for the 
      respective methods.
    
    + Bugfix in cvFit(): default names are retained if cvTool() returns a 
      1x1 matrix.
    
    + Bugfix in plot.cvSelect(): default handling of argument 'method' no 
      longer throws error in case of only one replication.
    
    + Added functionality to incorporate standard errors.
    
    + cvSelect() and cvTuning() now allow to specify criterion for selecting 
      the best model via the argument 'selectBest'.
    
    + New xyplot() and dotplot() methods for class "cv".



Changes in cvTools version 0.1.1

    + Added different types of obtaining CV folds.
    