`movvar`

Moving variance over a sliding window.

Category: Data Import and Analysis > Descriptive Statistics

Syntax

V = movvar(A, wlen)

Variance over a window of wlen elements centered on each element.

V = movvar(A, [nb na])

Window of nb elements before and na after the current one, plus the element itself.

Description

movvar computes var inside each window, normalized by the number of elements minus one. It is the square of movstd. The result is zero where the window is constant or holds a single element.

Windows shrink at the ends of the data, so the output has the same size as the input.

Input arguments

A: Vector or matrix; operates along the first non-singleton dimension.
wlen: Window length.

Output arguments

V: Array of the same size as A.

Example

Rolling variance

x = [3 1 4 1 5 9 2 6 5 3];
movvar(x, 3)

Output

ans = 1×10

    2.0000    2.3333    3.0000    4.3333   16.0000   12.3333   12.3333    4.3333    2.3333    2.0000

movvar