rumale 0.20.0 → 0.22.0

data/lib/rumale/model_selection/time_series_split.rb

@@ -0,0 +1,91 @@
+# frozen_string_literal: true
+
+require 'rumale/base/splitter'
+
+module Rumale
+  module ModelSelection
+    # TimeSeriesSplit is a class that generates the set of data indices for time series cross-validation.
+    # It is assumed that the dataset given are already ordered by time information.
+    #
+    # @example
+    #   cv = Rumale::ModelSelection::TimeSeriesSplit.new(n_splits: 5)
+    #   x = Numo::DFloat.new(6, 2).rand
+    #   cv.split(x, nil).each do |train_ids, test_ids|
+    #     puts '---'
+    #     pp train_ids
+    #     pp test_ids
+    #   end
+    #
+    #   # ---
+    #   # [0]
+    #   # [1]
+    #   # ---
+    #   # [0, 1]
+    #   # [2]
+    #   # ---
+    #   # [0, 1, 2]
+    #   # [3]
+    #   # ---
+    #   # [0, 1, 2, 3]
+    #   # [4]
+    #   # ---
+    #   # [0, 1, 2, 3, 4]
+    #   # [5]
+    #
+    class TimeSeriesSplit
+      include Base::Splitter
+
+      # Return the number of splits.
+      # @return [Integer]
+      attr_reader :n_splits
+
+      # Return the maximum number of training samples in a split.
+      # @return [Integer/Nil]
+      attr_reader :max_train_size
+
+      # Create a new data splitter for time series cross-validation.
+      #
+      # @param n_splits [Integer] The number of splits.
+      # @param max_train_size [Integer/Nil] The maximum number of training samples in a split.
+      def initialize(n_splits: 5, max_train_size: nil)
+        check_params_numeric(n_splits: n_splits)
+        check_params_numeric_or_nil(max_train_size: max_train_size)
+        @n_splits = n_splits
+        @max_train_size = max_train_size
+      end
+
+      # Generate data indices for time series cross-validation.
+      #
+      # @overload split(x, y) -> Array
+      #   @param x [Numo::DFloat] (shape: [n_samples, n_features])
+      #     The dataset to be used to generate data indices for time series cross-validation.
+      #     It is expected that the data will be ordered by time information.
+      #   @param y [Numo::Int32] (shape: [n_samples])
+      #     This argument exists to unify the interface between the K-fold methods, it is not used in the method.
+      # @return [Array] The set of data indices for constructing the training and testing dataset in each fold.
+      def split(x, _y)
+        x = check_convert_sample_array(x)
+
+        n_samples = x.shape[0]
+        unless (@n_splits + 1).between?(2, n_samples)
+          raise ArgumentError,
+                'The number of folds (n_splits + 1) must be not less than 2 and not more than the number of samples.'
+        end
+
+        test_size = n_samples / (@n_splits + 1)
+        offset = test_size + n_samples % (@n_splits + 1)
+
+        Array.new(@n_splits) do |n|
+          start = offset * (n + 1)
+          train_ids = if !@max_train_size.nil? && @max_train_size < test_size
+                        Array((start - @max_train_size)...start)
+                      else
+                        Array(0...start)
+                      end
+          test_ids = Array(start...(start + test_size))
+          [train_ids, test_ids]
+        end
+      end
+    end
+  end
+end

data/lib/rumale/pipeline/pipeline.rb

@@ -140,7 +140,7 @@ module Rumale
       def validate_steps(steps)
         steps.keys[0...-1].each do |name|
           transformer = steps[name]
-          next if transformer.nil? || %i[fit transform].all? { |m| transformer.class.method_defined?(m) }
+          next if transformer.nil? || (transformer.class.method_defined?(:fit) && transformer.class.method_defined?(:transform))
 
           raise TypeError,
                 'Class of intermediate step in pipeline should be implemented fit and transform methods: ' \

data/lib/rumale/probabilistic_output.rb

@@ -98,7 +98,7 @@ module Rumale
 
       def hessian_matrix(probs, df, sigma)
         sub = probs * (1 - probs)
-        h11 = (df * df * sub).sum + sigma
+        h11 = (df**2 * sub).sum + sigma
         h22 = sub.sum + sigma
         h21 = (df * sub).sum
         Numo::DFloat[[h11, h21], [h21, h22]]

data/lib/rumale/tree/base_decision_tree.rb

@@ -75,17 +75,10 @@ module Rumale
         node = Node.new(depth: depth, impurity: impurity, n_samples: n_samples)
 
         # terminate growing.
-        unless @params[:max_leaf_nodes].nil?
-          return nil if @n_leaves >= @params[:max_leaf_nodes]
-        end
-
+        return nil if !@params[:max_leaf_nodes].nil? && @n_leaves >= @params[:max_leaf_nodes]
         return nil if n_samples < @params[:min_samples_leaf]
         return put_leaf(node, y) if n_samples == @params[:min_samples_leaf]
-
-        unless @params[:max_depth].nil?
-          return put_leaf(node, y) if depth == @params[:max_depth]
-        end
-
+        return put_leaf(node, y) if !@params[:max_depth].nil? && depth == @params[:max_depth]
         return put_leaf(node, y) if stop_growing?(y)
 
         # calculate optimal parameters.

data/lib/rumale/tree/gradient_tree_regressor.rb

@@ -138,7 +138,7 @@ module Rumale
         nil
       end
 
-      def grow_node(depth, x, y, g, h)
+      def grow_node(depth, x, y, g, h) # rubocop:disable Metrics/AbcSize
         # intialize some variables.
         sum_g = g.sum
         sum_h = h.sum
@@ -146,17 +146,10 @@ module Rumale
         node = Node.new(depth: depth, n_samples: n_samples)
 
         # terminate growing.
-        unless @params[:max_leaf_nodes].nil?
-          return nil if @n_leaves >= @params[:max_leaf_nodes]
-        end
-
+        return nil if !@params[:max_leaf_nodes].nil? && @n_leaves >= @params[:max_leaf_nodes]
         return nil if n_samples < @params[:min_samples_leaf]
         return put_leaf(node, sum_g, sum_h) if n_samples == @params[:min_samples_leaf]
-
-        unless @params[:max_depth].nil?
-          return put_leaf(node, sum_g, sum_h) if depth == @params[:max_depth]
-        end
-
+        return put_leaf(node, sum_g, sum_h) if !@params[:max_depth].nil? && depth == @params[:max_depth]
         return put_leaf(node, sum_g, sum_h) if stop_growing?(y)
 
         # calculate optimal parameters.

data/lib/rumale/version.rb

@@ -3,5 +3,5 @@
 # Rumale is a machine learning library in Ruby.
 module Rumale
   # The version of Rumale you are using.
-  VERSION = '0.20.0'
+  VERSION = '0.22.0'
 end

data/rumale.gemspec

@@ -45,4 +45,5 @@ Gem::Specification.new do |spec|
   }
 
   spec.add_runtime_dependency 'numo-narray', '>= 0.9.1'
+  spec.add_runtime_dependency 'lbfgsb', '>=0.3.0'
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: rumale
 version: !ruby/object:Gem::Version
-  version: 0.20.0
+  version: 0.22.0
 platform: ruby
 authors:
 - yoshoku
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2020-08-01 00:00:00.000000000 Z
+date: 2020-11-22 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: numo-narray
@@ -24,6 +24,20 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: 0.9.1
+- !ruby/object:Gem::Dependency
+  name: lbfgsb
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.3.0
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.3.0
 description: |
   Rumale is a machine learning library in Ruby.
   Rumale provides machine learning algorithms with interfaces similar to Scikit-Learn in Python.
@@ -43,7 +57,7 @@ extensions:
 - ext/rumale/extconf.rb
 extra_rdoc_files: []
 files:
-- ".coveralls.yml"
+- ".github/workflows/build.yml"
 - ".gitignore"
 - ".rspec"
 - ".rubocop.yml"
@@ -135,10 +149,13 @@ files:
 - lib/rumale/model_selection/cross_validation.rb
 - lib/rumale/model_selection/function.rb
 - lib/rumale/model_selection/grid_search_cv.rb
+- lib/rumale/model_selection/group_k_fold.rb
+- lib/rumale/model_selection/group_shuffle_split.rb
 - lib/rumale/model_selection/k_fold.rb
 - lib/rumale/model_selection/shuffle_split.rb
 - lib/rumale/model_selection/stratified_k_fold.rb
 - lib/rumale/model_selection/stratified_shuffle_split.rb
+- lib/rumale/model_selection/time_series_split.rb
 - lib/rumale/multiclass/one_vs_rest_classifier.rb
 - lib/rumale/naive_bayes/base_naive_bayes.rb
 - lib/rumale/naive_bayes/bernoulli_nb.rb
@@ -206,7 +223,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.1.2
+rubygems_version: 3.1.4
 signing_key:
 specification_version: 4
 summary: Rumale is a machine learning library in Ruby. Rumale provides machine learning

data/.coveralls.yml

@@ -1 +0,0 @@
-service_name: travis-ci