Why is my plot not showing?

Only axes are appearing but plot is not showing. Please help me to detect the error i am making in the program. Apologies for a long code.

clear all; close all; clc;

  test_points=3;
  start_point=1000;
  x_rec(1)=x(start_point);
  x_real(1)=x(start_point);
  c1=ksaix(1);
  c2=ksaix(2);
  c3=ksaix(3);
  c4=ksaix(4);
  c5=ksaix(5);
  c6=ksaix(6);
  c7=ksaix(7);
  c8=ksaix(8);
  c9=ksaix(9);
  c10=ksaix(10);
  c11=ksaix(11);
  c12=ksaix(12);
  c13=ksaix(13);
  c14=ksaix(14);
  c15=ksaix(15);
  c16=ksaix(16);
  c17=ksaix(17);
  c18=ksaix(18);
  c19=ksaix(19);
  c20=ksaix(20);
  c21=ksaix(21);
  c22=ksaix(22);
  c23=ksaix(23);
  c24=ksaix(24);
  c25=ksaix(25);

error=0;
DN=0;
for i=4:test_points
x_rec(i)=c1+c2*x_rec(i-1)+c3*x_rec(i-1)^2+c4*x_rec(i-1)^3+c5*x_rec(i-1)^4+c6*x_rec(i-1)^5+c7*x_rec(i-1)^6+c8*x_rec(i-1)^7+c9*x_rec(i-1)^8+c10*x_rec(i-1)^9+c11*x_rec(i-1)^10+c12*x_rec(i-1)^11+c13*x_rec(i-1)^12+c14*x_rec(i-1)^13+c15*x_rec(i-1)^14+c16*x_rec(i-1)^15+c17*x_rec(i-1)^16+c18*x_rec(i-1)^17+c19*x_rec(i-1)^18+c20*x_rec(i-1)^19+c21*x_rec(i-1)^20+c22*x_rec(i-1)^21+c23*x_rec(i-1)^22+c24*x_rec(i-1)^23+c25*x_rec(i-1)^24;
error=error+(x_rec(i)-x_real(i))^2;
DN=DN+x_real(i)^2;
end
Error_x(count)=sqrt(error)/sqrt(DN);
end
figure
hold on
plot (Data,Error_x,'s-r')

[madhan ravi]

Please explain why someone should answer your question, when you’re going to edit the question completely after getting a proper answer for your homework?

[David Goodmanson]

Hello Heya,

Actually since you have put in the work and provided the code it would be fairly easy to take a look at this. Unfortunately the code is not plain vanilla Matlab since it utilizes cvx software. Could you provide a workaround?

The thing to do first, though, is carefully examine Data and Error_x. Probably the most common reason for the no-plot effect is that one of the x or y array is either the empty array or is full of NaNs.

[dpb]

It wouldn't be nearly so long if you would use MATLAB like MATLAB instead of Visual Basic...

for i=1:N_measurements
  for j=1:N_basis
    if j==1
      M(i,j)=1;
    elseif j==2
      M(i,j)=x(index(i));
    elseif j==3
      M(i,j)=x(index(i))^2;
    ...
    elseif j==24
      M(i,j)=x(index(i))^23;
    else 
      M(i,j)=x(index(i))^24;
    end    
  end
end

is

for i=1:N_measurements
  M(:,j)=x(index(i)).^[0:24];
end

keeping the I loop; it could be reduced by vectorizing the index variable over i as well.

Similarly in all the remainder with a zillion named variables instead of arrays.

[Heya :)]

Thank you!

[dpb]

  x_rec(i)=c1+c2*x_rec(i-1)   + c3*x_rec(i-1)^2 + c4*x_rec(i-1)^3 + c5*x_rec(i-1)^4 + ...
              c6*x_rec(i-1)^5 + c7*x_rec(i-1)^6 + c8*x_rec(i-1)^7 + c9*x_rec(i-1)^8 + ...
              c10*x_rec(i-1)^9+c11*x_rec(i-1)^10+c12*x_rec(i-1)^11+c13*x_rec(i-1)^12+ ...
              c14*x_rec(i-1)^13+c15*x_rec(i-1)^14+c16*x_rec(i-1)^15+c17*x_rec(i-1)^16+ ...
              c18*x_rec(i-1)^17+c19*x_rec(i-1)^18+c20*x_rec(i-1)^19+c21*x_rec(i-1)^20+ ...
              c22*x_rec(i-1)^21+c23*x_rec(i-1)^22+c24*x_rec(i-1)^23+c25*x_rec(i-1)^24;

is

x_rec(i)=polyval(flip(ksaix),x_rec(i-1));

NB: polyval is defined as

Y = C(1)*X^N + C(2)*X^(N-1) + ... + C(N)*X + C(N+1)  

so must reverse order of coefficient array.

[dpb]

As for the original question, anything that uses a polynomial to the 24th power is just asking for numerical troubles--either over- or underflow and imprecision.

As another said, probably you're ending up with NaN or Inf as results.

Answer 1

[Image Analyst]

Look at the first few lines:

clear all; close all; clc;
  test_points=3;
  start_point=1000;
  x_rec(1)=x(start_point);

Now you try to get x(start_point) but you never defined it. SO that throws an error. Or if you did (in another script), you blew it away with the clear all command.