Output Data¶

Output is controlled by the frequency variables already introduced in Global Control, together with the writer operators below. This page only covers particle-level output (the species/positions/velocities you defined earlier) — see the tip at the end for grid-level output.

Thermodynamic state (screen & file)¶

Printing the thermodynamic state to the screen (print_thermodynamic_state) and writing it to a .csv file (dump_thermodynamic_state) are both enabled by default; only their frequency and columns need to be set from global:

1
2
3
4
5
6

global:
  simulation_thermostate_screen_frequency: 10   # print to screen every 10 steps
  simulation_thermostate_file_frequency: 10     # append to file every 10 steps
  thermostate_file: "thermodynamic_state.csv"
  log_mode: mechanical                          # preset name, or a ';'-separated list of keywords
  # log_format: "%9.0f;%.6e"                    # optional printf formats, applied column by column

log_mode and log_format set in global are read by both the screen printer and the file writer, so the screen and the .csv file show the same columns. The file uses more significant digits by default.

Presets¶

Keywords¶

Instead of a preset, log_mode accepts any ;-separated list of the keywords below, printed in the given order, e.g. log_mode: "stp;pht;tmp;pxx;pyy;pzz;vol". An unknown keyword stops the run and prints the list of valid keywords.

When a two-temperature model is active, the ele and ite columns are added at the end automatically if they are not already in the list.

Column formats (log_format)¶

log_format is an optional ;-separated list of printf-style formats, applied in order to the active columns. Fewer formats than columns only change the first columns, and an empty entry keeps the default format of that column:

1
2
3

global:
  log_mode: "stp;pht;tmp;pre"
  log_format: "%9.0f;;%12.4f"     # step as integer, default time format, temperature with 4 decimals

File writer options¶

dump_thermodynamic_state also accepts:

thermostate_file:     string, default "thermodynamic_state.csv"  # Output file (also settable in global).
print_header:         bool, default true          # Write the column header line.
internal_units:       bool, default false         # Write values in internal units instead of eV, K, Pa, g/cm^3.
force_flush_file:     bool, default false         # Flush the file after every write.
force_append_thermo:  bool, default false         # Append to an existing file instead of overwriting it.
is_dump_virial:       bool, default false         # Append the 9 raw virial components S11 ... S33 (Pa).

print_thermodynamic_state accepts print_header and internal_units (default false for both).

Binary restart file¶

write_restart is nop by default; assign it one of the predefined aggregates from config_restart.msp (see Restarts) to activate restart writing, at the frequency set by simulation_restart_frequency:

1
2
3
4

global:
  simulation_restart_frequency: 1000

write_restart: write_restart_atoms

This writes an exaNBody-native, MPI-IO binary dump (particles, velocities, domain, species and timestep) through write_dump_atoms, which can later be re-read with read_dump_atoms (see Reading an atoms restart file). It also accepts compression_level (zlib level, default 6) and max_part_size (file partition size, default: system value).

1
2
3
4
5
6
7
8

write_restart_atoms:
  - timestep_file: "atoms_%09d.MpiIO"
  - message: { mesg: "Write restart " , endl: false }
  - print_restart_file:
      rebind: { mesg: filename }
      body:
        - message: { endl: true }
  - write_dump_atoms

Snapshot file¶

write_snapshot is likewise nop by default; assign it a predefined aggregate to write snapshots for visualization, at the frequency set by simulation_snapshot_frequency. Two common formats:

1
2
3
4

global:
  simulation_snapshot_frequency: 1000

write_snapshot: write_snapshot_xyz

write_snapshot_xyz writes a plain-text .xyz file (species name and position per atom) through write_xyz.

1
2
3
4
5
6
7
8

write_snapshot_xyz:
  - timestep_file: "exaStamp_%09d.xyz"
  - message: { mesg: "Write xyz file" , endl: false }
  - print_dump_file:
      rebind: { mesg: filename }
      body:
        - message: { endl: true }
  - write_xyz

1
2
3
4

global:
  simulation_snapshot_frequency: 1000

write_snapshot: write_snapshot_paraview

write_snapshot_paraview writes a Paraview/VTK file through write_paraview, including every available per-particle field (position, velocity, force, type, MPI rank, ...) by default — narrow it down with the fields: list of regular expressions if needed (e.g. fields: [ "type", "vx|vy|vz" ]).

1
2
3
4
5
6
7
8

write_snapshot_paraview:
  - timestep_file: "paraview/output_%09d"
  - message: { mesg: "Write paraview file" , endl: false }
  - print_dump_file:
      rebind: { mesg: filename }
      body:
        - message: { endl: true }
  - write_paraview