MATLAB: Error using ==> horzcat CAT arguments dimensions are not consistent. Help please !

axishorzcatMATLAB

Hello Matlab community !
I am working on kinematics for a 3DOF finger. I recently made a small program which moves the finger from one spot to another.For example from spot A(x1,y1) to B(x2,y2).
The problem i have with the program is that if x1=x2 or y1=y2 I get an error which says "
"??? Error using ==> horzcat
CAT arguments dimensions are not consistent.
Error in ==> move at 72
C=[ones(size(B))*A1,A2] "
This is the program i have made.
————————————————————
function move(X1,Y1,X2,Y2,phi)
%example  move(8,4,4,3,0)
%dimensions of the finger bones
l1=3.9;
l2=6.2;
l3=4.7;
%i do that cause i don't know if i will move from A->B or B->A
GB=abs(X1-X2);
AG=abs(Y1-Y2);
%it is like I  create a triangle ABG and i have, AG=Y=|Y1-Y2|
and GB=X=|X1-X2|
   B=20;
        %i do that cause i want to have 20 points from A->G and G->B
        if X1>X2;
            A1=Linspace(X1,X2,B);    
       elseif X1<X2;
            A1=Linspace(X1,X2,B);    
        else
            A1=[ones(size(B))*X1];  
        end
        A1=A1'
        if Y1>Y2;
             A2=Linspace(Y1,Y2,B);  
         elseif Y1<Y2;
             A2=Linspace(Y1,Y2,B);   
        else
            A2=[ones(size(B))*Y1];  
        end
        A2=A2'
        %--------------------------------------------------------------------
% C= [20 points GB , 20 points AG]
C=[ones(size(B))*A1,A2]
%----------------------------------------------------------------
%--------------------------kinematics-------------------------
i=1;
for i=1:B
      Px=C((i),1); 
      Py=C((i),2); 
      %respectively (x,y) 
      phi=phi*pi/180;
      sigma=(Px^2+Py^2+l3^2-l1^2-l2^2-2*l3*(Px*cos(phi)+Py*sin(phi)))/2/l1/l2;
      Q2=atan2(+sqrt(1-sigma^2),sigma);
      num1=(Py-l3*sin(phi)) * (l1+l2*cos(Q2)) - (Px-l3*cos(phi))*l2*sin(Q2);
      den=(l1^2+l2^2+2*l1*l2*cos(Q2));
      num2=(Px-l3*cos(phi))*(l1+l2*cos(Q2))+(Py-l3*sin(phi))*l2*sin(Q2);
      Q1=atan2(num1/den,num2/den);
      Q3=phi-Q1-Q2;
      Q1=Q1*180/pi;
      Q2=Q2*180/pi;
      Q3=Q3*180/pi;
      Q((i),1)=Q1;        
      Q((i),2)=Q2; 
end
grid on 
j=1;
for j=1:B,
    q1=Q((j),1); 
    q2=Q((j),2); 
    q3=Q((j),3); 
      q1=q1*pi/180;
      q2=q2*pi/180;
      q3=q3*pi/180;
      x=l1*cos(q1)+l2*cos(q1+q2)+l3*cos(q1+q2+q3);
      y=l1*sin(q1)+l2*sin(q1+q2)+l3*sin(q1+q2+q3);
      f=q1+q2+q3;
      x1=l1*cos(q1);
      x2=l1*cos(q1)+l2*cos(q1+q2);
      y1=l1*sin(q1);
      y2=l1*sin(q1)+l2*sin(q1+q2);
      line( [0 x1 x2 x], [0 y1 y2 y], [ 0 0 0 0 ])
      hold on  
      plot(0,0,'o')
      plot(x1,y1,'o')
      plot(x2,y2,'o')
      plot(x,y,'o','LineWidth',30)
      AXIS([-3 15.13 -3 15.13])
      pause(0.2)
end
end
Thanks a lot in advance !

Best Answer

When x1=x2 but y1~=y2, your A1 is a scalar but A2 is a column vector, hence it cannot be concatenated. Similar thing for x1~=x2 but y1=y2. But you should be able to easily figure this out if you simply put a breakpoint at the error out line and test the dimension of A1 and A2.
Related Question