This is the mail archive of the gsl-discuss@sources.redhat.com mailing list for the GSL project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Householder solver


Henri Lindén writes:
 > I am having problems with the Householder alogithm. Usually this is
 > used for overdetermined linear systems, i.e. systems in which m >
 > n, where A \in R^{m \times n}. Then how come i.e. the following
 > code: yields the following error message:
 >  gsl: hh.c:41: ERROR: System is underdetermined Aborted
 >  ???
 >  Usually the first dimension refers to row dimension and the second
 > to columns, which is exactly as things generally are done in
 > GSL. Thus my example system should by all means be overdetermined
 > with 5 equalities and 3 variables. However the Householder solver
 > seems to disagree!

Hi, 

Thanks for the bug report. 

That code looks to be a bit broken. It is only working for M=N and
gives the wrong message for M>N.  

Maybe somebody on the list could volunteer to fix it (??)  (also to
make it take a workspace argument of size N for the column norms
instead of allocating memory internally).

In the meantime I'd recommend using the QR or SVD routines for
overdetermined systems instead.

regards
-- 
Brian Gough

----------------------------------------------------------------------
Network Theory Ltd            Phone: 0117 3179309 (+44 117 3179309)
15 Royal Park                 WWW: http://www.network-theory.co.uk/
Bristol BS8 3AL               Email: bjg@network-theory.co.uk     
United Kingdom                
----------------------------------------------------------------------


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]