Trouble Creating a Vector using a for loop with matrices

I'm having trouble creating a vector of answers with a for loop involving matrices. When I call on w, it gives me different values than if I were to call on a specific part of w.

B1=[1 0; 0 -7.5];

D1=[0 0;0 10];

X1=[0 0;0 -2.5];

B2=[49.8 0;-0.127 -50];

D2=[25 0;0 25];

A2=[25 0;0 25];

B3=[49.8 0;-0.127 -50];

D3=[25 0;0 25];

A3=[25 0;0 25];

B4=[49.8 0;-0.127 -50];

D4=[25 0;0 25];

A4=[25 0;0 25];

B5=[7.5 0;0 1];

A5=[-10 0;0 0];

Y5=[2.5 0;0 0];

o=[0 0;0 0];

%M

M=[B1 D1 X1 o o;A2 B2 D2 o o;o A3 B3 D3 o;o o A4 B4 D4;o o Y5 A5 B5];

Mi=inv(M);

G=[.21;0;0;0;0;0;0;0;0;0.127];

CG=[.1;.1;.1;.1;.1;.1;.1;.1;.1;.1];

%find L

%L=[b1 o o o o;a2 b2 o o o;o a3 b3 o o;o o a4 b4 o;o o y5 a5 b5]

%U=[I -E1 -x1 o o;o I -E2 o o;o o I -E3 o;o o o I -E4;o o o o I]

n=2;

NJ=5;

w=zeros(2*NJ,1);

for j=1

L(1:(j+1),1:(j+1))=B1; %b(1)=B1

U(1:j+1,j+2:j+3)=-inv(L(1:j+1,1:j+1))*D1; %E(1)=D1*inv(-b(1))

U(1:j+1,2*j+3:2*j+4)=-inv(L(1:j+1,1:j+1))*X1; %x(1)=X1*inv(-b(1))

w(1:2)=inv(L(1:j+1,1:j+1))*G(1:2); %ksi(1) = b(1)^-1 * G(1)

end

for j=2:NJ-1

L(2*j-1:2*j,2*j-3:2*j-2)=A2; %a(2)=A2

L(2*j-1:2*j,2*j-1:2*j)=B2+L(2*j-1:2*j,2*j-3:2*j-2)*U(2*(j-1)-1:2*(j-1),2*(j-1)+1:2*(j-1)+2); %b(2)=B2+a(2)*E(1)

U(2*j-1:2*j,2*j+1:2*j+2)=-inv(L(2*j-1:2*j,2*j-1:2*j))*D2; %E(2)=-inv(b(2))*D(2)

w(2*j-1:2*j)=-inv(L(2*j-1:2*j,2*j-1:2*j))*((-G(2*j-1:2*j)+(L(2*j-1:2*j,2*j-3:2*j-2)*w((2*(j-1)-1):(2*(j-1)))))); %ksi(2)=inv(-b(2))*(-G(2)+a(2)*ksi(1))

end

for j=NJ

L(2*j-1:2*j,2*j-5:2*j-4)=Y5; %y(5)=Y5

L(2*j-1:2*j,2*j-3:2*j-2)=A5+L(2*j-1:2*j,2*j-5:2*j-4)*U(2*(j-2)-1:2*(j-2),2*(j-2)+1:2*(j-2)+2); %a(5)=A5+y(5)*E(NJ-2)

L(2*j-1:2*j,2*j-1:2*j)=B5+L(2*j-1:2*j,2*j-3:2*j-2)*U(2*(j-1)-1:2*(j-1),2*(j-1)+1:2*(j-1)+2); %b(5)=B5+a(5)*E(4)

w(2*j-1:2*j)=-inv(L(2*j-1:2*j,2*j-1:2*j))*(-G(2*j-1:2*j)+L(2*j-1:2*j,2*j-3:2*j-2)*w(2*(j-1)-1:2*(j-1))+L(2*j-1:2*j,2*j-5:2*j-4)*w(2*(j-2)-1:2*(j-2)))%-inv(b(5))*(-G(5)+a(5)*w(4)+y*(w(3))

end

%w=inv(L)*G

%C=inv(U)*w

[Stephen23]

Note: to make your code more efficient and robust you should probably be using mldivide instead of inv and *. Explicit matrix inversion is rarely required (because there are better methods for solving such systems of equations).

https://www.mathworks.com/help/matlab/ref/mldivide.html