[gmx-users] Problem with gmx clustsize ..(Not enough memory. Failed to calloc)

[Sergio Garay]

Thank you Mark for your answer! I have also tried to run clustsize
using different dt, from 1 ns to 10ns (the trajectory only have 10
frames) but it still give the same error, so I think that I am doing
something wrong. May be the problem is the way the program identify
each molecule.

Thank you very much, I will keep trying to solve this problem.

Sergio

>Hi,
>
>Clustering requires that you build a matrix in memory that contains the
>difference between each structure you intend to cluster. If you e.g. use
>every frame from your trajectory, then you will typically exhaust the
>available memory, as here. Reconsider how you will find out the period
>between statistically independent observations, and (if needed once you
>have done the former) access a node with more memory.
>
>Mark

>On Sat, May 12, 2018 at 4:18 PM Sergio Garay <sergio.alberto.garay at gmail.com <https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users>>
>wrote:

>* Dear Gromacs users
*>>* I am trying to analyse a MD simulation of 30 lipopeptides in water, in
*>* particular I would like to obtain the average number of these molecules in
*>* each assembly, along the time. I have tried g_clustsize with my simulation
*>* box with only the lipopeptides. For that I prepared a tpr file for these
*>* molecules only. But I have these error:
*>>* *g_clustsize -f 0-100ns.xtc -s micelle-alone.tpr -nc -mol -dt 1000*
*>>* Fatal error:
*>* Not enough memory. Failed to calloc -1082130432 elements of size 4 for
*>* grps[j].nm_ind
*>* (called from file
*>* /home/sgaray/Programas/gromacs-5.0.4/src/gromacs/fileio/tpxio.c,
*>* line 2510)
*>>* Could someone tell me what I am doing wrong?
*>>* Thank you in advance.*



-- 








*Sergio GarayDr. en Ciencias BiológicasFacultad de Bioquimica y Cs.
BiológicasUniversidad Nacional del LitoralSanta Fe - ArgentinaC.C. 242 -
Ciudad Universitaria - C.P. S3000ZAAArgentinaPh. +54 (342) 4575-213Fax. +54
(342) 4575-221 *