Do I need to reset Fsolve Parameters if I use Fsolve twice in my code ?

Hi All

If I have to use Fsolve twice in my code, do I need to reset any paramter ?

[x,fval,exitflag,output] = fsolve(fun,x0initial,options);

[dpb]

Impossible to say with no more information than that ... depends upon what is changed between the two calls.

It would seem there would be no reason for the second call unless something were to change; else't would simply be returning the same solution twice.

Need more context of what the object is for calling twice...

[Star Strider]

Each call to fsolve most likely begins a new instance, so it resets itself.

Note that fsolve will not integrate differential equations.

What are you doing?

[farzad]

Well I'm doing numerical integration. I'm solving two different problems of nl equations so I need to use fsolve twice. The variables don't relate.

[farzad]

Thank you i think problem is solved if fsolve resolves itself

[Star Strider]

If you want to estimate the parameters of a differential equaton — with repsect to fitting them to data or optimising them — there are more appropriate menthods than fsolve, since it is a root-finder.

[dpb]

"problem is solved if fsolve resolves itself"

What does this mean? How can it "resolve itself"? It operates on precisely what you provide it -- if you don't change anything in the call to it, then it'll simply do the same thing twice.

Also Star-S's comment is most pertinent with the shred of additional info provided -- but still not at all enough to comment fully.

Remember, we only know of your problem from what you tell us; you understand what it is you're trying to do so shorthand descriptions are meaningful to you--for us, "not so much!" as we don't have any a priori info.

[farzad]

Ah sorry I didn't clarify I did not use exactly the same formulation every time That was only an example

[dpb]

Well, again, all we see is what you post...

So, is the Q? closed? Or is there still some issue?

Answer 1

[Walter Roberson]

No. However if you do not change any of the parameters and the function does not involve random variables, then the second call would give the same result as the first call.

fsolve does not modify any of the input parameters.

I do sometimes see people using code such as

x = [0,10]
x = fzero(SomeFunction, x0, x);

and then get surprised when they try to use x as if it were still the vector instead of having been replaced with the specific solution.

[dpb]

Good point about the RVs, Walter.