I’m running tad and my dump file has a bunch of random stuff in the beginning. I’ve tried both xyz and dcd. I’m loading a dcd into a pdb in VMD. Has anyone experienced this?
I'm running tad and my dump file has a bunch of random stuff in the beginning. I've tried both xyz and dcd. I'm loading a dcd into a pdb in VMD. Has anyone experienced this?
------------------------------------------------
Not like this, but this can happen if you write to the same file on a
networked filesystem from multiple Processes.
Axel
Axel, I am writing from multiple file streams. I’m have 12 processors on two different computers and I’m using the command line argument -partition 12x1. I’m guessing that that’s probably messing up the dcd/xyz write. Is there a correct or better way to write the tad position information?
Currently I’m just declaring a dump as:
dump 1 all dcd 1000 simulation.dcd
Thanks for the help.
Axel, I am writing from multiple file streams. I'm have 12 processors on two different computers and I'm using the command line argument -partition 12x1. I'm guessing that that's probably messing up the dcd/xyz write. Is there a correct or better way to write the tad position information?
Currently I'm just declaring a dump as:
dump 1 all dcd 1000 simulation.dcd
Thanks for the help.
Have you looked at the dump documentation? If there is a way of having
a per partition output naming.
Axel
In the dump documentation there is this section for “%”
If a “" character appears in the filename, then one file is written for each processor and the "” character is replaced with the processor ID from 0 to P-1. For example, tmp.dump.% becomes tmp.dump.0, tmp.dump.1, …
tmp.dump.P-1, etc. This creates smaller files and can be a fast mode of output on parallel machines that support parallel I/O for output. This option is not available for the dcd, xtc, and xyz styles.
But the problem is that “%” doesn’t work for dcd or xyz.
I would like to be able to load a dcd or xyz file into a pdb in VMD if that’s possible.
Thanks
Try looking at the tad example in the examples dir.
You can define the filename with a world-style
variable (see the variable doc page), which will
generate a different file name for each replica,
You need to do this for all the replica-based run styles (tad, prd, temper, etc)
if you want separate output from each replica.
Steve