standard_deviation 0.0.1 → 1.0.0

data/README.md

@@ -18,13 +18,24 @@ Or install it yourself as:
 
 ## Usage
 
-Just call `standard_deviation` or `stdev` on a array with numbers.
+Just call `standard_deviation` or `stdev` on an array with numbers:
 
 ``` ruby
 [1, 2, 3, 4, 5].stdev => 1.5811388300841898
 
 [521.0, BigDecimal("1234.45"), 1_120].standard_deviation => 383.168958598336
 ```
+For population standard deviation, call `stdevp`:
+
+``` ruby
+[1, 2, 3, 4, 5].stdevp => 1.4142135623730951
+```
+
+Also, the API exposes the `sample_variance` and `population_variance` calculations:
+``` ruby
+[1, 2, 3, 4, 5].sample_variance => 2.5
+[1, 2, 3, 4, 5].population_variance => 2.0
+```
 
 ## Benchmarks
 
@@ -34,9 +45,9 @@ class Array
   def stdev_ruby
     total = inject :+
     mean  = total.to_f / size
-    variance = inject(0) { |variance, value| variance + (value - mean) ** 2 }
+    variance = inject(0) { |variance, value| variance + (value - mean) ** 2 } / (size - 1)
 
-    Math.sqrt(variance / (size - 1))
+    Math.sqrt(variance)
   end
 end
 

data/ext/standard_deviation/standard_deviation.c

@@ -1,28 +1,72 @@
 #include <ruby.h>
 #include <math.h>
 
-static VALUE stdev(VALUE self) {
-  int i, size;
-  double total, mean, variance;
-
-  size  = RARRAY_LEN(self);
-  total = variance = 0;
-  VALUE *array = RARRAY_PTR(self);
+static double sum(VALUE *array, int size) {
+  int i;
+  double total = 0;
 
   for (i = 0; i < size; i++) {
     total += NUM2DBL(array[i]);
   }
 
-  mean = total / size;
+  return total;
+}
+
+static double distance_from_mean(VALUE *array, int size) {
+  int i;
+  double mean, distance;
 
-  for (i = 0; i < size; i++) {
-    variance += pow((NUM2DBL(array[i]) - mean), 2);
+  mean = sum(array, size) / size;
+
+  for (i = 0, distance = 0; i < size; i++) {
+    distance += pow((NUM2DBL(array[i]) - mean), 2);
   }
 
-  return rb_float_new(sqrt(variance / (size - 1)));
+  return distance;
+}
+
+static double compute_sample_variance(VALUE *array, int size) {
+  return distance_from_mean(array, size) / (size - 1);
+}
+
+static double compute_population_variance(VALUE *array, int size) {
+  return distance_from_mean(array, size) / size;
+}
+
+static VALUE sample_variance(VALUE self) {
+  int size = RARRAY_LEN(self);
+  VALUE *array = RARRAY_PTR(self);
+
+  return rb_float_new(compute_sample_variance(array, size));
+}
+
+static VALUE population_variance(VALUE self) {
+  int size = RARRAY_LEN(self);
+  VALUE *array = RARRAY_PTR(self);
+
+  return rb_float_new(compute_population_variance(array, size));
+}
+
+static VALUE stdev(VALUE self) {
+  int size = RARRAY_LEN(self);
+  VALUE *array = RARRAY_PTR(self);
+
+  return rb_float_new(sqrt(compute_sample_variance(array, size)));
+}
+
+static VALUE stdevp(VALUE self) {
+  int size = RARRAY_LEN(self);
+  VALUE *array = RARRAY_PTR(self);
+
+  return rb_float_new(sqrt(compute_population_variance(array, size)));
 }
 
 void Init_standard_deviation() {
   rb_define_method(rb_cArray, "stdev", stdev, 0);
   rb_define_alias(rb_cArray, "standard_deviation", "stdev");
+  rb_define_method(rb_cArray, "stdevp", stdevp, 0);
+
+  rb_define_method(rb_cArray, "sample_variance", sample_variance, 0);
+  rb_define_alias(rb_cArray, "variance", "sample_variance");
+  rb_define_method(rb_cArray, "population_variance", population_variance, 0);
 }

data/lib/standard_deviation/version.rb

@@ -1,3 +1,3 @@
 module StandardDeviation
-  VERSION = "0.0.1"
+  VERSION = "1.0.0"
 end

data/spec/standard_deviation/standard_deviation_spec.rb

@@ -19,4 +19,61 @@ describe Array do
       it { should == 9.32559202767667 }
     end
   end
+
+  describe "#stdevp" do
+    subject { values.stdevp }
+
+    context "with integer values" do
+      let(:values) { [1, 2, 6, 3, -4, 23] }
+      it { should == 8.513061859414755 }
+    end
+
+    context "with float values" do
+      let(:values) { [1.0, 2.0, 6.0, 3.0, -4.0, 23.0] }
+      it { should == 8.513061859414755 }
+    end
+
+    context "with bigdecimal values" do
+      let(:values) { [BigDecimal("1.0"), BigDecimal("2.0"), BigDecimal("6.0"), BigDecimal("3.0"), BigDecimal("-4.0"), BigDecimal("23.0")] }
+      it { should == 8.513061859414755 }
+    end
+  end
+
+  describe "#sample_variance" do
+    subject { values.sample_variance }
+
+    context "with integer values" do
+      let(:values) { [1, 2, 6, 3, -4, 23] }
+      it { should == 86.96666666666665 }
+    end
+
+    context "with float values" do
+      let(:values) { [1.0, 2.0, 6.0, 3.0, -4.0, 23.0] }
+      it { should == 86.96666666666665 }
+    end
+
+    context "with bigdecimal values" do
+      let(:values) { [BigDecimal("1.0"), BigDecimal("2.0"), BigDecimal("6.0"), BigDecimal("3.0"), BigDecimal("-4.0"), BigDecimal("23.0")] }
+      it { should == 86.96666666666665 }
+    end
+  end
+
+  describe "#population_variance" do
+    subject { values.population_variance }
+
+    context "with integer values" do
+      let(:values) { [1, 2, 6, 3, -4, 23] }
+      it { should == 72.47222222222221 }
+    end
+
+    context "with float values" do
+      let(:values) { [1.0, 2.0, 6.0, 3.0, -4.0, 23.0] }
+      it { should == 72.47222222222221 }
+    end
+
+    context "with bigdecimal values" do
+      let(:values) { [BigDecimal("1.0"), BigDecimal("2.0"), BigDecimal("6.0"), BigDecimal("3.0"), BigDecimal("-4.0"), BigDecimal("23.0")] }
+      it { should == 72.47222222222221 }
+    end
+  end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: standard_deviation
 version: !ruby/object:Gem::Version
-  version: 0.0.1
+  version: 1.0.0
   prerelease: 
 platform: ruby
 authors:
@@ -13,7 +13,7 @@ date: 2012-04-13 00:00:00.000000000Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rake-compiler
-  requirement: &2156579940 !ruby/object:Gem::Requirement
+  requirement: &2161598240 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -21,10 +21,10 @@ dependencies:
         version: '0'
   type: :development
   prerelease: false
-  version_requirements: *2156579940
+  version_requirements: *2161598240
 - !ruby/object:Gem::Dependency
   name: rspec
-  requirement: &2156579420 !ruby/object:Gem::Requirement
+  requirement: &2161597660 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -32,7 +32,7 @@ dependencies:
         version: '0'
   type: :development
   prerelease: false
-  version_requirements: *2156579420
+  version_requirements: *2161597660
 description: ! "An implementation of the standard deviation calculation in C,\n                         with
   much better performance (50x-100x) than using pure ruby."
 email:
@@ -69,7 +69,7 @@ required_ruby_version: !ruby/object:Gem::Requirement
       version: '0'
       segments:
       - 0
-      hash: -2904390813022944632
+      hash: 2357727006440476187
 required_rubygems_version: !ruby/object:Gem::Requirement
   none: false
   requirements:
@@ -78,7 +78,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
       segments:
       - 0
-      hash: -2904390813022944632
+      hash: 2357727006440476187
 requirements: []
 rubyforge_project: 
 rubygems_version: 1.8.10