Curve fitter, how to start?

Hi, I just imported curve fitter to the my MATLAB. I have these coordinates from the xls file , and I wish to import them directly in curve fitter. But that doesn't open it. Can this be done easily?

Any help appreciated

Thanks

Answer 1

[Star Strider]

Try this —

T1 = readtable('data.xlsx');
Var1 = str2double(cellfun(@(x)regexp(x, '\d*', 'match'), T1.Var1));
T1 = removevars(T1,'Var1');
T1 = addvars(T1, Var1, 'Before','Var2')

T1 = 5×2 table
    Var1    Var2
    ____    ____

     39     167 
     40     192 
     41     240 
     42     343 
     43     497 

fitf1 = fit(T1.Var1, T1.Var2, 'poly1')

fitf1 = 
     Linear model Poly1:
     fitf1(x) = p1*x + p2
     Coefficients (with 95% confidence bounds):
       p1 =        81.1  (32.16, 130)
       p2 =       -3037  (-5045, -1030)

figure
plot(fitf1, T1.Var1, T1.Var2)
grid

fitf2 = fit(T1.Var1, T1.Var2, 'poly2')

fitf2 = 
     Linear model Poly2:
     fitf2(x) = p1*x^2 + p2*x + p3
     Coefficients (with 95% confidence bounds):
       p1 =       22.36  (14.35, 30.36)
       p2 =       -1752  (-2409, -1096)
       p3 =    3.45e+04  (2.105e+04, 4.795e+04)

figure
plot(fitf2, T1.Var1, T1.Var2)
grid

.

[Sergio]

Thanks, the exponential form was the desired! Thanks!

[Star Strider]

As always, my pleasure!

Not all the estimated parameters in the exponential or power functions were significantly different from zero (the confidence limits have opposite signs) so I only posted the models with significant parameters. (I tried several different models with 3 or fewer parameters, since there are only 4 data points, only models with 3 or fewer paramters will give reliable results in this instance.)