lammps/doc/run.html

178 lines
7.2 KiB
HTML
Raw Normal View History

<HTML>
<CENTER><A HREF = "http://lammps.sandia.gov">LAMMPS WWW Site</A> - <A HREF = "Manual.html">LAMMPS Documentation</A> - <A HREF = "Section_commands.html#comm">LAMMPS Commands</A>
</CENTER>
<HR>
<H3>run command
</H3>
<P><B>Syntax:</B>
</P>
<PRE>run N keyword values ...
</PRE>
<UL><LI>N = # of timesteps
<LI>zero or more keyword/value pairs may be appended
<LI>keyword = <I>upto</I> or <I>start</I> or <I>stop</I> or <I>pre</I> or <I>post</I> or <I>every</I>
<PRE> <I>upto</I> value = none
<I>start</I> value = N1
N1 = timestep at which 1st run started
<I>stop</I> value = N2
N2 = timestep at which last run will end
<I>pre</I> value = <I>no</I> or <I>yes</I>
<I>post</I> value = <I>no</I> or <I>yes</I>
<I>every</I> values = M c1 c2 ...
M = break the run into M-timestep segments and invoke one or more commands between each segment
c1,c2,...,cN = one or more LAMMPS commands, each enclosed in quotes
c1 = NULL means no command will be invoked
</PRE>
</UL>
<P><B>Examples:</B>
</P>
<PRE>run 10000
run 1000000 upto
run 100 start 0 stop 1000
run 1000 pre no post yes
run 100000 start 0 stop 1000000 every 1000 "print Protein Rg = $r"
run 100000 every 1000 NULL
</PRE>
<P><B>Description:</B>
</P>
<P>Run or continue dynamics for a specified number of timesteps.
</P>
<P>When the <A HREF = "run_style.html">run style</A> is <I>respa</I>, N refers to outer
loop (largest) timesteps.
</P>
<P>A value of N = 0 is acceptable; only the thermodynamics of the system
are computed and printed without taking a timestep.
</P>
<P>The <I>upto</I> keyword means to perform a run starting at the current
timestep up to the specified timestep. E.g. if the current timestep
is 10,000 and "run 100000 upto" is used, then an additional 90,000
timesteps will be run. This can be useful for very long runs on a
machine that allocates chunks of time and terminate your job when time
is exceeded. If you need to restart your script multiple times
(reading in the last restart file), you can keep restarting your
script with the same run command until the simulation finally
completes.
</P>
<P>The <I>start</I> or <I>stop</I> keywords can be used if multiple runs are being
performed and you want a <A HREF = "fix.html">fix</A> command that changes some
value over time (e.g. temperature) to make the change across the
entire set of runs and not just a single run. See the doc page for
individual fixes to see which ones can be used with the <I>start/stop</I>
keywords. The <A HREF = "pair_style.html">pair_style soft</A> potential also
changes its pair potential coefficients in this manner.
</P>
<P>For example, consider this fix followed by 10 run commands:
</P>
<PRE>fix 1 all nvt 200.0 300.0 1.0
run 1000 start 0 stop 10000
run 1000 start 0 stop 10000
...
run 1000 start 0 stop 10000
</PRE>
<P>The NVT fix ramps the target temperature from 200.0 to 300.0 during a
run. If the run commands did not have the start/stop keywords (just
"run 1000"), then the temperature would ramp from 200.0 to 300.0
during the 1000 steps of each run. With the start/stop keywords, the
ramping takes place over the 10000 steps of all runs together.
</P>
<P>The <I>pre</I> and <I>post</I> keywords can be used to streamline the setup,
clean-up, and associated output to the screen that happens before and
after a run. This can be useful if you wish to do many short runs in
succession (e.g. LAMMPS is being called as a library which is doing
other computations between successive short LAMMPS runs).
</P>
<P>By default (pre and post = yes), LAMMPS creates neighbor lists,
computes forces, and imposes fix constraints before every run. And
after every run it gathers and prints timings statistics. If a run is
just a continuation of a previous run (i.e. no settings are changed),
the initial computation is not necessary; the old neighbor list is
still valid as are the forces. So if <I>pre</I> is specified as "no" then
the initial setup is skipped, except for printing thermodynamic info.
Note that if <I>pre</I> is set to "no" for the very 1st run LAMMPS
performs, then it is overridden, since the initial setup computations
must be done.
</P>
<P>IMPORTANT NOTE: If your input script changes settings between 2 runs
(e.g. adds a <A HREF = "fix.html">fix</A> or <A HREF = "dump.html">dump</A> or
<A HREF = "compute.html">compute</A> or changes a <A HREF = "neigh_modify.html">neighbor</A> list
parameter), then the initial setup must be performed. LAMMPS does not
check for this, but it would be an error to use the <I>pre no</I> option in
this case.
</P>
<P>If <I>post</I> is specified as "no", the full timing summary is skipped;
only a one-line summary timing is printed.
</P>
<P>The <I>every</I> keyword provides a means of breaking a LAMMPS run into a
series of shorter runs. Optionally, one or more LAMMPS commands (c1,
c2, ..., cN) will be executed in between the short runs. If used, the
<I>every</I> keyword must be the last keyword, since it has a variable
number of arguments. Each of the trailing arguments is a single
LAMMPS command, and each command should be enclosed in quotes, so that
the entire command will be treated as a single argument. This will
also prevent any variables in the command from being evaluated until
it is executed multiple times during the run.
</P>
<P>The <I>every</I> keyword is a means to avoid listing a long series of runs
and interleaving commands in your input script. For example, a
<A HREF = "print.html">print</A> command could be invoked or a <A HREF = "fix.html">fix</A> could
be redefined, e.g. to reset a thermostat temperature. Or this could
be useful for invoking a command you have added to LAMMPS that wraps
some other code (e.g. as a library) to perform a computation
periodically during a long LAMMPS run. See <A HREF = "Section_modify.html">this
section</A> of the documentation for info about how
to add new commands to LAMMPS. See <A HREF = "Section_howto.html#4_10">this
section</A> of the documentation for ideas about
how to couple LAMMPS to other codes.
</P>
<P>With the <I>every</I> option, N total steps are simulated, in shorter runs
of M steps each. After each M-length run, the specified commands are
invoked. If only a single command is specified as NULL, then no
command is invoked. Thus these lines:
</P>
<PRE>variable q equal x[100]
run 6000 every 2000 "print Coord = $q"
</PRE>
<P>are the equivalent of:
</P>
<PRE>variable q equal x[100]
run 2000
print Coord = $q
run 2000
print Coord = $q
run 2000
print Coord = $q
</PRE>
<P>which does 3 runs of 2000 steps and prints the x-coordinate of a
particular atom between runs. Note that the variable "$q" will
be evaluated afresh each time the print command is executed.
</P>
<P>If the <I>pre</I> and <I>post</I> options are set to "no" when used with the
<I>every</I> keyword, then the 1st run will do the full setup and the last
run will print the full timing summary, but these operations will be
skipped for intermediate runs.
</P>
<P><B>Restrictions:</B> none
</P>
<P><B>Related commands:</B>
</P>
<P><A HREF = "minimize.html">minimize</A>, <A HREF = "run_style.html">run_style</A>,
<A HREF = "temper.html">temper</A>
</P>
<P><B>Default:</B>
</P>
<P>The option defaults are start = the current timestep, stop = current
timestep + N, pre = yes, and post = yes.
</P>
</HTML>