This documentation briefly introduces SOMD's command line interface (CLI).
After installing the SOMD package, the command somd will be available.
You could type somd -v to check the package version, and somd -h for the
help messages.
In order to execute a SOMD operation through the CLI, you will need to specify the require execution mode and the corresponding keywords through command:
somd [mode] --keyword_1 --keyword_2 ...The [mode] string has two valid options, which are run and select.
The run mode reads reads a TOML file and does the dynamics,
and the select model selects structures for the active learning task.
This mode taks the following keywords.
-
-i/--inputIf Mandatory: yes
Type:
strNumber of Required Parameters: 1
Default Value: None
Descriptions: Name of the input
TOMLfile.
This mode taks the following keywords.
-
-t/--trajectoriesIf Mandatory: yes
Type:
List[str](delimited by white spaces)Number of Required Parameters: more than 1
Default Value: None
Descriptions: Paths to the input HDF5 trajectories, with forces recorded.
-
-o/--outputIf Mandatory: yes
Type:
strNumber of Required Parameters: 1
Default Value: None
Descriptions: Name of the output trajectory file.
-
--msd_f_minIf Mandatory: no
Type:
floatNumber of Required Parameters: 1
Default Value: 50.0
Descriptions: The lower boundary of the force MSD used for identifying candidate structures. In unit of (kJ/mol/nm).
-
--msd_f_maxIf Mandatory: no
Type:
floatNumber of Required Parameters: 1
Default Value: 250.0
Descriptions: The upper boundary of the force MSD used for identifying candidate structures. In unit of (kJ/mol/nm).
-
--n_maxIf Mandatory: no
Type:
intNumber of Required Parameters: 1
Default Value:
infDescriptions: The max number of structures to select (default: inf).
-
--log_fileIf Mandatory: no
Type:
strNumber of Required Parameters: 1
Default Value:
somd_selection.jsonDescriptions: The name of the log file that contains the selection information.
-
--shuffleIf Mandatory: no
Number of Required Parameters: 0
Default Value:
TrueDescriptions: If perform randomly selection if the number of candidate structures is larger than
n_max.