[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
## RE: Integration / curve fitting

**From**: |
Windhorn, Allen E [ACIM/LSA/MKT] |

**Subject**: |
RE: Integration / curve fitting |

**Date**: |
Mon, 12 Aug 2019 22:03:54 +0000 |

Ruan,
From: Help-octave <help-octave-bounces+allen.windhorn=address@hidden> On Behalf
Of Ruan Pieterse
>* I’m in desperate need of help, I’m working on my final masters project,*
>* as part of this I have a non linear ODE which I need to integrate, at the*
>* same rate I need to fit this data to experimental data I’ve generated. *
>* At this stage I have a function which I’m passing to ODE45, with 4*
>* parameters which I need to optimise. I want to use lsqcurvefit but I have*
>* no idea how to incorporate this.*
Probably an optimization algorithm will be easier to work into the method?
Start with some reasonable values for your parameters and run the ODE --
then measure the LSQ error with the experimental data. This is the number
that you are trying to minimize. Feed the ODE error to the optimization
algorithm as the fitting function, and let it adjust the parameters for best
fit. Could be gradient method or genetic algorithm, not sure what will be
the best for this problem.
Can you give us your ODE and some of the data? This sounds like a very
interesting problem.
Regards,
Allen