forked from lijiext/lammps
334 lines
16 KiB
HTML
334 lines
16 KiB
HTML
<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>variable command
|
|
</H3>
|
|
<P><B>Syntax:</B>
|
|
</P>
|
|
<PRE>variable name style args ...
|
|
</PRE>
|
|
<UL><LI>name = name of variable to define
|
|
|
|
<LI>style = <I>index</I> or <I>loop</I> or <I>world</I> or <I>universe</I> or <I>uloop</I> or <I>equal</I> or <I>atom</I>
|
|
|
|
<PRE> <I>index</I> args = one or more strings
|
|
<I>loop</I> args = N = integer size of loop
|
|
<I>world</I> args = one string for each partition of processors
|
|
<I>universe</I> args = one or more strings
|
|
<I>uloop</I> args = N = integer size of loop
|
|
<I>equal</I> or <I>atom</I> args = one equation containing numbers, thermo keywords, math functions, group functions, atom vectors, compute references, other variables
|
|
numbers = 0.0, -5.4, 2.8e-4, etc
|
|
thermo keywords = vol, ke, press, etc from <A HREF = "thermo_style.html">thermo_style</A>
|
|
math functions = add(x,y), sub(x,y), mult(x,y), div(x,y),
|
|
neg(x), pow(x,y), exp(x), ln(x), sqrt(x)
|
|
group functions = mass(group), charge(group),
|
|
xcm(group,dim), vcm(group,dim), fcm(group,dim),
|
|
bound(group,xmin), gyration(group)
|
|
atom vectors for <I>equal</I> = mass[N], x[N], y[N], z[N],
|
|
vx[N], vy[N], vz[N],
|
|
fx[N], fy[N], fz[N]
|
|
atom vectors for <I>atom</I> = mass[], x[], y[], z[],
|
|
vx[], vy[], vz[],
|
|
fx[], fy[], fz[]
|
|
compute references = c_ID[0], c_ID[N]
|
|
fix references = f_ID[0], f_ID[N]
|
|
other variables = v_abc, v_x, etc
|
|
</PRE>
|
|
|
|
</UL>
|
|
<P><B>Examples:</B>
|
|
</P>
|
|
<PRE>variable x index run1 run2 run3 run4 run5 run6 run7 run8
|
|
variable LoopVar loop $n
|
|
variable beta equal div(temp,3.0)
|
|
variable b1 equal add(x[234],mult(0.5,vol))
|
|
variable b equal div(xcm(mol1,x),2.0)
|
|
variable b equal c_myTemp[0]
|
|
variable b atom div(mult(x<B></B>,y<B></B>),vol)
|
|
variable temp world 300.0 310.0 320.0 $<I>Tfinal</I>
|
|
variable x universe 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
|
|
variable x uloop 15
|
|
</PRE>
|
|
<P><B>Description:</B>
|
|
</P>
|
|
<P>This command assigns one or more strings to a variable name.
|
|
Variables can be used in several ways in LAMMPS. A variable can be
|
|
referenced elsewhere in an input script. For variable styles that
|
|
store multiple strings, the <A HREF = "next.html">next</A> command can be used to
|
|
increment which string is assigned to the variable. Variables can be
|
|
evaluated to produce a numeric value which can be output either
|
|
directly (see the <A HREF = "print.html">print</A>, <A HREF = "fix_print.html">fix print</A>, and
|
|
<A HREF = "run.html">run every</A> commands) or as part of thermodynamic output (see
|
|
the <A HREF = "thermo_style.html">thermo_style</A> command), fix output (see the
|
|
<A HREF = "fix_ave_spatial.html">fix ave/spatial</A> and <A HREF = "compute_variable_atom.html">compute
|
|
variable/atom</A> commands), or dump output
|
|
(see the <A HREF = "dump.html">dump custom</A> and <A HREF = "compute_variable_atom.html">compute
|
|
variable/atom</A> commands).
|
|
</P>
|
|
<P>In the discussion that follows, the "name" of the variable is the
|
|
arbitrary string that is the 1st argument in the variable command.
|
|
The "string" is one or more of the subsequent arguments. The "string"
|
|
can be simple text as in the 1st example above, it can contain other
|
|
variables as in the 2nd example, or it can be an equation as in the
|
|
3rd example. The "value" is the numeric quantity resulting from
|
|
evaluation of the string. Note that the same string can generate
|
|
different values when it is evaluated at different times during a
|
|
simulation.
|
|
</P>
|
|
<P>IMPORTANT NOTE: When a variable command is encountered in the input
|
|
script and the variable name has already been specified, the command
|
|
is ignored. This means variables can NOT be re-defined in an input
|
|
script (with 2 exceptions, read further). This is to allow an input
|
|
script to be processed multiple times without resetting the variables;
|
|
see the <A HREF = "jump.html">jump</A> or <A HREF = "include.html">include</A> commands. It also
|
|
means that using a command-line switch -var will override a
|
|
corresponding variable setting in the input script.
|
|
</P>
|
|
<P>There are two exceptions to this rule. First, variables of style
|
|
<I>equal</I> ARE redefined each time the command is encountered. This
|
|
allow them to be reset, e.g. in a loop. This means an <I>equal</I>-style
|
|
variable will re-define a command-line swich -var setting, so use an
|
|
<I>index</I>-style variable for these settings instead, as in bench/in.lj.
|
|
</P>
|
|
<P>Second, as described below, if a variable is iterated on to the end of
|
|
its list of strings via the <A HREF = "next.html">next</A> command, it is removed
|
|
from the list of active varaibles, and is thus available to be
|
|
re-defined in a subsequent variable command.
|
|
</P>
|
|
<HR>
|
|
|
|
<P><A HREF = "Section_commands.html#3_2">This section</A> of the manual explains how
|
|
occurrences of a variable name in an input script line are replaced by
|
|
the variable's string. The variable name can be referenced as $x if
|
|
the name "x" is a single character, or as ${LoopVar} if the name
|
|
"LoopVar" is one or more characters.
|
|
</P>
|
|
<P>As described below, for variable styles <I>index</I>, <I>loop</I>, <I>universe</I>,
|
|
and <I>uloop</I>, which string is assigned to a variable can be incremented
|
|
via the <A HREF = "next.html">next</A> command. When there are no more strings to
|
|
assign, the variable is "exhausted" and a flag is set that causes the
|
|
next <A HREF = "jump.html">jump</A> command encountered in the input script to be
|
|
skipped. This enables the construction of simple loops in the input
|
|
script that are iterated over and exited from.
|
|
</P>
|
|
<HR>
|
|
|
|
<P>For the <I>index</I> style, one or more strings are specified. Initially,
|
|
the 1st string is assigned to the variable. Each time a
|
|
<A HREF = "next.html">next</A> command is used with the variable name, the next
|
|
string is assigned. All processors assign the same string to the
|
|
variable.
|
|
</P>
|
|
<P><I>Index</I> style variables with a single string value can also be set by
|
|
using the command-line switch -var; see <A HREF = "Section_start.html#2_6">this
|
|
section</A> for details.
|
|
</P>
|
|
<P>The <I>loop</I> style is identical to the <I>index</I> style except that the
|
|
strings are the integers from 1 to N. This allows generation of a
|
|
long list of runs (e.g. 1000) without having to list N strings in the
|
|
input script. Initially, the string "1" is assigned to the variable.
|
|
Each time a <A HREF = "next.html">next</A> command is used with the variable name,
|
|
the next string ("2", "3", etc) is assigned. All processors assign
|
|
the same string to the variable.
|
|
</P>
|
|
<P>For the <I>world</I> style, one or more strings are specified. There must
|
|
be one string for each processor partition or "world". See <A HREF = "Section_start.html#2_6">this
|
|
section</A> of the manual for information on
|
|
running LAMMPS with multiple partitions via the "-partition"
|
|
command-line switch. This variable command assigns one string to each
|
|
world. All processors in the world are assigned the same string. The
|
|
next command cannot be used with <I>equal</I> style variables, since there
|
|
is only one value per world. This style of variable is useful when
|
|
you wish to run different simulations on different partitions, or when
|
|
performing a parallel tempering simulation (see the
|
|
<A HREF = "temper.html">temper</A> command), to assign different temperatures to
|
|
different partitions.
|
|
</P>
|
|
<P>For the <I>universe</I> style, one or more strings are specified. There
|
|
must be at least as many strings as there are processor partitions or
|
|
"worlds". See <A HREF = "Section_start.html#2_6">this page</A> for information on
|
|
running LAMMPS with multiple partitions via the "-partition"
|
|
command-line switch. This variable command initially assigns one
|
|
string to each world. When a <A HREF = "next.html">next</A> command is encountered
|
|
using this variable, the first processor partition to encounter it, is
|
|
assigned the next available string. This continues until all the
|
|
variable strings are consumed. Thus, this command can be used to run
|
|
50 simulations on 8 processor partitions. The simulations will be run
|
|
one after the other on whatever partition becomes available, until
|
|
they are all finished. <I>Universe</I> 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.
|
|
</P>
|
|
<P>The <I>uloop</I> style is identical to the <I>universe</I> style except that the
|
|
strings are the integers from 1 to N. This allows generation of long
|
|
list of runs (e.g. 1000) without having to list N strings in the input
|
|
script.
|
|
</P>
|
|
<HR>
|
|
|
|
<P>For the <I>equal</I> and <I>atom</I> styles, a single string is specified which
|
|
represents an equation that will be evaluated afresh each time the
|
|
variable is used. For <I>equal</I> style variables this equation computes
|
|
a scalar quantity, which becomes the value of the variable whenever it
|
|
is evaluated. For <I>atom</I> style variables it computes a quantity for
|
|
each atom, which is used by the "compute variable/atom" command and
|
|
its associated output.
|
|
</P>
|
|
<P>Note that <I>equal</I> and <I>atom</I> variables can produce different values at
|
|
different stages of the input script or at different times during a
|
|
run. For example, if an <I>equal</I> variable is used in a <A HREF = "fix_print.html">fix
|
|
print</A> command, different values could be printed each
|
|
timestep it was invoked.
|
|
</P>
|
|
<P>The next command cannot be used with <I>equal</I> or <I>atom</I> style
|
|
variables, since there is only one string.
|
|
</P>
|
|
<P>The equation for an <I>equal</I> or <I>atom</I> variable can contain a variety
|
|
of quantities. The syntax for each kind of quantity is simple, but
|
|
multiple quantities can be nested and combined in various ways to
|
|
build up formulas of arbitrary complexity. For example, this is a
|
|
valid (though strange) variable equation:
|
|
</P>
|
|
<PRE>variable x equal div(add(pe,c_MyTemp[0]),pow(vol,div(1,3)))
|
|
</PRE>
|
|
<P>Specifically, an equation can contain numbers, thermo keywords, math
|
|
functions, group functions, atom vectors, compute references, and
|
|
other variables. There is one difference between <I>equal</I> and <I>atom</I>
|
|
variables; the syntax of Atom vector references is different.
|
|
</P>
|
|
<DIV ALIGN=center><TABLE BORDER=1 >
|
|
<TR><TD >Number</TD><TD > 0.2, 1.0e20, -15.4, etc</TD></TR>
|
|
<TR><TD >Thermo keywords</TD><TD > vol, pe, ebond, etc</TD></TR>
|
|
<TR><TD >Math functions</TD><TD > add(x,y), sub(x,y), mult(x,y), div(x,y), neg(x), pow(x,y), exp(x), ln(x), sqrt(x)</TD></TR>
|
|
<TR><TD >Group functions</TD><TD > mass(ID), charge(ID), xcm(ID,dim), vcm(ID,dim), fcm(ID,dim) bound(ID,dir), gyration(ID)</TD></TR>
|
|
<TR><TD >Atom vectors for <I>equal</I></TD><TD > mass[N], x[N], y[N], z[N], vx[N], vy[N], vz[N], fx[N], fy[N], fz[N]</TD></TR>
|
|
<TR><TD >Atom vectors for <I>atom</I></TD><TD > mass[], x[], y[], z[], vx[], vy[], vz[], fx[], fy[], fz[]</TD></TR>
|
|
<TR><TD >Compute references</TD><TD > c_ID[0], c_ID[N]</TD></TR>
|
|
<TR><TD >Fix references</TD><TD > f_ID[0], f_ID[N]</TD></TR>
|
|
<TR><TD >Other variables</TD><TD > v_abc, v_x, etc
|
|
</TD></TR></TABLE></DIV>
|
|
|
|
<P>The thermo keywords allowed in the equation are those defined by the
|
|
"thermo_style custom" command. Note that many thermodyanmic
|
|
quantities are only computable after the first simulation has begun.
|
|
Likewise, many thermodynamic quantities (such as energies) are only
|
|
computed on timesteps when thermodyanmic output is being performed.
|
|
If the variable equation these quantities at other times, out-of-date
|
|
or invalid values may be used.
|
|
</P>
|
|
<P>Math functions take one or two arguments, each of which may be an
|
|
equation containing any of the quantities defined above. This allows
|
|
equations to be nested, as in the examples above.
|
|
</P>
|
|
<P>Group functions take one or two arguments. The first argument is the
|
|
group-ID. The <I>dim</I> argument is <I>x</I> or <I>y</I> or <I>z</I>. The <I>dir</I>
|
|
argument is <I>xmin</I>, <I>xmax</I>, <I>ymin</I>, <I>ymax</I>, <I>zmin</I>, or <I>zmax</I>. The
|
|
group functions mass() and charge() are the total mass and charge of
|
|
the group of atoms. Xcm() and vcm() return components of the position
|
|
and velocity of the center of mass of the group. Fcm() returns a
|
|
component of the total force on the group of atoms. Bound() returns
|
|
the min/max of a particular coordinate for all atoms in the group.
|
|
Gyration() computes the radius-of-gyration of the group of atoms. See
|
|
the <A HREF = "fix_gyration.html">fix gyration</A> command for the formula.
|
|
</P>
|
|
<P>For <I>equal</I> style variables, atom vectors take a single integer
|
|
argument from 1-N, which is the desired atom-ID, e.g. x[243]. For
|
|
<I>atom</I> style variables, atom vectors take no argument. Since <I>atom</I>
|
|
style variables compute one value per atom, a reference like x[]
|
|
means the x-coord of each atom will be used when evaluating the
|
|
variable.
|
|
</P>
|
|
<P>Compute references access scalar or vector quantities calculated by a
|
|
<A HREF = "compute.html">compute</A>. The ID in the reference should be replaced by
|
|
the actual ID of the compute defined elsewhere in the input script.
|
|
See the <A HREF = "compute.html">compute</A> command for details. Note that
|
|
per-atom quantities calculated by a compute cannot be accessed this
|
|
way, but only global scalar or vector quantities. However, there is a
|
|
<A HREF = "compute_sum.html">compute sum</A> command which sums per-atom quantities
|
|
into a global scalar or vector which can be accessed by a variable.
|
|
</P>
|
|
<P>If <I>c_ID[0]</I> is used as a keyword, then the scalar quantity
|
|
calculated by the compute is used. If <I>c_ID[N]</I> is used, then one
|
|
component of the vector quantity calculated by the compute is used. N
|
|
should be an integer from 1-M, where M is the length of the vector
|
|
calculated by the compute.
|
|
</P>
|
|
<P>Fix references access scalar or vector quantities calculated by a
|
|
<A HREF = "fix.html">fix</A>. See the doc pages for individual fixes to see which
|
|
ones compute a scalar or vector quantity. Since the fix may not be
|
|
invoked every timestep and compute its quantities, you should insure
|
|
the variable is only evaluated on appropriate timesteps. The ID in
|
|
the reference should be replaced by the actual ID of the fix defined
|
|
elsewhere in the input script. See the <A HREF = "fix.html">fix</A> command for
|
|
details. Note that per-atom quantities calculated by a fix cannot be
|
|
accessed this way, but only global scalar or vector quantities.
|
|
</P>
|
|
<P>If <I>f_ID[0]</I> is used as a keyword, then the scalar quantity
|
|
calculated by the fix is used. If <I>f_ID[N]</I> is used, then one
|
|
component of the vector quantity calculated by the fix is used. N
|
|
should be an integer from 1-M, where M is the length of the vector
|
|
calculated by the fix.
|
|
</P>
|
|
<P>The current values of other variables can be accessed by prepending a
|
|
"v_" to the variable name. This will cause the other variable to be
|
|
evaulated. Note that if you do something circular like this:
|
|
</P>
|
|
<PRE>variable a equal v_b
|
|
variable b equal v_a
|
|
print $a
|
|
</PRE>
|
|
<P>then LAMMPS will run for a while when the print statement is invoked.
|
|
</P>
|
|
<P>Note that there is a subtle difference between using a variable in a
|
|
<I>equal</I> or <I>atom</I> style equation in the form $x versus v_x.
|
|
</P>
|
|
<P>In the former case, as with any other input script command, the
|
|
variable's value is substituted for immediately when the line is read
|
|
from the input script. Thus if the current simulation box volume was
|
|
1000.0, then these lines:
|
|
</P>
|
|
<PRE>variable x equal vol
|
|
variable y equal mult($x,2)
|
|
</PRE>
|
|
<P>would associate the equation string "mult(1000.0,2)" with variable y.
|
|
</P>
|
|
<P>By contrast, these lines:
|
|
</P>
|
|
<PRE>variable x equal vol
|
|
variable y equal mult(v_x,2)
|
|
</PRE>
|
|
<P>would associate the equation string "mult(v_x,2)" with variable y.
|
|
</P>
|
|
<P>Thus if the variable y were evaluated periodically during a run where
|
|
the box volume changed, the resulting value would always be 2000.0 for
|
|
the first case, but would change dynamically for the second case.
|
|
</P>
|
|
<HR>
|
|
|
|
<P><B>Restrictions:</B>
|
|
</P>
|
|
<P>The use of atom vectors in <I>equal</I> style variables requires the atom
|
|
style to use a global mapping in order to look up the vector indices.
|
|
Only atom styles with molecular information create global maps unless
|
|
the <A HREF = "atom_modify.html">atom_modify map</A> command is used.
|
|
</P>
|
|
<P>All <I>universe</I>- and <I>uloop</I>-style variables must have the same number
|
|
of values.
|
|
</P>
|
|
<P><B>Related commands:</B>
|
|
</P>
|
|
<P><A HREF = "next.html">next</A>, <A HREF = "jump.html">jump</A>, <A HREF = "include.html">include</A>,
|
|
<A HREF = "temper.html">temper</A>, <A HREF = "fix_print.html">fix print</A>, <A HREF = "print.html">print</A>
|
|
</P>
|
|
<P><B>Default:</B> none
|
|
</P>
|
|
</HTML>
|