rfUtilities_1.0-1
=================

* Fixed bug, in functions requiring randomForest class object, where stop error is issued when 
    model is specified using formula interface.
	
* Added new function "rf.partial.ci" for Random Forests regression partial dependency plot with 
    confidence intervals.
	
* Added new function "rf.class.sensitivity" for sensitivity test of class to cross-assignment. 

rfUtilities_1.0-2
=================

* Fixed bug, in rf.modelSel where no model was returned when final=TRUE.

* Added new function  rf.imp.freq that prints/plots frequency of variable selection 
    across classes.

rfUtilities_2.0-0
=================

* Major restructuring of the package, adding object classes along with print, plot and  
    summary functions for classes.
	
* Added new function "accuracy" for a variety of classification accuracy statistics.  
    Function is also called from rf.crossValidation. 
	
* Added new function "occurrence.threshold" for testing the sensitivity of probability 
     thresholds.
	 
* Added new function "rf.effect.size" for calculating "parameter" effect size on  
    classification or regression models using the Cohen's f2 statistic. 
	
* Added new function "rf.regression.fit" providing fit statistics for random forests 
    regression models. 
	
* Rewrote and added class attribute to rf.significance function along with print, 
    summary and plot generics.
	
* Rewrote and added class attribute to rf.crossValidation function along with print, 
    summary and plot generics.
	
* Added users accuracy, producers accuracy and kappa to rf.crossValidation output.

rfUtilities_2.0-1
=================
* Bug fix - Thanks to Patrick Fekety for pointing out that the seed argument in rf.modelSel was not being honoured,
              This fix will stabilize stochasticity in the model selection.

rfUtilities_2.1-0
=================
* The rf.crossValidation now supports regression 
* Major changes to the multi.collinear function. Now performs column order and leave-one-out permutation test.
* Print and plot generics support change in rf.crossValidation
* rf.modelSel print generic now returns additional test statistics (test threshold, OOB, class error and number of parameters) and associated, sorted parameters 

rfUtilities_2.1-1
=================
* Fixed bug in rf.crossValidation where the rmse for regression was being reported incorrectly (thanks to Patrick Fekety for finding error)
* Added Mean Absolute Error (mae) and Mean Bias Error (mbe) to validation statistics in rf.crossValidation for regression
* Added additional information to print and summary for rf.crossValidation
* Added additional print capacity for "mae" and "mbe" statistics in rf.crossValidation

rfUtilities_2.1-2
=================
* Added new function "log.loss" for evaluation of estimate quality in binomial models using cross-entropy or log likelihood loss 
* Added new function "probability.calibration" for performing an isotonic regression calibration of posterior probabilities
* Added new function "rf.effectSize" for deriving parameter effect size based on partial dependency (Cafri & Bailey, 2016)

rfUtilities_2.1-3
=================
* Added new function "bivariate.partialDependence" for deriving bivariate partial dependence providing a graphical depiction of the marginal effect of two variables on the class probability (classification) or response (regression)
* Added new function "rf.combine" that is an extension of the randomForest::combine function, providing model fit statistics (with associated print and summary methods)
* Added optional confidence intervals to "rf.partial.prob" function. 
* Added new function "rf.unsupervised" That performs an unsupervised classification

rfUtilities_2.1-4
=================
* Fixed a bug in rf.classBalance so that y in the model object reflects the input y
* Fixed a bug in rf.crossValidation that threw sample size error when class imbalance data

rfUtilities_2.1-5
=================
* Fixed a bug in rf.crossValidation for sample size in regression
* Function enhancement in rf.crossValidation, added KS test for regression