datamix 0.0.3 → 0.0.4

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 2e25311621834c0792028b34973e716c1dc0bce2
-  data.tar.gz: df08942707700a403c16e55bf0cd39659520ec7c
+  metadata.gz: 2cf4ea62b0a5c3151b16327c1bf13aaa05c9af55
+  data.tar.gz: 0a45edcde7a6c198ceb68b5e69654b1e4cdc77f3
 SHA512:
-  metadata.gz: 643e11e9a83a4c498ca2c042b59780eed0e95a26f07cb6f2624e294b78812c6abeb6669bc897cccb21b65119d906f751b3f671e4696b7d90b802df9a0399d933
-  data.tar.gz: 1c9dc5162696b3b6de6110bee2d0cf183ceec3f913186bc7887554a6d716255212b4761514da3762866b1a668d41c0f55ac2d1a98183d5d86265d36948a82e58
+  metadata.gz: 947376b91089d8389c446101b06b0386e247aa4dd372eda9c88ebd399f15817c6326cc9b96a97311fd7b9635a4e9dc9dd0ed3216ae54f4b35a5617aa50868fb5
+  data.tar.gz: 9dfb88a50870a940da4454a1084bdc145af6ba30ff8e3c757e82e8382d22d6a0ec17ade3f356ed6d03c669ccdb27ecdb2c1b38498ccca6855fb23144cc6d4459

data/lib/datamix/refinements/array.rb

@@ -33,6 +33,53 @@ module DataMix
       offset -rows
     end
 
+    def resample(chunk_range, seed: nil)
+      generator = seed ? Random.new(seed) : Random.new
+
+      # Save min value, we will adjust the result to it later
+      min_value = min
+
+      # Split the array to chunks
+      chunks = []
+      clone = dup
+      while !clone.empty? do
+        seam = generator.rand(chunk_range) - 1
+        chunks.push clone.slice! 0..seam
+      end
+
+      # If the last chunk contains one element only, merge with the 
+      # previous chunk
+      if chunks.last.size == 1
+        chunks[chunks.size-2].push chunks.last.first
+        chunks = chunks.first chunks.size-1
+      end
+
+      # Shuffle the chunks
+      chunks = chunks.sample chunks.size, random: generator
+
+      # Adjust each chunk so that its beginning connects with the previous
+      # chunk normally. For this we calculate the median change value of the 
+      # array.
+      chunks.each_with_index do |chunk, i|
+        next if i == 0
+        connector = chunks[i-1].last
+
+        # Calculate an array of changes, and then take a random sample
+        diff = chunk.each_cons(2).map { |a,b| b-a }.sample random: generator
+
+        delta = diff + connector - chunk.first
+        chunks[i].map! { |val| val+delta }
+      end
+
+      # Merge chunks to a flat array
+      result = chunks.flatten
+
+      # Move the entire array up or down so that its min value is equal to 
+      # the original min value recorded at the beginning.
+      diff = result.min - min_value
+      result.map! { |val| val-diff }
+    end
+
     def round(decimals=0)
       map { |val| val ? val.round(decimals) : nil }
     end
@@ -62,4 +109,4 @@ module DataMix
     end
 
   end
-end
+end

data/lib/datamix/refinements/csv_table.rb

@@ -76,6 +76,17 @@ module DataMix
       delete from
     end
 
+    # Create a similar data table with resampled data
+    def resample(range, except: [], seed: nil)
+      except = [except] unless except.is_a? Array
+      cols = headers.reject { |h| except.include? h }
+      seed ||= Time.now.to_f
+
+      cols.each do |col|
+        by_col[col] = by_col[col].resample range, seed: seed
+      end
+    end
+
     # Rounds all values in a column
     def round(col, decimals: 0)
       by_col[col] = by_col[col].map { |val| val ? val.round(decimals) : nil }

data/lib/datamix/version.rb

@@ -1,3 +1,3 @@
 module DataMix
-  VERSION = "0.0.3"
+  VERSION = "0.0.4"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: datamix
 version: !ruby/object:Gem::Version
-  version: 0.0.3
+  version: 0.0.4
 platform: ruby
 authors:
 - Danny Ben Shitrit
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2017-03-10 00:00:00.000000000 Z
+date: 2017-03-30 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: terminal-table
@@ -184,7 +184,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.6.6
+rubygems_version: 2.5.1
 signing_key: 
 specification_version: 4
 summary: DSL for manipulating tabular data