Matlab: Optimizing energy computation code

Hello,

I really need to optimize this block of code:

for(i = 1:33)
E(frame,i)=sum(power(samp_borders(i):samp_borders(i+1)).^2);
end;

I wanted to use array indexing instead of the for loop, since it should be
faster, but everything I try does not work. power() is the array I'm
working on. It is power spectrum data.

The squaring and the summing 33 times is just taking way too long for what
I'm doing. (I'm computing the energy for 33 bands of a power spectrum, but
I'm doing it thousands of times.)

Thanks.

On 23 Jul., 09:43, "jfrog" <ajmcg...@gmail.com> wrote:
> Hello,
>
> I really need to optimize this block of code:
>
> for(i = 1:33)
> E(frame,i)=sum(power(samp_borders(i):samp_borders(i+1)).^2);
> end;
>
> I wanted to use array indexing instead of the for loop, since it should be
> faster, but everything I try does not work. power() is the array I'm
> working on. It is power spectrum data.
>
> The squaring and the summing 33 times is just taking way too long for what
> I'm doing. (I'm computing the energy for 33 bands of a power spectrum, but
> I'm doing it thousands of times.)
>
> Thanks.

As long as samp_borders is not equispaced, I see no possibility for
speeding
up this code, despite rewriting it in C.

Greetings, Uwe.

>On 23 Jul., 09:43, "jfrog" <ajmcg...@gmail.com> wrote:
>> Hello,
>>
>> I really need to optimize this block of code:
>>
>> for(i = 1:33)
>> E(frame,i)=sum(power(samp_borders(i):samp_borders(i+1)).^2);
>> end;
>>
>> I wanted to use array indexing instead of the for loop, since it should
be
>> faster, but everything I try does not work. power() is the array I'm
>> working on. It is power spectrum data.
>>
>> The squaring and the summing 33 times is just taking way too long for
what
>> I'm doing. (I'm computing the energy for 33 bands of a power spectrum,
but
>> I'm doing it thousands of times.)
>>
>> Thanks.
>
>As long as samp_borders is not equispaced, I see no possibility for
>speeding
>up this code, despite rewriting it in C.

Ok.  Thanks.  samp_borders is logarithmically spaced, though.  That
wouldn't help by any chance, would it?

On 2008-07-23, jfrog <ajmcgraw@gmail.com> wrote:
> Hello,
>
> I really need to optimize this block of code:
>
> for(i = 1:33)
> E(frame,i)=sum(power(samp_borders(i):samp_borders(i+1)).^2);
> end;
>
> I wanted to use array indexing instead of the for loop, since it should be
> faster, but everything I try does not work. power() is the array I'm
> working on. It is power spectrum data.

You can certainly move the squaring out of the loop, which will help
unless you're only using a very sparse subset of it, so:

power2 = power .^ 2;

Hoewver, assuming samp_borders is constant, you can probably do better
by constructing a mapping matrix from it and re-using that.

# you need to know length(power) in advance, but I don't know:
samp_borders_map = zeros(length(power), 33);
for (i = 1:33)
samp_borders_map(samp_borders(i):samp_borders(i+1),i) += 1;
end

Then your equation becomes:

E(frame)=(power.^2) * samp_borders_map;

-- 
Ben Jackson AD7GD
<ben@ben.com>
http://www.ben.com/