How can I simulate this code with 1000 simulations?

If I run a chain of n = 50 with four states, then how can I adjust the following code to run for 1000 times.

% I want to run this code 1000 times? 
n = 50; % length of the chain
X = zeros(n,1); % matrix to save the chain in future 
mu = [1/4 1/4 1/4 1/4]; % initial position of the chain
P = [0 1/2 0 1/2; 1/2 0 1/2 0; 0 1/2 0 1/2; 1/2 0 1/2 0]; % probability transition matrix 
U = rand; % generate a random number 
X_0 = min(find(U <= cumsum(mu))); % get the next state 
U = rand;
X(1) = min(find(U <= cumsum(P(X_0,:))));
% do this process for the entire length (n) of the chain
for i = 1:n-1
U = rand;
X(i + 1) = min(find(U <= cumsum(P(X(i),:))));
end

How can I run this chain e.g. 1000 times

The above code is just an example, in practice, I have P = cell array. It is difficult to provide the whole data here thats why I tried to come up with a small example.

I tried to do something like this that but it did not work.

n = 50; % length of the chain
simulations = 1000
X = zeros(n,simulations); % matrix to save the chain in future 
mu = [1/4 1/4 1/4 1/4]; % initial position of the chain
P = [0 1/2 0 1/2; 1/2 0 1/2 0; 0 1/2 0 1/2; 1/2 0 1/2 0]; % probability transition matrix 
for k = 1:simulations
U = rand(1,m); % generate a random number 
X_0 = min(find(U <= cumsum(mu),2)); % get the next state 
U = rand(1,m);
X(1,m) = min(find(U <= cumsum(P(X_0,:))));
% do this process for the entire length (n) of the chain
for i = 1:n-1
U = rand;
X(i + 1) = min(find(U <= cumsum(P(X(i),:))));
end
end 

[Mathieu NOE]

hello

P is not defined - cannot run your code

[Walter Roberson]

What is the point of mu having 10 entries? the cumsum will be 1 by the 4th, and rand() cannot exceed 1.

[susman]

I have just rephrased the questions. I tried to provide the small example to understand how does the code work. If its still not clear, I would try to rephrase agaib. Really sorry for that.

[Walter Roberson]

What is m? You have

U = rand(1,m);
X(1,m) = min(find(U <= cumsum(P(X_0,:))));

Your U is going to be a row vector, but P(X_0,:) is going to be a row vector, so you have <= between a row vector and a column vector, which is going to get you an m by size(P,2) -> m x 4 array. But then you try to store that into the scalar location X(1,m)