[backtrader source code analysis 7] analysis of the functions for calculating mean value, variance and standard deviation in mathsupport in backtrader (with low gold content)

The previous articles tried to optimize backtrader Several time processing functions to improve efficiency , Use cython It can improve the efficiency of a single function, but in the whole, after the efficiency decreases and leads to failure , I tried again numba To improve those time processing functions , And I failed . At present, there is no better method to improve those time conversion functions , Which great God has a good way to help , Help each other . You can pay .
Next, this article analyzes several mathematical support functions ： Average 、 variance 、 Standard deviation , On the whole, there is no gold content . In the next article , Focus on parsing a key source file ：metabase.py

#!/usr/bin/env python
# -*- coding: utf-8; py-indent-offset:4 -*-
from __future__ import (absolute_import, division, print_function,
                        unicode_literals)

import math

#  Look at the , These functions are mainly used to calculate some indicators , Not used in the main body , Comment on , Come back later to see if you need to use cython improvement , There is no need to improve for the time being .
#  But these functions can actually be considered numpy To improve the ,numpy Specific functions are provided to calculate the mean value , Calculate the standard deviation 

#  This calculation is the average , With a parameter bessel, Used to determine whether the value of the denominator subtracts one... When calculating the average value . Molecular use math.fsum For calculation and 
def average(x, bessel=False):
    '''
    Args:
      x: iterable with len

      oneless: (default ``False``) reduces the length of the array for the
                division.

    Returns:
      A float wi