class ABVariant < ActiveRecord::Base
  
  set_table_name :a_b_variants
  belongs_to :test, :class_name => 'ABTest', :foreign_key => 'a_b_test_id'
  
  after_destroy :save_test
  
  validates_uniqueness_of :name
  
  def confidence
    if compute_confidence?
      cumulative_normal_distribution(z_score(self.test.control))
    else
      'n/a'
    end
  end
  
  def confidence_ok?
    cumulative_normal_distribution(z_score(self.test.control)) >= 0.95
  end
  
  def conversion_rate
    if conversions > 0
      1.0 * conversions / visitors
    else
      0.0
    end
  end
  
  def conversion_rate_ok?
    conversion_rate > self.test.control.conversion_rate
  end
  
  def pretty_confidence
    pretty confidence
  end
  
  def pretty_conversion_rate
    pretty conversion_rate
  end
  
  def suggested_visitors
    size = sample_size(self.test.control)
    if conversion_rate == 0 || size < 100
      100
    else
      commafy size
    end
  end
  
  def suggested_visitors_ok?
    visitors > sample_size(self.test.control)
  end
  
  def suggested_visitors_with_commas
    commafy suggested_visitors
  end
  
  def visitors_with_commas
    commafy visitors
  end
  
  private
  
  def commafy(num)
    num.to_s.gsub(/(\d)(?=\d{3}+(?:\.|$))(\d{3}\..*)?/,'\1,\2')
  end
  
  def cumulative_normal_distribution(z)
    b1 =  0.319381530
    b2 = -0.356563782
    b3 =  1.781477937
    b4 = -1.821255978
    b5 =  1.330274429
    p  =  0.2316419
    c  =  0.39894228

    if z >= 0.0
      t = 1.0 / (1.0 + p * z)
      (1.0 - c * Math.exp(-z * z / 2.0) * t * (t * (t * (t * (t * b5 + b4) + b3) + b2) + b1))
    else
      t = 1.0 / (1.0 - p * z)
      (c * Math.exp(-z * z / 2.0) * t * (t * (t * (t * (t * b5 + b4) + b3) + b2) + b1))
    end
  end
  
  def compute_confidence?
    self.test &&
    self.test.control && self != self.test.control &&
    self.visitors > 0
  end
  
  def pretty(num)
    if num.respond_to?(:strip)
      num
    else
      sprintf("%.3f", num)[1..-1]
    end
  end
  
  def sample_size(control)
    # confidence level is 95% and the desired power is 80%
    variance = control.conversion_rate * (1 - control.conversion_rate)
    sensitivity = (self.conversion_rate - control.conversion_rate) ** 2
    if sensitivity != 0
      (16 * variance / sensitivity).to_i
    else
      0
    end
  end
  
  def save_test
    self.test.save if self.test
  end
  
  def z_score(control)
    cr1 = control.conversion_rate
    cr2 = self.conversion_rate

    v1 = control.visitors
    v2 = self.visitors
    
    if v1 == 0.0 || v2 == 0.0
      0.0
    else
      z = cr2 - cr1
      s = (cr2 * (1 - cr2)) / v2 + (cr1 * (1 - cr1)) / v1
      z / Math.sqrt(s)
    end
  end
end