ruby-statistics 2.0.4 → 2.0.5

checksums.yaml

@@ -1,7 +1,7 @@
 ---
-SHA1:
-  metadata.gz: 298bc7d8dff1aeabc7db9c11fe9d7987f16bde40
-  data.tar.gz: 1d796e62c18052f87fc2616b4c1a5f777080c1ab
+SHA256:
+  metadata.gz: 1e73ce22e1ad6da4f9d2925ed5884d6e72d915ebad14b56b9e81c64c422c14cf
+  data.tar.gz: 18fa7cafd8bbf457dd8963b2e666525964ad588257ffc8774cee2d1250f4b469
 SHA512:
-  metadata.gz: 98e8c58f34668e839be9689c74debd75bd7a6869372536d7e9927a63f77fca59ab05e06b413705f0d286094292cb566c01e6fe71145cdd7d2152fc930829910e
-  data.tar.gz: 37b78191adb8d659f21134346a8a415c5bd7bd8a7dd99b2c1f8d7793a2ea741c43e60d8235a7d5fcc2bc0284b24e8e58e8404e0b4b4401ee3bc60f7e1afc8b8b
+  metadata.gz: d362a9a2a5ea950ccc37ca5754dca0644040f49563afb14bbd10792ac2ef1f13853724cb693107fa6eb571972432026c5b3cf363ea4932513360f33513147401
+  data.tar.gz: cd5dd0a2b386fcaa0190d369a1f9d40248d4ed315e165df43403884cb33438902dd080a3f1579ba498eb88a5f2936ba91f6cd4a9e31a948ac31e2d124ba232fe

data/.travis.yml

@@ -1,8 +1,7 @@
 sudo: false
 language: ruby
 rvm:
-  - 2.2
-  - 2.3.1
-  - 2.4.0
-  - 2.5.0
+  - 2.3.7
+  - 2.4.4
+  - 2.5.1
 before_install: gem install bundler

data/lib/statistics/distribution/bernoulli.rb

@@ -0,0 +1,35 @@
+module Statistics
+  module Distribution
+    class Bernoulli
+      def self.density_function(n, p)
+        return if n != 0 && n != 1 # The support of the distribution is n = {0, 1}.
+
+        case n
+        when 0 then 1.0 - p
+        when 1 then p
+        end
+      end
+
+      def self.cumulative_function(n, p)
+        return if n != 0 && n != 1 # The support of the distribution is n = {0, 1}.
+
+        case n
+        when 0 then 1.0 - p
+        when 1 then 1.0
+        end
+      end
+
+      def self.variance(p)
+        p * (1.0 - p)
+      end
+
+      def self.skewness(p)
+        (1.0 - 2.0*p).to_f / Math.sqrt(p * (1.0 - p))
+      end
+
+      def self.kurtosis(p)
+        (6.0 * (p ** 2) - (6 * p) + 1) / (p * (1.0 - p))
+      end
+    end
+  end
+end

data/lib/statistics/distribution/geometric.rb

@@ -0,0 +1,76 @@
+module Statistics
+  module Distribution
+    class Geometric
+      attr_accessor :probability_of_success, :always_success_allowed
+
+      def initialize(p, always_success: false)
+        self.probability_of_success = p.to_f
+        self.always_success_allowed = always_success
+      end
+
+      def density_function(k)
+        k = k.to_i
+
+        if always_success_allowed
+          return if k < 0
+
+          ((1.0 - probability_of_success) ** k) * probability_of_success
+        else
+          return if k <= 0
+
+          ((1.0 - probability_of_success) ** (k - 1.0)) * probability_of_success
+        end
+      end
+
+      def cumulative_function(k)
+        k = k.to_i
+
+        if always_success_allowed
+          return if k < 0
+
+          1.0 - ((1.0 - probability_of_success) ** (k + 1.0))
+        else
+          return if k <= 0
+
+          1.0 - ((1.0 - probability_of_success) ** k)
+        end
+      end
+
+      def mean
+        if always_success_allowed
+          (1.0 - probability_of_success) / probability_of_success
+        else
+          1.0 / probability_of_success
+        end
+      end
+
+      def median
+        if always_success_allowed
+          (-1.0 / Math.log2(1.0 - probability_of_success)).ceil - 1.0
+        else
+          (-1.0 / Math.log2(1.0 - probability_of_success)).ceil
+        end
+      end
+
+      def mode
+        if always_success_allowed
+          0.0
+        else
+          1.0
+        end
+      end
+
+      def variance
+        (1.0 - probability_of_success) / (probability_of_success ** 2)
+      end
+
+      def skewness
+        (2.0 - probability_of_success) / Math.sqrt(1.0 - probability_of_success)
+      end
+
+      def kurtosis
+        6.0 + ((probability_of_success ** 2) / (1.0 - probability_of_success))
+      end
+    end
+  end
+end

data/lib/statistics/distribution/logseries.rb

@@ -0,0 +1,51 @@
+module Statistics
+  module Distribution
+    class LogSeries
+      def self.density_function(k, p)
+        return if k <= 0
+        k = k.to_i
+
+        left = (-1.0 / Math.log(1.0 - p))
+        right = (p ** k).to_f
+
+        left * right / k
+      end
+
+      def self.cumulative_function(k, p)
+        return if k <= 0
+
+        # Sadly, the incomplete beta function is converging
+        # too fast to zero and breaking the calculation on logs.
+        # So, we default to the basic definition of the CDF which is
+        # the integral (-Inf, K) of the PDF, with P(X <= x) which can
+        # be solved as a summation of all PDFs from 1 to K. Note that the summation approach
+        # only applies to discrete distributions.
+        #
+        # right = Math.incomplete_beta_function(p, (k + 1).floor, 0) / Math.log(1.0 - p)
+        # 1.0 + right
+
+        result = 0.0
+        1.upto(k) do |number|
+          result += self.density_function(number, p)
+        end
+
+        result
+      end
+
+      def self.mode
+        1.0
+      end
+
+      def self.mean(p)
+        (-1.0 / Math.log(1.0 - p)) * (p / (1.0 - p))
+      end
+
+      def self.variance(p)
+        up = p + Math.log(1.0 - p)
+        down = ((1.0 - p) ** 2) * (Math.log(1.0 - p) ** 2)
+
+        (-1.0 * p) * (up / down.to_f)
+      end
+    end
+  end
+end

data/lib/statistics/distribution/negative_binomial.rb

@@ -0,0 +1,51 @@
+module Statistics
+  module Distribution
+    class NegativeBinomial
+      attr_accessor :number_of_failures, :probability_per_trial
+
+      def initialize(r, p)
+        self.number_of_failures = r.to_i
+        self.probability_per_trial = p
+      end
+
+      def probability_mass_function(k)
+        return if number_of_failures < 0 || k < 0 || k > number_of_failures
+
+        left = Math.combination(k + number_of_failures - 1, k)
+        right = ((1 - probability_per_trial) ** number_of_failures) * (probability_per_trial ** k)
+
+        left * right
+      end
+
+      def cumulative_function(k)
+        return if k < 0 || k > number_of_failures
+        k = k.to_i
+
+        1.0 - Math.incomplete_beta_function(probability_per_trial, k + 1, number_of_failures)
+      end
+
+      def mean
+        (probability_per_trial * number_of_failures)/(1 - probability_per_trial).to_f
+      end
+
+      def variance
+        (probability_per_trial * number_of_failures)/((1 - probability_per_trial) ** 2).to_f
+      end
+
+      def skewness
+        (1 + probability_per_trial).to_f / Math.sqrt(probability_per_trial * number_of_failures)
+      end
+
+      def mode
+        if number_of_failures > 1
+          up = probability_per_trial * (number_of_failures - 1)
+          down = (1 - probability_per_trial).to_f
+
+          (up/down).floor
+        elsif number_of_failures <= 1
+          0.0
+        end
+      end
+    end
+  end
+end

data/lib/statistics/version.rb

@@ -1,3 +1,3 @@
 module Statistics
-  VERSION = "2.0.4"
+  VERSION = "2.0.5"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: ruby-statistics
 version: !ruby/object:Gem::Version
-  version: 2.0.4
+  version: 2.0.5
 platform: ruby
 authors:
 - esteban zapata
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2018-05-18 00:00:00.000000000 Z
+date: 2018-07-04 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -137,10 +137,14 @@ files:
 - lib/math.rb
 - lib/statistics.rb
 - lib/statistics/distribution.rb
+- lib/statistics/distribution/bernoulli.rb
 - lib/statistics/distribution/beta.rb
 - lib/statistics/distribution/binomial.rb
 - lib/statistics/distribution/chi_squared.rb
 - lib/statistics/distribution/f.rb
+- lib/statistics/distribution/geometric.rb
+- lib/statistics/distribution/logseries.rb
+- lib/statistics/distribution/negative_binomial.rb
 - lib/statistics/distribution/normal.rb
 - lib/statistics/distribution/poisson.rb
 - lib/statistics/distribution/t_student.rb
@@ -173,7 +177,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.5.2.1
+rubygems_version: 2.7.3
 signing_key: 
 specification_version: 4
 summary: A ruby gem for som specific statistics. Inspired by the jStat js library.