Abstract initialization of your code

[clbustos]

On many classes, the initialize method is always the same

          def initialize
            super
            set_options(self.class.options) if self.class.options 
            self.class.data.setClassIndex(self.class.class_index) if self.class.class_index
            buildClassifier(self.class.data)
          end

You should create a base module, like RubyMining, with a method that run this procedure
Example:

        module RubyMining
        def init_classifier
            super
            set_options(self.class.options) if self.class.options 
            self.class.data.setClassIndex(self.class.class_index) if self.class.class_index
            buildClassifier(self.class.data)
         end
         end
         class Base < NaiveBayes
         include ::RubyMining
         def initialize
         super
          init_classifier
         end
         end

[clbustos]

You could use metaprogramming, to create automagicly classes...

module Weka::Classifier::Trees
      java_import 'weka.classifiers.trees.J48'
      java_import 'weka.classifiers.trees.RandomForest'
      create_classifier :FT
      create_classifier :J48
end

[rjpbonnal]

Good point.
I would dedicate a section of the second part of GSoC to refactoring.
@alberto Claudio is right, first finish the implementation planned, then
start a strong refactoring.

I think the students should notice how tedious is re writing the same thing
all the times :)

On Fri, Jul 26, 2013 at 8:58 AM, Claudio Bustos notifications@github.comwrote:

You could use metaprogramming, to create automagicly classes...

module Weka::Classifier::Trees
java_import 'weka.classifiers.trees.J48'
java_import 'weka.classifiers.trees.RandomForest'
create_classifier :FT
create_classifier :J48end

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/6#issuecomment-21604638
.

Ra