From: Jan Saam (jan.saam_at_charite.de)
Date: Tue Sep 05 2006 - 19:12:07 CDT

Hi Axel (and John),

I believe that a TCL interface to general matrix/linear algebra package
such as BLAS would be among the the best imaginable improvements of VMD
since it would allow for so many serious types of data analysis! We need
SVD and stuff like that!
I've been thinking about it myself, but same as you I never found the time.
Maybe we should urge John or coordinate the work somehow...

Actually, in one of the paratool related plugins (hesstrans) we used a
freeware C++ based Matrix library (newmat10) which I found very handy, but
we are just using it in the C++ code without a TCL interface. It would a
moderate amount of work to wrap it in TCL and since it is rather small we
could ship it with the plugin. But John has once pointed out that this lib
has issues since it does not compile on some machines.

So we should probably indeed use one of the less convenient but ubiquitous
ones.
What do you guys think?

Jan

---------------------------
Jan Saam
Institute of Biochemistry
Charite Berlin
Monbijoustr. 2
10117 Berlin
Germany

+49 30 450-528-446
saam_at_charite.de

> hi,
>
> a generic matrix inversion is not available.
> to undo many transformation matrices, you
> may get away with a transpose+rescale (it
> has been a looong time since i needed to
> deal with it seriously).
>
> but you are in the market for real inversion, you
> may want to have a look at:
> http://www.hume.com/la/
> if your code is not performance critical.
>
> other than that, the next best option would be
> to implement a wrapper plugin for BLAS/LAPACK...
> i've been thinking about this for a while,
> but there was rarely enough contigouus time
> (and if there was, there were more pressing projects)
> to try something like that in a serious manner.
>
> cheers,
> axel.
>
> On 9/5/06, Neelanjana Sengupta <senguptan_at_gmail.com> wrote:
>> Hi,
>> Does VMD have a routine to find the inverse of a matrix? I am referring
>> to
>> the 'trans' matrix operations, used to perform a variety pf coordianate
>> transformations:
>> http://www.ks.uiuc.edu/Research/vmd/current/ug/node176.html
>>
>> I did not see anything like "transinverse"...
>>
>> Thanks and regards,
>> Neela
>>
>
>
> --
> =======================================================================
> Axel Kohlmeyer akohlmey_at_cmm.chem.upenn.edu http://www.cmm.upenn.edu
> Center for Molecular Modeling -- University of Pennsylvania
> Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
> tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425
> =======================================================================
> If you make something idiot-proof, the universe creates a better idiot.
>