How to fix the upper asymptote in sigmoid?

Question

Srilatha Raghavan 2020-7-31

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/573415-how-to-fix-the-upper-asymptote-in-sigmoid

评论： Srilatha Raghavan 2020-8-4

采纳的回答： Shae Morgan

在 MATLAB Online 中打开

Hello, I have a sigmoid equation. My qestion is in a comment here:

https://www.mathworks.com/matlabcentral/answers/505984-fitting-a-sigmoid-curve-using-curve-fitting-tool-box

The problem I am facing is, this sigmoid equation is a power curve and needs to reach a certain fixed height at the rated speed (variable 'a'). But I am not able to adjust my upper assymptode anything above 650, even by forcing my a to be an independent variable and also by replacing a by 900. Since, I am new to sigmoid, I would like some help to rewrite this piece of code to make my sigmoid reach 'a'. Also, I def need my xon, xoff and the everything about the curve from below to remain the same.

TIA

xon =3;
xrated = 12.5;
off=25;
a=900;
ft = fittype(@(b,c,x) mysigmoid(b,c,x,xon,xrated,a));
mdl = fit(x',y',ft,'start',[5 20]);
function yhat= mysigmoid(b,c,xon,xrated,a)
x= max(xon,min(x,xrated))
yhat = a./(1+exp(-(x-c)/b))- a./(1+exp(-(xon-c)/b));
end

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Shae Morgan 2020-7-31

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/573415-how-to-fix-the-upper-asymptote-in-sigmoid#answer_473539

Your call to mysigmoid has 6 input arguments (b,c,x,xon,xrated,a; the last of which is your "a"), but your defined function only accepts 5 (b,c,xon,xrated,a). you're missing the 'x' input argument in your function definition.

This is likely leading to your issues with manually setting the "height" of the sigmoid function.

4 个评论
显示 2更早的评论隐藏 2更早的评论

Shae Morgan 2020-8-3

your code still doesn't run after reading in these X and Y values. Please provide a code that will run so it can be used.

ex. missing 'b', 'c', and 'x' values.

Srilatha Raghavan 2020-8-4

I figured this out! If the data that is used to initialize the fit does not reach the required upper symptote, we cannot expect the model to do what want it to with our data. It was a basic understanding of the sigmoid and curve fitting toolbox. Anyways thank you.

请先登录，再进行评论。