Codeminer42/cm42-central

module Statistics
  def self.mean(enumerable)
    values = to_float(enumerable)
    return 0.0 if total(values) == 0.0
    sum(values) / total(values)
  end

  # assumes that enumerable is the entire population
  # sample_size is < than enumerable.size than the variance will use N - 1 to adjust for sample
  def self.variance(enumerable, sample_size = enumerable.size)
    slice = slice_to_sample_size(enumerable, sample_size)
    total = slice.size > 1 && (total(slice) < total(enumerable)) ? total(slice) - 1 : sample_size
    mean  = mean(slice)
    sum(slice.map { |sample| (mean - sample)**2 }) / total
  end

  def self.standard_deviation(enumerable, sample_size = enumerable.size)
    return 0 if enumerable.empty?
    values = to_float(enumerable)

    Math.sqrt(variance(values, sample_size))
  end

  def self.volatility(enumerable, sample_size = enumerable.size)
    slice = slice_to_sample_size(enumerable, sample_size)
    return 0 if slice.empty?
    mean = mean(slice)
    return 0 if mean.zero?

    standard_deviation(enumerable, sample_size) / mean
  end

  def self.to_float(enumerable)
    enumerable.map(&:to_f)
  end

  def self.sum(enumerable)
    enumerable.reduce(:+) || 0
  end

  def self.total(enumerable)
    enumerable.size.to_f
  end

  def self.slice_to_sample_size(enumerable, sample_size = 3)
    enumerable.reverse.take(sample_size)
  end
end