Also, the covariance matrix discussed in Numerical Recipes in the Nonlinear models section that discusses the Levenberg-Marquardt method.

@mmcdara 

It seems that Origin 2022 does non-linear fitting and provides a covariance matrix.  I also find in Bevington's "Data Reduction..." chapter 8 a discussion about how one gets the parameter uncertainties from the covariance or error matrix terms.  Am I using the wrong name for what I am looking for (I want the inverse of the curvature matrix)?