#include <Linreg.hh>
Inheritance diagram for RazorBack::Linreg_:


Public Methods | |
| Linreg_ () | |
| Init to empty. | |
| Linreg_ (size_t NM, size_t NP) | |
| Init to desired sizes. More... | |
| double | lin_reg (const Rmatdbl_ &Xmeas, const Vecdbl_ &Ymeas, const Vecdbl_ &Userwgt) |
| lin_reg(Xmeas, Ymeas [,Userwgt]): multivariate linear regression calculating the parameters P in the equation Y=P[0]*X[0]+...+P[Np-1]*X[Np-1]. More... | |
| double | lin_reg (const Rmatdbl_ &Xmeas, const Vecdbl_ &Ymeas) |
Protected Methods | |
| void | adjust_dimensions (const Rmatdbl_ &Xmeas, const Vecdbl_ &Ymeas, const Vecdbl_ &Userwgt) |
| void | setup_weight (const Vecdbl_ &YW) |
| void | setup_xtx (const Rmatdbl_ &Xmeas, const Vecdbl_ &Ymeas) |
| double | quality (const Rmatdbl_ &Xmeas, const Vecdbl_ &Ymeas) |
Private Attributes | |
| Vecdbl_ | W |
| weight vector. | |
Definition at line 24 of file Linreg.hh.
|
|
Init to empty.
|
|
|
Init to desired sizes. Actual regression calls may change the settings |
|
|
lin_reg(Xmeas, Ymeas [,Userwgt]): multivariate linear regression calculating the parameters P in the equation Y=P[0]*X[0]+...+P[Np-1]*X[Np-1]. Xmeas is an Nm x Np matrix holding Nm rows of Np independent vars, Ymeas is an Nm-long vector holding the dependent vars. Userwgt is an optional Nm-long vector of weights. Throws Dimexc_ and Toofewexc_. Return value: the residual deviation. Invoke param(), sdev(), correl() for the detailed results. |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
weight vector.
|
1.2.6 written by Dimitri van Heesch,
© 1997-2001