to find average value in a particular range of a matrix

hai

i am having a matrix with values of size 1*700 i have to find the avg value of first 1 to 100 ,avg value of 100 - 200 and so on upto 600 -700 is there any function to find the average of 1 -100 and 100 - 200, 200 - 300,300 - 400,400 - 500,500 - 600,600 - 700,

Answer 1

[Andrei Bobrov]

variant 1 for the intervals: 1:100, 101:200,...,601:700

out = mean(reshape(yourmtx,100,[]));

variant 2 for the intervals: 1:100, 101:200,...,601:700

out = blkproc(yourmtx,[1 100],@mean)

variant 3 for the intervals: 1:100, 100:200,...,600:700

idx = bsxfun(@plus,0:100,(0:100:600)')';
idx(1) = 1;
mv = yourmtx(idx);;
mv(1)=nan;
out = nanmean(mv);

variant 4 for the intervals: 1:100, 100:200,...,600:700

out = mean(reshape(yourmtx,100,[]));
out(2:end) = mean([yourmtx(100:100:600);out(2:end)]);

[Daniel]

Hello andrei.

I am having more or less the same problem.

But I need to extract mean values within a certain range of a matrix. This range is given by each column's min and max value, and a step.

Basically I've done the next:

1. I have a 2x720 matrix with data, let's call it A.

2. I calculate a = max(A), b = min(A).

3. I apply the histcnd function (which counts frequency of values within certain edges or range).

4. The edges = {b(1):2:a(1),b(2):10:a(2)}, that is from min and max value of each column of A, with a certain step (2 and 10).

This returns an answer matrix, whose dimension depends on the steps given.

This matrix needs to be divised by certain factors, but then, I have to re-multiply it to the values of the first matrix A. The problem? Matrixs dimension do not fit to do so. The solution would be to obtain a new matrix from A, containing mean values within each column given range (the ones used on 'edges').

Is this posible?

Thanks a lot!