how to run written code for 1000 times by for loop i addition to inner for loops inside code?

i have written my code but i need to run it for 1000 times because i need to take mean of some variables such as Qfog after 1000 times because i used many rand & randi in my code, how can i do this in presecence og pre-allocated variables. the code is attached

clear
Nu=(1:1:10);   NF=(1:1:4);    Zu=(1:1:10); Cf=10^6*[4 10 15 7]; % different computing capacity on eacg fog
yt=0.1;   ye=0.1;  delta=0.05;

%wireless interference model Rayleigh
W =10*10^6;
% from file new try.m
N = 10; % number of users distributed in the ciurcle
R = 100;
C=[40 40]; % center of circle
%x0 = [200 400 0 400] ; % Center of the circle in the x direction.
%y0= [200 200 200 0]; % Center of the circle in the y direction.

%generate the circle boundary
t = linspace(0, 2*pi, 50);
x = R*cos(t) + C(1);
y = R*sin(t) + C(2);

% generate random points / users inside
th = 2*pi*rand(N,1);
r  = R*rand(N,1);

xR = r.*cos(th) + C(1);
yR = r.*sin(th) + C(2);
Xf=[20 40 0 40];Yf=[20 20 20 0];

% finding distance between each random point xR,yR and centers of fogs Xf,Yf
x_coor=zeros(10,4);y_coor=zeros(10,4);

for i=1:4%columns
    Xf(1,i)=Xf(1,i);
    Yf(1,i)=Yf(1,i);  
    x_coor(:,i)=xR(:,1)-Xf(1,i);
    y_coor(:,i)=yR(:,1)-Yf(1,i);
end
d=sqrt(x_coor.^2+y_coor.^2);% distance between fog & user

n=1000;
result=zeros(n,1);
for p=1:n
    
%%generating average sojourn time in 4 fogs
mu=30;var=100;
%rng (1,'twister')
ts=mu+sqrt(var)*randn(10,4);

Du=randi([10*10^6,15*10^6],10,1);

fu=randi([2*10^9,5*10^9],10,1);

Culocal=randi([0.5*10^9,0.8*10^9],10,1);

Tulocal=fu./Culocal;

k=10^-26;

Elocal=k*Culocal'.^2*fu;

Qlocal=0.1.*Elocal+0.1.*Tulocal;

SINR_dB=randi([-2,2],10,4);
SINR=10.^(SINR_dB./10);

Ruf=10*10^6*(log2(1.+SINR));

tup=zeros(10,4); texe=zeros(10,4);   %%tup = Du./Ruf;
for rr=1:10 %calculating tup for each fog based on Du of 10 UEs (j=num of fog)
   tup(rr,:)=Du(rr,1)./Ruf(rr,:); 
   texe(rr,:)=fu(rr,1)./Cf(1,:);  
end
Tuf=texe+tup;  %computing time on fog Tuf=texe+tup.%texe=fu./Cf;
pu=0.1;   Eup=tup*pu;  % calculating uplink energy.

%%%%%calculating mean Q on fog.
Qmig=delta.*Du;
part1=1-exp(-Tuf./ts);
%% for loop to get Qfog due to dimensiion
Qfog=zeros(10,4); revenu=zeros(10,4);

for v=1:10    %%rows of num. of users
    Qmig(:,1)=Qmig(:,1);
    Qfog(v,:)=(Qmig(v,1).*part1(v,:))+(yt.*Tuf(v,:)+ye.*Eup(v,:));  %Qfog=(Qmig*(1-exp(-Tuf'/ts')))+yt.*Tuf+ye.*Eup;
    revenu(v,:)=Qlocal(v,1)-Qfog(v,:); %revenu=(Qlocal-Qfog);
end
result(p)=mean(Qfog);

end

%revenu1=max(revenu);
%plot(Nu,revenu1)
%% algorithm 1 steps
Qlocal_rep=repmat(Qlocal,1,4); %repeat dimension for comparison in Bf
%Bf=find((tup<ts)&(Qfog<Qlocal_rep));% gives indices in double format
%Nu1=Nu(Bf);
Bf=zeros(10,4);
for xx=1:10
    for yy=1:4
        if Qfog(xx,yy)<Qlocal_rep(xx,yy);
            Bf(xx,yy)=1;
        else
            Bf(xx,yy)=0;
        end
    end
end

cc=Qlocal_rep-Qfog;

[Julius Muschaweck]

You could just enclose your complete code into a function, returning the results you're interested in, and run that function 1000 times? Your preallocated variables will be local within the function, and the rand and randi functions will return different random numbers in each call.

Like

someseedvalue = 42;
rng(someseedvalue);
for i =1:1000
    test(i) = tmp();
end
result = mean(test)

result = 4.9026

function rv = tmp()
    a = 10;
    rv = a * rand(1);
end

The call to rng (init random number generator) ensures you get reproducible results. If you omit this call, you will get different result each time you run the loop.

BTW: Your way of creating random user locations does not equidistribute them inside the circle. Your users will crowd around the circle center. In your particular case, you could achieve equidistributed users by replacing

r  = R*rand(N,1);

with

r  = R*sqrt(rand(N,1));

[heba raouf]

thanks sir -- i will try to learn converting code to function and tell u if there any problem

Answer 1

[Pritam Das]

You can run an another for loop to get the 1000 values of your desired variable. Put a for loop for the whole code that you want to run for thousant times and put index on the variables you need. suppose you need 1000 values of 'cc'.

for i=1:1000

...Codes...

cc(i)=Qlocal_rep-Qfog;

end

[heba raouf]

thanks sir i will tyr and tell you

[heba raouf]

i tried your solution sir and this is the error message:

In an assignment A(I) = B, the number of elements in B and I must be the same.

Error in smal_scale_1000 (line 86)

result(p)=mean(Qfog);

i dont understand why?

can u tell me ?

thanks for your effort

[Pritam Das]

If you pass index to variable A i.e. A(I), number of elements in 'A' will be equal to 'I'. If you want 1000 values in 'A' then only do A(I). And to avoid dimension error you can initialize the size of A.

As Qfog is a matrix mean(Qfog) will give you a row vector. Use the output as 'result(P,:)'.