fitting a Gaussian curve convoluted with exponential decay

8 次查看(过去 30 天)
I have to fit a gaussian curve convoluted with an expoential tail to the attached data. For the curve i have the below equation but it does not produce the desired shape. The attached image shows the curve shape I am aiming to achieve using these values and curve equation the exponential section dominates such that the gaussian component cannot even be seen. How would i go about getting the desired curve shape?
y2=0.5*exp(-a*(x2-(0.5*o^2*a))).*(1+erf((x2-o^2*a)/sqrt(2)*o));
using values;
a=exp(-1);
o=0.009931621960609;
where a is a place holder for the decay constant (it may be the issue if it is a very incorrect guess but i am unsure) and o is teh standard deviation of the data.
  1 个评论
Idris Phillips-Fry
Idris Phillips-Fry 2020-12-10
The data file may not open as i have since discovered it is damaged. However that is not of great consequence as i mainly wish to know the general form of the equation i will need to form the desired curve shape shown in the image. Any assistance would be greatly appreciated.

请先登录,再进行评论。

采纳的回答

Jeff Miller
Jeff Miller 2020-12-11
It sounds like you want what is often called the "ex-Gaussian" distribution. Its density function is on Wikipedia.
Cupid will fit it for you, if you don't want to write the code yourself.

更多回答(0 个)

类别

Help CenterFile Exchange 中查找有关 Interpolation 的更多信息

产品


版本

R2020b

Community Treasure Hunt

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

Start Hunting!

Translated by