[SciPy-dev] scipy.test() failures

Neil Killeen nkilleen@unimelb.edu...
Mon Oct 26 21:04:53 CDT 2009


On 27/10/2009, at 11:45 , David Warde-Farley wrote:

> On Tue, Oct 27, 2009 at 10:18:15AM +1100, Neil Killeen wrote:
>>>>
>>>> ERROR: Failure: ImportError (/usr/local/lib/python2.4/site- 
>>>> packages/
>>>> scipy/linalg/fblas.so: undefined symbol: srotmg_)
>>>>
>>>> and
>>>>
>>>> ImportError: cannot import name {linalg, calc_lwork}
>>>>
>>>> Perhaps the latter are due to the former ?
>>>>
>>>>
>
>> yes from source
>>
>>>
>>> What about lapack?
>>
>> i am using the system version, /usr/lib  The FAQ states that the
>> srotmg error is a BLAS
>> error and that I should install from netlib which I did
>
> Disclaimer: I usually build ATLAS and LAPACK from source, I've
> never really played with the netlib BLAS much.

I'm a bit confused by all the various packagings (bear in mind I am
just installing this for end users, I'm not a consumer myself).

By going to netlib, which I thought was the home of BLAS,
I *did* install from source.

ATLAS seems to be repackaging/reimpl of LAPACK and BLAS ?

Installing ATLAS seemed rather a lot of bother so I was trying
the easy path,

>
> liblapack.a may have certain BLAS functions inside the archive (I  
> think)
> if you installed ATLAS packages from your distribution.
> I'd recommend a clean LAPACK as well, and building numpy and scipy
> against the same set of libraries, just to be safe... From  
> show_config()
> it looks like everything should be in order, I think.

ok i will try that.   I gather though buiding from the reference  
implementations
may bring poor performance and that ATLAS seeks to redress that ?

Do you recommend I bite the bullet and try to make an ATLAS install ?

>
> Are you getting the exact same errors after building against your
> source built BLAS?

yes

>
> David
> _______________________________________________
> Scipy-dev mailing list
> Scipy-dev@scipy.org
> http://mail.scipy.org/mailman/listinfo/scipy-dev


cheers
Neil
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.scipy.org/pipermail/scipy-dev/attachments/20091027/3ae113f5/attachment-0001.html 


More information about the Scipy-dev mailing list