minlmoptimize function wrap

arima · **Joined:** Wed Dec 07, 2016 5:00 am **Posts:** 3

Hi there,

I have a problem related with using minlmoptimize method for curve fitting.
I tried to create a vector of fi[i] of each data point, as shown below, then pass them into the minlmoptimize function.

Code:

        public static void function_fvec(double[] vars, double[] fi, object obj)
        {                        
            double[] tempfuncs = new double[10];
            
            for (int i = 0; i < 10; i++)
            {
                tempfuncs[i] = Math.Abs(data_y[i] - simu(vars, data_x[i])); //compare input data and calculated results.
            }

            fi= tempfuncs;            
        }

However, it seems does not work. Can someone help me out?

Thank you!

Sergey.Bochkanov · **Joined:** Fri May 07, 2010 7:06 am **Posts:** 927

Hi!

You should rewrite elements of fi, not create new array. As you may see, fi[] is passed by value, so when you create tempfuncs[] there is no way to tell optimizer that it should work with newly allocated array.

arima · **Joined:** Wed Dec 07, 2016 5:00 am **Posts:** 3

Hello Mr.Bochkanov,

Thank you for replying in person! Now the problem is fixed, your solver works wonderfully!

Maybe another quick question, is there a way to get the correlation matrix of the fitting varables?

Many thanks!

Sergey.Bochkanov · **Joined:** Fri May 07, 2010 7:06 am **Posts:** 927

No, LM optimizer does not support covariance matrix estimation. Only linear LS solver provides such option.

forum.alglib.net

Forum rules

minlmoptimize function wrap

Who is online