How to correct this error?"Subscripted assignment dimension mismatch."

Hi, I would like to create a column vector which is called ub with for loop but I got this error: Subscripted assignment dimension mismatch.

Error in Test_WSVM (line 74) ub(i,1)= C*(7/15)*ones(n,1); » Here is my code

    if true
      clc;clear;close all;

%Load Data

load Unblanced_Data.mat

    x = unblanced_data;

 %%Data Normalaization

   m=size(x,1);
   xMean = repmat(mean(x),m,1);
   xStd = repmat(std(x),m,1);
   x_norm = (x - xMean)./(xStd);

 %%Training SVM

   x_1=[x_norm(1:6400,:); x_norm(8001:13600,:)];

   y_1=[-ones(6400,1);ones(5600,1) ];

   TrainInputs = (x_1)';
   TrainTargets =(y_1)';

   ClassFault = find(TrainTargets == 1);
   ClassNormal = find(TrainTargets == -1);

   n=numel(TrainTargets);

%% Design SVM

   C=10;

   sigma=1.5;
   Kernel=@(xi,xj) exp(-1/(2*sigma^2)*norm(xi-xj)^2);

   H=zeros(n,n);

   for i=1:n
      for j=i:n
           H(i,j)=TrainTargets(i)*TrainTargets(j)*Kernel(TrainInputs(:,i),TrainInputs(:,j));
           H(j,i)=H(i,j);
       end
   end

f=-ones(n,1);

Aeq=(y_1)';
beq=0;

lb=zeros(n,1);

ub = zeros(n,1);

for i=1:n

      if (TrainTargets(1,i) == 1)

       ub(i,1)= C*(8/15)*ones(n,1); 

      else

       ub(i,1)= C*(7/15)*ones(n,1); 

      end
  end

 options = optimset('Algorithm', 'interior-point-convex','Display',...
     'off','MaxIter',20);

 alpha=quadprog(H,f,[],[],Aeq,beq,lb,ub,[],options)';
    end

Any helps would be appreciated.

Answer 1

[James Tursa]

ub(i,1) is a scalar element, but ones(n,1) is a vector if n>1. You can't assign a vector to a scalar element.

[Fizo Sd]

Thanks Mr James Tursa for your reply. n =12000 and TrainTargets is a row vector (1*12000). I tried ub(i,:) but I had this error again . I would like to have ub vector after for loop and vector size should be 12000*1 . Would you please guide me to correct my mistakes?