Effacer les filtres
Effacer les filtres

nlinfit and lsqcurvefit for problems with multiple variables?

2 vues (au cours des 30 derniers jours)
wesleynotwise
wesleynotwise le 25 Juin 2017
Commenté : wesleynotwise le 26 Juin 2017
Am I right to say that the nlinfit and lsqcurvefit are unable to solve non-linear problem with multiple variables?

Réponse acceptée

John D'Errico
John D'Errico le 25 Juin 2017
Modifié(e) : John D'Errico le 25 Juin 2017
No. You are not right. Not even remotely correct. They are both completely capable of solving a problem with multiple variables.
Your question is ambiguous though. Are you asking about independent predictor variables? The answer is: no problems.
Are you asking about multiple parameters to estimate? Again, no problem.
What you don't say is why you think there is any problem at all.
You might be confused because these functions expect a model of the form
fun(params,xdata)
That does not mean that params is a scalar. Nor does it mean that xdata is a vector.
If you have multiple parameters to estimate, params can be a vector or array of any shape or size. If your parameters are split into several distinct variables, then what stops you from combining them into one vector? Nothing.
If you have multiple independent variables, then note that xdata may be an array.

Plus de réponses (0)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by