How to insert values into parameter and not function
Mostra commenti meno recenti
Hi everybody,
I hope somebody can help me out: I want to create a constraint function for fmincon with an input x which comprises of six symbolic variables x1 - x6. The variable 'parameter' depends on x which is why I want to insert their values into 'parameter'. The only way I found out is to create a function as shown below:
function [c, ceq] = constraints(x)
load ('File.mat','parameter');
parameter_function = matlabFunction(parameter);
c(1) = abs(parameter_function(x)); %Constraint
ceq = [];
end
This, however, takes a lot of time to run the code and I am quite sure because of the function creation. For this reason I would like to ask if anybody knows a way to insert the x values into 'parameter' without needing to create the function.
Thanks in advance!
Nicolas
6 Commenti
f = @(x) parameter_function(x)
or more simply
f = @parameter_function
will, in general, give a function handle that takes x as input and then passes it to another function. You should be able to pass f into anywhere that expects a function of 1 argument which will be x.
The whole concept is slow: every time the optimizer calls this function (and that will be many times) then the function load-s some data and converts this using matlabFunction: that is going to be a total waste of time. Much more efficient would be to simply load it once and add it as an input parameter to the function.
If that data is changing then it will still be faster to pass it as a parameter, rather than using a slow file I/O call.
Does parameter_function really need to be generated anew on each call?
Matt J
il 18 Ago 2017
You should probably square c(1) to keep things differentiable
c(1) = parameter_function(x).^2;
Nicolas Ochmann
il 18 Ago 2017
Stephen23
il 18 Ago 2017
@Nicolas Ochmann: function handles can be passed as arguments, so there is no reason why you cannot generate that function and pass it as a parameter (see the links I gave).
Nicolas Ochmann
il 18 Ago 2017
Risposta accettata
Più risposte (0)
Categorie
Scopri di più su MATLAB Mobile Fundamentals in Centro assistenza e File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!