looking for help in indexing of variables

Hi all, I have a matrix C, 3 by 3 it is a random variables with 4 iterations, I want to calculte for each element its instantanous value devided by it's average up to each iteration, the output should be as a matrix M, I did that but with numbering the variables I am looking for a better way to do that, if some one could help will be very thankful.

C =
   14.4894   37.8764    6.6557
   12.5079   23.9724   22.0415
   26.9530   18.5281   19.4084
M =
     1     1     1
     1     1     1
     1     1     1

C =
    2.9873   16.6287   11.1656
   24.6604    8.9310   10.9217
   55.4434   37.2271    7.7009
M =
    0.2865    0.4460    1.2911
    1.4268    0.4376    0.5727
    1.5811    1.4904    0.4835

C =
   27.9885   24.7014   16.6055
   26.2066    4.5133    4.3492
    8.3868    0.7076   23.5264
M =
    2.0480    0.8376    1.5458
    1.3162    0.2780    0.2758
    0.1312    0.0209    1.4968

C =
   18.8612    1.7305    7.2938
   21.1609   27.3332   18.3864
   26.2062    7.6808    8.8939
M =
    1.1109    0.0503    0.6400
    0.9715    1.8540    1.3399
    0.6439    0.3157    0.5003

x1=[];x2=[];x3=[];x4=[];x5=[];x6=[];x6=[];x7=[];x8=[];x9=[];
for i=1:4
  F=random('exp',1,3,3);
C=20*log2(1+F)
x1=[x1;F(1)]; x2=[x2;F(4)];x3=[x3;F(7)];x4=[x4;F(2)];x5=[x5;F(5)];x6=[x6;F(8)];x7=[x7;F(3)];x8=[x8;F(6)];x9=[x9;F(9)];
M1=F(1)/mean(x1);M2=F(4)/mean(x2);M3=F(7)/mean(x3);M4=F(2)/mean(x4);M5=F(5)/mean(x5);M6=F(8)/mean(x6);M7=F(3)/mean(x7);M8=F(6)/mean(x8);M9=F(9)/mean(x9);
M=[M1,M2,M3;M4,M5,M6;M7,M8,M9]

  
end

Answer 1

[Matt J]

  F=random('exp',1,3,3,4);
  C=20*log2(1+F);
  M=C./movmean(C,[4,0],3)

[Ashraf Sherif]

Undefined function 'movmean' for input arguments of type 'double'.

Error in Untitled5 (line 3)

M=C./movmean(C,[4,0],3)

>>

[Matt J]

Upgrade!

Or replace with

M=bsxfun(@times,  C./cumsum(C,3) , reshape(1:4,1,1,[])   );

[Ashraf Sherif]

Thank you very much Matt J that is what I want