fix ave/chunk command¶
fix ID group-ID ave/chunk Nevery Nrepeat Nfreq chunkID value1 value2 ... keyword args ...
ID, group-ID are documented in fix command
ave/chunk = style name of this fix command
Nevery = use input values every this many timesteps
Nrepeat = # of times to use input values for calculating averages
Nfreq = calculate averages every this many timesteps
chunkID = ID of compute chunk/atom command
one or more input values can be listed
value = vx, vy, vz, fx, fy, fz, density/mass, density/number, temp, c_ID, c_ID[I], f_ID, f_ID[I], v_name
vx,vy,vz,fx,fy,fz = atom attribute (velocity, force component) density/number, density/mass = number or mass density temp = temperature c_ID = per-atom vector calculated by a compute with ID c_ID[I] = Ith column of per-atom array calculated by a compute with ID, I can include wildcard (see below) f_ID = per-atom vector calculated by a fix with ID f_ID[I] = Ith column of per-atom array calculated by a fix with ID, I can include wildcard (see below) v_name = per-atom vector calculated by an atom-style variable with name
zero or more keyword/arg pairs may be appended
keyword = norm or ave or bias or adof or cdof or file or overwrite or title1 or title2 or title3
norm arg = all or sample or none = how output on Nfreq steps is normalized all = output is sum of atoms across all Nrepeat samples, divided by atom count sample = output is sum of Nrepeat sample averages, divided by Nrepeat none = output is sum of Nrepeat sample sums, divided by Nrepeat ave args = one or running or window M one = output new average value every Nfreq steps running = output cumulative average of all previous Nfreq steps window M = output average of M most recent Nfreq steps bias arg = bias-ID bias-ID = ID of a temperature compute that removes a velocity bias for temperature calculation adof value = dof_per_atom dof_per_atom = define this many degrees-of-freedom per atom for temperature calculation cdof value = dof_per_chunk dof_per_chunk = define this many degrees-of-freedom per chunk for temperature calculation file arg = filename filename = file to write results to overwrite arg = none = overwrite output file with only latest output format arg = string string = C-style format string title1 arg = string string = text to print as 1st line of output file title2 arg = string string = text to print as 2nd line of output file title3 arg = string string = text to print as 3rd line of output file
fix 1 all ave/chunk 10000 1 10000 binchunk c_myCentro title1 "My output values" fix 1 flow ave/chunk 100 10 1000 molchunk vx vz norm sample file vel.profile fix 1 flow ave/chunk 100 5 1000 binchunk density/mass ave running fix 1 flow ave/chunk 100 5 1000 binchunk density/mass ave running
If you are trying to replace a deprecated fix ave/spatial command with the newer, more flexible fix ave/chunk and compute chunk/atom commands, you simply need to split the fix ave/spatial arguments across the two new commands. For example, this command:
fix 1 flow ave/spatial 100 10 1000 y 0.0 1.0 vx vz norm sample file vel.profile
could be replaced by:
compute cc1 flow chunk/atom bin/1d y 0.0 1.0 fix 1 flow ave/chunk 100 10 1000 cc1 vx vz norm sample file vel.profile
Use one or more per-atom vectors as inputs every few timesteps, sum the values over the atoms in each chunk at each timestep, then average the per-chunk values over longer timescales. The resulting chunk averages can be used by other output commands such as thermo_style custom, and can also be written to a file.
In LAMMPS, chunks are collections of atoms defined by a compute chunk/atom command, which assigns each atom to a single chunk (or no chunk). The ID for this command is specified as chunkID. For example, a single chunk could be the atoms in a molecule or atoms in a spatial bin. See the compute chunk/atom page and the Howto chunk page for details of how chunks can be defined and examples of how they can be used to measure properties of a system.
Note that only atoms in the specified group contribute to the summing and averaging calculations. The compute chunk/atom command defines its own group as well as an optional region. Atoms will have a chunk ID = 0, meaning they belong to no chunk, if they are not in that group or region. Thus you can specify the “all” group for this command if you simply want to use the chunk definitions provided by chunkID.
Each specified per-atom value can be an atom attribute (position, velocity, force component), a mass or number density, or the result of a compute or fix or the evaluation of an atom-style variable. In the latter cases, the compute, fix, or variable must produce a per-atom quantity, not a global quantity. Note that the compute property/atom command provides access to any attribute defined and stored by atoms. If you wish to time-average global quantities from a compute, fix, or variable, then see the fix ave/time command.
The per-atom values of each input vector are summed and averaged independently of the per-atom values in other input vectors.
Computes that produce per-atom quantities are those which have the word atom in their style name. See the doc pages for individual fixes to determine which ones produce per-atom quantities. Variables of style atom are the only ones that can be used with this fix since all other styles of variable produce global quantities.
Note that for values from a compute or fix, the bracketed index I can be specified using a wildcard asterisk with the index to effectively specify multiple values. This takes the form “*” or “*n” or “n*” or “m*n”. If N = the size of the vector (for mode = scalar) or the number of columns in the array (for mode = vector), then an asterisk with no numeric values means all indices from 1 to N. A leading asterisk means all indices from 1 to n (inclusive). A trailing asterisk means all indices from n to N (inclusive). A middle asterisk means all indices from m to n (inclusive).
Using a wildcard is the same as if the individual columns of the array had been listed one by one. E.g. these 2 fix ave/chunk commands are equivalent, since the compute property/atom command creates, in this case, a per-atom array with 3 columns:
compute myAng all property/atom angmomx angmomy angmomz fix 1 all ave/chunk 100 1 100 cc1 c_myAng[*] file tmp.angmom fix 2 all ave/chunk 100 1 100 cc1 c_myAng c_myAng c_myAng file tmp.angmom
This fix works by creating an array of size Nchunk by Nvalues on each processor. Nchunk is the number of chunks which is defined by the compute chunk/atom command. Nvalues is the number of input values specified. Each processor loops over its atoms, tallying its values to the appropriate chunk. Then the entire array is summed across all processors. This means that using a large number of chunks will incur an overhead in memory and computational cost (summing across processors), so be careful to define a reasonable number of chunks.
The Nevery, Nrepeat, and Nfreq arguments specify on what timesteps the input values will be accessed and contribute to the average. The final averaged quantities are generated on timesteps that are a multiples of Nfreq. The average is over Nrepeat quantities, computed in the preceding 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. Also, the timesteps contributing to the average value cannot overlap, i.e. Nrepeat*Nevery can not exceed Nfreq.
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. Similarly for timesteps 190,192,194,196,198,200 on timestep 200, etc. If Nrepeat=1 and Nfreq = 100, then no time averaging is done; values are simply generated on timesteps 100,200,etc.
Each input value can also be averaged over the atoms in each chunk. The way the averaging is done across the Nrepeat timesteps to produce output on the Nfreq timesteps, and across multiple Nfreq outputs, is determined by the norm and ave keyword settings, as discussed below.
To perform per-chunk averaging within a Nfreq time window, the number of chunks Nchunk defined by the compute chunk/atom command must remain constant. If the ave keyword is set to running or window then Nchunk must remain constant for the duration of the simulation. This fix forces the chunk/atom compute specified by chunkID to hold Nchunk constant for the appropriate time windows, by not allowing it to re-calculate Nchunk, which can also affect how it assigns chunk IDs to atoms. This is particularly important to understand if the chunks defined by the compute chunk/atom command are spatial bins. If its units keyword is set to box or lattice, then the number of bins Nchunk and size of each bin will be fixed over the Nfreq time window, which can affect which atoms are discarded if the simulation box size changes. If its units keyword is set to reduced, then the number of bins Nchunk will still be fixed, but the size of each bin can vary at each timestep if the simulation box size changes, e.g. for an NPT simulation.
The atom attribute values (vx,vy,vz,fx,fy,fz) are self-explanatory. As noted above, any other atom attributes can be used as input values to this fix by using the compute property/atom command and then specifying an input value from that compute.
The density/number value means the number density is computed for each chunk, i.e. number/volume. The density/mass value means the mass density is computed for each chunk, i.e. total-mass/volume. The output values are in units of 1/volume or density (mass/volume). See the units command page for the definition of density for each choice of units, e.g. gram/cm^3. If the chunks defined by the compute chunk/atom command are spatial bins, the volume is the bin volume. Otherwise it is the volume of the entire simulation box.
The temp value means the temperature is computed for each chunk, by the formula KE = DOF/2 k T, where KE = total kinetic energy of the chunk of atoms (sum of 1/2 m v^2), DOF = the total number of degrees of freedom for all atoms in the chunk, k = Boltzmann constant, and T = temperature.
The DOF is calculated as N*adof + cdof, where N = number of atoms in the chunk, adof = degrees of freedom per atom, and cdof = degrees of freedom per chunk. By default adof = 2 or 3 = dimensionality of system, as set via the dimension command, and cdof = 0.0. This gives the usual formula for temperature.
Note that currently this temperature only includes translational degrees of freedom for each atom. No rotational degrees of freedom are included for finite-size particles. Also no degrees of freedom are subtracted for any velocity bias or constraints that are applied, such as compute temp/partial, or fix shake or fix rigid. This is because those degrees of freedom (e.g. a constrained bond) could apply to sets of atoms that are both included and excluded from a specific chunk, and hence the concept is somewhat ill-defined. In some cases, you can use the adof and cdof keywords to adjust the calculated degrees of freedom appropriately, as explained below.
Also note that a bias can be subtracted from atom velocities before they are used in the above formula for KE, by using the bias keyword. This allows, for example, a thermal temperature to be computed after removal of a flow velocity profile.
Note that the per-chunk temperature calculated by this fix and the compute temp/chunk command can be different. The compute calculates the temperature for each chunk for a single snapshot. This fix can do that but can also time average those values over many snapshots, or it can compute a temperature as if the atoms in the chunk on different timesteps were collected together as one set of atoms to calculate their temperature. The compute allows the center-of-mass velocity of each chunk to be subtracted before calculating the temperature; this fix does not.
If a value begins with “c_”, a compute ID must follow which has been previously defined in the input script. If no bracketed integer is appended, the per-atom vector calculated by the compute is used. If a bracketed integer is appended, the Ith column of the per-atom array calculated by the compute is used. Users can also write code for their own compute styles and add them to LAMMPS. See the discussion above for how I can be specified with a wildcard asterisk to effectively specify multiple values.
If a value begins with “f_”, a fix ID must follow which has been previously defined in the input script. If no bracketed integer is appended, the per-atom vector calculated by the fix is used. If a bracketed integer is appended, the Ith column of the per-atom array calculated by the fix is used. Note that some fixes only produce their values on certain timesteps, which must be compatible with Nevery, else an error results. Users can also write code for their own fix styles and add them to LAMMPS. See the discussion above for how I can be specified with a wildcard asterisk to effectively specify multiple values.
If a value begins with “v_”, a variable name must follow which has been previously defined in the input script. Variables of style atom can reference thermodynamic keywords and various per-atom attributes, or invoke other computes, fixes, or variables when they are evaluated, so this is a very general means of generating per-atom quantities to average within chunks.
Additional optional keywords also affect the operation of this fix and its outputs.
The norm keyword affects how averaging is done for the per-chunk values that are output every Nfreq timesteps.
It the norm setting is all, which is the default, a chunk value is summed over all atoms in all Nrepeat samples, as is the count of atoms in the chunk. The averaged output value for the chunk on the Nfreq timesteps is Total-sum / Total-count. In other words it is an average over atoms across the entire Nfreq timescale. For the density/number and density/mass values, the volume (bin volume or system volume) used in the final normalization will be the volume at the final Nfreq timestep. For the temp values, degrees of freedom and kinetic energy are summed separately across the entire Nfreq timescale, and the output value is calculated by dividing those two sums.
If the norm setting is sample, the chunk value is summed over atoms for each sample, as is the count, and an “average sample value” is computed for each sample, i.e. Sample-sum / Sample-count. The output value for the chunk on the Nfreq timesteps is the average of the Nrepeat “average sample values”, i.e. the sum of Nrepeat “average sample values” divided by Nrepeat. In other words it is an average of an average. For the density/number and density/mass values, the volume (bin volume or system volume) used in the per-sample normalization will be the current volume at each sampling step.
If the norm setting is none, a similar computation as for the sample setting is done, except the individual “average sample values” are “summed sample values”. A summed sample value is simply the chunk value summed over atoms in the sample, without dividing by the number of atoms in the sample. The output value for the chunk on the Nfreq timesteps is the average of the Nrepeat “summed sample values”, i.e. the sum of Nrepeat “summed sample values” divided by Nrepeat. For the density/number and density/mass values, the volume (bin volume or system volume) used in the per-sample sum normalization will be the current volume at each sampling step.
The ave keyword determines how the per-chunk values 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, which is the default, then the chunk 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 chunk values produced on timesteps that are multiples of Nfreq are summed and averaged in a cumulative sense before being output. Each output chunk value is thus the average of the chunk value produced on that timestep with all preceding values for the same chunk. 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 chunk 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 chunk 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 chunk values on steps 8000,9000,10000. Outputs on early steps will average over less than M values if they are not available.
The bias keyword specifies the ID of a temperature compute that removes a “bias” velocity from each atom, specified as bias-ID. It is only used when the temp value is calculated, to compute the thermal temperature of each chunk after the translational kinetic energy components have been altered in a prescribed way, e.g. to remove a flow velocity profile. See the doc pages for individual computes that calculate a temperature to see which ones implement a bias.
The adof and cdof keywords define the values used in the degree of freedom (DOF) formula described above for temperature calculation for each chunk. They are only used when the temp value is calculated. They can be used to calculate a more appropriate temperature for some kinds of chunks. Here are 3 examples:
If spatially binned chunks contain some number of water molecules and fix shake is used to make each molecule rigid, then you could calculate a temperature with 6 degrees of freedom (DOF) (3 translational, 3 rotational) per molecule by setting adof to 2.0.
If compute temp/partial is used with the bias keyword to only allow the x component of velocity to contribute to the temperature, then adof = 1.0 would be appropriate.
If each chunk consists of a large molecule, with some number of its bonds constrained by fix shake or the entire molecule by fix rigid/small, adof = 0.0 and cdof could be set to the remaining degrees of freedom for the entire molecule (entire chunk in this case), e.g. 6 for 3d, or 3 for 2d, for a rigid molecule.
The file keyword allows a filename to be specified. Every Nfreq timesteps, a section of chunk info will be written to a text file in the following format. A line with the timestep and number of chunks is written. Then one line per chunk is written, containing the chunk ID (1-Nchunk), an optional original ID value, optional coordinate values for chunks that represent spatial bins, the number of atoms in the chunk, and one or more calculated values. More explanation of the optional values is given below. The number of values in each line corresponds to the number of values specified in the fix ave/chunk command. The number of atoms and the value(s) are summed or average quantities, as explained above.
The overwrite keyword will continuously overwrite the output file with the latest output, so that it only contains one timestep worth of output. This option can only be used with the ave running setting.
The format keyword sets the numeric format of each value when it is printed to a file via the file keyword. Note that all values are floating point quantities. The default format is %g. You can specify a higher precision if desired, e.g. %20.16g.
The title1 and title2 and title3 keywords allow specification of the strings that will be printed as the first 3 lines of the output file, assuming the file keyword was used. LAMMPS uses default values for each of these, so they do not need to be specified.
By default, these header lines are as follows:
# Chunk-averaged data for fix ID and group name # Timestep Number-of-chunks # Chunk (OrigID) (Coord1) (Coord2) (Coord3) Ncount value1 value2 ...
In the first line, ID and name are replaced with the fix-ID and group name. The second line describes the two values that are printed at the first of each section of output. In the third line the values are replaced with the appropriate value names, e.g. fx or c_myCompute.
The words in parenthesis only appear with corresponding columns if the chunk style specified for the compute chunk/atom command supports them. The OrigID column is only used if the compress keyword was set to yes for the compute chunk/atom command. This means that the original chunk IDs (e.g. molecule IDs) will have been compressed to remove chunk IDs with no atoms assigned to them. Thus a compressed chunk ID of 3 may correspond to an original chunk ID or molecule ID of 415. The OrigID column will list 415 for the third chunk.
The CoordN columns only appear if a binning style was used in the compute chunk/atom command. For bin/1d, bin/2d, and bin/3d styles the column values are the center point of the bin in the corresponding dimension. Just Coord1 is used for bin/1d, Coord2 is added for bin/2d, Coord3 is added for bin/3d. For bin/sphere, just Coord1 is used, and it is the radial coordinate. For bin/cylinder, Coord1 and Coord2 are used. Coord1 is the radial coordinate (away from the cylinder axis), and coord2 is the coordinate along the cylinder axis.
Note that if the value of the units keyword used in the compute chunk/atom command is box or lattice, the coordinate values will be in distance units. If the value of the units keyword is reduced, the coordinate values will be in unitless reduced units (0-1). This is not true for the Coord1 value of style bin/sphere or bin/cylinder which both represent radial dimensions. Those values are always in distance units.
Restart, fix_modify, output, run start/stop, minimize info¶
This fix computes a global array of values which can be accessed by various output commands. The values can only be accessed on timesteps that are multiples of Nfreq since that is when averaging is performed. The global array has # of rows = the number of chunks Nchunk as calculated by the specified compute chunk/atom command. The # of columns = M+1+Nvalues, where M = 1 to 4, depending on whether the optional columns for OrigID and CoordN are used, as explained above. Following the optional columns, the next column contains the count of atoms in the chunk, and the remaining columns are the Nvalue quantities. When the array is accessed with a row I that exceeds the current number of chunks, than a 0.0 is returned by the fix instead of an error, since the number of chunks can vary as a simulation runs depending on how that value is computed by the compute chunk/atom command.
The array values calculated by this fix are treated as “intensive”, since they are typically already normalized by the count of atoms in each chunk.
The option defaults are norm = all, ave = one, bias = none, no file output, and title 1,2,3 = strings as described above.