Im using nlinfit function and I need to know the R square for the fitted solution

Im using nlinfit function and I need to know the R square for the fitted solution

Answer 1

[the cyclist]

rng default

N = 100;
x = randn(N,1);
y = 2 + 3*x + 5*randn(N,1);

modelfun = @(F,x) F(1) + F(2)*x;
beta0 = [0 0];

[beta,R] = nlinfit(x,y,modelfun,beta0);

Rsquared = 1 - sum(R.^2)/sum(((y-mean(y)).^2))

Rsquared = 0.3732

but I would recommend using fitnlm instead:

mdl = fitnlm(x,y,modelfun,beta0)

mdl = 
Nonlinear regression model:
    y ~ F1 + F2*x

Estimated Coefficients:
          Estimate      SE       tStat      pValue  
          ________    _______    _____    __________

    F1     1.5965     0.50649    3.152     0.0021516
    F2     3.3261     0.43546    7.638    1.4894e-11


Number of observations: 100, Error degrees of freedom: 98
Root Mean Squared Error: 5.04
R-Squared: 0.373,  Adjusted R-Squared 0.367
F-statistic vs. constant model: 58.3, p-value = 1.49e-11

mdl.Rsquared

ans = struct with fields:
    Ordinary: 0.3732
    Adjusted: 0.3668