weka 0.3.0-java → 0.4.0-java

data/lib/weka/core/converters.rb

@@ -11,6 +11,8 @@ module Weka
                     :CSVSaver,
                     :JSONLoader,
                     :JSONSaver,
+                    :C45Loader,
+                    :C45Saver,
                     include_concerns: false
     end
   end

data/lib/weka/core/dense_instance.rb

@@ -1,13 +1,13 @@
 module Weka
   module Core
-    java_import "weka.core.DenseInstance"
+    java_import 'weka.core.DenseInstance'
 
     class DenseInstance
-      java_import "java.util.Date"
-      java_import "java.text.SimpleDateFormat"
+      java_import 'java.util.Date'
+      java_import 'java.text.SimpleDateFormat'
 
       def initialize(data, weight: 1.0)
-        if data.kind_of?(Integer)
+        if data.is_a?(Integer)
           super(data)
         else
           super(weight, to_java_double(data))
@@ -38,8 +38,8 @@ module Weka
         end
       end
 
-      alias :values       :to_a
-      alias :values_count :num_values
+      alias values       to_a
+      alias values_count num_values
 
       private
 
@@ -61,7 +61,7 @@ module Weka
           format_date(value, attribute.date_format)
         elsif attribute.numeric?
           value
-        elsif attribute.nominal?
+        elsif attribute.nominal? || attribute.string?
           attribute.value(value)
         end
       end

data/lib/weka/core/instances.rb

@@ -6,13 +6,13 @@ require 'weka/concerns/serializable'
 
 module Weka
   module Core
-    java_import "weka.core.Instances"
-    java_import "weka.core.FastVector"
+    java_import 'weka.core.Instances'
+    java_import 'weka.core.FastVector'
 
     class Instances
       include Weka::Concerns::Serializable
 
-      DEFAULT_RELATION_NAME = 'Instances'
+      DEFAULT_RELATION_NAME = 'Instances'.freeze
 
       class << self
         def from_arff(file)
@@ -26,6 +26,16 @@ module Weka
         def from_json(file)
           Loader.load_json(file)
         end
+
+        # Loads instances based on a given *.names file (holding the attribute
+        # values) or a given *.data file (holding the attribute values).
+        # The respective other file is loaded from the same directory.
+        #
+        # See http://www.cs.washington.edu/dm/vfml/appendixes/c45.htm for more
+        # information about the C4.5 file format.
+        def from_c45(file)
+          Loader.load_c45(file)
+        end
       end
 
       def initialize(relation_name: DEFAULT_RELATION_NAME, attributes: [], &block)
@@ -48,13 +58,33 @@ module Weka
       end
 
       def add_attributes(&block)
-        self.instance_eval(&block) if block
+        instance_eval(&block) if block
         self
       end
 
-      alias :with_attributes  :add_attributes
-      alias :instances_count  :num_instances
-      alias :attributes_count :num_attributes
+      alias with_attributes       add_attributes
+      alias instances_count       num_instances
+      alias attributes_count      num_attributes
+      alias has_string_attribute? check_for_string_attributes
+
+      ## Check if the instances has any attribute of the given type
+      # @param [String, Symbol, Integer] type type of the attribute to check
+      #   String and Symbol argument are converted to corresponding type
+      #   defined in Weka::Core::Attribute
+      #
+      # @example Passing String
+      #   instances.has_attribute_type?('string')
+      #   instances.has_attribute_type?('String')
+      #
+      # @example Passing Symbol
+      #   instances.has_attribute_type?(:String)
+      #
+      # @example Passing Integer
+      #   instances.has_attribute_type?(Attribute::STRING)
+      def has_attribute_type?(type)
+        type = map_attribute_type(type) unless type.is_a?(Integer)
+        check_for_attribute_type(type)
+      end
 
       def each
         if block_given?
@@ -96,26 +126,39 @@ module Weka
         Saver.save_json(file: file, instances: self)
       end
 
+      # Creates a file with the istances's attribute values and a *.data file
+      # with the actual data.
+      #
+      # You should choose another file extension than .data (preferably
+      # *.names) for the file, else it will just be overwritten with the
+      # automatically created *.data file.
+      #
+      # See http://www.cs.washington.edu/dm/vfml/appendixes/c45.htm for more
+      # information about the C4.5 file format.
+      def to_c45(file)
+        Saver.save_c45(file: file, instances: self)
+      end
+
       def numeric(name, class_attribute: false)
-        attribute = Attribute.new(name.to_s)
+        attribute = Attribute.new_numeric(name)
         add_attribute(attribute)
         self.class_attribute = name if class_attribute
       end
 
       def nominal(name, values:, class_attribute: false)
-        attribute = Attribute.new(name.to_s, Array(values).map(&:to_s))
+        attribute = Attribute.new_nominal(name, values)
         add_attribute(attribute)
         self.class_attribute = name if class_attribute
       end
 
       def string(name, class_attribute: false)
-        attribute = Attribute.new(name.to_s, [])
+        attribute = Attribute.new_string(name)
         add_attribute(attribute)
         self.class_attribute = name if class_attribute
       end
 
       def date(name, format: 'yyyy-MM-dd HH:mm', class_attribute: false)
-        attribute = Attribute.new(name.to_s, format)
+        attribute = Attribute.new_date(name, format)
         add_attribute(attribute)
         self.class_attribute = name if class_attribute
       end
@@ -129,10 +172,10 @@ module Weka
         end
       end
 
-      alias :add_numeric_attribute :numeric
-      alias :add_string_attribute  :string
-      alias :add_nominal_attribute :nominal
-      alias :add_date_attribute    :date
+      alias add_numeric_attribute numeric
+      alias add_string_attribute  string
+      alias add_nominal_attribute nominal
+      alias add_date_attribute    date
 
       def class_attribute
         classAttribute if class_attribute_defined?
@@ -187,7 +230,7 @@ module Weka
         return if attribute_names.include?(name.to_s)
 
         error   = "\"#{name}\" is not defined."
-        hint    = "Only defined attributes can be used as class attribute!"
+        hint    = 'Only defined attributes can be used as class attribute!'
         message = "#{error} #{hint}"
 
         raise ArgumentError, message
@@ -198,14 +241,31 @@ module Weka
       end
 
       def instance_from(instance_or_values, weight:)
-        if instance_or_values.kind_of?(Java::WekaCore::Instance)
+        if instance_or_values.is_a?(Java::WekaCore::Instance)
           instance_or_values.weight = weight
           instance_or_values
         else
           data = internal_values_of(instance_or_values)
+
+          # string attribute has unlimited range of possible values.
+          # Check the return index, if it is -1 then add the value to
+          # the attribute before creating the instance
+          data.map!.with_index do |value, index|
+            if value == -1 && attribute(index).string?
+              attribute(index).add_string_value(instance_or_values[index].to_s)
+            else
+              value
+            end
+          end
+
           DenseInstance.new(data, weight: weight)
         end
       end
+
+      def map_attribute_type(type)
+        return -1 unless Attribute::TYPES.include?(type.downcase.to_sym)
+        Attribute.const_get(type.upcase)
+      end
     end
 
     Java::WekaCore::Instances.__persistent__ = true

data/lib/weka/core/loader.rb

@@ -18,6 +18,16 @@ module Weka
           load_with(Converters::JSONLoader, file: file)
         end
 
+        # Takes either a *.names or a *.data file and loads the respective other
+        # file from the same directory automatically.
+        # Returns a Weka::Core::Instances object.
+        #
+        # See http://www.cs.washington.edu/dm/vfml/appendixes/c45.htm for more
+        # information about the C4.5 file format.
+        def load_c45(file)
+          load_with(Converters::C45Loader, file: file)
+        end
+
         private
 
         def load_with(loader_class, file:)
@@ -27,6 +37,5 @@ module Weka
         end
       end
     end
-
   end
 end

data/lib/weka/core/saver.rb

@@ -18,6 +18,27 @@ module Weka
           save_with(Converters::JSONSaver, file: file, instances: instances)
         end
 
+        # Saves the given `instances` into a file with the given name and a
+        # *.data file in the same directory.
+        # The file with the given file name includes the instances's attribute
+        # values, the *.data file holds the actual data.
+        #
+        # Example:
+        #
+        #   Weka::Core::Saver.save_c45(
+        #     file: './path/to/example.names',
+        #     instances: instances
+        #   )
+        #
+        # creates an example.names file and an example.data file in the
+        # ./path/to/ directory.
+        #
+        # See: http://www.cs.washington.edu/dm/vfml/appendixes/c45.htm for more
+        # information about the C4.5 file format.
+        def save_c45(file:, instances:)
+          save_with(Converters::C45Saver, file: file, instances: instances)
+        end
+
         private
 
         def save_with(saver_class, file:, instances:)
@@ -29,6 +50,5 @@ module Weka
         end
       end
     end
-
   end
 end

data/lib/weka/core/serialization_helper.rb

@@ -3,10 +3,9 @@ module Weka
     java_import 'weka.core.SerializationHelper'
 
     class SerializationHelper
-
       class << self
-        alias :deserialize :read
-        alias :serialize   :write
+        alias deserialize read
+        alias serialize   write
       end
     end
   end

data/lib/weka/filters/filter.rb

@@ -4,6 +4,5 @@ module Weka
 
     class Filter
     end
-
   end
 end

data/lib/weka/filters/supervised/attribute.rb

@@ -16,10 +16,9 @@ module Weka
                       :PartitionMembership
 
         class AttributeSelection
-          alias :use_evaluator :set_evaluator
-          alias :use_search    :set_search
+          alias use_evaluator set_evaluator
+          alias use_search    set_search
         end
-
       end
     end
   end

data/lib/weka/filters/utils.rb

@@ -1,17 +1,14 @@
-require 'active_support/concern'
-
 module Weka
   module Filters
     module Utils
-      extend ActiveSupport::Concern
-
-      included do
-        def filter(instances)
-          set_input_format(instances)
-          Filter.use_filter(instances, self)
+      def self.included(base)
+        base.class_eval do
+          def filter(instances)
+            set_input_format(instances)
+            Filter.use_filter(instances, self)
+          end
         end
       end
-
     end
   end
 end

data/lib/weka/jars.rb

@@ -1,19 +1,16 @@
-require 'active_support/concern'
-
 module Weka
   module Jars
-    extend ActiveSupport::Concern
-
-    included do
-      require 'lock_jar'
+    def self.included(base)
+      base.class_eval do
+        require 'lock_jar'
 
-      lib_path = File.expand_path('../../', File.dirname(__FILE__))
-      lockfile = File.join(lib_path, 'Jarfile.lock')
-      jars_dir = File.join(lib_path, 'jars')
+        lib_path = File.expand_path('../../', File.dirname(__FILE__))
+        lockfile = File.join(lib_path, 'Jarfile.lock')
+        jars_dir = File.join(lib_path, 'jars')
 
-      LockJar.install(lockfile, local_repo: jars_dir)
-      LockJar.load(lockfile, local_repo: jars_dir)
+        LockJar.install(lockfile, local_repo: jars_dir)
+        LockJar.load(lockfile, local_repo: jars_dir)
+      end
     end
-
   end
 end

data/lib/weka/version.rb

@@ -1,3 +1,3 @@
 module Weka
-  VERSION = "0.3.0"
+  VERSION = '0.4.0'.freeze
 end

data/weka.gemspec

@@ -9,8 +9,8 @@ Gem::Specification.new do |spec|
   spec.authors       = ['Paul Götze']
   spec.email         = ['paul.christoph.goetze@gmail.com']
 
-  spec.summary       = %q{Machine Learning & Data Mining with JRuby.}
-  spec.description   = %q{A JRuby wrapper for the Weka library (http://www.cs.waikato.ac.nz/ml/weka/)}
+  spec.summary       = 'Machine Learning & Data Mining with JRuby.'
+  spec.description   = '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'
 
@@ -23,7 +23,6 @@ Gem::Specification.new do |spec|
   spec.require_paths = ['lib']
 
   spec.add_runtime_dependency     'lock_jar',         '~> 0.13'
-  spec.add_runtime_dependency     'activesupport',    '~> 4.0'
 
   spec.add_development_dependency 'bundler',          '~> 1.6'
   spec.add_development_dependency 'rake',             '~> 10.0'

metadata

@@ -1,99 +1,85 @@
 --- !ruby/object:Gem::Specification
 name: weka
 version: !ruby/object:Gem::Version
-  version: 0.3.0
+  version: 0.4.0
 platform: java
 authors:
 - Paul Götze
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2016-02-10 00:00:00.000000000 Z
+date: 2016-12-22 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
-  name: lock_jar
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '0.13'
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '0.13'
+  name: lock_jar
   prerelease: false
   type: :runtime
-- !ruby/object:Gem::Dependency
-  name: activesupport
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '4.0'
+        version: '0.13'
+- !ruby/object:Gem::Dependency
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '4.0'
-  prerelease: false
-  type: :runtime
-- !ruby/object:Gem::Dependency
+        version: '1.6'
   name: bundler
+  prerelease: false
+  type: :development
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '1.6'
+- !ruby/object:Gem::Dependency
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.6'
+        version: '10.0'
+  name: rake
   prerelease: false
   type: :development
-- !ruby/object:Gem::Dependency
-  name: rake
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '10.0'
+- !ruby/object:Gem::Dependency
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '10.0'
+        version: '3.0'
+  name: rspec
   prerelease: false
   type: :development
-- !ruby/object:Gem::Dependency
-  name: rspec
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '3.0'
+- !ruby/object:Gem::Dependency
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '3.0'
+  name: shoulda-matchers
   prerelease: false
   type: :development
-- !ruby/object:Gem::Dependency
-  name: shoulda-matchers
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '3.0'
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '3.0'
-  prerelease: false
-  type: :development
 description: A JRuby wrapper for the Weka library (http://www.cs.waikato.ac.nz/ml/weka/)
 email:
 - paul.christoph.goetze@gmail.com
@@ -178,7 +164,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project:
-rubygems_version: 2.4.8
+rubygems_version: 2.6.6
 signing_key:
 specification_version: 4
 summary: Machine Learning & Data Mining with JRuby.