movave

Very efficient moving average filter

Description

Very efficient moving average filter implemented using convolution.

Usage:

Smoothed Data = movave(Data Vector, Averaging Window Size in Samples)

See also: slidefilter.m by the same author

• MATLAB release: MATLAB 7.0.4 (R14SP2)

[Duane Hanselman]

This function is not much more than filter(ones(1,N)/N,1,data(:)). While this may be commonly done, this simple MATLAB statement is less confusing than the submission. Plus the simple MATLAB statement runs faster, which is important to this author. I do not see a need for this function. It could have marginal value if the help text conformed to MATLAB standards, which it clearly does not.

[Hazem Baqaen]

Less confusing than Output = movave(input, window) ?? I submit that most people who need a moving average filter don't realize or don't know how to do it with convolution. This is a useful function. In this case, its simplicity is a strength, not a weakness.

[Jos van der Geest]

movave has some arithmetic problems
A = [1 1 4 1 1 1 1 1] ;
a = movave(A,3)
% -> 0.33 0.67 2.0 2.0 2.0 1.0 1.0 1.0
% rounded, but if you take a close look:
fprintf('%.20f\n',a) % ->
0.33333333333333331000
0.66666666666666663000
2.00000000000000000000
1.99999999999999980000
1.99999999999999980000
1.00000000000000000000
1.00000000000000000000
1.00000000000000000000

where as it should be simply integers (besides the edges)
[. 1 2 2 2 1 1 1 .]

Take a look at <runmean> on the FEX

[Maciej Bartkowiak]

Useless. Can be easily replaced by filter or smooth - the standard Matlab functions.