Getting x value for a given y value from a fitted model

Hi,

I have my code:

curve = @(a, b, c, d, e, x) a./(b+exp(-c*(x-d)))+e;
lower_bound = [-Inf -Inf -Inf 0, 0];
upper_bound = [Inf Inf Inf max(x), max(y)];
starting_param = [max(y) 1 1 x(round(length(x)/2)), min(y)];
sigm=fit(x, y, curve, 'StartPoint', starting_param, 'Lower', lower_bound, 'Upper', upper_bound) ;

I'm getting my fit result in the sigm which is cfit type.

My question is: how to obtain x value from a given y value out of sigm? If I write ans=sigm(0.5) I'm getting y value for a x=0.5 What I want is to get x value for y=0.5. How to code it?

thanks for help!

[Daniel Pereira]

Why not fitting x=g(y) instead of y=f(x)?

Answer 1

[dpb]

A) Evaluate the function over a range of x and then use interp1 to interpolate over the response to find a given x that gives a desired value.

B) Use fsolve() Recast the function to form f(x)-y = 0 where y is your target value.