ipsuite.calculators package

Submodules

ipsuite.calculators.ase_geoopt module

class ipsuite.calculators.ase_geoopt.ASEGeoOpt(*args, **kwargs)[source]

Bases: IPSNode

Geometry optimization using ASE optimizers.

Parameters

datalist[ase.Atoms]

Input atomic configurations to optimize.

data_idint, default=-1

Index of the configuration to optimize from the data list.

modelNodeWithCalculator

A model node that implements ‘get_calculator’ method.

optimizerstr, default=”FIRE”

Name of the ASE optimizer to use (e.g., “FIRE”, “BFGS”, “LBFGS”).

repeattuple[int, int, int], default=(1, 1, 1)

Repeat the unit cell in x, y, z directions.

run_kwargsdict, default={“fmax”: 0.05}

Keyword arguments passed to the optimizer run method.

init_kwargsdict, default={}

Keyword arguments passed to the optimizer initialization.

maxstepint, optional

Maximum number of optimization steps to perform.

sampling_rateint, default=1

How often to sample atomic configurations during optimization.

dump_rateint, default=1000

How many configurations to cache before writing to file.

checkslist, optional

List of check nodes for monitoring optimization progress.

constraintslist, optional

List of constraint nodes to apply during optimization.

Attributes

frameslist[ase.Atoms]

Optimized atomic configurations from the trajectory.

traj_filePath

Path to the HDF5 trajectory file containing optimization steps.

Examples

>>> model = ips.MACEMPModel()
>>> with project:
...     data = ips.AddData(file="ethanol.xyz")
...     geoopt = ips.ASEGeoOpt(data=data.frames, model=model, optimizer="FIRE")
>>> project.repro()
>>> print(f"Optimization completed with {len(geoopt.frames)} steps")
Optimization completed with 74 steps
checks: list = None
constraints: list = None
data: List[Atoms]
data_id: int = -1
dump_rate: int = 1000
property frames: List[Atoms]
get_atoms() → Atoms[source]
init_kwargs: dict
log_file: Path = PosixPath('$nwd$/opt.log')
maxstep: int = None
model: NodeWithCalculator
model_outs: Path = PosixPath('$nwd$/model_outs')
optimizer: str = 'FIRE'
repeat: tuple[int, int, int] = (1, 1, 1)
run()[source]
run_kwargs: dict
sampling_rate: int = 1
traj_file: Path = PosixPath('$nwd$/structures.h5')

ipsuite.calculators.ase_md module

ipsuite.calculators.ase_standard module

class ipsuite.calculators.ase_standard.EMTCalculator(**kwargs)[source]

Bases: EMT

calculate(atoms=None, properties=None, system_changes=['positions', 'numbers', 'cell', 'pbc', 'initial_charges', 'initial_magmoms'])[source]

Do the calculation.

properties: list of str

List of what needs to be calculated. Can be any combination of ‘energy’, ‘forces’, ‘stress’, ‘dipole’, ‘charges’, ‘magmom’ and ‘magmoms’.

system_changes: list of str

List of what has changed since last calculation. Can be any combination of these six: ‘positions’, ‘numbers’, ‘cell’, ‘pbc’, ‘initial_charges’ and ‘initial_magmoms’.

Subclasses need to implement this, but can ignore properties and system_changes if they want. Calculated properties should be inserted into results dictionary like shown in this dummy example:

self.results = {'energy': 0.0,
                'forces': np.zeros((len(atoms), 3)),
                'stress': np.zeros(6),
                'dipole': np.zeros(3),
                'charges': np.zeros(len(atoms)),
                'magmom': 0.0,
                'magmoms': np.zeros(len(atoms))}

The subclass implementation should first call this implementation to set the atoms attribute and create any missing directories.

class ipsuite.calculators.ase_standard.EMTSinglePoint[source]

Bases: object

This is a testing Node! It uses ASE’S EMT calculator with default arguments. The calculator accept all elements and implements energy, forces, making it very useful for creating dummy data.

get_calculator(**kwargs)[source]

Get an EMT ase calculator.

class ipsuite.calculators.ase_standard.LJSinglePoint(epsilon: float = 1.0, sigma: float = 1.0, rc: float = 10.0)[source]

Bases: object

This is a testing Node! It uses ASE’S Lennard-Jones calculator with default arguments. The calculator accept all elements and implements energy, forces and stress, making it very useful for creating dummy data.

epsilon: float = 1.0
get_calculator(**kwargs)[source]

Get an LJ ase calculator.

rc: float = 10.0
sigma: float = 1.0

ipsuite.calculators.cp2k module

CP2K interface without ASE calculator.

This interface is less restrictive than CP2K Single Point.

class ipsuite.calculators.cp2k.CP2KSinglePoint(*args, **kwargs)[source]

Bases: IPSNode

Node for running CP2K Single point calculations.

Parameters

cp2k_shellstr, default=None

The cmd to run cp2k. If None, the environment variable IPSUITE_CP2K_SHELL is used.

cp2k_paramsstr

The path to the cp2k yaml input file. cp2k-input-tools is used to generate the input file from the yaml file.

cp2k_fileslist[str]

Additional dependencies for the cp2k calculation.

wfn_restart_filestr, optional

The path to the wfn restart file.

wfn_restart_nodestr, optional

A cp2k Node that has a wfn restart file.

cp2k_directory: Path = PosixPath('$nwd$/cp2k')
cp2k_files: list[str] = None
cp2k_params: str = 'cp2k.yaml'
cp2k_shell: str | None = None
data: list[Atoms]
failed_configs: dict = NOT_AVAILABLE
failure_policy: Literal['skip', 'fail'] = 'fail'
property frames: list[Atoms]
get_calculator(directory: str | None = None, idx: int = 0)[source]
get_input_script()[source]

Return the input script.

output_file: Path = PosixPath('$nwd$/structures.h5')
run()[source]

ZnTrack run method.

Raises

RuntimeError

If the cp2k_shell is not set.

wfn_restart_file: str = None
wfn_restart_node: Node = None

ipsuite.calculators.lammps module

class ipsuite.calculators.lammps.LammpsSimulator(*args, **kwargs)[source]

Bases: ProcessSingleAtom

Can perform LAMMPS Simulations.

Parameters

lmp_exe: str

This is the name or path of the LAMMPS executable. Either path to executable, “lmp” or “lamp_<machine>”. See https://docs.lammps.org/Run_basics.html for more information

lmp_params: str

Path to file. To be able to change parameters with DVC and not have to change them manually in the input script, a params file in yaml format and corresponding template file must be provided.

lmp_template: str

Path to file. In combination with the params file this will be the input script for the LAMMPS simulation

skiprun: bool, optional

Whether to skip running LAMMPS or not, by default False

Returns

None

This function does not return anything. Instead, it creates a LAMMPS input script based on the specified template and parameter files, runs the LAMMPS simulation using the specified executable, and saves the simulation output to the specified directory.

create_input_script()[source]
data: List[ase.Atoms]
dump_file: Path = PosixPath('$nwd$/NPT.lammpstraj')
fill_atoms_with_life()[source]
property frames: list[Atoms]
get_atoms()[source]
classmethod get_template(filename: str = 'lmp.jinja2', name='lammps_npt.jinja2') → None[source]

Get the template file.

Parameters

filename: str, optional

The name where to write the template to.

name: str, optional

The name of the template, by default “lammps_npt.jinja2”

Returns

lmp_directory: str = PosixPath('$nwd$/lammps')
lmp_exe: str = 'lmp_serial'
lmp_params: str
lmp_template: str
log_file: Path = PosixPath('$nwd$/NPT.log')
run()[source]
skiprun: bool = False

ipsuite.calculators.orca module

class ipsuite.calculators.orca.OrcaSinglePoint(*args, **kwargs)[source]

Bases: IPSNode

Node for running ORCA Single point calculations.

Parameters

orca_shellstr, default=None

The cmd to run orca. If None, the environment variable IPSUITE_ORCA_SHELL is used.

data: list[Atoms]
property frames: list[Atoms]
get_calculator(directory: str = None)[source]
orca_directory: Path = PosixPath('$nwd$/orca')
orca_shell: str | None = None
orcablocks: str = '%pal nprocs 16 end'
orcasimpleinput: str = 'B3LYP def2-TZVP'
output_file: str = PosixPath('$nwd$/structures.h5')
run()[source]

ipsuite.calculators.xtb module

class ipsuite.calculators.xtb.xTBSinglePoint(*args, **kwargs)[source]

Bases: ProcessAtoms

Node for labeling date with xTB and obtaining ASE calculators.

Installation: conda install conda-forge::tblite-python

Attributes

method: str

xTB method to be used. Only “GFN1-xTB” supports PBC.

accuracy: float = 1.0
charge: int = None
data: list[ase.Atoms]
electronic_temperature: float = 300.0
get_calculator(**kwargs)[source]

Get an xtb ase calculator.

initial_guess: str = 'sad'
max_iterations: int = 250
method: str = 'GFN1-xTB'
mixer_damping: float = 0.4
multiplicity: int = None
run()[source]
spin_polarization: float = None

Module contents

Calculators can be used for labeling a given set of data and running molecular dynamics. For all cases, ASE calculators are used.

class ipsuite.calculators.ASEGeoOpt(*args, **kwargs)[source]

Bases: IPSNode

Geometry optimization using ASE optimizers.

Parameters

datalist[ase.Atoms]

Input atomic configurations to optimize.

data_idint, default=-1

Index of the configuration to optimize from the data list.

modelNodeWithCalculator

A model node that implements ‘get_calculator’ method.

optimizerstr, default=”FIRE”

Name of the ASE optimizer to use (e.g., “FIRE”, “BFGS”, “LBFGS”).

repeattuple[int, int, int], default=(1, 1, 1)

Repeat the unit cell in x, y, z directions.

run_kwargsdict, default={“fmax”: 0.05}

Keyword arguments passed to the optimizer run method.

init_kwargsdict, default={}

Keyword arguments passed to the optimizer initialization.

maxstepint, optional

Maximum number of optimization steps to perform.

sampling_rateint, default=1

How often to sample atomic configurations during optimization.

dump_rateint, default=1000

How many configurations to cache before writing to file.

checkslist, optional

List of check nodes for monitoring optimization progress.

constraintslist, optional

List of constraint nodes to apply during optimization.

Attributes

frameslist[ase.Atoms]

Optimized atomic configurations from the trajectory.

traj_filePath

Path to the HDF5 trajectory file containing optimization steps.

Examples

>>> model = ips.MACEMPModel()
>>> with project:
...     data = ips.AddData(file="ethanol.xyz")
...     geoopt = ips.ASEGeoOpt(data=data.frames, model=model, optimizer="FIRE")
>>> project.repro()
>>> print(f"Optimization completed with {len(geoopt.frames)} steps")
Optimization completed with 74 steps
checks: list = None
constraints: list = None
data: List[Atoms]
data_id: int = -1
dump_rate: int = 1000
property frames: List[Atoms]
get_atoms() → Atoms[source]
init_kwargs: dict
log_file: Path = PosixPath('$nwd$/opt.log')
maxstep: int = None
model: NodeWithCalculator
model_outs: Path = PosixPath('$nwd$/model_outs')
optimizer: str = 'FIRE'
repeat: tuple[int, int, int] = (1, 1, 1)
run()[source]
run_kwargs: dict
sampling_rate: int = 1
traj_file: Path = PosixPath('$nwd$/structures.h5')
class ipsuite.calculators.CP2KSinglePoint(*args, **kwargs)[source]

Bases: IPSNode

Node for running CP2K Single point calculations.

Parameters

cp2k_shellstr, default=None

The cmd to run cp2k. If None, the environment variable IPSUITE_CP2K_SHELL is used.

cp2k_paramsstr

The path to the cp2k yaml input file. cp2k-input-tools is used to generate the input file from the yaml file.

cp2k_fileslist[str]

Additional dependencies for the cp2k calculation.

wfn_restart_filestr, optional

The path to the wfn restart file.

wfn_restart_nodestr, optional

A cp2k Node that has a wfn restart file.

cp2k_directory: Path = PosixPath('$nwd$/cp2k')
cp2k_files: list[str] = None
cp2k_params: str = 'cp2k.yaml'
cp2k_shell: str | None = None
data: list[Atoms]
failed_configs: dict = NOT_AVAILABLE
failure_policy: Literal['skip', 'fail'] = 'fail'
property frames: list[Atoms]
get_calculator(directory: str | None = None, idx: int = 0)[source]
get_input_script()[source]

Return the input script.

output_file: Path = PosixPath('$nwd$/structures.h5')
run()[source]

ZnTrack run method.

Raises

RuntimeError

If the cp2k_shell is not set.

wfn_restart_file: str = None
wfn_restart_node: Node = None
class ipsuite.calculators.EMTSinglePoint[source]

Bases: object

This is a testing Node! It uses ASE’S EMT calculator with default arguments. The calculator accept all elements and implements energy, forces, making it very useful for creating dummy data.

get_calculator(**kwargs)[source]

Get an EMT ase calculator.

class ipsuite.calculators.LJSinglePoint(epsilon: float = 1.0, sigma: float = 1.0, rc: float = 10.0)[source]

Bases: object

This is a testing Node! It uses ASE’S Lennard-Jones calculator with default arguments. The calculator accept all elements and implements energy, forces and stress, making it very useful for creating dummy data.

epsilon: float = 1.0
get_calculator(**kwargs)[source]

Get an LJ ase calculator.

rc: float = 10.0
sigma: float = 1.0
class ipsuite.calculators.LammpsSimulator(*args, **kwargs)[source]

Bases: ProcessSingleAtom

Can perform LAMMPS Simulations.

Parameters

lmp_exe: str

This is the name or path of the LAMMPS executable. Either path to executable, “lmp” or “lamp_<machine>”. See https://docs.lammps.org/Run_basics.html for more information

lmp_params: str

Path to file. To be able to change parameters with DVC and not have to change them manually in the input script, a params file in yaml format and corresponding template file must be provided.

lmp_template: str

Path to file. In combination with the params file this will be the input script for the LAMMPS simulation

skiprun: bool, optional

Whether to skip running LAMMPS or not, by default False

Returns

None

This function does not return anything. Instead, it creates a LAMMPS input script based on the specified template and parameter files, runs the LAMMPS simulation using the specified executable, and saves the simulation output to the specified directory.

create_input_script()[source]
data: List[ase.Atoms]
dump_file: Path = PosixPath('$nwd$/NPT.lammpstraj')
fill_atoms_with_life()[source]
property frames: list[Atoms]
get_atoms()[source]
classmethod get_template(filename: str = 'lmp.jinja2', name='lammps_npt.jinja2') → None[source]

Get the template file.

Parameters

filename: str, optional

The name where to write the template to.

name: str, optional

The name of the template, by default “lammps_npt.jinja2”

Returns

lmp_directory: str = PosixPath('$nwd$/lammps')
lmp_exe: str = 'lmp_serial'
lmp_params: str
lmp_template: str
log_file: Path = PosixPath('$nwd$/NPT.log')
run()[source]
skiprun: bool = False
class ipsuite.calculators.MixCalculator(*args, **kwargs)[source]

Bases: IPSNode

Combine multiple models or calculators into one.

Attributes:
calculators: list[NodeWithCalculator]

List of calculators to combine.

method: str

choose from “mean” or “sum” to combine the calculators.

calculators: list[NodeWithCalculator]
get_calculator(**kwargs) → Calculator[source]

Property to return a model specific ase calculator object.

Returns

calc:

ase calculator object

method: str = 'sum'
run() → None[source]
class ipsuite.calculators.OrcaSinglePoint(*args, **kwargs)[source]

Bases: IPSNode

Node for running ORCA Single point calculations.

Parameters

orca_shellstr, default=None

The cmd to run orca. If None, the environment variable IPSUITE_ORCA_SHELL is used.

data: list[Atoms]
property frames: list[Atoms]
get_calculator(directory: str = None)[source]
orca_directory: Path = PosixPath('$nwd$/orca')
orca_shell: str | None = None
orcablocks: str = '%pal nprocs 16 end'
orcasimpleinput: str = 'B3LYP def2-TZVP'
output_file: str = PosixPath('$nwd$/structures.h5')
run()[source]
class ipsuite.calculators.PlumedModel(*args, **kwargs)[source]

Bases: IPSNode

Plumed interface.

Parameters

data: list[ase.Atoms]

List of ase atoms objects used to initialize the calculator.

data_id: int

Index of the ase atoms object to use for initialization.

model: NodeWithCalculator

The node that provides the calculator to compute unbiased energy and forces.

config: str | Path

Path to the plumed input file.

temperature: float

Temperature of the simulation in Kelvin.

timestep: float

Timestep of the simulation in fs.

Example

An example config file for plumed can look like this:

hoh-c: DISTANCE ATOMS=48,3
c-r1: DISTANCE ATOMS=2,3
metad: METAD ARG=hoh-c,c-r1 PACE=100 HEIGHT=0.75 SIGMA=0.5,0.5                BIASFACTOR=10 TEMP=400 FILE=HILLS GRID_MIN=1.15,1.15                GRID_MAX=8.0,8.0 GRID_BIN=200,200
PRINT ARG=hoh-c,c-r1 STRIDE=10 FILE=COLVAR

References

[1] Plumed manual: https://www.plumed.org/doc-master/user-doc/html/index.html [2] Plumed : https://www.plumed.org/

config: str | Path
data: list[Atoms]
data_id: int = -1
get_calculator(directory: str | Path) → Plumed[source]
model: NodeWithCalculator
run()[source]
temperature: float
timestep: float
class ipsuite.calculators.xTBSinglePoint(*args, **kwargs)[source]

Bases: ProcessAtoms

Node for labeling date with xTB and obtaining ASE calculators.

Installation: conda install conda-forge::tblite-python

Attributes

method: str

xTB method to be used. Only “GFN1-xTB” supports PBC.

accuracy: float = 1.0
charge: int = None
data: list[ase.Atoms]
electronic_temperature: float = 300.0
get_calculator(**kwargs)[source]

Get an xtb ase calculator.

initial_guess: str = 'sad'
max_iterations: int = 250
method: str = 'GFN1-xTB'
mixer_damping: float = 0.4
multiplicity: int = None
run()[source]
spin_polarization: float = None