finding the mean for large data

FIR (view profile)

on 30 Nov 2012

I have 3 rows ans 192 column,i want to find mean for each 9 columns in row wise,

for ex if i have rand(3,192)

taking 3x9 , i need mean in row wise so i will have 3x1 matrix after finding mean

for each 3x9 i want to perform

Jos (10584)

Jos (10584) (view profile)

on 30 Nov 2012

192 is not divisible by 9, so what to do with the last columns?

Image Analyst

Image Analyst (view profile)

on 30 Nov 2012

It's not "large" either. Before I read it I thought that he was talking hundreds of millions of elements, not five hundred and something. (And there is a "bug" for the case of calculating the mean for large numbers of singles)

Products

No products are associated with this question.

Andrei Bobrov (view profile)

on 30 Nov 2012
Edited by Andrei Bobrov

Andrei Bobrov (view profile)

on 30 Nov 2012
```A = rand(3,192);
n = 9;
```
```s = size(A);
out = squeeze(nanmean(reshape([A nan(s(1),mod(-s(2),n))],s(1),n,[]),2));
```

or

```out = blockproc(A,[1 n],@(x)mean(x.data));
```

or

```k = s(2) - rem(s(2),n);
out = [squeeze(mean( reshape(A(:,1:k),s(1),[]),2 )) , mean(A(:,k+1:end),2)];
```

Jan Simon (view profile)

on 30 Nov 2012
```A = rand(3, 192);
B = reshape(A(:, 1:189), 3, 9, []);
C = squeeze(mean(B, 2));
```