The manual is up to date. It is because the code you downloaded is not the
latest.
I have been busy with sonething else. I think I will review the code and
manual and make a new version soon. There have been a few interesting new
features added since v2.4.1.
Zheyong
Zheyong
Florencia Carusela <flor.caru@xxxxxxxxx> 于 2020年4月24日周五 20:52写道：
Zheyong, I used the formats defined in the manual. May be they are not all
updated.
Thanks again!
Florencia
On Fri, Apr 24, 2020 at 2:42 PM Bruce Fan <brucenju@xxxxxxxxx> wrote:
Your experience indicates thatI need to improve the code to make it
report more reasonable error:-)
Zheyong
Florencia Carusela <flor.caru@xxxxxxxxx> 于 2020年4月24日周五 20:27写道：
Thanks! Now works..
F.
On Fri, Apr 24, 2020 at 2:22 PM Bruce Fan <brucenju@xxxxxxxxx> wrote:
N M cutoff 0 has_velocity has_layer number_of_grouping_methods
7 items, not 6.
Zheyong
Bruce Fan <brucenju@xxxxxxxxx> 于 2020年4月24日周五 20:19写道：
2638 50 3.0 0 1 0 1
Zheyong
Florencia Carusela <flor.caru@xxxxxxxxx> 于 2020年4月24日周五 19:59写道：
Hi Zheyong
If I use:
2638 50 3.0 1 0 1
0 0 0 50.5 60.5 16.5
This is the output:
---------------------------------------------------------------
Started initializing positions and related parameters.
---------------------------------------------------------------
Number of atoms is 2638.
Maximum number of neighbors is 50.
Initial cutoff for neighbor list is 3 A.
*Use triclinic box.*
*Do not specify initial velocities here.*
Specify layer indices here.
Have no grouping method.
Use free boundary conditions along x.
Use free boundary conditions along y.
ERROR: invalid boundary conditions along z.
Exit
*___________________________________________________________*
If I use:
2638 50 3.0 0 1 1
0 0 0 50.5 60.5 16.5
--> I set an orthogonal box ( I need this)
Output:
---------------------------------------------------------------
Started initializing positions and related parameters.
---------------------------------------------------------------
Number of atoms is 2638.
Maximum number of neighbors is 50.
Initial cutoff for neighbor list is 3 A.
*Use orthogonal box.*
*Specify initial velocities here.*
Specify layer indices here.
Have no grouping method.
Use free boundary conditions along x.
Use free boundary conditions along y.
ERROR: invalid boundary conditions along z.
Exit
*_____________________________________________________*
However, look the following. Suppose that now I set x and z free,
and y periodic. And an orthogonal box
2638 50 3.0 0 1 1
0 1 0 50.5 60.5 16.5
Output
---------------------------------------------------------------
Started initializing positions and related parameters.
---------------------------------------------------------------
Number of atoms is 2638.
Maximum number of neighbors is 50.
Initial cutoff for neighbor list is 3 A.
Use orthogonal box.
Specify initial velocities here.
Specify layer indices here.
Have no grouping method.
*Use periodic boundary conditions along x.* * ->> this
should be free !!!!! *
*Use free boundary conditions along y. ->> this
should be periodic !!!!!*
ERROR: invalid boundary conditions along z.
May be the boundary conditions are not been read correctly o may be I
am not setting them correctly...? May be this because my file includes
velocities?
Thanks !
Florencia
On Fri, Apr 24, 2020 at 1:22 PM Bruce Fan <brucenju@xxxxxxxxx> wrote:
In GPUMD-v2.4.1 or the earlier stage of the developing version (not
released), the first two lines in xyz.in should be:
N M cutoff 0 has_velocity *has_layer* number_of_grouping_methods
pbc_x pbc_y pbc_z L_x L_y L_z
However, at some point for the developing version, we have removed
the item *has_layer*, which is the format you used. It is likely
you are using v2.4.1 or an earlier copy of the developing version.
So please try to change your first line
2638 50 3.0 0 1 1
to
2638 50 3.0 0 1 *0* 1
Welcome to discuss if this does not solve your problem.
Zheyong
On Fri, Apr 24, 2020 at 3:35 PM Bruce Fan <brucenju@xxxxxxxxx>
wrote:
thanks, I will check it soon.
Zheyong
Florencia Carusela <flor.caru@xxxxxxxxx> 于 2020年4月24日周五 14:58写道：
Hi Zheyong
Sorry but I cannot find the info of the GPUMD version I installed
in any file. I downloaded from the main page in April 2019...
probably it
is the v2.4.1
<https://github.com/brucefan1983/GPUMD/releases/tag/v2.4-correction>
but I am not 100% sure.
F.
On Fri, Apr 24, 2020 at 4:41 AM Bruce Fan <brucenju@xxxxxxxxx>
wrote:
Hi Florencia,
Could you report your GPUMD version? This is important
because the xyz.in file format has been modified a couple of
times during the last few versions.
Zheyong
On Thu, Apr 23, 2020 at 11:10 PM Florencia Carusela <
flor.caru@xxxxxxxxx> wrote:
Hi
I am trying to run heat_lan ensemble with the xyz.in file that
attach (It has only some lines, the first 340).
I need free boundary in all directions. However, I always obtain
this error message and I cannot understand the reason:
ERROR: invalid boundary conditions along z.
I tried different types and combinations (in x-y-z) of boundary
conditions in the second line of xyz.in and I always get the
same
error (always z).
Could you please tell me why? May be something is not correct in
my xyz.in? (I do not have this problem if I don't include
velocities.)
Thanks a lot.
Florencia
