Despite checking for quite some time online and in the docu, I cannot find an "easy" way to get the covariance matrix from FindMinimum.

When FindMinimum has reached the local minimum, it returns the set of best fit parameters, however, in order to access the uncertainties in those, one needs the covariance matrix at this point.

It seems to me like an obvious problem that many people must have encountered before.

Also, I cannot easily re-cast the problem into the form for NonlinearModelFit, since it is a more complicated chi^2 function and structured data I try to minimize.

Thanks for any help!

  • $\begingroup$ If you switched from minimizing a $\chi^2$ function to maximizing a likelihood (or log likelihood), there would be more options. An alternative would be to use a bootstrap approach. Without a specific example (including how the samples were selected), it's unlikely that more specific advice can be given (although I'd be happy to be wrong about that). $\endgroup$ – JimB May 15 at 15:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.