[R] Singularity problem

Berend Hasselman bhh at xs4all.nl
Wed Mar 16 21:47:09 CET 2011


On 16-03-2011, at 21:11, David Winsemius wrote:

> 
> On Mar 16, 2011, at 1:32 PM, Berend Hasselman wrote:
>> .....
>> svd(a) indicates the problem.
>> 
>> largest singular value / smallest singular value=1e17  (condition number)
>> --> reciprocal condition number=1e-17
>> and the standard solve can't handle that.
> 
> Actually it can if you relax the default tolerance settings:
> 
> > solve(a, tol=1e-21)
>      [,1]   [,2]
> [1,]  1e-20 -1e-20
> [2,] -1e-21  1e-03
> > a%*%solve(a, tol=1e-21)
>    [,1] [,2]
> [1,]    1    0
> [2,]    0    1


Lesson: Use ? before replying:-)

Berend



More information about the R-help mailing list