How to sample values from inner for loop vector to an outer for loop vector

Hi,

I'm trying to model the difference of strength in a cable consisting of N wires in a cable due to different strain values. For this I have created two for loops. In the inner for loop I calculate the strength of the cables for each strain step. These values are sampled in a vector. I want to use Monte Carlo simulations and run the inner for loop K times. For each simulation of K I want to sample the "jobb" vector into a large vector containing every "jobb" vector value. Can this be done?

my script:

clc; clear all; close all;  
E = 7.75*1e4;
Area = 38.48;
mu = 0.4343;
su = 0.3295; 
my = 0.8686;
sy = 0.1318;
k = 8;                                                                      
N = 200;                                                                  
K = 10;                                                                                                                             
for simulations = 1:K;
    P = rand(N,1);                                                          
    epsilon_y = logninv(P,my,sy);                                           
    epsilon_u_y = logninv(P,mu,su);                                         
    epsilon_u = (epsilon_y + epsilon_u_y);                                  
    n = numel(P);                                                                                                                               
    g = 1:k;                                                                
    epsilon_0 = 1;                                                          
    jobb = zeros(k,1);                                                      
    model = zeros(n,1);                                                     

for epsilon = 1:k;                                            

      for ii = 1:n;                                                           
      if epsilon_y(ii) >= epsilon                                             
          ep = epsilon;                                                       
      elseif epsilon_y(ii) < epsilon && epsilon < epsilon_u(ii)
          ep = epsilon_y(ii);
      elseif epsilon_u(ii) <= epsilon
          ep = 0;
      end
      model(ii) = ((ep/100)*E);                                               

      end
      p = sum(model)*Area;                                                    

      jobb(epsilon) = p;                                                     

end
large_vector_that_I_want_to_find = ..;

end

Best regards

Trond Oesten

Answer 1

[Guillaume]

Rather than storing the result in a big vector, I'd use a matrix, where each column is a simulation:

jobb = zeros(k, K);
for simulation = 1:K
   %initialise everything but jobb as normal
   for epsilon = 1:k
      %calculate everything as normal
      jobb(epsilon, simulation) = p;
   end
end

If you want to convert the matrix into a big vector, then it's simply

large_vector_that_I_want_to_find = jobb(:);

[Trond Oesten]

That worked. Thank you!