[Rd] bug in det using method="qr" (PR#1244) (PR#4450)

wb at arb-phys.uni-dortmund.de wb at arb-phys.uni-dortmund.de
Wed Nov 12 13:37:13 MET 2003


I just detected, that det() is not working on complex matrices any more,
due to the fix to the bug reports noted above. I am not happy with this,
as determinants are perfectly usable on complex matrices.

AFAIUI the bugs resulted from less than optimal behaviour of qr() in
certain cases. IMHO this is due to the unhappy decision to use a default for
parameter tol to decide whether the the decomposition is rank deficient.

A better fix for (PR#1244) should be considered. I propose, using qr method
with tol=0 as default, for det(). Even more preferrable, tol=0 can be made the
default for qr(), forcing all applications to set a reasonable tol for their
own.

wbk
-- 
Dipl.-Math. Wilhelm Bernhard Kloke
Institut fuer Arbeitsphysiologie an der Universitaet Dortmund
Ardeystrasse 67, D-44139 Dortmund, Tel. 0231-1084-257



More information about the R-devel mailing list