How to create matrix from a function

Hi guys, I need to create a matrix (in my script is "array"), deriving from a function, depending on three variables, which alternate.

clear all
Var; %from which I derive T1 and m1
beta= 0.05;
k0 = 1.67*10^13; 
R = 0.0083;
for j = 1:length(T1)
    alpha1_tga(j) = (m1(1,1)-m1(j))./(m1(1,1)-m1(end,end));
    alpha1_tga(j) = 1 - alpha1_tga(j);
end

E1= 150;
E2 = 300;
Estep = 5;
sigma1 = 10;
sigma2 = 70;
sigmastep = 5;
[alpha_daem,h2,array] = solvedaem(T1,beta,k0,R,E1,Estep,E2,sigma1,sigmastep,sigma2,alpha1_tga);

Here, there's the function.

function [alpha_daem,h2,array] = solvedaem(T1,beta,k0,R,E1,Estep,E2,sigma1,sigmastep,sigma2,alpha1_tga)
  for E0 = E1:Estep:E2
    for sigma = sigma1:sigmastep:sigma2
      h2 = 0;
      array = [];
      for j = 1:length(T1)
        f = @(x) (exp(-k0./beta.*R.*T1(j).^2./x.*exp(-x./R./T1(j))))./(sigma.*sqrt(2*pi)).*exp(-(x-E0).^2./(2.*sigma.^2));
        a = E0-25; b = E0+25; %lower and upper limits
        nSimp = 100; %number of intervals
        hSimp = (b - a)/nSimp; %length of each interval
        s = 0; 
        for i=0:nSimp
          if i==0 || i==nSimp
            p = 1;
          elseif mod(i,2) ~= 0
            p = 4;
          else
            p = 2;
          end
          x = a + i*hSimp;
          s = s + p*f(x);
        end
        alpha_daem(j) = hSimp/3*s;
        h2 = h2+(alpha1_tga(j) - alpha_daem(j)).^2;
      end
      array = [E0,sigma,h2];
    end
  end
end

I mean that I need to create the matrix "array", in which the first column in equal to E1 ( from 150 to 300 with step 5), the second column equal to sigma (from 10 to 70 with step 5, that should repeat for each value of E1), and third column equal to h2.

Could anyone help me?

Answer 1

[Abhishek Gupta]

Hi,

As per my understanding, you want to create a matrix containing E1, sigma, and h2 values. The simple way would be as follows:-

Approach 1 - size of the output matrix is known

output = zeros(size(E1:Estep:E2,2)*size(sigma1:sigmastep:sigma2,2), 3);  % initialize the matrix 

count = 1;
for E0 = E1:Estep:E2
    for sigma = sigma1:sigmastep:sigma2
       %-------%
    end
    output(count, :) = [E0,sigma,h2];   % append the array 
    count = count + 1;
end

Approach 2 - size of the output matrix is unknown

output = []; % initialize the matrix 

for E0 = E1:Estep:E2
    for sigma = sigma1:sigmastep:sigma2
       %-------%
    end
    output = [output; [E0,sigma,h2]];   % append the array 
end

[Gabriella Gonnella]

The point is that I need to create a matrix in which, for every value of E (that are 30), correspond all the values of sigma (that are 13). So every value of E should be repeated 13 times before going on in counting.

Using your code, at the end of run I have a matrix with E from E1 to E2, and sigma only equal to sigma2.

[Abhishek Gupta]

As far as I understand, you want a 30x2 matrix (column 1 - E, column 2 - sigma), where you want to repeat sigma (13x1) to fill up the 30x1 column. You can do the same as follow: -

sigma   = rand(13,1);   % replace this with your sigma
len = 30;       % length of the matrix
extendedSigma = [repmat(X, floor(len / numel(X)), 1); ...
             X(1:mod(len, numel(X)))];

output = [E sigma]; % desired matrix -- Col 1 - E, Col 2 - sigma

For more information, check out the link below: -

https://www.mathworks.com/matlabcentral/answers/313282-repeating-a-vector-up-to-a-given-length

[Gabriella Gonnella]

I need to find three values of E and sigma which minimize h2. For this reason I would create the matrix with all the values inside. Could be a faster way to find them?