How do I calculate the error of a fitted curve?

23 views (last 30 days)
Say I have some data points, which I do not have error bars for, and I fit a Gaussian (or possibly Lorentzian) curve to these data points. I want to find a parameter representing how accurate this fit is, with respect to the x coordinate, if I assume that each data point has no error. How could I go about doing this kind of fit in Matlab? I'm happy with some simple error parameter like the standard error of the data points with respect to the fit, but I'm not sure of the best Matlab function to use and whether Matlab has this information without doing extra calculations.

Accepted Answer

Star Strider
Star Strider on 15 Jan 2019
See if the Statistics and Machine Learning Toolbox nlpredci (link) function will do what you want.
  1 Comment
Star Strider
Star Strider on 16 Jan 2019
@Janna Hinchliff — I replied to your classmate’s Question on a similar topic, and to the Comment you posted to it.
If my Answer to either Question helped you solve your problem, please Accept my Answer here!

Sign in to comment.

More Answers (0)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by