From 721cd1679857397a48c5d0fb42280faf513668ce Mon Sep 17 00:00:00 2001 From: sjplimp Date: Thu, 18 Oct 2007 22:27:53 +0000 Subject: [PATCH] git-svn-id: svn://svn.icms.temple.edu/lammps-ro/trunk@1061 f3b2605a-c512-4ea7-a41b-209d697bcdaa --- doc/fix_ave_spatial.html | 191 ++++++++++++++++++++++++--------------- doc/fix_ave_spatial.txt | 189 +++++++++++++++++++++++--------------- doc/fix_ave_time.html | 19 ++-- doc/fix_ave_time.txt | 19 ++-- 4 files changed, 253 insertions(+), 165 deletions(-) diff --git a/doc/fix_ave_spatial.html b/doc/fix_ave_spatial.html index 62569cd0eb..7e514e1a6f 100644 --- a/doc/fix_ave_spatial.html +++ b/doc/fix_ave_spatial.html @@ -31,8 +31,6 @@
  • delta = thickness of spatial layers in dim (distance units) -
  • file = filename to write results to (NULL = no file) -
  • style = density or compute or fix 
      density arg = mass or number
@@ -43,25 +41,32 @@
  • zero or more keyword/value pairs may be appended -
    keyword = norm or units
+
  • keyword = norm or units or file or ave 
+
+
      units value = box or lattice or reduced
   norm value = all or sample
-  units value = box or lattice or reduced 
+  file arg = filename
+    filename = file to write results to
+  ave args = one or running or window M
+    one = output new average value every Nfreq steps
+    running = output cummulative average of all previous Nfreq steps
+    window M = output average of M most recent Nfreq steps 
 
    
 
 
 
    Examples:
 
    
-
    fix 1 all ave/spatial 10000 1 10000 z lower 2.0 centro.profile compute myCentro
-fix 1 flow ave/spatial 100 10 1000 y 0.0 1.0 vel.profile compute Vx norm sample
-fix 1 flow ave/spatial 100 5 1000 y 0.0 2.5 dens.profile density mass 
+
    fix 1 all ave/spatial 10000 1 10000 z lower 0.02 compute myCentro units reduced
+fix 1 flow ave/spatial 100 10 1000 y 0.0 1.0 compute Vx norm sample file vel.profile
+fix 1 flow ave/spatial 100 5 1000 y 0.0 2.5 density mass ave running 
 
    
 
    Description:
 
    
 
    Calculate one or more instantaneous per-atom quantities every few
 timesteps, average them by layer in a chosen dimension and over a
-longer timescale.  The resulting averages can be written to a file
-and/or used by other output commands such as
-thermo_style custom.
+longer timescale.  The resulting averages can be used by other output
+commands such as thermo_style
+custom, and also written to a file.
 
    
 
    This fix can be used to spatially average per-atom properties
 (velocity, force) or per-atom quantities calculated by a
@@ -70,14 +75,14 @@ and/or used by other output commands suc
 equation you define (see the compute
 variable/atom command).
 
    
-
    The density styles means to simply count the number of atoms in each
-layer, either by mass or number.  The compute style allows
-specification of a compute which will be invoked to
-calculate the desired property.  The compute can be previously defined
-in the input script or it can be a compute defined by a dump
-custom command.
+
    For style density, the number of atoms in each layer is counted,
+either by mass or by number.  The compute style allows specification
+of a compute which will be invoked to calculate the
+desired property.  The compute can be previously defined in the input
+script or it can be a compute defined by a dump custom
+command.
 
    
-
    For the compute style, the fix ave/spatial command accesses the
+
    For style compute, the fix ave/spatial command accesses the
 per-atom scalar or vector values stored by the compute.  Thus it must
 be a "per-atom" compute with the word "atom" in its style name, rather
 than a "global" compute.  See the fix ave/time
@@ -95,28 +100,41 @@ per-atom compute styles and add them to LAMMPSdump custom command can also be used to
 directly output per-atom quantities calculated by a per-atom compute.
 
    
-
    For the fix style, the fix ave/spatial command accesses the per-atom
+
    For style fix, the fix ave/spatial command accesses the per-atom
 scalar or vector values stored by another fix.  The fix
 ave/atom command is an example of such a fix.
 
    
-
    In all cases, the calculated property is averaged over atoms in each
-layer, where the layers are in a particular dim and have a thickness
-given by delta.  Every Nfreq steps, when a property is calculated
-for the first time (after a previous write), the number of layers and
-the layer boundaries are computed.  Thus if the simlation box changes
-size during a simulation, the number of layers and their boundaries
-may also change.  Layers are defined relative to a specified origin,
-which may be the lower/upper edge of the box (in dim) or its center
-point, or a specified coordinate value.  Starting at the origin,
-sufficient layers are created in both directions to completely cover
-the box.  On subsequent timesteps every atom is mapped to one of the
-layers.  Atoms beyond the lowermost/uppermost layer are counted in the
-first/last layer.
+
    The Nevery, Nrepeat, and Nfreq arguments specify on what
+timesteps the per-atom property will be evaluated in order to
+contribute to the average.  The final averaged values are computed
+every Nfreq timesteps.  The average is over Nrepeat values,
+computed in the preceeding portion of the simulation every Nevery
+timesteps.  Nfreq must be a multiple of Nevery and Nevery must
+be non-zero even if Nrepeat is 1.
+
    
+
    For example, if Nevery=2, Nrepeat=6, and Nfreq=100, then values on
+timesteps 90,92,94,96,98,100 will be used to compute the final average
+on timestep 100.  Similary for timesteps 190,192,194,196,198,200 on
+timestep 200, etc.
+
    
+
    The per-atom property is also averaged over atoms in each layer, where
+the layers are in a particular dim and have a thickness given by
+delta.  Every Nfreq steps, when an averaging is being performed and
+the per-atom property is calculated for the first time, the number of
+layers and the layer boundaries are computed.  Thus if the simlation
+box changes size during a simulation, the number of layers and their
+boundaries may also change.  Layers are defined relative to a
+specified origin, which may be the lower/upper edge of the box (in
+dim) or its center point, or a specified coordinate value.  Starting
+at the origin, sufficient layers are created in both directions to
+completely cover the box.  On subsequent timesteps every atom is
+mapped to one of the layers.  Atoms beyond the lowermost/uppermost
+layer are counted in the first/last layer.
 
    
 
    For orthogonal simulation boxes, the layers are "slices" aligned with
 the xyz coordinate axes.  For non-orthogonal (triclinic) simulation
-boxes, the layers are "tilted slices" that align with the tilted faces
-of the box.  See the region prism command for a
+boxes, the layers are "tilted slices" that are parallel to the tilted
+faces of the box.  See the region prism command for a
 discussion of the geometry of tilted boxes in LAMMPS.  As described
 there, a tilted simulation box has edge vectors a,b,c.  In that
 nomenclature, layers in the x dimension have faces with normals in the
@@ -124,12 +142,26 @@ nomenclature, layers in the x dimension have faces with normals in the
 cross "c" direction.  And layers in z have faces normal to the "a"
 cross "b" direction.  Note that in order to define the thickness and
 position of these tilted layers in an unambiguous fashion, the units
-option must be set to reduced.
+option must be set to reduced when using a non-orthogonal simulation
+box, as discussed below.
+
    
+
    For the compute and fix keywords, the per-atom calculation
+performed by the compute or fix is on the group defined by that
+command.  However, only atoms in the fix group are included in the
+layer averaging.
+
    
+
    Note that some computes perform costly calculations, involving the
+creation or use of neighbor lists.  If the compute is invoked too
+often by fix ave/spatial, it can slow down a simulation.
+
    
+
    
+
+
    Additional optional keywords also affect the operation of this fix.
 
    
 
    The units keyword determines the meaning of the distance units used
 for the layer thickness delta and for origin if it is a coordinate
 value.  For orthogonal simulation boxes, any of the 3 options may be
-used.  For non-orthongal (triclinic) simulation boxes, only the
+used.  For non-orthogonal (triclinic) simulation boxes, only the
 reduced option may be used.
 
    
 
    A box value selects standard distance units as defined by the
@@ -149,52 +181,64 @@ the lower "b" cross "c" plane of the simulation box and an origin of
 A delta value of 0.1 means there will be 10 layers from 0.0 to 1.0,
 regardless of the current size or shape of the simulation box.
 
    
-
    The Nevery, Nrepeat, and Nfreq arguments specify how the
-property will be time-averaged.  The final averaged value(s) are
-computed every Nfreq timesteps.  The average is over Nrepeat
-values, computed in the preceeding portion of the simulation every
-Nevery timesteps.  Thus if Nevery=2, Nrepeat=6, and Nfreq=100, then
-values on timesteps 90,92,94,96,98,100 will be used to compute the
-final average on timestep 100.  Similary for timesteps
-190,192,194,196,198,200 on timestep 200, etc.
-
    
-
    The norm keyword also affects how time-averaging is done.  For an
-all setting, a layer quantity is summed over all atoms in all
-Nfreq/Nevery samples, as is the count of atoms in the layer.  The
-printed value for the layer is Total-quantity / Total-count.
-In other words it is an average over the entire Nfreq timescale.
+
    The norm keyword affects how time-averaging is done within for the
+output produced every Nfreq timesteps.  For an all setting, a
+layer quantity is summed over all atoms in all Nrepeat samples, as
+is the count of atoms in the layer.  The printed value for the layer
+is Total-quantity / Total-count.  In other words it is an average over
+the entire Nfreq timescale.
 
    
 
    For a sample setting, the quantity is summed over atoms for only a
 single sample, as is the count, and a "average sample value" is
 computed, i.e. Sample-quantity / Sample-count.  The printed value for
-the layer is the average of the M "average sample values", where M =
-Nfreq/Nevery.  In other words it is an average of an average.
+the layer is the average of the Nrepeat "average sample values", In
+other words it is an average of an average.
 
    
-
    If file output is specified, each time info is written to the file, it
-is in the following format.  A line with the timestep and number of
-layers is written.  Then one line per layer is written, containing the
-layer ID (1-N), the coordinate of the center of the layer, the number
-of atoms in the layer, and one or more calculated values.  The number
-of atoms and the value(s) are average quantities.  If the value of the
-units keyword is box or lattice, the "coord" is printed in box
-units.  If the value of the units keyword is reduced, the "coord"
-is printed in reduced units (0-1).
+
    The file keyword allows a filename to be specified.  Every Nfreq
+timesteps, layer info will be written to a text file in the following
+format.  A line with the timestep and number of layers is written.
+Then one line per layer is written, containing the layer ID (1-N), the
+coordinate of the center of the layer, the number of atoms in the
+layer, and one or more calculated values.  The number of atoms and the
+value(s) are average quantities.  If the value of the units keyword
+is box or lattice, the "coord" is printed in box units.  If the
+value of the units keyword is reduced, the "coord" is printed in
+reduced units (0-1).
 
    
-
    If the density keyword is used, or the compute or fix keyword
-with a compute/fix that calculates a single quantity per atom, then a
-single value will be printed for each layer.  If the compute or
-fix keyword is used with a compute/fix that calculates N quantities
-per atom, then N values per line will be written, each of them
-averaged independently.
+
    If the style is density, or the style is compute or fix with a
+compute/fix that calculates a single quantity per atom, then a single
+value will be printed for each layer.  If the style is compute or
+fix with a compute/fix that calculates N quantities per atom, then N
+values per line will be written, each of them averaged independently.
 
    
-
    For the compute and fix keywords, the calculation performed by the
-compute or fix is on the group defined by the that command.  However,
-only atoms in the fix group are included in the layer averaging.
+
    The ave keyword determines how the layer values produced every
+Nfreq steps are averaged with layer values produced on previous
+steps that were multiples of Nfreq, before they are accessed by
+another output command or written to a file.
 
    
-
    Note that some computes perform costly calculations, involving use of
-or creation of neighbor lists.  If the compute is invoked too often by
-fix ave/spatial, it can slow down a simulation.
+
    If the ave setting is one, then the layuer values produced on
+timesteps that are multiples of Nfreq are independent of each other;
+they are output as-is without further averaging.
 
    
+
    If the ave setting is running, then the layer values produced on
+timesteps that are multiples of Nfreq are summed and averaged in a
+cummulative sense before being output.  Each output layer value is
+thus the average of the layer value produced on that timestep with all
+preceeding values for the same layer.  This running average begins
+when the fix is defined; it can only be restarted by deleting the fix
+via the unfix command, or re-defining the fix by
+re-specifying it.
+
    
+
    If the ave setting is window, then the layer values produced on
+timesteps that are multiples of Nfreq are summed and averaged within
+a moving "window" of time, so that the last M values for the same
+layer are used to produce the output.  E.g. if M = 3 and Nfreq = 1000,
+then the output on step 10000 will be the average of the individual
+layer values on steps 8000,9000,10000.  Outputs on early steps will
+average over less than M values if they are not available.
+
    
+
    
+
 
    Restart, fix_modify, output, run start/stop, minimize info:
 
    
 
    No information about this fix is written to binary restart
@@ -224,6 +268,7 @@ minimization.
 
    
 
    Default:
 
    
-
    The option defaults are norm = all and units = lattice.
+
    The option defaults are units = lattice, norm = all, no file output,
+and ave = one.
 
    
 
diff --git a/doc/fix_ave_spatial.txt b/doc/fix_ave_spatial.txt
index 8ce58792d8..a84b9c9245 100644
--- a/doc/fix_ave_spatial.txt
+++ b/doc/fix_ave_spatial.txt
@@ -20,7 +20,6 @@ Nfreq = timestep frequency at which the average value is computed :l
 dim = {x} or {y} or {z} :l
 origin = {lower} or {center} or {upper} or coordinate value (distance units) :l
 delta = thickness of spatial layers in dim (distance units) :l
-file = filename to write results to (NULL = no file) :l
 style = {density} or {compute} or {fix} :l
   {density} arg = {mass} or {number}
     {mass} = compute mass density
@@ -28,24 +27,30 @@ style = {density} or {compute} or {fix} :l
   {compute} arg = compute-ID that stores or calculates per-atom quantities
   {fix} arg = fix-ID that stores or calculates per-atom quantities :pre
 zero or more keyword/value pairs may be appended :l
-keyword = {norm} or {units}
+keyword = {norm} or {units} or {file} or {ave} :l
+  {units} value = {box} or {lattice} or {reduced}
   {norm} value = {all} or {sample}
-  {units} value = {box} or {lattice} or {reduced} :pre
+  {file} arg = filename
+    filename = file to write results to
+  {ave} args = {one} or {running} or {window M}
+    one = output new average value every Nfreq steps
+    running = output cummulative average of all previous Nfreq steps
+    window M = output average of M most recent Nfreq steps :pre
 :ule
 
 [Examples:]
 
-fix 1 all ave/spatial 10000 1 10000 z lower 2.0 centro.profile compute myCentro
-fix 1 flow ave/spatial 100 10 1000 y 0.0 1.0 vel.profile compute Vx norm sample
-fix 1 flow ave/spatial 100 5 1000 y 0.0 2.5 dens.profile density mass :pre
+fix 1 all ave/spatial 10000 1 10000 z lower 0.02 compute myCentro units reduced
+fix 1 flow ave/spatial 100 10 1000 y 0.0 1.0 compute Vx norm sample file vel.profile
+fix 1 flow ave/spatial 100 5 1000 y 0.0 2.5 density mass ave running :pre
 
 [Description:]
 
 Calculate one or more instantaneous per-atom quantities every few
 timesteps, average them by layer in a chosen dimension and over a
-longer timescale.  The resulting averages can be written to a file
-and/or used by other "output commands"_Section_howto.html#4_15 such as
-"thermo_style custom"_thermo_style.html.
+longer timescale.  The resulting averages can be used by other "output
+commands"_Section_howto.html#4_15 such as "thermo_style
+custom"_thermo_style.html, and also written to a file.
 
 This fix can be used to spatially average per-atom properties
 (velocity, force) or per-atom quantities calculated by a
@@ -54,14 +59,14 @@ This fix can be used to spatially average per-atom properties
 equation you define (see the "compute
 variable/atom"_compute_variable_atom.html command).
 
-The {density} styles means to simply count the number of atoms in each
-layer, either by mass or number.  The {compute} style allows
-specification of a "compute"_compute.html which will be invoked to
-calculate the desired property.  The compute can be previously defined
-in the input script or it can be a compute defined by a "dump
-custom"_dump.html command.
+For style {density}, the number of atoms in each layer is counted,
+either by mass or by number.  The {compute} style allows specification
+of a "compute"_compute.html which will be invoked to calculate the
+desired property.  The compute can be previously defined in the input
+script or it can be a compute defined by a "dump custom"_dump.html
+command.
 
-For the {compute} style, the fix ave/spatial command accesses the
+For style {compute}, the fix ave/spatial command accesses the
 per-atom scalar or vector values stored by the compute.  Thus it must
 be a "per-atom" compute with the word "atom" in its style name, rather
 than a "global" compute.  See the "fix ave/time"_fix_ave_time.html
@@ -79,28 +84,41 @@ per-atom compute styles and "add them to LAMMPS"_Section_modify.html.
 Note that the "dump custom"_dump.html command can also be used to
 directly output per-atom quantities calculated by a per-atom compute.
 
-For the {fix} style, the fix ave/spatial command accesses the per-atom
+For style {fix}, the fix ave/spatial command accesses the per-atom
 scalar or vector values stored by another fix.  The "fix
 ave/atom"_fix_ave_atom.html command is an example of such a fix.
 
-In all cases, the calculated property is averaged over atoms in each
-layer, where the layers are in a particular {dim} and have a thickness
-given by {delta}.  Every Nfreq steps, when a property is calculated
-for the first time (after a previous write), the number of layers and
-the layer boundaries are computed.  Thus if the simlation box changes
-size during a simulation, the number of layers and their boundaries
-may also change.  Layers are defined relative to a specified {origin},
-which may be the lower/upper edge of the box (in {dim}) or its center
-point, or a specified coordinate value.  Starting at the origin,
-sufficient layers are created in both directions to completely cover
-the box.  On subsequent timesteps every atom is mapped to one of the
-layers.  Atoms beyond the lowermost/uppermost layer are counted in the
-first/last layer.
+The {Nevery}, {Nrepeat}, and {Nfreq} arguments specify on what
+timesteps the per-atom property will be evaluated in order to
+contribute to the average.  The final averaged values are computed
+every {Nfreq} timesteps.  The average is over {Nrepeat} values,
+computed in the preceeding portion of the simulation every {Nevery}
+timesteps.  {Nfreq} must be a multiple of {Nevery} and {Nevery} must
+be non-zero even if {Nrepeat} is 1.
+
+For example, if Nevery=2, Nrepeat=6, and Nfreq=100, then values on
+timesteps 90,92,94,96,98,100 will be used to compute the final average
+on timestep 100.  Similary for timesteps 190,192,194,196,198,200 on
+timestep 200, etc.
+
+The per-atom property is also averaged over atoms in each layer, where
+the layers are in a particular {dim} and have a thickness given by
+{delta}.  Every Nfreq steps, when an averaging is being performed and
+the per-atom property is calculated for the first time, the number of
+layers and the layer boundaries are computed.  Thus if the simlation
+box changes size during a simulation, the number of layers and their
+boundaries may also change.  Layers are defined relative to a
+specified {origin}, which may be the lower/upper edge of the box (in
+{dim}) or its center point, or a specified coordinate value.  Starting
+at the origin, sufficient layers are created in both directions to
+completely cover the box.  On subsequent timesteps every atom is
+mapped to one of the layers.  Atoms beyond the lowermost/uppermost
+layer are counted in the first/last layer.
 
 For orthogonal simulation boxes, the layers are "slices" aligned with
 the xyz coordinate axes.  For non-orthogonal (triclinic) simulation
-boxes, the layers are "tilted slices" that align with the tilted faces
-of the box.  See the "region prism"_region.html command for a
+boxes, the layers are "tilted slices" that are parallel to the tilted
+faces of the box.  See the "region prism"_region.html command for a
 discussion of the geometry of tilted boxes in LAMMPS.  As described
 there, a tilted simulation box has edge vectors a,b,c.  In that
 nomenclature, layers in the x dimension have faces with normals in the
@@ -108,12 +126,26 @@ nomenclature, layers in the x dimension have faces with normals in the
 cross "c" direction.  And layers in z have faces normal to the "a"
 cross "b" direction.  Note that in order to define the thickness and
 position of these tilted layers in an unambiguous fashion, the {units}
-option must be set to {reduced}.
+option must be set to {reduced} when using a non-orthogonal simulation
+box, as discussed below.
+
+For the {compute} and {fix} keywords, the per-atom calculation
+performed by the compute or fix is on the group defined by that
+command.  However, only atoms in the fix group are included in the
+layer averaging.
+
+Note that some computes perform costly calculations, involving the
+creation or use of neighbor lists.  If the compute is invoked too
+often by fix ave/spatial, it can slow down a simulation.
+
+:line
+
+Additional optional keywords also affect the operation of this fix.
 
 The {units} keyword determines the meaning of the distance units used
 for the layer thickness {delta} and for {origin} if it is a coordinate
 value.  For orthogonal simulation boxes, any of the 3 options may be
-used.  For non-orthongal (triclinic) simulation boxes, only the
+used.  For non-orthogonal (triclinic) simulation boxes, only the
 {reduced} option may be used.
 
 A {box} value selects standard distance units as defined by the
@@ -133,51 +165,63 @@ the lower "b" cross "c" plane of the simulation box and an {origin} of
 A {delta} value of 0.1 means there will be 10 layers from 0.0 to 1.0,
 regardless of the current size or shape of the simulation box.
 
-The {Nevery}, {Nrepeat}, and {Nfreq} arguments specify how the
-property will be time-averaged.  The final averaged value(s) are
-computed every {Nfreq} timesteps.  The average is over {Nrepeat}
-values, computed in the preceeding portion of the simulation every
-{Nevery} timesteps.  Thus if Nevery=2, Nrepeat=6, and Nfreq=100, then
-values on timesteps 90,92,94,96,98,100 will be used to compute the
-final average on timestep 100.  Similary for timesteps
-190,192,194,196,198,200 on timestep 200, etc.
-
-The {norm} keyword also affects how time-averaging is done.  For an
-{all} setting, a layer quantity is summed over all atoms in all
-Nfreq/Nevery samples, as is the count of atoms in the layer.  The
-printed value for the layer is Total-quantity / Total-count.
-In other words it is an average over the entire Nfreq timescale.
+The {norm} keyword affects how time-averaging is done within for the
+output produced every {Nfreq} timesteps.  For an {all} setting, a
+layer quantity is summed over all atoms in all {Nrepeat} samples, as
+is the count of atoms in the layer.  The printed value for the layer
+is Total-quantity / Total-count.  In other words it is an average over
+the entire {Nfreq} timescale.
 
 For a {sample} setting, the quantity is summed over atoms for only a
 single sample, as is the count, and a "average sample value" is
 computed, i.e. Sample-quantity / Sample-count.  The printed value for
-the layer is the average of the M "average sample values", where M =
-Nfreq/Nevery.  In other words it is an average of an average.
+the layer is the average of the {Nrepeat} "average sample values", In
+other words it is an average of an average.
 
-If file output is specified, each time info is written to the file, it
-is in the following format.  A line with the timestep and number of
-layers is written.  Then one line per layer is written, containing the
-layer ID (1-N), the coordinate of the center of the layer, the number
-of atoms in the layer, and one or more calculated values.  The number
-of atoms and the value(s) are average quantities.  If the value of the
-{units} keyword is {box} or {lattice}, the "coord" is printed in box
-units.  If the value of the {units} keyword is {reduced}, the "coord"
-is printed in reduced units (0-1).
+The {file} keyword allows a filename to be specified.  Every {Nfreq}
+timesteps, layer info will be written to a text file in the following
+format.  A line with the timestep and number of layers is written.
+Then one line per layer is written, containing the layer ID (1-N), the
+coordinate of the center of the layer, the number of atoms in the
+layer, and one or more calculated values.  The number of atoms and the
+value(s) are average quantities.  If the value of the {units} keyword
+is {box} or {lattice}, the "coord" is printed in box units.  If the
+value of the {units} keyword is {reduced}, the "coord" is printed in
+reduced units (0-1).
 
-If the {density} keyword is used, or the {compute} or {fix} keyword
-with a compute/fix that calculates a single quantity per atom, then a
-single value will be printed for each layer.  If the {compute} or
-{fix} keyword is used with a compute/fix that calculates N quantities
-per atom, then N values per line will be written, each of them
-averaged independently.
+If the style is {density}, or the style is {compute} or {fix} with a
+compute/fix that calculates a single quantity per atom, then a single
+value will be printed for each layer.  If the style is {compute} or
+{fix} with a compute/fix that calculates N quantities per atom, then N
+values per line will be written, each of them averaged independently.
 
-For the {compute} and {fix} keywords, the calculation performed by the
-compute or fix is on the group defined by the that command.  However,
-only atoms in the fix group are included in the layer averaging.
+The {ave} keyword determines how the layer values produced every
+{Nfreq} steps are averaged with layer values produced on previous
+steps that were multiples of {Nfreq}, before they are accessed by
+another output command or written to a file.
 
-Note that some computes perform costly calculations, involving use of
-or creation of neighbor lists.  If the compute is invoked too often by
-fix ave/spatial, it can slow down a simulation.
+If the {ave} setting is {one}, then the layuer values produced on
+timesteps that are multiples of {Nfreq} are independent of each other;
+they are output as-is without further averaging.
+
+If the {ave} setting is {running}, then the layer values produced on
+timesteps that are multiples of {Nfreq} are summed and averaged in a
+cummulative sense before being output.  Each output layer value is
+thus the average of the layer value produced on that timestep with all
+preceeding values for the same layer.  This running average begins
+when the fix is defined; it can only be restarted by deleting the fix
+via the "unfix"_unfix.html command, or re-defining the fix by
+re-specifying it.
+
+If the {ave} setting is {window}, then the layer values produced on
+timesteps that are multiples of {Nfreq} are summed and averaged within
+a moving "window" of time, so that the last M values for the same
+layer are used to produce the output.  E.g. if M = 3 and Nfreq = 1000,
+then the output on step 10000 will be the average of the individual
+layer values on steps 8000,9000,10000.  Outputs on early steps will
+average over less than M values if they are not available.
+
+:line
 
 [Restart, fix_modify, output, run start/stop, minimize info:]
 
@@ -208,4 +252,5 @@ minimization"_minimize.html.
 
 [Default:]
 
-The option defaults are norm = all and units = lattice.
+The option defaults are units = lattice, norm = all, no file output,
+and ave = one.
diff --git a/doc/fix_ave_time.html b/doc/fix_ave_time.html
index b8c560827e..d91ed43122 100644
--- a/doc/fix_ave_time.html
+++ b/doc/fix_ave_time.html
@@ -37,8 +37,6 @@
     scalar = single scalar value from fix or compute
     vector = vector of values from fix or compute
     both = both a single value and vector of values from fix or compute
-  file arg = filename
-    filename = file to write results to
   ave args = one or running or window M
     one = output new average value every Nfreq steps
     running = output cummulative average of all previous Nfreq steps
@@ -102,7 +100,7 @@ timestep 200, etc.
 
    
 
    
 
-
    Additional optional keywords also affect the averaging.
+
    Additional optional keywords also affect the operation of this fix.
 
    
 
    The type keyword chooses whether the scalar and/or vector quantities
 produced by the compute or fix are used.  For a setting of scalar a
@@ -117,8 +115,9 @@ and/or N vector quantities are written to the file in a
 self-explanatory text format.
 
    
 
    The ave keyword determines how the scalar and/or vector values
-produced every Nfreq steps are averaged with each other before they
-are accessed by another output command or written to a file.
+produced every Nfreq steps are averaged with values produced on
+previous steps that were multiples of Nfreq, before they are
+accessed by another output command or written to a file.
 
    
 
    If the ave setting is one, then the values produced on timesteps
 that are multiples of Nfreq are independent of each other; they are
@@ -127,10 +126,10 @@ output as-is without further averaging.
 
    If the ave setting is running, then the values produced on
 timesteps that are multiples of Nfreq are summed and averaged in a
 cummulative sense before being output.  Each output value is thus the
-average of the value on that timestep with all preceeding values.
-This running average begins when the fix is defined; it can only be
-restarted by deleting the fix via the unfix command, or
-re-defining the fix by re-specifying it.
+average of the value produced on that timestep with all preceeding
+values.  This running average begins when the fix is defined; it can
+only be restarted by deleting the fix via the unfix
+command, or re-defining the fix by re-specifying it.
 
    
 
    If the ave setting is window, then the values produced on
 timesteps that are multiples of Nfreq are summed and averaged within
@@ -176,6 +175,6 @@ ave/spatial
 
    
 
    Default: none
 
    
-
    The option defaults are style = scalar, no file output, and ave = one.
+
    The option defaults are type = scalar, no file output, and ave = one.
 
    
 
diff --git a/doc/fix_ave_time.txt b/doc/fix_ave_time.txt
index 08dec5bc9d..61dd38a934 100644
--- a/doc/fix_ave_time.txt
+++ b/doc/fix_ave_time.txt
@@ -26,8 +26,6 @@ keyword = {type} or {file} or {ave} :l
     scalar = single scalar value from fix or compute
     vector = vector of values from fix or compute
     both = both a single value and vector of values from fix or compute
-  {file} arg = filename
-    filename = file to write results to
   {ave} args = {one} or {running} or {window M}
     one = output new average value every Nfreq steps
     running = output cummulative average of all previous Nfreq steps
@@ -90,7 +88,7 @@ timestep 200, etc.
 
 :line
 
-Additional optional keywords also affect the averaging.
+Additional optional keywords also affect the operation of this fix.
 
 The {type} keyword chooses whether the scalar and/or vector quantities
 produced by the compute or fix are used.  For a setting of {scalar} a
@@ -105,8 +103,9 @@ and/or N vector quantities are written to the file in a
 self-explanatory text format.
 
 The {ave} keyword determines how the scalar and/or vector values
-produced every {Nfreq} steps are averaged with each other before they
-are accessed by another output command or written to a file.
+produced every {Nfreq} steps are averaged with values produced on
+previous steps that were multiples of {Nfreq}, before they are
+accessed by another output command or written to a file.
 
 If the {ave} setting is {one}, then the values produced on timesteps
 that are multiples of {Nfreq} are independent of each other; they are
@@ -115,10 +114,10 @@ output as-is without further averaging.
 If the {ave} setting is {running}, then the values produced on
 timesteps that are multiples of {Nfreq} are summed and averaged in a
 cummulative sense before being output.  Each output value is thus the
-average of the value on that timestep with all preceeding values.
-This running average begins when the fix is defined; it can only be
-restarted by deleting the fix via the "unfix"_unfix.html command, or
-re-defining the fix by re-specifying it.
+average of the value produced on that timestep with all preceeding
+values.  This running average begins when the fix is defined; it can
+only be restarted by deleting the fix via the "unfix"_unfix.html
+command, or re-defining the fix by re-specifying it.
 
 If the {ave} setting is {window}, then the values produced on
 timesteps that are multiples of {Nfreq} are summed and averaged within
@@ -164,4 +163,4 @@ ave/spatial"_fix_ave_spatial.html
 
 [Default:] none
 
-The option defaults are style = scalar, no file output, and ave = one.
+The option defaults are type = scalar, no file output, and ave = one.