Asked by Dan
on 29 Sep 2013

%Number of components K = zeros(1,111);

%sampling frequency/timing interval fs = 8000; Fs = 10*fs; T = 2;

%frequency components fk =100:30:3400;

%amplitude components Ak = rand(1,length(K));

%phase components phik = 2*pi*rand(1,length(K));

for t = 0:(1/Fs):T %time interval K = Ak.*cos(2*pi.*fk*t + phik); y = t/(1/Fs)+1; xt(1,y) = sum(K,2); end

I'm trying to run this code but for some reason there is an error at y = 14 which is no where near the end of the loop. I'm pretty sure that xt(1,14) is a valid index. Any reason why this might be happening?

*No products are associated with this question.*

Answer by dpb
on 29 Sep 2013

Edited by dpb
on 30 Sep 2013

The answer is in the following--

>> fs = 8000; >> Fs = 10*fs; >> t=0:1/Fs:2; >> y = t/(1/Fs)+1; >> all(fix(y)==y) ans = 0 >> neq=find(fix(y)~=y); >> neq(1) ans = 14 >> y(14)-fix(y(14)) ans = 1.7764e-15 >> length(neq) ans = 39209 >> length(y) ans = 160001 >>

Moral: Don't use FP calculations as indices...

Answer by Wayne King
on 29 Sep 2013

Edited by Wayne King
on 29 Sep 2013

I'm not sure what you are trying to do, create a superposition of sine waves with different frequencies and a random phase and amplitudes? If so, I'll show you how to do that.

But your immediate issue here is that y=0 when t = 0 and then you try to access the zero-th column of xt(), but that can't exist.

You can do this.

Fs = 8000; t = (0:1/Fs:2-1/Fs)';

fk = 100:30:3400; X = zeros(length(t),length(fk));

phi = -pi+2*pi*rand(1,length(fk));

Ak = rand(1,length(fk));

tmat = repmat(t,1,length(fk)); omega = 2*pi*fk; omega = repmat(omega,size(tmat,1),1); phi = repmat(phi,size(tmat,1),1); freq_phi = omega.*tmat+phi; Ak = repmat(Ak,size(tmat,1),1); sigz = Ak.*cos(freq_phi); sig = sum(sigz,2);

Now plot the resulting signal

plot(t,sig)

Opportunities for recent engineering grads.

## 0 Comments