Create a Forum Topic by Logging In or Creating an Account

Q: curvefit for multiple curves with a same function

Thu, 01/14/2010 - 04:22 pm

I want to figure out a decent way to fit multiple curves with a same function for a special case.

Here is the situation: I have 4 pairs of XY data with different x setting, which corresponds to different states of an experiment process. The simulation will generates results for all 4 states simultaneously. This simulation costs significant time. Of course I can use global fitting for the 4 curves, however it will do the simulation for each set of the data and it is not time efficient.

I am now using a cheating way. I stitch all four sets of data together and create a info wave, which contains the stitching information, saying the nth set of data contains N data points. And then in the fit function, I recall this wave to get the simulated value for the stitched data points. Therefore I can avoid using global fitting so that the simulation will only do once.

Does anyone have better idea how to do this? Comments and suggestions are welcome.

andyfaff

If you install the Gencurvefit XOP (curvefitting with genetic optimisation), there is a copy of Global Fit 2.ipf included with it that allows you to simulate the system before you do the fit.

January 14, 2010 at 10:34 pm - Permalink