Learning functions

Hello UKman,

I'm not certain what you're after but the attached altered MC sheet may help. It is not an elegant solution but it will help to find peaks.

Basically, what I did is as follows:

• use a fitting function to create a polynomial
  
• extract the coefficients of the fitting function
  

You need to create a polynomial with this coefficients which than can be evaluated as shown in the quicksheet "Finding roots of a polynomial"

Raiko

Don't smooth data that you are going to fit a function to using least squares. Least squares fitting handles noise very well, and smoothing will either have almost no effect or it will bias the parameter estimates. It's better to do the least squares fit using minerr rather than genfit. It's easier to set up and more flexible.

If the heights and/or standard deviations of the two distributions are supposed to be the same you can easily modify the worksheet so that those parameters are scalars rather than two element vectors.