setting constraints in polynomial fit

actionbrandon · **Joined:** Mon Aug 13, 2012 6:22 pm **Posts:** 3

I'm sure there's a simple answer to this, but I'm stuck. I have a small set of that data that I am doing a polynomial fit on (data below). When I run (m=4):

Code:

alglib.polynomialfit(x, y, m, out info, out p, out rep);
alglib.polynomialbar2pow(p, out coeffs);

I get the results I am anticipating: {25.426138,0.092663,-0.000658,0.000001}

I would like to be able to weight this data, but when I run the polynomialfitwc function, I get different results. I set all w=1 (an array of equal length to x where all elements =1), and:

Code:

double[] xc = new double[] { 0 }; //constraints = 0
double[] yc = new double[] { 0 };
int[] dc = new int[] { 0 };

These settings give me the following coefficients: {0.000000,1.116681,-0.011579,0.000035}

What am I doing wrong? I'd like to set up the wc function so it returns that same as polynomialfit() if the weights are 1.

Thanks in advance. This package is really impressive.

Data:

X_1 Y_1
6 26
28 27.5
50 28.4
69 29.35
93 29.4
116 29.35
137 29.15
158 28.82
180 28.5

actionbrandon · **Joined:** Mon Aug 13, 2012 6:22 pm **Posts:** 3

Came back to this and realized this is (as I figured) a simple syntax error. I'll post here in case this saves anybody some time in the future:

empty constraints should be set as:

Code:

            double[] xc = new double[0]; //constraints = 0
            double[] yc = new double[0];
            int[] dc = new int[0];

forum.alglib.net

Forum rules

setting constraints in polynomial fit

Who is online