Output Files#
Trajectories#
trajectory.son#
This file contains metadata and calculation results for a set of related calculations that, e.g., come from a geometry optimization, a phonopy calculation, or a molecular dynamics run. The file format is son, a slight extension to json allowing to add data sequentially.
Example trajectory.son
{"MD": {
"type": "molecular-dynamics",
...},
"calculator": {
"calculator": "Aims",
"calculator_parameters": {
"xc": "pbesol",
"k_grid": [2, 2, 2],
...}},
"atoms": {
"cell":
[[ 8.33141234000000e+00, -8.33141234000000e+00, 0.00000000000000e+00],
[ 8.33141234000000e+00, 8.33141234000000e+00, 0.00000000000000e+00],
[ 0.00000000000000e+00, 0.00000000000000e+00, 1.24971185100000e+01]],
"positions":
[[ 0.00000000000000e+00, 0.00000000000000e+00, 0.00000000000000e+00],
[ 4.16570617000000e+00, 0.00000000000000e+00, 0.00000000000000e+00],
…}
“primitive”: { …
…}
===
{"atoms": {
"info": {
"nsteps": 0,
"dt": 4.91134739423203e-01,
"aims_uuid": "D985353A-F8FD-4635-A939-E129A7E6E146"},
"positions":
[[ 0.00000000000000e+00, 0.00000000000000e+00, 0.00000000000000e+00],
[ 4.16570617000000e+00, 0.00000000000000e+00, 0.00000000000000e+00], …}
---
...
trajectory.nc#
A NetCDF file containing and xarray.Dataset with post-processed data
Load trajectory.nc
>>> import xarray as xr
>>>
>>> ds = xr.open_dataset('trajectory.nc')
>>>
>>> print(ds)
<xarray.Dataset>
Dimensions: (I: 180, a: 3, b: 3, time: 10001)
Coordinates:
* time (time) float64 0.0 2.0 4.0 6.0 ... 2e+04 2e+04 2e+04
Dimensions without coordinates: I, a, b
Data variables:
positions (time, I, a) float64 ...
displacements (time, I, a) float64 ...
velocities (time, I, a) float64 ...
momenta (time, I, a) float64 ...
forces (time, I, a) float64 ...
energy_kinetic (time) float64 ...
energy_potential (time) float64 ...
stress (time, a, b) float64 ...
stress_kinetic (time, a, b) float64 ...
stress_potential (time, a, b) float64 ...
temperature (time) float64 ...
cell (time, a, b) float64 ...
positions_reference (I, a) float64 ...
lattice_reference (a, b) float64 ...
pressure (time) float64 ...
pressure_kinetic (time) float64 ...
pressure_potential (time) float64 ...
Attributes:
name: trajectory
system_name: O3Sb2
natoms: 180
time_unit: fs
timestep: 1.9999999999999978
nsteps: 10000
symbols: ['O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O',...
masses: [ 15.999 15.999 15.999 15.999 15.999 15.999 15.99...
atoms_reference: {"pbc": [true, true, true],\n"cell": \n[[ 1.50410061436...
atoms_primitive: {"pbc": [true, true, true],\n"cell": \n[[ 5.01366871456...
atoms_supercell: {"pbc": [true, true, true],\n"cell": \n[[ 1.50410061436...
volume: 3051.2387320953862
raw_metadata: {"MD": {\n "type": "molecular-dynamics",\n "md-type":...
hash: 097714462c68b9f8cbdf08e6a29c0bfda7922c01
Numerical Data#
.dat files#
Files with .dat suffix are 1D or 2D arrays that can be read with numpy.loadtxt:
Example frequencies.dat
0.000000000000000000e+00
0.000000000000000000e+00
0.000000000000000000e+00
9.328537621518436795e-01
1.319254442145902706e+00
1.615750112078766065e+00
1.865707524303687359e+00
2.085924425237942970e+00
2.285015721907639907e+00
2.468099064244776208e+00
2.638508884291805412e+00
2.798561286455531594e+00
Load frequencies.dat
>>> import numpy as np
>>>
>>> data = np.loadtxt('frequencies.dat')
>>>
>>> print(data)
[0. 0. 0. 0.93285376 1.31925444 1.61575011
1.86570752 2.08592443 2.28501572 2.46809906 2.63850888 2.79856129]
.csv files#
Files .csv suffix are standard comma-separated values that can be parsed, e.g., with pandas.
Example trajectory.csv
time,temperature,energy_kinetic,energy_potential,pressure_kinetic,pressure_potential,pressure
0.0,286.2001210048815,6.658958660176626,-13092419.8403197,0.0014549191863471355,-0.0001421961672687167,0.0013127230190784188
1.9999999999999978,289.5694460043815,6.7373520438187935,-13092419.9182694,0.0014720473956910023,-0.00012147676349475005,0.0013505706321962523
3.9999999999999956,292.46310420107915,6.80467818694143,-13092419.9839958,0.0014867575181546967,,
5.999999999999994,294.9143609337456,6.86171106725956,-13092420.0383063,0.0014992186605137458,,
Load trajectory.csv
>>> import pandas as pd
>>>
>>> df = pd.read_csv('trajectory.csv')
>>>
>>> print(df)
time temperature energy_kinetic energy_potential pressure_kinetic pressure_potential pressure
0 0.0 286.200121 6.658959 -1.309242e+07 0.001455 -0.000142 0.001313
1 2.0 289.569446 6.737352 -1.309242e+07 0.001472 -0.000121 0.001351
2 4.0 292.463104 6.804678 -1.309242e+07 0.001487 NaN NaN
[3 rows x 7 columns]
.json files#
These are plain JSON files that can be parsed with the python builtin json module
Example md_describe.json
{
"time": {
"count": 10001.0,
"mean": 9999.999999999987,
"std": 5774.368710084239,
"min": 0.0,
"25%": 4999.999999999994,
"50%": 9999.999999999987,
"75%": 14999.999999999984,
"max": 19999.999999999975
},
"temperature": {
"count": 10001.0,
"mean": 309.14540993328325,
"std": 14.54050396748833,
"min": 257.7711666913883,
"25%": 299.2299218510854,
"50%": 309.2638007517673,
"75%": 319.0663317618751,
"max": 355.05955146050223
},
"energy_kinetic": {
"count": 10001.0,
"mean": 7.19282192299974,
"std": 0.3383108800851539,
"min": 5.997508095931853,
"25%": 6.962120325100335,
"50%": 7.1955764975375205,
"75%": 7.4236499467457895,
"max": 8.261096699662168
},
"energy_potential": {
"count": 10001.0,
"mean": -13092416.82839676,
"std": 2.2319201670206765,
"min": -13092420.901731301,
"25%": -13092419.756083699,
"50%": -13092415.5125696,
"75%": -13092415.1902583,
"max": -13092414.1603322
},
"pressure_kinetic": {
"count": 10001.0,
"mean": 0.0015715632359058732,
"std": 7.391771228878895e-05,
"min": 0.00131039852390555,
"25%": 0.001521157129150454,
"50%": 0.0015721650842653186,
"75%": 0.001621996965507344,
"max": 0.0018049711226602926
},
"pressure_potential": {
"count": 1032.0,
"mean": 0.00027321549417593297,
"std": 0.002184381533766492,
"min": -0.006481330360441526,
"25%": -0.001147865102920749,
"50%": 0.00023375660603530354,
"75%": 0.001799169124854924,
"max": 0.00778187950662052
},
"pressure": {
"count": 1032.0,
"mean": 0.001844596382543767,
"std": 0.0021801900840728618,
"min": -0.004976051568943327,
"25%": 0.00041292059984573813,
"50%": 0.0018072058785148773,
"75%": 0.0033777412146081182,
"max": 0.009323241222005936
}
}
Load md_describe.json
>>> import json
>>>
>>> data = json.load(open('md_describe.json'))
>>>
>>> pprint(data)
{'energy_kinetic': {'25%': 6.962120325100335,
'50%': 7.1955764975375205,
'75%': 7.4236499467457895,
'count': 10001.0,
'max': 8.261096699662168,
'mean': 7.19282192299974,
'min': 5.997508095931853,
'std': 0.3383108800851539},
'energy_potential': {'25%': -13092419.756083699,
'50%': -13092415.5125696,
'75%': -13092415.1902583,
'count': 10001.0,
'max': -13092414.1603322,
'mean': -13092416.82839676,
'min': -13092420.901731301,
'std': 2.2319201670206765},
'pressure': {'25%': 0.00041292059984573813,
'50%': 0.0018072058785148773,
'75%': 0.0033777412146081182,
'count': 1032.0,
'max': 0.009323241222005936,
'mean': 0.001844596382543767,
'min': -0.004976051568943327,
'std': 0.0021801900840728618},
'pressure_kinetic': {'25%': 0.001521157129150454,
'50%': 0.0015721650842653186,
'75%': 0.001621996965507344,
'count': 10001.0,
'max': 0.0018049711226602926,
'mean': 0.0015715632359058732,
'min': 0.00131039852390555,
'std': 7.391771228878895e-05},
'pressure_potential': {'25%': -0.001147865102920749,
'50%': 0.00023375660603530354,
'75%': 0.001799169124854924,
'count': 1032.0,
'max': 0.00778187950662052,
'mean': 0.00027321549417593297,
'min': -0.006481330360441526,
'std': 0.002184381533766492},
'temperature': {'25%': 299.2299218510854,
'50%': 309.2638007517673,
'75%': 319.0663317618751,
'count': 10001.0,
'max': 355.05955146050223,
'mean': 309.14540993328325,
'min': 257.7711666913883,
'std': 14.54050396748833},
'time': {'25%': 4999.999999999994,
'50%': 9999.999999999987,
'75%': 14999.999999999984,
'count': 10001.0,
'max': 19999.999999999975,
'mean': 9999.999999999987,
'min': 0.0,
'std': 5774.368710084239}}
Force Constants#
FORCE_CONSTANTS#
These are force constants in the phonopy format in the compact form (n_primitive, n_supercell, 3, 3). They can be parsed with phonopy.file_IO.parse_FORCE_CONSTANTS.
FORCE_CONSTANTS_remapped#
These are force constants mapped to (3 * n_supercell, 3 * n_supercell) shape. They can be parsed with numpy.loadtxt similar to .dat. files.
Green-Kubo results#
greenkubo.nc#
A NetCDF file containing and xarray.Dataset with post-processed data
Load trajectory.nc
>>> import xarray as xr
>>>
>>> ds = xr.open_dataset('greenkubo.nc')
>>>
>>> print(ds)
<xarray.Dataset> Size: 34MB
Dimensions: (time: 3001, a: 3, b: 3,
s: 6, q: 108, ia: 6,
q_ir: 24, nq: 9,
q_int: 8000, i: 2, J: 216)
Coordinates:
* time (time) float64 24kB 0.0 .....
* nq (nq) int64 72B 4 6 ... 18 20
Dimensions without coordinates: a, b, s, q, ia, q_ir, q_int, i, J
Data variables: (12/42)
heat_flux (time, a) float64 72kB -1....
heat_flux_autocorrelation (time, a, b) float64 216kB ...
heat_flux_autocorrelation_filtered (time, a, b) float64 216kB ...
heat_flux_autocorrelation_cumtrapz (time, a, b) float64 216kB ...
heat_flux_autocorrelation_cumtrapz_filtered (time, a, b) float64 216kB ...
kappa (a, b) float64 72B 1.423 ....
... ...
interpolation_w_sq (s, q_int) float64 384kB 0...
interpolation_tau_sq (s, q_int) float64 384kB 0...
force_constants (i, J, a, b) float64 31kB ...
thermal_conductivity_corrected (a, b) float64 72B 2.324 ....
volume (time) float64 24kB 5.873e...
temperature (time) float64 24kB 307.5 ...
Attributes: (12/19)
name: trajectory
system_name: CuI
natoms: 216
time_unit: fs
timestep: 5.0
nsteps: 12000
... ...
sigma: 0.6311824798505474
st_size: 749635900
hash_raw: 4967e1958c2f78b579e3669a6582ab830b075df8
gk_window_fs: 1195.5434782608695
gk_prefactor: 1134529729.719537
filter_prominence: 0.2