Basic function to apply the harvest algorithm to the training data set, computing whether we can harvest any nodes based on the classic classification tree algorithm.
Usage
harfunc(rpart.object, data, varname, sig = 0.95)
Arguments
rpart.object
classification result of training data from traditional classification tree(rpart function).
data
original training data where 'y' stores classmembership
varname
the name of each explaanatory variables
sig
significance level (default 0.95)
Value
the list of orginial result of classification, likelihood improvment and harvested classification result.