On 15/07/10 17:39, Daniel O'Donovan wrote:
> Hi Justin,
>
> Tim says that the first thing to try would be to delete old ensembles and restraint sets - this should help.
>
> Brian Smith has been having similar issues and so switched to a 32 bit system. The problem (I think) lies in the fact that Python has no float (32 bit) / double (64 bit) distinction without explicit coding. Mr. Python will assume that you want to store large arrays to double precision (never mind where they come from) and this effectively doubles the memory foot-print!
>
> We are aware of this issue and we're trying to think of innovative ways to avoid it. I'm pretty sure this isn't a 'leak' but if you have evidence of memory leaking then that would be very helpful,
>
> Thanks, Dan
>
> Daniel O'Donovan
> [log in to unmask]
Hi Dan,
no I don't think that something is leaking, more what you said, that
something in either the datastructure or the way data is used, is going
wrong. But I will clean up everything. Hope that works.
Thanks, justin
--
Justin Lecher
Institute for Neuroscience and Biophysics
ISB 3 - Institute for structural biochemistry
Research Centre Juelich GmbH,
52425 Juelich,Germany
phone: +49 2461 61 5385
|