There seems to be a large number of questions surrounding the fitting of functions to data sets. Many of these have the same problems with invalid starting values, missing x-values, insufficient amount of data points, missing restraints. Many are closed off-topic, others receive simple answers and are never closed.

  • +1 In such a canonical Q&A there should also be links to answers like this one about fitting to complex data using a custom NormFunction
    – Sascha
    Feb 28, 2017 at 10:20
  • +1, it will take some effort to outline common problems but I think it is worth it. Btw. related topic showing that fitting is only part of a problem: “Mathematica can't solve this” as a reason for closing a question.
    – Kuba Mod
    Feb 28, 2017 at 10:53
  • @Kuba I've started a sketch for the answer related to various methods of fitting, do you believe this should include differential fitting too? Perhaps someone else could add an additional answer for that to the CQ.
    – Feyre
    Feb 28, 2017 at 11:51
  • @Feyre You can add that point in the question and skip in your answer if you are not sure. Someone else will come and improve / add an answer.
    – Kuba Mod
    Feb 28, 2017 at 11:52
  • OK. I'll bite: What is a "canonical question" ?
    – JimB
    Feb 28, 2017 at 16:57
  • Nevermind. I just did what I should have done before writing the comment: did a search for "canonical" on this site.
    – JimB
    Feb 28, 2017 at 17:05

1 Answer 1


A preliminary Q&A is now posted here:

What are some common issues with fitting functions to data? If additional info is required feel free to add it to the answer, or post a new answer if it concerns different MMA functions.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .