A tiny problem ...

[Dr.-Ing. Dieter Jurzitza]

Dear listmembers,
I am using a peace of code as shown below:

"signalfilt" is a low pass filtered vector of 44100 complex elements, "noise" 
a similar but uncorrelated one of the same size.

"rmsval" computes the rms value of an arbitrary (complex) input vector

****************** SNIPP **********************
NUMEL=5000;

atten=zeros(1,NUMEL);
b=signalfilt+noise;
b=b/rmsval(b);

for i=1:NUMEL
  b=shift(b,1,2);
  atten(i)=rmsval(signalfilt-b);
endfor
****************** SNAPP **********************

What I actually do here is some kind of correlation measurement I do not plan 
to dive too deep into right now. The loop takes about 6s on my computer given 
the dimensions of the vectors and numbers as listed above.

The shift operation is the most costly one in the loop, say about 10% longer 
than the rmsval of the vector difference below.

If I had to implement this in "C" I would put b two times one behind the other 
into memory, rather than actually shifting I would only increment the start 
address pointing to that new "b" as used within the loop by one for each loop, 
thereby avoiding any real shift at all. This way I could replace a shift of 
44100 elements by one single pointer increment.

However, I cannot see a way to do this in octave, as there are no pointers 
AFAIK but maybe one much more experienced user in the list has a good idea for 
me.

By the way, I am using octave 4.2, but this should not have a significant 
impact IMHO.

Thank you very much for your efforts,
regards




Dieter Jurzitza
-- 
-----------------------------------------------------------
Dr.-Ing. Dieter Jurzitza                    76131 Karlsruhe