Project

General

Profile

Bug #492

gmx_rmpbc and periodic molecules

Added by Berk Hess over 9 years ago. Updated over 9 years ago.

Status:
Closed
Priority:
Normal
Assignee:
Erik Lindahl
Category:
analysis tools
Target version:
Affected version - extra info:
Affected version:
Difficulty:
uncategorized
Close

Description

gmx_rmpbc should not try to make periodic molecules whole.
gmx_rmpbc currently only gets t_idef passed, not t_inputrec,
so it can not know if periodic molecules are present.
Since gmx_rmpbc gets called 50 times, I do not want to do such
a large change in the beta release stage.
We should fix this properly by having a single trajectory reading class
that also determines the consistency of the tpr and trx file.

Berk

History

#1 Updated by Erik Lindahl over 9 years ago

Hi,

I think we discussed this before, but I can't remember why it's really necessary to have the entry in the inputrec. Couldn't we simply construct a bond graph to automatically determine whether there are periodic molecules that undergo a net non-zero box shift before connecting back to the first atom?

#2 Updated by David van der Spoel over 9 years ago

Yes that's possible, but I agree with Berk: not now. This code is the worst we have in gromacs as regards bug history. Mea culpa.

#3 Updated by Erik Lindahl over 9 years ago

Not now. Too complex change at this stage.

Also available in: Atom PDF