rumale 0.18.4 → 0.18.5

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: a4ec0f04029a88ea7950bb74bf9e0bd970cf8b546253da4ef4545244ad0adf68
-  data.tar.gz: ed927fb58359fe49d4834593ec2034535f5ac8d37c773d6cb5a53227c4777f61
+  metadata.gz: 90ea96f8e47e9710da8f9e3d2c0d88b45e5ec86f349fb4dc1d9058867750c88c
+  data.tar.gz: dde564e3cb15d6abbbb35c2629d1b9c53c79a456f29a60a5cfe7780aa06dd47b
 SHA512:
-  metadata.gz: 58727ddc9c5c6f9c12ac8231f57295988795bf07ec6f748543fff924a4d8085472a210789992ba2a3cc4827300823d141313da1fd20a10be3b5970b40871e4e3
-  data.tar.gz: cc1af7ff552a9fe516e0588081ad6645cb03d4c8dd774492ea35c3e1a28eece7383dc88604c41e1cc1417f5d05af1324a7b97c121cb7ee76bcc697e4173195be
+  metadata.gz: aaaad994a319b406686962301ad28a67e9d92b3fd15930f9792eb0e6249342d98b08a822737771f19501928036cfdf64a41dede62e9a3fe114aefa13a77f72c6
+  data.tar.gz: f71fe52af3418d4ff74641a6f651fda783eca32450c36023de276ed41f0e326e035f3996f9433e2514569f8bcb496276bc582d0b507a5c603c90164ad388fad4

data/.rubocop.yml

@@ -15,10 +15,26 @@ AllCops:
 Style/Documentation:
   Enabled: false
 
+Style/HashEachMethods:
+  Enabled: true
+
+Style/HashTransformKeys:
+  Enabled: true
+
+Style/HashTransformValues:
+  Enabled: true
+
+Lint/RaiseException:
+  Enabled: true
+
+Lint/StructNewOverride:
+  Enabled: true
+
 Layout/LineLength:
   Max: 145
   IgnoredPatterns: ['(\A|\s)#']
 
+
 Metrics/ModuleLength:
   Max: 200
 

data/CHANGELOG.md

@@ -1,3 +1,7 @@
+# 0.18.5
+- Add functions for calculation of cosine similarity and distance to [Rumale::PairwiseMetric](https://yoshoku.github.io/rumale/doc/Rumale/PairwiseMetric.html).
+- Refactor some codes with Rubocop.
+
 # 0.18.4
 - Add transformer class for [KernelFDA](https://yoshoku.github.io/rumale/doc/Rumale/KernelMachine/KernelFDA.html).
 - Refactor [KernelPCA](https://yoshoku.github.io/rumale/doc/Rumale/KernelMachine/KernelPCA.html).

data/lib/rumale/clustering/hdbscan.rb

@@ -134,6 +134,7 @@ module Rumale
         res
       end
 
+      # rubocop:disable Metrics/AbcSize, Metrics/MethodLength, Metrics/PerceivedComplexity
       def condense_tree(hierarchy, min_cluster_size)
         n_edges = hierarchy.size
         root = 2 * n_edges
@@ -262,6 +263,7 @@ module Rumale
         end
         res
       end
+      # rubocop:enable Metrics/AbcSize, Metrics/MethodLength, Metrics/PerceivedComplexity
     end
   end
 end

data/lib/rumale/ensemble/gradient_boosting_regressor.rb

@@ -125,12 +125,10 @@ module Rumale
         n_outputs = @estimators.first.is_a?(Array) ? @estimators.size : 1
         if n_outputs > 1
           multivar_predict(x)
+        elsif enable_parallel?
+          parallel_map(@params[:n_estimators]) { |n| @estimators[n].predict(x) }.reduce(&:+) + @base_predictions
         else
-          if enable_parallel?
-            parallel_map(@params[:n_estimators]) { |n| @estimators[n].predict(x) }.reduce(&:+) + @base_predictions
-          else
-            @estimators.map { |tree| tree.predict(x) }.reduce(&:+) + @base_predictions
-          end
+          @estimators.map { |tree| tree.predict(x) }.reduce(&:+) + @base_predictions
         end
       end
 

data/lib/rumale/evaluation_measure/function.rb

@@ -42,6 +42,8 @@ module Rumale
       conf_mat
     end
 
+    # rubocop:disable Metrics/MethodLength, Metrics/AbcSize
+
     # Output a summary of classification performance for each class.
     #
     # @example
@@ -69,7 +71,8 @@ module Rumale
       y_true = Rumale::Validation.check_convert_label_array(y_true)
       y_pred = Rumale::Validation.check_convert_label_array(y_pred)
       # calculate each evaluation measure.
-      supports = y_true.bincount
+      classes = y_true.to_a.uniq.sort
+      supports = Numo::Int32.asarray(classes.map { |label| y_true.eq(label).count })
       precisions = Rumale::EvaluationMeasure::PrecisionRecall.precision_each_class(y_true, y_pred)
       recalls = Rumale::EvaluationMeasure::PrecisionRecall.recall_each_class(y_true, y_pred)
       fscores = Rumale::EvaluationMeasure::PrecisionRecall.f_score_each_class(y_true, y_pred)
@@ -83,7 +86,7 @@ module Rumale
       weighted_recall = (Numo::DFloat.cast(recalls) * weights).sum
       weighted_fscore = (Numo::DFloat.cast(fscores) * weights).sum
       # output reults.
-      target_name ||= y_true.to_a.uniq.sort.map(&:to_s)
+      target_name ||= classes.map(&:to_s)
       if output_hash
         res = {}
         target_name.each_with_index do |label, n|
@@ -107,9 +110,8 @@ module Rumale
           fscore: weighted_fscore,
           support: sum_supports
         }
-        res
       else
-        width = ['weighted avg'.size, target_name.map(&:size).max].max
+        width = [12, target_name.map(&:size).max].max # 12 is 'weighted avg'.size
         res = +''
         res << "#{' ' * width}  precision    recall  f1-score   support\n"
         res << "\n"
@@ -136,8 +138,9 @@ module Rumale
         fscore_str = format('%#10s', format('%.2f', weighted_fscore))
         res << format("%##{width}s ", 'weighted avg')
         res << "#{precision_str}#{recall_str}#{fscore_str}#{supports_str}\n"
-        res
       end
+      res
     end
+    # rubocop:enable Metrics/MethodLength, Metrics/AbcSize
   end
 end

data/lib/rumale/kernel_machine/kernel_fda.rb

@@ -81,7 +81,7 @@ module Rumale
         within_mat = centered_kernel_mat.dot(centered_kernel_mat.transpose) + @params[:reg_param] * Numo::DFloat.eye(n_samples)
 
         # calculate projection matrix.
-        eig_vals, eig_vecs = Numo::Linalg.eigh(
+        _, eig_vecs = Numo::Linalg.eigh(
           between_mat, within_mat,
           vals_range: (n_samples - n_components)...n_samples
         )

data/lib/rumale/nearest_neighbors/vp_tree.rb

@@ -84,14 +84,12 @@ module Rumale
           else
             node
           end
+        elsif dist + tau >= node.threshold
+          node.right.n_samples < k ? node : search(q, node.right, k, tau)
+        elsif dist - tau <= node.threshold
+          node.left.n_samples < k ? node : search(q, node.left, k, tau)
         else
-          if dist + tau >= node.threshold
-            node.right.n_samples < k ? node : search(q, node.right, k, tau)
-          elsif dist - tau <= node.threshold
-            node.left.n_samples < k ? node : search(q, node.left, k, tau)
-          else
-            node
-          end
+          node
         end
         # :nocov:
       end

data/lib/rumale/pairwise_metric.rb

@@ -54,6 +54,39 @@ module Rumale
         err_mat.class.maximum(err_mat, 0)
       end
 
+      # Calculate the pairwise cosine simlarities between x and y.
+      #
+      # @param x [Numo::DFloat] (shape: [n_samples_x, n_features])
+      # @param y [Numo::DFloat] (shape: [n_samples_y, n_features])
+      # @return [Numo::DFloat] (shape: [n_samples_x, n_samples_x] or [n_samples_x, n_samples_y] if y is given)
+      def cosine_similarity(x, y = nil)
+        y_not_given = y.nil?
+        x = Rumale::Validation.check_convert_sample_array(x)
+        y = Rumale::Validation.check_convert_sample_array(y) unless y_not_given
+        x_norm = Numo::NMath.sqrt((x**2).sum(1))
+        x_norm[x_norm.eq(0)] = 1
+        x /= x_norm.expand_dims(1)
+        if y_not_given
+          x.dot(x.transpose)
+        else
+          y_norm = Numo::NMath.sqrt((y**2).sum(1))
+          y_norm[y_norm.eq(0)] = 1
+          y /= y_norm.expand_dims(1)
+          x.dot(y.transpose)
+        end
+      end
+
+      # Calculate the pairwise cosine distances between x and y.
+      #
+      # @param x [Numo::DFloat] (shape: [n_samples_x, n_features])
+      # @param y [Numo::DFloat] (shape: [n_samples_y, n_features])
+      # @return [Numo::DFloat] (shape: [n_samples_x, n_samples_x] or [n_samples_x, n_samples_y] if y is given)
+      def cosine_distance(x, y = nil)
+        dist_mat = 1 - cosine_similarity(x, y)
+        dist_mat[dist_mat.diag_indices] = 0 if y.nil?
+        dist_mat.clip(0, 2)
+      end
+
       # Calculate the rbf kernel between x and y.
       #
       # @param x [Numo::DFloat] (shape: [n_samples_x, n_features])

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 = '0.18.4'
+  VERSION = '0.18.5'
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: rumale
 version: !ruby/object:Gem::Version
-  version: 0.18.4
+  version: 0.18.5
 platform: ruby
 authors:
 - yoshoku
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2020-04-11 00:00:00.000000000 Z
+date: 2020-04-18 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: numo-narray
@@ -180,8 +180,6 @@ files:
 - LICENSE.txt
 - README.md
 - Rakefile
-- bin/console
-- bin/setup
 - ext/rumale/extconf.rb
 - ext/rumale/rumale.c
 - ext/rumale/rumale.h

data/bin/console

@@ -1,14 +0,0 @@
-#!/usr/bin/env ruby
-
-require 'bundler/setup'
-require 'rumale'
-
-# You can add fixtures and/or initialization code here to make experimenting
-# with your gem easier. You can also use a different console, if you like.
-
-# (If you use this, don't forget to add pry to your Gemfile!)
-# require 'pry'
-# Pry.start
-
-require 'irb'
-IRB.start(__FILE__)

data/bin/setup

@@ -1,8 +0,0 @@
-#!/usr/bin/env bash
-set -euo pipefail
-IFS=$'\n\t'
-set -vx
-
-bundle install
-
-# Do any other automated setup that you need to do here