Function skewness

pattern/metrics.py:605–613

Returns the degree of asymmetry of the given list of values: > 0.0 => relatively few values are higher than mean(list), < 0.0 => relatively few values are lower than mean(list), = 0.0 => evenly distributed on both sides of the mean (= normal distribution).

(list)

    return sum([(x-m)**k for x in list]) / (len(list) or 1)
    
def skewness(list):
    """ Returns the degree of asymmetry of the given list of values:
        > 0.0 => relatively few values are higher than mean(list),
        < 0.0 => relatively few values are lower than mean(list),
        = 0.0 => evenly distributed on both sides of the mean (= normal distribution).
    """
    # Distributions with skew and kurtosis between -1 and +1 
    # can be considered normal by approximation.
    return moment(list, 3) / (moment(list, 2) ** 1.5 or 1)

skew = skewness