how do I go about creating and storing data for a monte carlo simulation?

Hi all,

This is my ham-fisted attempt to assign memory for the above. Let's say that I want my simulated data to be normally distributed random numbers (they're not, but right now I just care about having a structure in place) arranged in 3 series of 226 observations. So I create a cell that contains the 226x3 (empty, for now) arrays. Let's say I want 4 simulations, so 4 four of those (that's rubbish,but, again, just to keep things simple).

Then I create a 3 dimensional array of the right dimensions with said normal, random data. At this point I ought to extract each 226x3 lots of normal random data and shove it in the cell. Presumably, at that point, I can just put my model in that second loop and do what I need to do.

I can't seem to be able to do the extracting bit. M only contains the last 226x3 block from the 3 dimensional array (rather than all 4).

clc;
clear;

par.muorig = 5; %mean
par.sigmaorig = 5; %stdev

Nsim = 4; % number of simulations
Lseries= 226;
M=cell(4,1);

%allocate memory for matrices-which perhaps I can store in cells
for imatrices=1:Nsim
M{imatrices}=zeros(Lseries,3);    
end

for isim= 1:Nsim

data(:,:,isim)=normrnd(par.muorig,par.sigmaorig,Lseries,3);


   ff=1
   for rr=1:Nsim
       M=data(:,:,isim); %pull out each array and shove it in the cells
   ff=ff+1   

   %my model here
   end

   
end

Any suggestions on how to do this? I look forward to hearing from you!

Answer 1

[Fede C 2018 London]

I think this does it. For those who might need to do something similar:

clc;
clear;

par.muorig = 5;
par.sigmaorig = 5;

Nsim = 4; % number of simulations
Lseries= 226; %sample size
M=cell(4,1); 

for imatrices=1:Nsim
M{imatrices}=zeros(Lseries,3);   

for isim= 1: Nsim
    M{isim}=normrnd(par.muorig,par.sigmaorig,Lseries,3);

end
%my model here

end