Re: [Rd] NaN and linear algebra

From: <>
Date: Wed 23 Mar 2005 - 20:13:11 GMT

I just tested this problem on a Windows 2000 Pro P4-Xeon system. I tried this in patched 2.0.1 (2005-2-28) and 2.1.0-alpha (2005-3-23) with both generic BLAS and precompiled ATLAS BLAS dll (downloaded from CRAN windows binaries). In all 4 combinations I get

> d<-matrix(NaN,3,3)
> f<-solve(d)
> f

     [,1] [,2] [,3]

[1,]  NaN  NaN  NaN
[2,]  NaN  NaN  NaN
[3,]  NaN  NaN  NaN

> det(d)

[1] NaN

which, I guess, is what is expected.

Hope this helps,


Andy Jaworski
Process Laboratory
3M Corporate Research Laboratory

Tel: (651) 733-6092
Fax: (651) 736-3122
             Prof Brian Ripley                                             
   >                                                     To 
             Sent by:                  Simon Urbanek                       
             r-devel-bounces@s         <>       
                                       Martin Maechler                     
             03/23/2005 01:28           
             PM                                                    Subject 
                                       Re: [Rd] NaN and linear algebra     

On Wed, 23 Mar 2005, Simon Urbanek wrote:

> On Mar 22, 2005, at 6:19 PM, Bill Northcott wrote:
>> On 23/03/2005, at 12:55 AM, Simon Urbanek wrote:
>> You may prefer the error, but it is not in the sprit of robust
>> ie
>> > d<-matrix(NaN,3,3)
>> > f<-solve(d)
>> Error in solve.default(d) : Lapack routine dgesv: system is exactly
>> singular
>> > f
>> Error: Object "f" not found
> For the record: this is the behavior on all platforms I tested (Mac,
> IRIX) - the only platform with a different result is reportedly Windows.
> what I was saying is that this is not "IEEE-754 non-compliance of MacOS
X" as
> you put it.

Hmm: all the platforms I have tested using R's internal blas and some with external BLASes pass this test. Only those with ATLAS and Goto BLAS' fail for me.

>> Clearly det(d) returning 0 is wrong. As a result based on a computation

>> including a NaN, it should return NaN.
> That's exactly what I was pointing out, and, yes, this is a bug in
> @Martin: I guess the most simple test for this would be probably
> is.nan(det(matrix(NaN,2,2)))
>>> Many functions in R will actually bark at NaN inputs (e.g. qr, eigen,
>>> - maybe you're saying that we should check for NaNs in solve before
>>> proceeding and raising an error?
>> However, this problem is in the Apple library not R.
> Since you're referencing the "solve" problem here, again, it's not.
> As I was explaining the the previous e-mail, there are at least two
> completely separate issues - handling of NaNs in solve and determinant of
> matrices. The latter is a bug in vecLib, the first one is not an OS X
> specific problem.

No, but it does appears to be an external BLAS-specific problem.

> Many R functions will object to taking NaN inputs such as qr.solve which
> was the previous implementation of solve. At any rate this is error
> handling in R and not Lapack. The "example" on the top doesn't make any
> sense because when you type "f" you know that it doesn't exist and both
> are R errors. If you used that in a script you have both choices: abort
> and that point or continue with whatever fallback you choose.

Brian D. Ripley,        
Professor of Applied Statistics,
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

______________________________________________ mailing list

______________________________________________ mailing list
Received on Thu Mar 24 07:19:08 2005

This archive was generated by hypermail 2.1.8 : Mon 24 Oct 2005 - 22:26:32 GMT