cepLs 3 cepLs Cacluates a Linear Least Squares transformation on a given dataset. It offers three The follwing is an example of how to create a least squares object. cepLs ls; An implementation of the <command>cepLs</command> class which calculates a Linear Least Squares transformation. </para> <para> <command>const cepLs & cepDoVCV(cepMatrix<double> &data, cepMatrix<double> &matP)</command> Calculates one iteration of the VCV least squares algoritum. <para><itemizedlist> <listitem><para>data:- The data matrix</para></listitem> <listitem><para>matP:- The given weighting matrix</para></listitem> </itemizedlist></para> </para> <para> <command>const cepLs & cepDoVCV(cepMatrix<double> &data)</command> Caculates the VCV least squares algorithum, automatically re-weighting the weighting matrix (P matrix) untill the residuals are stable. <para><itemizedlist> <listitem><para>data:- The data matrix</para></listitem> </itemizedlist></para> </para> <para> <command>const cepLs & cepDoRW(cepMatrix<double> &matA, cepMatrix<double> &matP)</command> Calculates one iteration of the Random Walk least squares algorithum <para><itemizedlist> <listitem><para>data:- The data matrix</para></listitem> <listitem><para>matP:- The given weighting matrix</para></listitem> </itemizedlist></para> </para> <para> <command>const cepMatrix<double> &getResidual()</command> Returns a cepMatrix which is the residuals for this caluclation. </para> <para> <command>double getB1()</command> Returns the value B1 in the least squares solution y=B1*x +B2. </para> <para> <command>double getB2()</command> Returns the value B2 in the solution y=B1*x +B2 </para> <para> <command>cepError getError()</command> Returns any error that may have occoured SEEALSOcepDataset SEEALSOcepMatrix SEEALSOcepError This documentation was generated for the Geodetic Data Manipulation System 0.1 by autodocbook (http://www.stillhq.com). The Geodetic Data Manipulation System is under development by Michael Still (mikal@stillhq.com), Daniel Fernandez, Kristy Van Der Vlist, Blake Swadling, and Nick Wheatstone. All code is Copyright 2002 Michael Still, Daniel Fernandez, Kristy Van Der Vlist, Blake Swadling and Nick Wheatstone, and is released under the GNU GPL. Code submissions are welcomed. Contact Michael at the address above if you have something you would like to contribute. There are no known bugs in the Geodetic Data Manipulation System. If you find one, please contact mikal@stillhq.com and let me know.