[gmx-users] Inaccurate time frame

[Hsin-Lin Chiang]

  Hi, Mark

Thank you for your reply.
I know -skip -sep is more robust.
Since the definition talked about nr-th frame but not nr-th ps.

-skip     int     1     Only write every nr-th frame
-[no]sep     bool     no     Write each frame to a separate .gro, .g96 
or .pdb file

The problem is that -skip 1000 -sep didn't consist of several time 
frames in the same file but just write nr-th frame.
That's why I use -split 1000 -timestep 1 according these definitions 
found in manual.

-split     time     0     Start writing new file when t MOD split = 
first time (ps)
-timestep     time     0     Change time step between input frames (ps)
With -dt it is possible to reduce the number of frames in the output. 
This option relies on the accuracy of the times in your input 
trajectory, so if these are inaccurate use the -timestep option to 
modify the time (this can be done simultaneously).

Would you please teach how to use -skip -sep to get the same kind of 
file which include all of 1000ps time frames together in a 1ns gro file?

I'll appreciate to any helps.

Sincerely yours,
Hsin-Lin
> >/    Hi,
> />/
> />/  My time unit is 1ps and today I have 300ns data generated by parallel simulation.
> />/  I use trjconv -split 1000 on my trajectory but get the truncated end at t= 5000.00000
> />/  Theoretically it should stop at t= 1000.000
> />/  I found that I don't have t= 1000.00000 frame but have t= 1000.00006, 2000.00012, 3000.00024, and t= 4000.00024.
> />/
> />/  I know I can add -timestep 1 to solve this problem and let file can be truncated at t= 1000.00006.
> />/
> />/  How does this kind of inaccurate time frames happen?
> />/  Is this trajectory a wrong result?
> />/
> /I just wrote a FAQ+wiki page for this, since it gets asked a bit. Seehttp://www.gromacs.org/Documentation/Floating_Point_Arithmetic
>
> trjconv -skip -sep is a more robust approach here
>
> Mark
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20101207/0810810e/attachment.html>