how to store an output in a function (or bring out the stored output into the workspace) for n-number of model run

I have a hydrological model which runs N number of times to calibrate the parameters.

For each model run, I have a following function where it grabs the modeled simulated output, as well as, NSE output for that run.

function [SimRR, NSE]=HYPE(x)

SimRR = dlmread('output.txt','\t','I2..I100');

NSE_num = importdata('subbasin2.txt','\t');
NSE=NSE_num.data([2]); %NSE value is located in the second spot


end

Everything runs fine. Here, I'd like to store each NSE value for N number of times. (eg. if I'm running the model 3000 times, then I would be having 3000 NSE values stored in an array).

I tried using 'assignin' to bring the value to the base and it only stores NSE value of the last run. If there is a way to store an output generated from a function, please let me know.

If you have any further questions, feel free to comment and I'll provide you with more details.

Answer 1

[Ken Atwell]

Assuming NSE is a scalar double (that is, a single number), When you call HYPE, you can append the NSE result to an (intially empty) vector:

NSE=[];
for i=1:3000
    % ... 
    [SimRR, NSE(end+1)]=HYPE(x);
end
% NSE should be length 3000

[Jung]

Is there a way to store the values within a function (maybe a table ) and call it out later?

The structure of this hydro model script that I'm using is a little different that what I'm used to so I try not to touch the initial interface as much as possible.

[Ken Atwell]

You can using a persistent variable who's value will "survive" across calls to the function. That said, preserving state inside a function is considered bad form -- very nearly a global variable -- that is best avoided if you can.