lammps/doc/next.txt

97 lines
3.0 KiB
Plaintext

"LAMMPS WWW Site"_lws - "LAMMPS Documentation"_ld - "LAMMPS Commands"_lc :c
:link(lws,http://lammps.sandia.gov)
:link(ld,Manual.html)
:link(lc,Section_commands.html#comm)
:line
next command :h3
[Syntax:]
next variables :pre
variables = one or more variable names :ul
[Examples:]
next x
next a t x myTemp :pre
[Description:]
This command is used with variables defined by the
"variable"_variable.html command. It assigns the next value to the
variable from the variable's list, so that when a variable is
subsequently substituted for in an input script command, the new value
is used. If a variable name is a single lower-case character from "a"
to "z", it can be used in an input script command as $a or $z. If it
is multiple letters, it can be used as ${myTemp}.
All variables in a single next command must be the same style:
{index}, {loop}, or {universe}. {Equal}- or {world}-style variables
cannot be incremented by a next command.
When any of the variables in the next command has no more values, a
flag is set that causes the input script to skip the next
"jump"_jump.html command encountered. This enables a loop containing
a next command to exit.
When the next command is used with {index}- or {loop}-style variables,
the next value is assigned to the variable for all processors. When
the next command is used with {universe}-style variables, the next
value is assigned to whichever processor partition executes the
command first. All processors in the partition are assigned the same
value. Running LAMMPS on multiple partitions of processors via the
"-partition" command-line switch is described in "this
section"_Section_start.html#2_4 of the manual. {Universe}-style
variables are incremented using the files "tmp.lammps.variable" and
"tmp.lammps.variable.lock" which you will see in your directory during
such a LAMMPS run.
Here is an example of running a series of simulations using the next
command with an {index}-style variable. If this input script is named
in.polymer, 8 simulations would be run using data files from
directories run1 thru run8.
variable d index run1 run2 run3 run4 run5 run6 run7 run8
shell cd $d
read_data data.polymer
run 10000
shell cd ..
clear
next d
jump in.polymer :pre
If the variable "d" were of style {universe}, and the same in.polymer
input script were run on 3 partitions of processors, then the first 3
simulations would begin, one on each set of processors. Whichever
partition finished first, it would assign variable "d" the 4th value
and run another simulation, and so forth until all 8 simulations were
finished.
Jump and next commands can also be nested to enable multi-level loops.
For example, this script will run 15 simulations in a double loop.
variable i loop 3
variable j loop 5
clear
...
read_data data.polymer.$i$j
print Running simulation $i.$j
run 10000
next j
jump in.script
next i
jump in.script
[Restrictions:] none
[Related commands:]
"jump"_jump.html, "include"_include.html, "shell"_shell.html,
"variable"_variable.html,
[Default:] none