HI Experts,
Can anyone here tell my why my manuelly calculated stderr of my parameters not the same as in the build-in package of Nonlinearfit?
Of course as the data is added with noise by random numbers not all generated data will find a fit. But even the added dataset, which will find a fit,
its not the same.
COuld the reason be the Pseudoinvers of a Matrix? But what are the basics of the build-in calculation. Does that not use the Pseudoinvers too?
Thanks
walter