Sample Standard Deviation - Rapid Calculation Methods

Rapid Calculation Methods

It has been suggested that this article be merged into Algorithms for calculating variance. (Discuss)

The following two formulas can represent a running (continuous) standard deviation. A set of two power sums s1 and s2 are computed over a set of N values of x, denoted as x1, ..., xN:

Given the results of these three running summations, the values N, s1, s2 can be used at any time to compute the current value of the running standard deviation:

Where :

Similarly for sample standard deviation,

In a computer implementation, as the three sj sums become large, we need to consider round-off error, arithmetic overflow, and arithmetic underflow. The method below calculates the running sums method with reduced rounding errors. This is a "one pass" algorithm for calculating variance of n samples without the need to store prior data during the calculation. Applying this method to a time series will result in successive values of standard deviation corresponding to n data points as n grows larger with each new sample, rather than a constant-width sliding window calculation.

For k = 1, ..., n:

\begin{align} A_0 &= 0\\ A_k &= A_{k-1}+\frac{x_k-A_{k-1}}{k} \end{align}

where A is the mean value.

\begin{align} Q_0 &= 0\\ Q_k &= Q_{k-1}+\frac{k-1}{k} (x_k-A_{k-1})^2 = Q_{k-1}+ (x_k-A_{k-1})(x_k-A_k) \end{align}

Sample variance:

Population variance:

Read more about this topic:  Sample Standard Deviation

Famous quotes containing the words rapid, calculation and/or methods:

    “Missionaries, whether of philosophy or religion, rarely make rapid way, unless their preachings fall in with the prepossessions of the multitude of shallow thinkers, or can be made to serve as a stalking-horse for the promotion of the practical aims of the still larger multitude, who do not profess to think much, but are quite certain they want a great deal.”
    —Thomas Henry Huxley (1825–95)

    ““To my thinking” boomed the Professor, begging the question as usual, “the greatest triumph of the human mind was the calculation of Neptune from the observed vagaries of the orbit of Uranus.”
    “And yours,” said the P.B.”
    —Samuel Beckett (1906–1989)

    “Generalization, especially risky generalization, is one of the chief methods by which knowledge proceeds... Safe generalizations are usually rather boring. Delete that “usually rather.” Safe generalizations are quite boring.”
    —Joseph Epstein (b. 1937)