BTW, if you need a full example, I think Zhen Li @zz <958497401@xxxxxx>
will be happy to provide one. The above preprint has just been accepted for
On Thu, Nov 21, 2019 at 3:06 PM Bruce Fan <brucenju@xxxxxxxxx> wrote:
This parameter in the xyz.in file refers to the number of grouping
methods you will *define* in this file. It can be 0, 1, 2, up to 10 (?).
I cannot remember if there is a limit :-)
If it is 0, you don't need to write any group induces in this file. This
is the case for ex1, ex2, and ex3.
If it is 1, you need to add a group index for each atom (each line
starting from line 3), after the mass (assuming you don't have velocities
here). This is the case for ex4.
Someone also have used "2", which requires adding two group indices for
each atom in the following form:
type x y z mass a b
Here, "a" and "b" indicate that the current atom is in group "a" for
grouping method "0" and in group "b" for grouping method "1". (We always
count from 0!) For example, in our recent preprint [arXiv:1905.11024
<https://arxiv.org/abs/1905.11024>], we have defined two grouping
methods, one used for the usual NEMD simulation (we need to define a group
index for the heat source, a group index for the sink, and some in the
middle), and the other for outputting the temperature profile in a greater
detail (even the temperature distribution in the heat source and sink). If
you check Fig. 4 and Fig. 5 of the above preprint, you will see that the
points in Fig. 4 are very dense (which are from grouping method "1") and
the points in Fig. 5 are very sparse (which are from grouping method "0").
In the run.in file, we have something like:
emsemble heat_lan 300 200 5 1 16 # this will use the grouping method
"0" by default
compute 1 10 1000 temperature # the first parameter "1" means that the
grouping method "1" will be used here
On Thu, Nov 21, 2019 at 2:43 PM Florencia Carusela <flor.caru@xxxxxxxxx>
Could you clarify me (may be with an example from 0-2) what this
parameter refers to? I think I am setting it incorrectly .