Averaging data

Hi,

How to average 1s long samples of a data set? I have a data in a form t = [0 1 2 3...5000...10000 ...10000] d = [5 5 5 4... 1... 6...8 ]

where t represents milliseconds and d represents specific values for each millisecond. I'd like to have t in seconds so then values from d should be averaged every 999ms. I mean e.g. for t=0s d=5, for t=1s, d should be an average of values form 1 to 999ms and so on.

Thanks for help.

Answer 1

[Wayne King]

One of many ways:

If the spacing between elements in your d vector is 1 msec, then you can reshape the data vector into a matrix with columns equal to 1000 rows and take the average of the columns.

Of course, if the vector does not reshape evenly in equal-length columns, you have to think about what to do at the end.

Assume two seconds of data:

x = randn(2000,1);
x1 = reshape(x,1000,2);
mean(x1)

Another possibility is to create a simple function to do it and call that function from the command line. You have to save the function as nway.m in a folder that is on your MATLAB path.

function B = nway(A,n)
% Compute average of every N elements of A and put them in B.
if ((mod(numel(A),n) == 0) && (n>=1 && n<=numel(A)))
    B = ones(1,numel(A)/n);
    k = 1;
    for i = 1 : numel(A)/n
         B(i) = mean(A(k + (0:n-1)));
         k = k + n;
    end
else
    error('n <= 0 or does not divide number of elements evenly');
end

then,

>>meanz = nway(x,1000);

gives you the same result that I showed above.

[Marcel]

Thanks very much for your help