so I've had to choose between Port and LM for a non-linear regression.

Initially, on the first 60 datasets, Port solved all (with bounded limits), LM chocked on one and the other 59 results were the same (same starting estimates for params).

Now after 90 more datasets of using Port, it finally had a convergence failure and so I wrote a tryCatch function to use Port, then fallback to LM if that failed.

Is one more prone to choke than another?
As far as I know, SigmaPlot uses LM, but it doesn't choke (convergence failure) on the 1 dataset that failed to converge in nlsLM. In fact, SigmaPlot (which i'm replacing with R and automating away) has never choked as far as I know.



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.