Fitting experimental data with experimental references

Hi, I have an iron compound that I know that has Fe, Fe2+ and Fe3+. In a XANES experimental data of my compound all the contributions overlap. I already have the expected shape of the Fe, Fe2+ and Fe3+ species. Thus, I need to fit the experimetnal data of my compound using the threee references, and calculate the relative amount of both species.

I believe it is an optimiation problem. Could some one help me to get the script to do it in matlab.

So I far his is what I got

prob = optimproblem('ObjectiveSense','min');

w1=optimvar("w1");

w2=optimvar("w2");

w3=optimvar("w3");

Yt2=Yls*w1+Yhs*w2+Y3*w3;

Ymin=Yt2-Yt;

prob.Objective =Ymin ;

show(prob)

initialGuess.w3 = 0.1;

initialGuess.w2 = 0.5;

initialGuess.w1 = 0.4;

[sol,optval] = solve(prob,initialGuess)

w1Opt = sol.w1

w2Opt = sol.w2

w3Opt = sol.w3

However, the values that I got are absurd, too high, like 30 or negative la -50, and they values should renge between 0-1

Answer 1

[Torsten]

w = [Yls,Yhs,Y3]\Yt

if Yls, Yhs, Y3 and Yt are column vectors of the same size.

[Ramón]

thanks you, it works.