Hi Jason,
Ok. I am assuming some things. I assume that you are a very basic and new user of MATLAB. So The code which I am giving is an extremely basic code just to give you a feel of the steps. There can be more optimized and complicated and slick codes but I have made this basic code for you for your understanding.
a(:,1) = randi(1000,[1,100]); % Suppose it denotes the time taken
a(:,2) = randi(5000,[1,100]); % Suppose it denotes constraint A
a(:,3)= randi(20000,[1,100]); % Suppose it denotes constraint B
[p,q] = sort(a(:,1),'descend'); % Please look the syntax up in the help file
a(:,[1:end]) = a(q,[1:end]); % What I have done is that I have just arranged the observations in the order of their decreasing time
work1 = []; % Suppose I have 3 workstations and I have declared three vectors to store the elements to work on them
work2=[];
work3=[];
for i = 1: size(a,1)
if a(i,2)+a(i,3)>2000; % IF the sum of constraints A and B is more than 2000 then work at workstation 1
work1= horzcat(work1,a(i,1));
else
if a(i,2)+a(i,3)<2000 && a(i,2)+a(i,3)>1000 % Similarly the two conditions
work2 = horzcat(work2,a(i,1));
else
work3=horzcat(work3,a(i,1));
end
end
end
Of Course the number of if statements involved would be large depending upon the constraints and their inter-relationships. In Pattern Recognition also we work on large number of constraints and in many cases it is impossible to escape the if-else statements. Dont be afraid of them. In some cases you may reduce the number of if-else statements but it takes luck and conditions. I hope this code would give you an idea of the problem. I hope I have captured ur problem and given u an apt reply. If you need more help, mail back
Happy to Help
UJJWAL
