weighted_sampler 1.0.3 → 1.1.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: cf378756434bce6feff8b33723b735706d4324977a34f0f86e2c8ebb82bf33d2
-  data.tar.gz: 3665445824ffe81a739c4fa3c67183bf217fee8233e7d33680e2cfdb0dbbacd0
+  metadata.gz: 418175c950561e4d8afe70b6d19be4c9957107d79416f6b5600dc83bb51f57f2
+  data.tar.gz: 1d1f5bac8a18c12fa86027ca024a3705fe39a1f65c0e39b69401854e34ea9626
 SHA512:
-  metadata.gz: 3c90d9fb14d935b7bacb06965a10e31515b0cb4e44a535b08e5e4d93e8ff78e712913079f30fedb1bd0d588531f9cafba4008b05a25d2e8624133338bdc6ee01
-  data.tar.gz: ec6ca14ac634db2989b08a8c35c4bb59c4bdfc00e554aa34dc7b99b2ee25c9ed66c96a9a64368b07ede953866ba77520fa4aba75890f68fdc8ccc875a54b92e1
+  metadata.gz: '0182411300d8c5e8355d376658b0d6cc58ac8ca973c9b06962ff45280c45b70ce4d32754f3d0afab212349e1dee0f05596805d462eecb302ddb7b66ee9426ab4'
+  data.tar.gz: 3df0c7ebecc6b17f2f12f58a059772889d93f852cee6ba302cd03969124be5ca3b2c5c694796077b0d705f38bae8060eec9a83ea5d8f9603767024c471e60e16

data/lib/weighted_sampler.rb

@@ -1,6 +1,8 @@
 # frozen_string_literal: true
 
 require 'weighted_sampler/version'
+# require 'pry'
+
 module WeightedSampler
 
   # sum of floats are never stable enough to guarantee exact equality to 1
@@ -12,33 +14,34 @@ module WeightedSampler
       @random = Random.new(seed) unless seed.nil?
 
       if enum.is_a?(Hash)
-        @p_ranges = normalized_ranges(enum.values, skip_normalization)
+        @p_margins = normalized_margins(enum.values, skip_normalization)
         @keys = enum.keys
       elsif enum.is_a?(Array)
-        @p_ranges = normalized_ranges(enum, skip_normalization)
+        @p_margins = normalized_margins(enum, skip_normalization)
         @keys = [*0...enum.size]
       end
 
-      return unless @p_ranges.nil? || @keys.nil? || @keys.empty?
+      return unless @p_margins.nil? || @keys.nil? || @keys.empty?
+
       raise ArgumentError, 'input structure must be a non-empty Hash or Array'
     end
 
     def sample
       pick = @random ? @random.rand : rand
 
-      idx = @p_ranges.index { |range| range.include? pick }
-      @keys[idx] if idx
+      idx = @p_margins.find_index { |margin| pick < margin }
+      idx ||= @p_margins.count - 1 # safe assignment if last margin was not good enough
+
+      @keys[idx]
     end
 
     private
 
-    def normalized_ranges(array, skip_normalization)
+    def normalized_margins(array, skip_normalization)
       raise ArgumentError, 'weights can be only positive' if array.any?(&:negative?)
 
-      probabilities = array
-      probabilities = normalize_probabilities(probabilities) unless skip_normalization
-
-      array_to_ranges probabilities
+      probabilities = skip_normalization ? array : normalize_probabilities(array)
+      incremental_margins probabilities
     end
 
     def normalize_probabilities(array)
@@ -47,18 +50,20 @@ module WeightedSampler
       array.map { |el| el / sum }
     end
 
-    def array_to_ranges(array)
+    # convert probs like [0.1, 0.2, 0.3, 0.4]
+    # to incremental margins [0.1, 0.3, 0.6, 1.0]
+    def incremental_margins(array)
       start = 0.0
-      ranges = array.map do |v|
-        p_start = start
-        start += v
+      margins = array.map do |v|
+        res = v + start
+        start = res
 
-        (p_start...v + p_start)
+        res
       end
 
       raise 'normalized probabilities total is not 1' if (start - 1.0).abs > ERROR_ALLOWANCE
 
-      ranges
+      margins
     end
 
   end

data/lib/weighted_sampler/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module WeightedSampler
-  VERSION = '1.0.3'.freeze
+  VERSION = '1.1.0'.freeze
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: weighted_sampler
 version: !ruby/object:Gem::Version
-  version: 1.0.3
+  version: 1.1.0
 platform: ruby
 authors:
 - Oleksiy Babich
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2018-07-26 00:00:00.000000000 Z
+date: 2019-01-02 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -85,8 +85,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubyforge_project: 
-rubygems_version: 2.7.6
+rubygems_version: 3.0.1
 signing_key: 
 specification_version: 4
 summary: Weighted Sampler helps you to pick a random samples from a collection with