pca 0.1.0 → 0.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 8dc374097222849bba01f272a8df8d019bdefad9
-  data.tar.gz: 70af4ff18c1cb5796b54308ea43f1b202fcca8ed
+  metadata.gz: 437676549bb33931b4565d105e4a73d5acfc1f35
+  data.tar.gz: da5c93ae3bcee85fa79a7aab62dc14d6da729856
 SHA512:
-  metadata.gz: 6c2508f2c7a86e677330736ea733e8fedb525301ae3d1c1e6ad00e95dc7be07ba50aa4a4ef5e27cde958bb0091fd0310e8d8ae44ec2cc0987928c26da7c0bd3d
-  data.tar.gz: f708f797cdb4811680ed4a4390db9b918555fec1c8033c44356ebe64950697b6d09855cb93fec8fcb56dab3930e66a17171fcf8e0953f7518b5259f9a87bcbc2
+  metadata.gz: bce79847afa2af0cba8305a5eeeccaae29ae651dd9c26a9cef9154971e0680225e61801e179b769dd893544bc4b145b0fea8456cc159984ad882690665a08b13
+  data.tar.gz: a5b103e1813788e5235c04b7cc90cbee3141d9ec49bfb342a2ba0a7e1019307464b1ea65914b1254b70f274d9021aeb0980b7933323f0f9bf6ec4e2e163b4f2f

data/lib/pca.rb

@@ -1,10 +1,11 @@
 require 'gsl'
 
 class PCA
-  attr_reader :components, :singular_values, :mean, :explained_variance, :explained_variance_ratio
+  attr_reader :components, :singular_values, :mean, :std, :explained_variance, :explained_variance_ratio
 
   def initialize opts = {}
     @n_components = opts[:components]
+    @scale_data = opts[:scale_data]
   end
 
   def fit x
@@ -14,7 +15,7 @@ class PCA
   end
 
   def transform x
-    x = prepare_data x, use_saved_mean: true
+    x = prepare_data x, use_saved_mean_and_std: true
     _transform x
   end
 
@@ -26,7 +27,8 @@ class PCA
 
   def inverse_transform x
     x = ensure_matrix x
-    out = x * @components.transpose
+    out = x * @components
+    out.size2.times {|col| out.col(col).mul! @std[col] } if @scale_data
     out.size2.times {|col| out.col(col).add! @mean[col] }
     out
   end
@@ -34,22 +36,26 @@ class PCA
   private
     def prepare_data x, opts = {}
       x = ensure_matrix x
-      @mean = calculate_mean(x) unless opts[:use_saved_mean]
+      @mean = calculate_mean(x) unless opts[:use_saved_mean_and_std]
       mean_normalize x
+      if @scale_data
+        @std = calculate_std(x) unless opts[:use_saved_mean_and_std]
+        scale(x)
+      end
       x
     end
 
     def _fit x
       covariance_matrix = (x.transpose * x) / x.size1
       u, v, s = covariance_matrix.SV_decomp
-      @components = slice_n u
+      @components = slice_n(u).transpose
       @singular_values = slice_n s
       @explained_variance = @singular_values**2 / x.size1
       @explained_variance_ratio = @explained_variance / @explained_variance.sum
     end
 
     def _transform x
-      x * @components
+      x * @components.transpose
     end
 
     def ensure_matrix x
@@ -71,6 +77,14 @@ class PCA
       x.size2.times {|col| x.col(col).sub! @mean[col] }
     end
 
+    def calculate_std x
+      x.size2.times.map {|col| x.col(col).sd }
+    end
+
+    def scale x
+      x.size2.times {|col| x.col(col).div! @std[col] }
+    end
+
     def slice_n x
       return x unless @n_components
       case x

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: pca
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.2.0
 platform: ruby
 authors:
 - Geoff Buesing
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-03-31 00:00:00.000000000 Z
+date: 2015-04-01 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rb-gsl
@@ -24,7 +24,7 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
-description: Principal Component Analysis
+description: Principal Component Analysis (PCA). Uses GSL for calculations.
 email: gbuesing@gmail.com
 executables: []
 extensions: []
@@ -54,6 +54,6 @@ rubyforge_project:
 rubygems_version: 2.4.5
 signing_key: 
 specification_version: 4
-summary: Principal Component Analysis
+summary: Principal Component Analysis (PCA)
 test_files: []
 has_rdoc: