How can I fix this problem?

Hello everyone, I wrote an analysis code of a conveyor line (it looks like crank-connecting rod mechanism) but I saw these mistakes in picture. How can I fix the problem? You can see my code in attach. Thanks a lot.

clc;
clear;
clear all;

global L1 L2 L3 L4 L5 theta1 theta2 theta3 theta4 w1

L1=1; L2=4; L3=3; L4=3.2; L5=6.5; %metre cinsinden uzuv uzunlukları
theta1=165*(pi/180); theta2=(0:1:360)*(pi/180); theta3=(129.732)*(pi/180); theta4=(14.12)*(pi/180); %radyan cinsinden açılar
ar=6:length(theta2)-5

N = length(theta2); %adım sayısı
x = zeros (4,N); %bilinmeyen açılar bulunur bu matrise kaydedilir

for i = 1:N
    theta2 = theta2(i); %radyan cinsinden Teta_2 açısı

    if i == 1
        x0 = 1*[165*pi/180;0*pi/180;129.732*pi/180;14.12*pi/180]
    else
        x0 = x(:,i-1);
    end
    opt=optimset('TolFun',1e-6,'TolX',1e-6,'Display','off');
    x(:,i) = fsolve(@fonk1,x0,opt);
end

theta1 = x(1,:);
theta2 = x(2,:);
theta3 = x(3,:);
theta4 = x(4,:);
[theta1(ar)' theta2(ar)' theta3(ar)' theta4(ar)']*180/pi; %derece cinsinden açılar

% ANÄ°MASYON
%figure

 for i = 1:N;
     A = L1*[cos(theta2(i)) sin(theta2(i))];
     B = L2*[cos(theta3(i)) sin(theta3(i))];
     C = L3*[cos(theta1(i)) sin(theta1(i))];
     D = L4*[cos(theta3(i)) sin(theta3(i))];
     E = L5*[cos(theta4(i)) sin(theta4(i))];
     Ax(i)=A(1); Ay(i)=A(2);
     Bx(i)=B(1); By(i)=B(2);
     Cx(i)=C(1); Cy(i)=Cy(2);
     Dx(i)=D(1); Dy(i)=Dy(2);
     Ex(i)=E(1); Ey(i)=Ey(2);

     AB = [A;B]';line(AB(1,:),AB(2,:))
     BC = [B;C]';line(BC(1,:),BC(2,:))
     CD = [C;D]';line(CD(1,:),CD(2,:))
     DE = [D;E]';line(DE(1,:),DE(2,:))
     BE = [B;E]';line(BE(1,:),BE(2,:))

     text(A(1),A(2),'A')
     text(B(1),B(2),'B')
     text(C(1),C(2),'C')
     text(D(1),D(2),'D')
     text(E(1),E(2),'E')

     axis([-5000 5000 -5000 5000])
     drawnow
     if i<N, clf;
     end
 end

 k = 180/pi; %radyandan dereceye dönüştürme
 hold on;
 plot (Bx,By,'k.' , Ax,Ay,'k.')
 figure;
 plot (k*theta3,Bx,'k.' , k*theta3,By,'k.')
 xlabel('theta3(derece)');
 legend('B_x' , 'B_y')
 figure;
 plot (k*theta3,Ax,'k.' , k*theta3,Ay,'c.')
 xlabel('theta3(derece)');
 legend('A_x' , 'A_y')
 figure;

 k = 180/pi; %radyandan dereceye dönüştürme
 plot(k*theta1(ar),k*theta2(ar));
 xlabel('theta2(derece)');
 ylabel('theta1(derece)');

% AÇISAL HIZLAR
 W1 = 1*ones(1,N);
 for i = 1:n
     w1 = W1(i);
     theta1=theta1(i);
     theta2=theta2(i);
     theta3=theta3(i);
     theta4=theta4(i);
     if i == 1
         x0 = [0 0 0 0] %Başlangıç değerleri
     else
         x0 = x(:, i-1);
     end
 end

 W1 = x(1,:); W2 = x(2,:); W3 = x(3,:); W4 = x(4,:); W5 = x(5,:);
 [(180/pi)*theta1(ar)' W1(ar)' W2(ar)' W3(ar)' W4(ar)' W5(ar)'] %açısal hızlar (rad/s)

 fu=gradient(theta3,theta1);
 W3=W1.*fu;

 fu=gradient(theta2,theta1);
 W2=W1.*fu

 figure;
 subplot(1,2,1);
 plot(theta1(ar)*k, W3(ar), 'k.');
 legend('W_3')
 subplot(1,2,2);
 plot(theta1(ar)*k,W2(ar), 'k.');
 legend('W_2')

function F = fonk1(x) %konum analizi
global L1 L2 L3 L4 L5 theta1 theta2 theta3 theta4 w1
W1=x(1); W2=x(2); W3=x(3); W4=x(4); W5=x(5);
j = sqrt(-1);

CD1 = L1*j*W1*exp(j*theta2) + L2*j*W2*exp(j*theta3) + L3*j*W3*exp(j*theta1);
CD2 = L3*j*W3*exp(j*theta1) + L4*j*W4*exp(j*theta4) + L5*j*W5*exp(j*theta4);

F = [real(CD1);
    imag(CD1);
    real(CD2);
    imag(CD2)];
end

function F = fonk2(x) %hız analizi
W1=x(1); W2=x(2); W3=x(3); W4=x(4); W5=x(5);
j = sqrt(-1);

CD1 = L1*j*W1*exp(j*theta2) + L2*j*W2*exp(j*theta3) + L3*j*W3*exp(j*theta1);
CD2 = L3*j*W3*exp(j*theta1) + L4*j*W4*exp(j*theta4) + L5*j*W5*exp(j*theta4);

F = [real(CD1);
    imag(CD1);
    real(CD2);
    imag(CD2)];
end

Answer 1

[Torsten]

You call "fsolve" with a vector of initial values of length 4, but within your function fonk1, you expect a vector x of length 5 as input. This is not consistent.

Answer 2

[Paul]

The variable x is defined at the top of the code to have 4 rows. In the loop that calls fsolve, x0 is defined with four elements on the first iteration and is assigned from a column of x on subsequent iterations. So fsolve is expecting that fonk1 operates on a 4-element vector input. But fonk1 tries to assign x(5) to W5, which results in the error because the input from fsolve to fonk1 only has four elements.