VMD-L Mailing List
From: John Stone (johns_at_ks.uiuc.edu)
Date: Thu Jun 07 2007 - 11:11:43 CDT
- Next message: lily jin: "Re: release memory in TCL"
- Previous message: Lambert van Eijck: "Re: traj not visualized but pdb is visualized"
- In reply to: Erkan Tuzel: ".xyz file, namelist index too large error"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
Hi,
The warning you're getting is the result of running out of index
space for unique atom names. The actual limit is much greater than 111,
but 111 was the value of the index at the time the problem occured, the
error message text is more helpful for me than for users I suppose :-)
The code in VMD 1.8.6 should allow at least 32,000 unique atom names/types
in a molecule.
What does your XYZ file look like?
Most likely you've got sequence numbers or some other non-atom-name
data being read as if it was the atom name, which would indeed run VMD
short on name indices. VMD 1.8.5 used twice as much memory for storing
various indices, allowing problematic files like this to be loaded.
VMD 1.8.6 has been made much more memory efficient, but the cost of this
is that files that contain serial numbers in atom name fields are going
to cause trouble.
The solution is either to:
a) write your XYZ files with atom names/types rather than serial numbers in
the atom name fields...
or
b) recompile VMD changing the nameindex type from "short" to "int", which
would give you at least 2 billion unique atom names/types, which is what
VMD 1.8.5 did. We felt this was wasteful in general, and that it would be
best for people to correct their data files, as in a) above.
If you can send me a sample of one of your XYZ files, I should be able
to determine why you're running into this problem.
Cheers,
John Stone
vmd_at_ks.uiuc.edu
On Wed, Jun 06, 2007 at 07:39:48PM -0500, Erkan Tuzel wrote:
> Hi,
>
> I have been using the XYZ file format for quite a while to display
> atoms in a 2d plane. However, after I have upgraded the version
> 1.8.6, VMD gives the following
> error while reading the data file:
>
> Info) Using plugin xyz for structure file test.xyz
> ERROR) BaseMolecule: Cannot add atom; namelist index value too large.
> ERROR) Recompile VMD with larger index types.
> ERROR) Atom namelist index values at time of overflow:
> ERROR) nameindex: 111
> ERROR) typeindex: 111
> ERROR) resnameindex: 0
> ERROR) segnameindex: 0
> ERROR) altlocindex: 0
> ERROR) chainindex: 0
> ERROR) MolFilePlugin: file load aborted
>
> Unless I reduce the number of atoms to the number specified, it does
> not seem to work. I normally have around 2000-3000 atoms in a file
> and the files work well with version 1.8.5. I use Mac OS X version
> 10.4 with PowerPC. Does anybody know a solution to this? Is this a
> bug? 111 atoms seems to be pretty small.
>
> Erkan
>
>
>
-- NIH Resource for Macromolecular Modeling and Bioinformatics Beckman Institute for Advanced Science and Technology University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801 Email: johns_at_ks.uiuc.edu Phone: 217-244-3349 WWW: http://www.ks.uiuc.edu/~johns/ Fax: 217-244-6078
- Next message: lily jin: "Re: release memory in TCL"
- Previous message: Lambert van Eijck: "Re: traj not visualized but pdb is visualized"
- In reply to: Erkan Tuzel: ".xyz file, namelist index too large error"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]