Multiple linear regression/weighted regression

Posted by Boris Skorodumov on
URL: http://quantlib.414.s1.nabble.com/Multiple-linear-regression-weighted-regression-tp5716.html

Dear QuantLib Users,

I just wondering does QuantLib.0.9.7 has capabilities to calculate multiple linear regression and weighted linear regression with several predictors?
I see, there is a <ql/math/linearleastsquaresregression.hpp> which is capable to carry out something like y~f(x). I was wondering whether I can do something like y ~ b0 + b1*x1+b2*x2+.... LinearLeastSquaresRegression accepts only one vector x and one vector y.

Also for the case of y ~ b0 + b1*x, errors of estimated coefficients somewhat different from those in R for example.

Here is an example:

 x          y
2.40    7.80
1.80    5.50
2.50    8.00
3.00    9.00
2.10    6.50
1.20    4.00
2.00    6.30
2.70    8.40
3.60    10.20

R will give:

           Estimate   Std Error
b0        0.9448     0.3654  
b1        2.6853     0.1487 


while LinearLeastSquaresRegression will give:

          Estimate   Std Error
b0        0.9448     1.2380  
b1        2.6853     0.5034 

I used the following code to test it:


int main()
{
 try
 {
    std::vector<double> x,y;
    x.push_back(2.4);
    x.push_back(1.8);
    x.push_back(2.5);
    x.push_back(3.0);
    x.push_back(2.1);
    x.push_back(1.2);
    x.push_back(2.0);
    x.push_back(2.7);
    x.push_back(3.6);

    y.push_back(7.8);
    y.push_back(5.5);
    y.push_back(8.0);
    y.push_back(9.0);
    y.push_back(6.5);
    y.push_back(4.0);
    y.push_back(6.3);
    y.push_back(8.4);
    y.push_back(10.2);

    std::vector<boost::function1<Real, Real> > v;
    v.push_back(constant<Real, Real>(1.0));
    v.push_back(identity<Real>());

    LinearLeastSquaresRegression<> m(x, y, v);

     for (unsigned int i=0; i < v.size(); ++i)
        cout << m.error()[i] <<"  " << m.a()[i] <<  endl;

        return 0;
    } catch (std::exception& e)
    {
        std::cout << e.what() << std::endl;
        return 1;
    } catch (...)
    {
        std::cout << "unknown error" << std::endl;
        return 1;
    }
}



Thank you in advance for suggestions,

Boris.

------------------------------------------------------------------------------
This SF.net email is sponsored by:
SourcForge Community
SourceForge wants to tell your story.
http://p.sf.net/sfu/sf-spreadtheword
_______________________________________________
QuantLib-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/quantlib-users