Nested For loop problem

Hello all, I am surprised by the thing which is happening to me today. As the following part of code I am using nested for loop.The thing which is troubling me that while running the value of variable Sum_error reaches to the order of 1e27.

Sum_err=0;

last_sum=0;

for k = 1:4;

      clear Sum_err;
      Sum_err=0;
      for n=1:2001
          Env(n)=Peak*exp(-(Ts(n)-Tp)/Tau(k));
          if (Env(n)<=AMt(n))
              Env(n)=AMt(n);
              Peak=AMt(n);
              Tp=Ts(n);
          end
          Sum_err=Sum_err+(Envelope(n)-Env(n))^2;

      end
      MSE(k)=Sum_err/2001;
  end

but when I iterate outer loop just once then the true value(which is expected logically) of Sum_err and MSE come likethis :

Sum_err=0;

last_sum=0;

for k = 5;

      clear Sum_err;
      Sum_err=0;
      for n=1:2001
          Env(n)=Peak*exp(-(Ts(n)-Tp)/Tau(k));
          if (Env(n)<=AMt(n))
              Env(n)=AMt(n);
              Peak=AMt(n);
              Tp=Ts(n);
          end
          Sum_err=Sum_err+(Envelope(n)-Env(n))^2;

      end
      MSE(k)=Sum_err/2001;
  end

Please answer this soon as you figure it out. I will be really grateful to you.

Moreover the full code is here for the sake of completeness.

%Matlab Assignment_1

%Code for simulation of envelope detection

clc;
clear all;
close all;
%AM signal Generation

Ka=.5;

Tm = 5e-4;          

%Time period of message signal
Ts =[0:Tm/1000:2*Tm];       
%Simulation time

Mt = cos(4000*pi*Ts);

%plot(Ts,Mt);

Ct = cos(80000*pi*Ts);

%plot(Ts,Ct);

Envelope=(Ka*Mt+1);

AMt = (Ka*Mt+1).*Ct;

%plot(Ts,AMt);

Tau = [2.5:.5:50]*1e-5;   %96 values

MSE = zeros(1,96);

%Envelope detection
%Tau=1e-4;
Env=zeros(1,2001);

Peak=0.0000;

Tp=0.0000;

Sum_err=0;

last_sum=0;

for k = 1:4;

      clear Sum_err;
      Sum_err=0;
      for n=1:2001
          Env(n)=Peak*exp(-(Ts(n)-Tp)/Tau(k));
          if (Env(n)<=AMt(n))
              Env(n)=AMt(n);
              Peak=AMt(n);
              Tp=Ts(n);
          end
          Sum_err=Sum_err+(Envelope(n)-Env(n))^2;

      end
      MSE(k)=Sum_err/2001;
  end

hold on;
plot(Tau,MSE);
%plot(Ts,Env,'color','red');

[Chaowei Chen]

I don't exactly understand what is your question,

from your code,

for k = 1:4;

clear Sum_err;

Sum_err=0;

...

end

It is expected that Sum_err is of course zero.

Answer 1

[Oleg Komarov]

You don't need the clear:

clear Sum_err;
Sum_err = 0;

Reinitialize Tp and Peak as well:

for k = 1:...
    Sum_err = 0;
    Tp      = 0;
    Peak    = 0;
    ...

You also don't need to declare them outside of the loop.

[Pratap Bhanu Solanki]

Thanks,

It really worked.