CurveFitting given only derivative

Hello Matlabbers,

I have given a derivative dF/dX = c*exp(F)+F

I also have an observation, F over x.

How can I curvefit the constant c so that the result would fit my observation dataset. I cannot put a derivate as model function in lsqcurvfit, right?

Thank you so much for your help!

Jan

[Torsten]

What is dF in exp(dF) ? Does it mean "d*F" and d is known ?

[Jaromir Slatan]

I edited the function, it is only F

Answer 1

[Torsten]

You don't need to put a derivative in lsqcurvefit since there is an analytical solution for your ODE from above:

https://www.wolframalpha.com/input/?i=y'(x)%3Dc*exp(y),y(a)%3Db

Best wishes

Torsten.

[Jaromir Slatan]

Hello, thank you Torsten! And what can I do if the function also contains a "+F"?

[Torsten]

Here is a worked-out example we always refer to if the simulated data for lsqcurvefit have to be obtained by numerically solving an ODE:

https://de.mathworks.com/matlabcentral/answers/43439-monod-kinetics-and-curve-fitting

Best wishes

Torsten.