Dear Users,
In my new project I want to convert a structure containing more than 60,000 atoms using msi2lmp tool. I get this error:
Unable to resolve atom number of atom 30657 conn 0 …
Please also look at the photo attached.
Do you have any experience how to solve this error? Thanks.
I found the codes, but could not exactly comprehend them:
http://redmine.scorec.rpi.edu/anonsvn/lammps-cuda/tools/msi2lmp/src/ReadMdfFile.c
Dear Users,
In my new project I want to convert a structure containing more than
60,000 atoms using msi2lmp tool. I get this error:Unable to resolve atom number of atom 30657 conn 0 ..............
Please also look at the photo attached.
Do you have any experience how to solve this error? Thanks.
no, but the problem is obviously due to the fact that the "unique name"
msi2lmp tries to generate is overflowing the "tempname" variable. so you
should try to a name shorter than XXXX for your molecule and similar
things.
Dear Users,
In my new project I want to convert a structure containing more than
60,000 atoms using msi2lmp tool. I get this error:Unable to resolve atom number of atom 30657 conn 0 ..............
Please also look at the photo attached.
Do you have any experience how to solve this error? Thanks.
I found the codes, but could not exactly comprehend them:http://redmine.scorec.rpi.edu/anonsvn/lammps-cuda/tools/msi2lmp/src/ReadMdfFile.c
this seems to be an over 5 year old version of the source.
the memory corruption that is causing your problem seems to be happening
elsewhere, though.
can you provide the mdf and car files that are giving the problems?
axel.
Dear Prof. Axel, thanks for your help.
The files are attached.
I used the precompiled version of msi2lmp.exe included in lammps download package; I got that error with the latest version in windows, showing an unsuccessful termination of the process (Something is wrong with the mdf file). Nevertheless, with a newly complied msi2lmp.exe file in linux, I only noticed exactly the same tempname overrun errors, but the process could finally terminate successfully and the data file was produced.
CP.car (4.84 MB)
CP.mdf (5.44 MB)
Dear Prof. Axel, thanks for your help.
The files are attached.
I used the precompiled version of msi2lmp.exe included in lammps download
package; I got that error with the latest version in windows, showing an
unsuccessful termination of the process (Something is wrong with the mdf
file). Nevertheless, with a newly complied msi2lmp.exe file in linux, I only
noticed exactly the same tempname overrun errors, but the process could
finally terminate successfully and the data file was produced.
you have to be extremely careful with this data file. the warning and
errors indicate some memory corruption is happening when trying to
determine the bond topology..
the difference between windows and linux is likely due to differences
in the memory layout imposed by the operating system.
Dear Prof. Axel, thanks for your help.
The files are attached.
thanks a lot for the files. i made some tests and found that the code
reading the mdf files was not fully honoring the compile time
constants MAX_NAME and MAX_STRING and was using some other values.
after a little refactoring this limitation could be lifted and the
updated msi2lmp seems to be able to read your mdf file without
complaints.
i will build and upload new rpms and windows installers later today or
tomorrow after i have merged in steve's latest patches.
axel.
Thank you very much for your time and help 