rb-libsvm 1.2.0 → 1.3.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 430498716ebb529469fa73ece0ffe044bac7c04a
-  data.tar.gz: d1effb4c07f3fa77418991d79ddf58ed5887ee8c
+  metadata.gz: a6628f92d9d2d643e194b46b8e225c5cade354eb
+  data.tar.gz: 58b9eee2167ec2d4eb60a6010c3d05fab7836c94
 SHA512:
-  metadata.gz: 6546735dc65a8bff5215960cec78755eab7ddd4ecffe047bd1e4a6db485bfbcb970fed54b61581b259478b0451cbcca21b936ed2bc01f3c48af32240bd0be553
-  data.tar.gz: 07c833fd03503d38e4b9a5c0a6ac6958e32b3ffcf8983f6ffe8663dd1914d9de9ee164f8725373d3957b9e264708783811929d49d69fce5bbfa461560d36400d
+  metadata.gz: 64ec0c4d492262ce7dbc4b28132c998a886e04b558beec111ee3118a7608732447d54641e5f9a400d7ed311d8cf8babfc07185e482359f998cd0d1174344fc78
+  data.tar.gz: 35956f1af2012ad765a3312b926f92e60af0c469924512db1325d0e52d97c6f9850cd8639fb02723615a6976a54d1c637ef51dd2baa232ee3c7eb6641f0091ce

data/.gitignore

@@ -11,3 +11,5 @@ tags
 *~
 /.rbx
 /vendor
+/.yardoc/
+/doc/

data/.travis.yml

@@ -2,6 +2,6 @@ language: ruby
 rvm:
   - 1.8.7
   - 1.9.3
-  - 2.0.0
-  - 2.1.1
+  - 2.1.5
+  - 2.2.0
   - ruby-head

data/README.md

@@ -25,7 +25,7 @@ this gem. You should install the original package if you need them.
 It is helpful to consult the [README][] of the LIBSVM package for
 reference when configuring the training parameters.
 
-Currently this package includes libsvm version 3.18.
+Currently this package includes libsvm version 3.20.
 
 ## Dependencies
 

data/ext/libsvm/libsvm.c

@@ -85,7 +85,7 @@ static struct svm_node *example_to_internal(VALUE example_ary)
   if(x == 0) {
     rb_raise(rb_eNoMemError, "on Libsvm::Node allocation" " %s:%i", __FILE__,__LINE__);
   }
-  /* loop it's element nodes */
+  /* loop its element nodes */
   for(j = 0; j < example_ary_len; ++j) {
     node = rb_ary_entry(example_ary,j);
     Data_Get_Struct(node,struct svm_node,node_struct);
@@ -122,12 +122,12 @@ static struct svm_node **examples_ary_to_internal(VALUE examples_ary)
    call-seq:
      problem.set_examples(labels, examples_array)
 
-     double *y;            // class (aka. label) of the example
-     struct svm_node **x;  // examples
+   Sets the examples and their label for a training set.
 
-   This method sets the contents of an SVM Problem, which consists
-   of lables (or classifications) and examples (or feature vectors).
-   If those 2 don't match in length and ArgumentError is raised.
+   The indices of the arrays are supposed to correspond. If they don't
+   match in length an ArgumentError is raised.
+
+   @return [Integer] number of examples in the traning set
 */
 static VALUE cProblem_examples_set(VALUE obj,VALUE labels_ary,VALUE examples_ary)
 {
@@ -167,10 +167,14 @@ static VALUE cProblem_examples_set(VALUE obj,VALUE labels_ary,VALUE examples_ary
 
 /*
    call-seq:
-   labels, array_of_arrays = problem.examples
+     problem.examples
+
+   Returns the array of labels and the array of corresponding examples
+   contained in this training set.
+
+     labels, array_of_examples = problem.examples
 
-   double *y; // class/label of the example
-   struct svm_node **x;
+   @return [[Array<Float>, Array<Array<Node>>]]
 */
 static VALUE cProblem_examples(VALUE problem) {
   struct svm_problem *prob;
@@ -249,21 +253,6 @@ rx_def_accessor(cSvmParameter,struct svm_parameter,double,cache_size);
 rx_def_accessor(cSvmParameter,struct svm_parameter,double,eps);
 rx_def_accessor_as(cSvmParameter,struct svm_parameter,double,C,c);
 
-/*  Label weight.
-
-    nr_weight, weight_label, and weight are used to change the penalty
-    for some classes (If the weight for a class is not changed, it is
-    set to 1). This is useful for training classifier using unbalanced
-    input data or with asymmetric misclassification cost.
-
-    nr_weight is the number of elements in the array weight_label and
-    weight. Each weight[i] corresponds to weight_label[i], meaning that
-    the penalty of class weight_label[i] is scaled by a factor of weight[i].
-
-    If you do not want to change penalty for any of the classes,
-    just set nr_weight to 0.
-
-*/
 static VALUE cSvmParameter_label_weights_set(VALUE obj,VALUE weight_hash) {
   struct svm_parameter *param;
   int i;
@@ -316,8 +305,20 @@ rx_def_accessor(cSvmParameter,struct svm_parameter,double,p);
 rx_def_accessor(cSvmParameter,struct svm_parameter,int,shrinking);
 rx_def_accessor(cSvmParameter,struct svm_parameter,int,probability);
 
-/*  Libsvm::Model  */
-
+/* call-seq:
+ *   Model.train(problem, parameter)
+ *
+ * Train a model with given training set (problem) and training
+ * parameter object.
+ *
+ * Library function
+ * svm_train[https://github.com/cjlin1/libsvm/blob/master/README#L313]
+ *
+ * @param [Libsvm::Problem] the training set (problem)
+ * @param [Libsvm::SvmParameter] training parameter object
+ *
+ * @return [Libsvm::Model] the trained model
+ */
 static VALUE cModel_class_train(VALUE obj,VALUE problem,VALUE parameter) {
   const struct svm_problem *prob;
   const struct svm_parameter *param;
@@ -336,6 +337,19 @@ static VALUE cModel_class_train(VALUE obj,VALUE problem,VALUE parameter) {
   return Data_Wrap_Struct(cModel, 0, model_free, model);
 }
 
+/* call-seq:
+ *   model.predict(example)
+ *
+ * Classify an example and return the class (label).
+ *
+ * This is the class (label) value for a classifier model or the
+ * funtion value for a regression model. 1 or -1 for one-class model.
+ *
+ * Library function
+ * svm_predict[https://github.com/cjlin1/libsvm/blob/master/README#L511].
+ *
+ * @return [Float] predicted class label
+ */
 static VALUE cModel_predict(VALUE obj,VALUE example) {
   struct svm_node *x;
   struct svm_model *model;
@@ -350,6 +364,22 @@ static VALUE cModel_predict(VALUE obj,VALUE example) {
   return rb_float_new(class);
 }
 
+/* call-seq:
+ *   model.predict_probability(example)
+ *
+ * Classify an example and return both the label (or regression
+ * value), as well as the array of probability found for each class.
+ *
+ * The first element of the returned array contains the label. The
+ * second element is another array which contains the probability
+ * value found for each model class.
+ *
+ * Library function
+ * svm_predict_probability[https://github.com/cjlin1/libsvm/blob/master/README#L591].
+ *
+ * @return [Array<Float, Array<Float>>] predicted label and
+ *   probability per class
+ */
 static VALUE cModel_predict_probability(VALUE obj,VALUE example) {
   struct svm_node *x;
   struct svm_model *model;
@@ -381,6 +411,15 @@ static VALUE cModel_predict_probability(VALUE obj,VALUE example) {
   return target;
 }
 
+/* call-seq:
+ *   model.save(filename)
+ *
+ * Saves the model to file `filename`. The format is the LIBSVM
+ * internal model representation.
+ *
+ * Library function
+ * svm_save_model[https://github.com/cjlin1/libsvm/blob/master/README#L621].
+ */
 static VALUE cModel_save(VALUE obj, VALUE filename)
 {
   const struct svm_model *model;
@@ -397,6 +436,14 @@ static VALUE cModel_save(VALUE obj, VALUE filename)
   return Qnil;
 }
 
+/* Type of the model. Integer value, one of the constants in the
+ * {Libsvm::SvmType} module.
+ *
+ * Library function
+ * svm_get_svm_type[https://github.com/cjlin1/libsvm/blob/master/README#L533].
+ *
+ * @return [C_SVC, NU_SVC, ONE_CLASS, EPSILON_SVR, NU_SVR]
+ */
 static VALUE cModel_svm_type(VALUE obj)
 {
   const struct svm_model *model;
@@ -404,20 +451,47 @@ static VALUE cModel_svm_type(VALUE obj)
   return INT2NUM(svm_get_svm_type(model));
 }
 
-static VALUE cModel_classes(VALUE obj)
+/* Number of classes the model is configured and trained to
+ * predict. For single-class or regression models 2 is returned.
+ *
+ * Library function
+ * svm_get_nr_class[https://github.com/cjlin1/libsvm/blob/master/README#L538].
+ *
+ * @return [Integer] the number of classes
+ */
+static VALUE cModel_classes_count(VALUE obj)
 {
   const struct svm_model *model;
   Data_Get_Struct(obj, struct svm_model, model);
   return INT2NUM(svm_get_nr_class(model));
 }
 
-static VALUE cModel_support_vectors(VALUE obj)
+/*
+ * Number of the support vectors the model contains.
+ *
+ * This method binds to the function
+ * svm_get_nr_sv[https://github.com/cjlin1/libsvm/blob/master/README#L555].
+ *
+ * @return [Integer] the number of support vectors
+ */
+static VALUE cModel_support_vectors_count(VALUE obj)
 {
   const struct svm_model *model;
   Data_Get_Struct(obj, struct svm_model, model);
   return INT2NUM(svm_get_nr_sv(model));
 }
 
+/* call-seq:
+ *   Model.load(filename)         # => model
+ *
+ * Load a {Libsvm::Model} from a file
+ *
+ * This load a model from file `filename`. Format is the LIBSVM
+ * internal file representation of the model.
+ *
+ * @param filename [String] name of the model file
+ * @return [Libsvm::Model] the model
+ */
 static VALUE cModel_class_load(VALUE cls, VALUE filename)
 {
   struct svm_model *model;
@@ -427,6 +501,17 @@ static VALUE cModel_class_load(VALUE cls, VALUE filename)
   return Data_Wrap_Struct(cModel, 0, model_free, model);
 }
 
+/* call-seq:
+ *   Model.cross_validation(problem, parameter, num_fold)
+ *
+ * Perform a cross-validation with num_fold split of the training data
+ * contained in problem.
+ *
+ * @param problem [Libsvm::Problem] the training set
+ * @param parameter [Libsvm::SvmParameter] training parameters object
+ * @param num_fold [Integer] the number of splits to devide the traning set into
+ * @return [Array<Integer>] the labels for each instance in training set
+ */
 static VALUE cModel_class_cross_validation(VALUE cls, VALUE problem, VALUE parameter, VALUE num_fold)
 {
   const struct svm_problem *prob;
@@ -499,16 +584,36 @@ void Init_libsvm_ext() {
   rb_define_singleton_method(cModel, "load", cModel_class_load, 1);
   rb_define_method(cModel, "save", cModel_save, 1);
   rb_define_method(cModel, "svm_type", cModel_svm_type, 0);
-  rb_define_method(cModel, "classes", cModel_classes, 0);
-  rb_define_method(cModel, "support_vectors", cModel_support_vectors, 0);
+  rb_define_method(cModel, "classes_count", cModel_classes_count, 0);
+  rb_define_method(cModel, "support_vectors_count", cModel_support_vectors_count, 0);
   rb_define_method(cModel, "predict", cModel_predict, 1);
   rb_define_method(cModel, "predict_probability", cModel_predict_probability, 1);
 
+  /**
+   * Module with constants for values that are allowed for
+   * {Libsvm::SvmParameter#kernel_type}. The value controls what kind
+   * of kernel is used when training the model.
+   */
   mKernelType = rb_define_module_under(mLibsvm, "KernelType");
+  /**
+   * A linear kernel
+   */
   rb_define_const(mKernelType, "LINEAR", INT2NUM(LINEAR));
+  /**
+   * A polynomial kernel
+   */
   rb_define_const(mKernelType, "POLY", INT2NUM(POLY));
+  /**
+   * A redial basis function kernel
+   */
   rb_define_const(mKernelType, "RBF", INT2NUM(RBF));
+  /**
+   * A sigmoid kernel
+   */
   rb_define_const(mKernelType, "SIGMOID", INT2NUM(SIGMOID));
+  /**
+   * A precomputed kernel
+   */
   rb_define_const(mKernelType, "PRECOMPUTED", INT2NUM(PRECOMPUTED));
 
   mSvmType = rb_define_module_under(mLibsvm,"SvmType");

data/lib/libsvm.rb

@@ -3,8 +3,36 @@ require 'libsvm/libsvm_ext'
 require 'libsvm/node'
 
 module Libsvm
+
+  # Module with constants for values that allowed for
+  # {Libsvm::SvmParameter#svm_type}. The value controls which kind of
+  # model is being trained.
+  #
+  # See the relevant section in the LIBSVM
+  # README[https://github.com/cjlin1/libsvm/blob/master/README#389]
+  # for more information.
+  module SvmType; end
+
   module CoreExtensions
     module Collection
+      # @!method to_example
+      #
+      # Converts this collection into an array of feature.
+      #
+      #   { 1 => 0.4, 3 => 0.5, 5 => 0.9 }.to_example
+      #   # => [#<Libsvm::Node: index=1, value=0.4>, #<Libsvm::Node: index=3, value=0.5>, #<Libsvm::Node: index=5, value=0.9>]
+      #   { 1 => 0.4, 3 => 0.5, 5 => 0.9 }.to_example
+      #   # => [#<Libsvm::Node: index=1, value=0.4>, #<Libsvm::Node: index=3, value=0.5>, #<Libsvm::Node: index=5, value=0.9>]
+      #   [0.4, 0.5, 0.9].to_example
+      #   # => [#<Libsvm::Node: index=0, value=0.4>, #<Libsvm::Node: index=1, value=0.5>, #<Libsvm::Node: index=2, value=0.9>]
+      #   [ [1, 0.4], [3, 0.5], [5, 0.9] ].to_example
+      #   # => [#<Libsvm::Node: index=1, value=0.4>, #<Libsvm::Node: index=3, value=0.5>, #<Libsvm::Node: index=5, value=0.9>]
+      #
+      # An array of features is the type used for classification and model learning.
+      #
+      # @note In an upcoming major release this core extension will become no longer included in Hash and Array, but rather be optional.
+      #
+      # @return [Array<Libsvm::Node>]
       def to_example
         Node.features(self)
       end
@@ -15,6 +43,7 @@ end
 class Hash
   include Libsvm::CoreExtensions::Collection
 end
+
 class Array
   include Libsvm::CoreExtensions::Collection
 end

data/lib/libsvm/model.rb

@@ -0,0 +1,9 @@
+module Libsvm
+  class Model
+    # @deprecated Use {#classes_count} instead
+    alias_method :classes, :classes_count
+
+    # @deprecated Use {#support_vectors_count} instead
+    alias_method :support_vectors, :support_vectors_count
+  end
+end

data/lib/libsvm/node.rb

@@ -1,15 +1,58 @@
 module Libsvm
+  # Represents a feature
+  #
+  # A feature has an index and a value.
+  #
+  # An array of features (Libsvm::Node instances) represents an
+  # example which can be classified, or in greater number can
+  # constitute the main part of a training set (Libsvm::Prolem).
+  #
+  # This class represents the struct
+  # svm_node[https://github.com/cjlin1/libsvm/blob/master/README#L357].
+  #
+  # @see Libsvm::Problem Libsvm::Problem, the training set class
   class Node
     class << self
+      # Create an array of features from collection or variable number
+      # of float values.
+      #
+      # Indices are converted to Integer, values are converted to
+      # Float. When passed a variable number of Float values or an
+      # array of Floast, the index is implied by the arguments
+      # position. Collection of tuples of hash arguments afford sparse
+      # feature arrays. These can represent features which don't have
+      # consecutive indices starting at zero.
+      #
+      # @overload features(0.3, 0.7, 0.8, ...)
+      # @param vararg variable number of value arguments, or
+      #
+      # @overload features([0.3, 0.7, 0.8, ...])
+      # @param array an array of values, or
+      #
+      # @overload features([[0, 0.3], [1, 0.7], [2, 0.8], ...])
+      # @param array an array of [index, value] tuples, or
+      #
+      # @overload features(0 => 0.3, 1 => 0.7, 2 => 0.8, 99 => 0.1)
+      # @param hash a hash from index to value
+      #
+      # @return [Array<Libsvm::Node>] example, i.e. an array of features
       def features(*vargs)
         array_of_nodes = []
         if vargs.size == 1
-          if vargs.first.class == Array
-            vargs.first.each_with_index do |value, index|
-              array_of_nodes << Node.new(index.to_i, value.to_f)
+          argument = vargs.first
+          if argument.class == Array
+            case argument.first
+            when Array
+              argument.each do |pair|
+                array_of_nodes << Node.new(pair.first.to_i, pair.last.to_f)
+              end
+            else
+              argument.each_with_index do |value, index|
+                array_of_nodes << Node.new(index.to_i, value.to_f)
+              end
             end
-          elsif vargs.first.class == Hash
-            vargs.first.each do |index, value|
+          elsif argument.class == Hash
+            argument.each do |index, value|
               array_of_nodes << Node.new(index.to_i, value.to_f)
             end
           else
@@ -23,18 +66,44 @@ module Libsvm
         array_of_nodes
       end
 
+      # Create a feature node.
+      #
+      #   Libsvm::Node[0, 1.1] # => #<Libsvm::Node: index=0, value=1.1>
+      #
+      # @return [Libsvm::Node]
       def [](index, value)
         new(index, value)
       end
     end
 
+    # @!attribute index
+    #   The index identifies the feature for which this node represents a value
+
+    # @!attribute value
+    #   The value of this feature in this instance
+
+    # Create a new feature node.
+    #
+    #   Libsvm::Node.new(0, 1.1) # => #<Libsvm::Node: index=0, value=1.1>
     def initialize(index=0, value=0.0)
       self.index = index
       self.value = value
     end
 
+    # Compare features for equality.
+    #
+    # Nodes with equal index and value are equal.
+    #
+    # @return [Boolean]
     def ==(other)
       index == other.index && value == other.value
     end
+
+    def inspect # :nodoc:
+      vars = %w(index value).map { |name|
+        "#{name}=#{send(name)}"
+      }.join(", ")
+      "#<#{self.class}: #{vars}>"
+    end
   end
 end

data/lib/libsvm/problem.rb

@@ -0,0 +1,20 @@
+module Libsvm
+  # This file only contains documentation comments. See
+  # ext/libsvm/libsvm.c for implementation.
+
+  # Represents a training set (alternatively called problem).
+  #
+  # In contains labels and examples in equal number.
+  #
+  # This class represents the struct
+  # svm_problem[https://github.com/cjlin1/libsvm/blob/master/README#L319].
+  class Problem
+
+    # @!attribute [r] l
+    #
+    # The number of labels and examples in this traning set. (The name
+    # is lower case L.)
+    #
+    # @return [Integer]
+  end
+end

data/lib/libsvm/svm_parameter.rb

@@ -0,0 +1,115 @@
+module Libsvm
+  # This file only contains documentation comments. See
+  # ext/libsvm/libsvm.c for implementation.
+
+  # Represents the learning parameter struct.
+  #
+  # The parameters in this object control how the SVM model is trained
+  # specifically.
+  #
+  # This class represents the struct
+  # svm_parameter[https://github.com/cjlin1/libsvm/blob/master/README#L366].
+  class SvmParameter
+
+    # @!attribute svm_type
+    #
+    # Type of SVM to use. This parameter controls if the model
+    # classifies or performs regression, if a one-class model is
+    # built, or if the NU variant of SVM is used.
+    #
+    # @see Libsvm::SvmType Libsvm::SvmType for value constants
+    # @return [C_SVC, NU_SVC, ONE_CLASS, EPSILON_SVR, NU_SVR]
+
+    # @!attribute kernel_type
+    #
+    # Type of kernel to use.
+    #
+    # @see Libsvm::KernelType Libsvm::KernelType for kernel_type constants
+    # @return [LINEAR, POLY, RBF, SIGMOID, PRECOMPUTED]
+
+
+    # @!attribute degree
+    #
+    # Degree parameter, relevant for POLY kernel type.
+    #
+    # @see Libsvm::KernelType Libsvm::KernelType for kernel_type constants
+    # @return [Integer]
+
+    # @!attribute gamma
+    #
+    # Gamma parameter, relevant for kernel types RBF, SIGMOID and POLY.
+    #
+    # @see Libsvm::KernelType Libsvm::KernelType for kernel_type constants
+    # @return [Float]
+
+    # @!attribute coef0
+    #
+    # Coeffectient 0, relevant for kernel types SIGMOID and POLY.
+    #
+    # @see Libsvm::KernelType Libsvm::KernelType for kernel_type constants
+    # @return [Float]
+
+    # @!attribute cache_size
+    #
+    # Size of the kernel cache, in megabytes.
+    #
+    # @return [Integer]
+
+    # @!attribute eps
+    #
+    # Stopping criterion parameter.
+    #
+    # @return [Float]
+
+    # @!attribute c
+    #
+    # The parameter C to use for this model.
+    #
+    # @return [Float]
+
+    # @!attribute label_weights
+    #
+    # Hash with indices of labels as keys and weights as values.
+    #
+    # These weights are used to change the penalty for specific labels
+    # (classes). If the weight for a label is not changed, it is set
+    # to 1.0.
+    #
+    # @return [Hash]
+
+    # @!attribute nu
+    #
+    # The nu parameter in nu-SVM, nu-SVR, and one-class-SVM (i.e. when
+    # the Libsvm::SvmParameter#svm_type is given as NU_SVC, NU_SVR or
+    # ONE_CLASS).
+    #
+    # @see Libsvm::SvmType Libsvm::SvmType for svm_type constants
+    # @return [Float]
+
+    # @!attribute p
+    #
+    # The epsilon in epsilon-insensitive loss function of epsilon-SVM
+    # regression (i.e. when the Libmsvm::SvmParameter#svm_type is
+    # given as Libsvm::SvmType::EPSILON_SVR).
+    #
+    # @return [Float]
+
+    # @!attribute shrinking
+    #
+    # Integer interpreted as boolean value.
+    #
+    # Controls if shrinking heuristics is used.
+    #
+    # @return [Integer] 0 meaning false, 1 true
+
+    # @!attribute probability
+    #
+    # Integer interpreted as boolean value.
+    #
+    # Controls if the model can create probability values for
+    # classifications in addition to the classification.
+    #
+    # @return [Integer] 0 meaning false, 1 true
+
+  end
+end

data/lib/libsvm/version.rb

@@ -1,3 +1,3 @@
 module Libsvm
-  VERSION = "1.2.0"
+  VERSION = "1.3.0"
 end

data/spec/core_ext_spec.rb

@@ -0,0 +1,15 @@
+require 'spec_helper'
+
+describe "Create examples" do
+  describe "from a hash" do
+    example = { 11 => 0.11, 21 => 0.21, 101 => 0.99 }.to_example.sort_by(&:index)
+    example.should == Node.features({11 => 0.11, 21 => 0.21, 101 => 0.99 }).sort_by(&:index)
+  end
+
+  describe "from an array of tuples" do
+    it "can create example from array of pairs" do
+      example = [ [11, 0.11], [21, 0.21], [101, 0.99] ].to_example
+      example.should == Node.features({11 => 0.11, 21 => 0.21, 101 => 0.99 }).sort_by(&:index)
+    end
+  end
+end

data/spec/model_spec.rb

@@ -98,7 +98,7 @@ describe "A Libsvm model" do
     end
 
     it "can be asked for its number of classes (aka. labels)" do
-      expect(@model.classes).to eq(2)
+      expect(@model.classes_count).to eq(2)
     end
   end
 
@@ -108,9 +108,9 @@ describe "A Libsvm model" do
     end
   end
 
-  describe "support_vectors" do
+  describe "support_vectors_count" do
     it "returns count" do
-      expect(@model.support_vectors).to eq(3)
+      expect(@model.support_vectors_count).to eq(3)
     end
   end
 
@@ -124,7 +124,7 @@ describe "A Libsvm model" do
     end
 
     it "produces probabilities for each class" do
-      expect(probabilities.length).to eq(@model.classes)
+      expect(probabilities.length).to eq(@model.classes_count)
     end
 
     it "can predict probability" do

data/spec/node_spec.rb

@@ -47,12 +47,14 @@ describe "A Node" do
     ary = Node.features([0.1, 0.2, 0.3, 0.4, 0.5])
     ary.map {|n| n.class.should == Node}
     ary.map {|n| n.value }.should == [0.1, 0.2, 0.3, 0.4, 0.5]
+    ary.map {|n| n.index }.should == [0, 1, 2, 3, 4]
   end
 
   it "class can create nodes from variable parameters" do
     ary = Node.features(0.1, 0.2, 0.3, 0.4, 0.5)
     ary.map {|n| Node.should === n }
     ary.map {|n| n.value }.should == [0.1, 0.2, 0.3, 0.4, 0.5]
+    ary.map {|n| n.index }.should == [0, 1, 2, 3, 4]
   end
 
   it "class can create nodes from hash" do

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: rb-libsvm
 version: !ruby/object:Gem::Version
-  version: 1.2.0
+  version: 1.3.0
 platform: ruby
 authors:
 - C. Florian Ebeling
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-11-16 00:00:00.000000000 Z
+date: 2015-01-18 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rake-compiler
@@ -67,9 +67,13 @@ files:
 - ext/libsvm/svm.cpp
 - ext/libsvm/svm.h
 - lib/libsvm.rb
+- lib/libsvm/model.rb
 - lib/libsvm/node.rb
+- lib/libsvm/problem.rb
+- lib/libsvm/svm_parameter.rb
 - lib/libsvm/version.rb
 - rb-libsvm.gemspec
+- spec/core_ext_spec.rb
 - spec/model_spec.rb
 - spec/node_spec.rb
 - spec/parameter_spec.rb
@@ -100,6 +104,7 @@ signing_key:
 specification_version: 4
 summary: Ruby bindings for LIBSVM
 test_files:
+- spec/core_ext_spec.rb
 - spec/model_spec.rb
 - spec/node_spec.rb
 - spec/parameter_spec.rb