[gmx-users] RE:g_cluster crashes on large structures?
K.A. Feenstra
Feenstra at chem.vu.nl
Mon Aug 12 09:43:57 CEST 2002
Mikko Huhtala wrote:
>
> Sorry to reply to my own message. I realized that the problem was that
> I was trying to select the whole system for output. g_cluster works on
> the trajectory, if solvent is not output. So I guess the short answer
> is that I should be in the market fr more memory (and g_cluster is not
> a particularly memory efficient program?).
Right ;-) Inevitably, if you want to cluster conformations in a
trajectory, you need access to each and all of the frames. This
means that you will want to read all of them into memory. So,
the out-of-memory you encountered was not for the 55095*4 bytes
mentioned in the error message, but this was already the n-th time
this was done for n frames in the trajectory (where n would be
approx 1G/(55095*4)...).
Of course, in this particular case were you cluster on C-a coordinates,
but you want output for the whole system it would be possible to
re-write the program to only use memory for the C-a coordinates
during clustering and re-read the whole trajectory for output only,
but I don't feel this is a general case.
--
Groetjes,
Anton
________ ___________________________________________________________
| | Anton Feenstra |
| . | Dept. of Pharmacochemistry - Vrije Universiteit Amsterdam |
| |---- | De Boelelaan 1083 - 1081 HV Amsterdam - The Netherlands |
| |---- | Tel: +31 20 44 47608 - Fax: +31 20 44 47610 |
| ' __ | Feenstra at chem.vu.nl - http://www.chem.vu.nl/afdelingen/FAR|
| / \ |-----------------------------------------------------------|
| ( ) | Dept. of Biophysical Chemistry - University of Groningen |
| \__/ | Nijenborgh 4 - 9747 AG Groningen - The Netherlands |
| __ | Tel +31 50 363 4327 - Fax +31 50 363 4800 |
| / \ | K.A.Feenstra at chem.rug.nl - http://md.chem.rug.nl/~anton |
| ( ) |-----------------------------------------------------------|
| \__/ | "If You See Me Getting High, Knock Me Down" |
| | (Red Hot Chili Peppers) |
|________|___________________________________________________________|
More information about the gromacs.org_gmx-users
mailing list