src/cobalt/webdriver_benchmarks/container_util.py

"""Base class for WebDriver tests."""

from __future__ import absolute_import
from __future__ import division
from __future__ import print_function

import math


def mean(list_values):
  """Returns the mean of a list of numeric values.

  Args:
      list_values: A list of numeric values.
  Returns:
      Appropriate value.
  """
  if not list_values:
    return None

  return sum(list_values) / len(list_values)


def percentile(list_values, target_percentile):
  """Returns the percentile of a list.

  This method interpolates between two numbers if the percentile lands between
  two data points.

  Args:
      list_values: A sortable list of values
      target_percentile: A number ranging from 0-100.
  Returns:
      Appropriate value.
  Raises:
    RuntimeError: Raised on invalid args.
  """
  if not list_values:
    return None
  if target_percentile > 100 or target_percentile < 0:
    raise RuntimeError("target_percentile must be 0-100")
  sorted_values = sorted(list_values)

  if target_percentile == 100:
    return sorted_values[-1]
  fractional, index = math.modf(
      (len(sorted_values) - 1) * (target_percentile * 0.01))
  index = int(index)

  if len(sorted_values) == index + 1:
    return sorted_values[index]

  return sorted_values[index] * (
      1 - fractional) + sorted_values[index + 1] * fractional


def sample_variance(list_values):
  """Returns the variance of a list of numeric values.

  Args:
      list_values: A list of numeric values.
  Returns:
      Appropriate value.
  """
  if not list_values:
    return None
  if len(list_values) <= 1:
    return 0

  mean_of_values = mean(list_values)
  dif_squared_sum = 0
  for value in list_values:
    dif = value - mean_of_values
    dif_squared_sum += dif * dif

  return dif_squared_sum / (len(list_values) - 1)


def sample_standard_deviation(list_values):
  """Returns the standard deviation of a list of numeric values.

  Args:
      list_values: A list of numeric values.
  Returns:
      Appropriate value.
  """
  if not list_values:
    return None

  return math.sqrt(sample_variance(list_values))