weka 0.1.0-java → 0.2.0-java

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: bedc1a7e9457ec42c3343b05a68e7e2598e9f0f9
-  data.tar.gz: ac3ade8d787979d9127d72daa6879b38cb58dfd4
+  metadata.gz: e635e174779d58d634091b4a7d0f523cd56e8548
+  data.tar.gz: 5695cfe2a3b173e8d5b96c4b9234a10da7d9205e
 SHA512:
-  metadata.gz: b5f1f0754f075545e460504e3a343b030bd130f27edd78c4ff312f96f134505df2a0f32a3a1b7461a1c588953764766c5d3acbf06acfa7d4ed76f44781a096a6
-  data.tar.gz: 1a457c4160b77baea4b7f471af5229062f48a791aae8f3a7a36dc901fb07b9582ffc3fdaf77a689598a411d054b698623ef583311b22df93c4f813d1b4419e7c
+  metadata.gz: 7e081b08900c2d77b9070ed4d963b4a5dcc9e2a52acc1bbc3f32cefd168abc1630f6a9f03331896be298b394bb222c5d258a9d8c11fe6e5d34251ca4697f5fbe
+  data.tar.gz: 56b23d5ba46094ebd5f4f74eb90cbaca74887c06726aea639616520dc13f6c910b2dc4edb2601f914d87c1a35b374a06fd6591a9d5b1791e793380261844bc27

data/README.md

@@ -23,6 +23,13 @@ Or install it yourself as:
 
 ## Usage
 
+* [Instances](#instances)
+* [Filters](#filters)
+* [Attribute selection](#attribute-selection)
+* [Classifiers](#classifiers)
+* [Clusterers](#clusterers)
+* [Serializing objects](#serializing-objects)
+
 Start using Weka's Machine Learning and Data Mining algorithms by requiring the gem:
 
 ```ruby
@@ -233,6 +240,9 @@ discretize = Discretize.new
 
 # apply a filter chain on instances
 filtered_data = instances.apply_filter(normalize).apply_filter(discretize)
+
+# or even shorter
+filtered_data = instances.apply_filters(normalize, discretize)
 ```
 
 #### Setting Filter options
@@ -655,6 +665,43 @@ sunny,80,90,TRUE,cluster1
 ...
 ```
 
+### Serializing Objects
+
+You can serialize objects with the `Weka::Core::SerializationHelper` class:
+
+```ruby
+# writing an Object to a file:
+Weka::Core::SerializationHelper.write('path/to/file.model', classifier)
+
+# load an Object from a serialized file:
+object = Weka::Core::SerializationHelper.read('path/to/file.model')
+```
+
+Instead of `.write` and `.read` you can also call the aliases `.serialize` and `.deserialize`.
+
+Serialization can be helpful if the training of e.g. a classifier model takes
+some minutes. Instead of running the whole training on instantiating a classifier you
+can speed up this process tremendously by serializing a classifier once it was trained and later load it from the file again.
+
+Classifiers, Clusterers, Instances and Filters also have a `#serialize` method
+which you can use to directly serialize an Instance of these, e.g. for a Classifier:
+
+```ruby
+instances  = Weka::Core::Instances.from_arff('weather.arff')
+instances.class_attribute = :play
+
+classifier = Weka::Core::Trees::RandomForest.build do
+  train_with_instances instances
+end
+
+# store trained model as binary file
+classifier.serialize('randomforest.model')
+
+# load Classifier from binary file
+loaded_classifier = Weka::Core::SerializationHelper.deserialize('randomforest.model')
+# => #<Java::WekaClassifiersTrees::RandomForest:0x197db331>
+```
+
 ## Development
 
 After checking out the repo, run `bin/setup` to install dependencies.

data/lib/weka/class_builder.rb

@@ -11,7 +11,7 @@ module Weka
 
       def build_class(class_name, weka_module: nil, include_concerns: true)
         java_import java_class_path(class_name, weka_module)
-        define_class(class_name, include_concerns: include_concerns)
+        define_class(class_name, weka_module, include_concerns: include_concerns)
       end
 
       def build_classes(*class_names, weka_module: nil, include_concerns: true)
@@ -52,15 +52,23 @@ module Weka
         self.name.scan('::').count == 1
       end
 
-      def define_class(class_name, include_concerns: true)
+      def define_class(class_name, weka_module, include_concerns: true)
         module_eval <<-CLASS_DEFINITION, __FILE__, __LINE__ + 1
           class #{class_name}
             #{'include Concerns' if include_concerns}
+            #{include_serializable_for(class_name, weka_module)}
             #{include_utils}
           end
         CLASS_DEFINITION
       end
 
+      def include_serializable_for(class_name, weka_module)
+        class_path   = java_class_path(class_name, weka_module)
+        serializable = Weka::Core::SerializationHelper.serializable?(class_path)
+
+        "include Weka::Concerns::Serializable" if serializable
+      end
+
       def include_utils
         return unless utils_defined?
         "include #{utils}"

data/lib/weka/concerns.rb

@@ -3,6 +3,7 @@ require 'weka/concerns/buildable'
 require 'weka/concerns/describable'
 require 'weka/concerns/optionizable'
 require 'weka/concerns/persistent'
+require 'weka/concerns/serializable'
 
 module Weka
   module Concerns

data/lib/weka/concerns/serializable.rb

@@ -0,0 +1,17 @@
+require 'active_support/concern'
+require 'weka/core/serialization_helper'
+
+module Weka
+  module Concerns
+    module Serializable
+      extend ActiveSupport::Concern
+
+      included do
+        def serialize(filename)
+          Weka::Core::SerializationHelper.write(filename, self)
+        end
+      end
+
+    end
+  end
+end

data/lib/weka/core.rb

@@ -4,3 +4,4 @@ require 'weka/core/saver'
 require 'weka/core/attribute'
 require 'weka/core/dense_instance'
 require 'weka/core/instances'
+require 'weka/core/serialization_helper'

data/lib/weka/core/instances.rb

@@ -2,6 +2,7 @@ require 'weka/core/converters'
 require 'weka/core/loader'
 require 'weka/core/saver'
 require 'weka/core/dense_instance'
+require 'weka/concerns/serializable'
 
 module Weka
   module Core
@@ -9,6 +10,7 @@ module Weka
     java_import "weka.core.FastVector"
 
     class Instances
+      include Weka::Concerns::Serializable
 
       DEFAULT_RELATION_NAME = 'Instances'
 
@@ -163,6 +165,12 @@ module Weka
         filter.filter(self)
       end
 
+      def apply_filters(*filters)
+        filters.inject(self) do |filtered_instances, filter|
+          filter.filter(filtered_instances)
+        end
+      end
+
       private
 
       def add_attribute(attribute)

data/lib/weka/core/serialization_helper.rb

@@ -0,0 +1,13 @@
+module Weka
+  module Core
+    java_import 'weka.core.SerializationHelper'
+
+    class SerializationHelper
+
+      class << self
+        alias :deserialize :read
+        alias :serialize   :write
+      end
+    end
+  end
+end

data/lib/weka/version.rb

@@ -1,3 +1,3 @@
 module Weka
-  VERSION = "0.1.0"
+  VERSION = "0.2.0"
 end

data/weka.gemspec

@@ -10,7 +10,7 @@ Gem::Specification.new do |spec|
   spec.email         = ['paul.christoph.goetze@gmail.com']
 
   spec.summary       = %q{Machine Learning & Data Mining with JRuby.}
-  spec.description   = %q{A wrapper for the Weka library (http://www.cs.waikato.ac.nz/ml/weka/)}
+  spec.description   = %q{A JRuby wrapper for the Weka library (http://www.cs.waikato.ac.nz/ml/weka/)}
   spec.homepage      = 'https://github.com/paulgoetze/weka-jruby'
   spec.license       = 'MIT'
 

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: weka
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.2.0
 platform: java
 authors:
 - Paul Götze
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2015-12-26 00:00:00.000000000 Z
+date: 2016-01-19 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: lock_jar
@@ -94,7 +94,7 @@ dependencies:
         version: '3.0'
   prerelease: false
   type: :development
-description: A wrapper for the Weka library (http://www.cs.waikato.ac.nz/ml/weka/)
+description: A JRuby wrapper for the Weka library (http://www.cs.waikato.ac.nz/ml/weka/)
 email:
 - paul.christoph.goetze@gmail.com
 executables:
@@ -138,6 +138,7 @@ files:
 - lib/weka/concerns/describable.rb
 - lib/weka/concerns/optionizable.rb
 - lib/weka/concerns/persistent.rb
+- lib/weka/concerns/serializable.rb
 - lib/weka/core.rb
 - lib/weka/core/attribute.rb
 - lib/weka/core/converters.rb
@@ -145,6 +146,7 @@ files:
 - lib/weka/core/instances.rb
 - lib/weka/core/loader.rb
 - lib/weka/core/saver.rb
+- lib/weka/core/serialization_helper.rb
 - lib/weka/exceptions.rb
 - lib/weka/filters.rb
 - lib/weka/filters/filter.rb