Info

此问题已关闭。 请重新打开它进行编辑或回答。

Curve fitting for two types of behaviours ( aiming for exponentials)

1 次查看(过去 30 天)
Hello, I have a set of points made out of an expriment, I am trying to somehow fit it with a curve which will be an exp function ( and not power) i tried cftool, but it has been unsucsessfull so far,
the problem is the the point are acting as a gaussian at the first few points and after, they behave as an exponent ( or power function)
any tips how i can fit a curve that would incloude or 1) 2 exponents 2) stich 2 solutions ( for first region and second) but with a stich that its derivative wont be huge. 3) other methods of approach?
Thanks you very much Eyal
  3 个评论
Image Analyst
Image Analyst 2015-5-31
What do you want to do? Do you want to create a parameterized model like you use polyfit() for? Do you just want a smoothed version of the data like you use conv() and sgolayfilt() for? Your data sounds like it might be a log-normal distribution but we can't see since you didn't attach any data or screenshots. Don't make us work blind.
Star Strider
Star Strider 2015-5-31
Are you fitting a model of the process that produced your data, or a probability distribution of it?
If you can create a mathematical model of that process, that’s the first step to doing a parametric model fit.
If you are doing a probability distribution, start with the histfit function.

回答(0 个)

此问题已关闭。

标签

Community Treasure Hunt

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

Start Hunting!

Translated by