Discover MakerZone

MATLAB and Simulink resources for Arduino, LEGO, and Raspberry Pi

Learn more

Discover what MATLAB® can do for your career.

Opportunities for recent engineering grads.

Apply Today

Apply boundary condition (upper and lower limit) to parameters during the fitting using the function nlinfit?

Asked by Manuel on 19 Feb 2013

Does anyone knows how to apply boundary condition to the parameters during the fitting using the function nlinfit?

I know that this could be done using the function fit, but i need to use the function nlinfit. (in that regards see the post in the link attached)

Thank you very much in advance for the help.

http://mathworks.com/matlabcentral/answers/63043-fit-model-with-3-independent-variables-and-many-parameters-to-data#answer_74613

0 Comments

Manuel

Products

1 Answer

Answer by the cyclist on 19 Feb 2013
Edited by the cyclist on 20 Feb 2013

I don't have time right now to try to craft an example, but I think one way to handle this would be to put a term in your fit that is like

Inf*(F > 50)^2 + Inf*(F < 0)^2

which would automatically make the fit horrible if the parameter F were outside the range [0,50].

You might need to use some huge (but finite) number in place of Inf.

I'll try to make an example if I can.

EDIT TO INCLUDE EXAMPLE:

So, here is an example of including a boundary condition on the parameter F(3) of my model. (I zero out the entire function if F(3) is not within the bounds.)

I think you need to well understand what your function looks like to do this, and I am not 100% sure this is the best way. You will notice that this example spits out warning about the Jacobian being ill-conditioned. I expect this is because F(3) does end up being at the border. The fit does do what I would expect, compared to the unconstrained fit.

% Here is an example of using nlinfit(). For simplicity, none of
% of the fitted parameters are actually nonlinear!
% Define the data to be fit
x=(0:1:10)'; % Explanatory variable
y = 5 + 3*x + 7*x.^2; % Response variable (if response were perfect)
y = y + 2*randn((size(x)));% Add some noise to response variable
% Define function that will be used to fit data
% (F is a vector of fitting parameters)
% f = @(F,x) F(1) + F(2).*x + F(3).*x.^2;
f = @(F,x) (F(1) + F(2).*x + F(3).*x.^2) * (1 - (F(3) > 4)|(F(3) < 0));
F_fitted = nlinfit(x,y,f,[1 1 1]);
% Display fitted coefficients
disp(['F = ',num2str(F_fitted)])
% Plot the data and fit
figure
plot(x,y,'*',x,f(F_fitted,x),'g');
legend('data','fit')

0 Comments

the cyclist

Contact us