Hi Sebastien and Carlos, About 6 months ago, I asked in roottalk (http://root.cern.ch/root/roottalk/roottalk03/4730.html) who is using this routine . Nobody reacted. Reason was that the internals of the matrix package have been rewritten . In particular, the storage of the array elements . It used to be column-wise, now it is row-wise !!!! If one accesses these elements directly (through GetElements()) , one has to carefully consider the implications of this change . For instance, if your matrix is symmetric, you do not care . In order to avoid mishap, we decided to change the name of the routine and make the old one private, so that the user is forced to look at it . Replacing GetElements by GetMatrixArray, can not be a big deal fo the user, however using the methods on non-symmetric matrices, not being aware of the storage change is a disaster. Eddy --- Sebastien Greder <greder@in2p3.fr> wrote: > > Hi Carlos, > > you're right I just noticed this also and it is > very annoying, > breaking backward compatibility ; it has been > replaced by > SetMatrixArray and moved up in TMatrixDBase class. > Could the developpers take it back since > SetMatrixArray uses > the same logic ; SetElements would just have to > make a call > to the latter. > > thanks, > > sebastien. > > On Thu, 20 May 2004, [ISO-8859-1] Carlos Muņoz > Camacho wrote: > > > > > Hi, > > > > What happened to TMatrixD::SetElements(const > Double_t* elements, > > Option_t* option) in ROOT 4.00/04 ? It doesn't > seem to be there any > > more and I wouldn't like to change all my code > which extensively relies on > > this method. > > > > Carlos > > > > > > > --------------------------------------------------------------------- > | Sebastien GREDER tel : > +33 (0)3 88 10 63 54 | > | D0 Strasbourg, | > | Institut de Recherches Subatomiques fax : > +33 (0)3 88 10 62 34 | > | 23, rue du loess - BP 28 secr : > +33 (0)3 88 10 66 40 | > | F-67037- STRASBOURG cedex 2 e-mail : > greder@in2p3.fr, | > | greder@fnal.gov | > > --------------------------------------------------------------------- > > > > >
This archive was generated by hypermail 2b29 : Sun Jan 02 2005 - 05:50:08 MET