Plumed

class mlptrain.sampling.plumed.PlumedAverageCV(name: str, atom_groups: Sequence | None = None)

Bases: _PlumedCV

Class used to initialise a PLUMED collective variable as an average between multiple degrees of freedom

__init__(name: str, atom_groups: Sequence | None = None)

PLUMED collective variable as an average between multiple degrees of freedom (distances, angles, torsions),

e.g. [(0, 1), (2, 3), (4, 5)] gives ζ = 1/3 * (r_01 + r_23 + r_45)

which are used to generate DOFs

class mlptrain.sampling.plumed.PlumedBias(cvs: Sequence[_PlumedCV] | _PlumedCV | None = None, filename: str | None = None)

Bases: ASEConstraint

@DynamicAttrs Class for storing collective variables and parameters used in biased simulations

__init__(cvs: Sequence[_PlumedCV] | _PlumedCV | None = None, filename: str | None = None)

Class for storing collective variables and parameters used in biased simulations, parameters are not initialised with the object and have to be defined seperately using PlumedBias methods.

Can be initialised from a complete PLUMED input file as well.

adjust_forces(atoms, forces) → None: Adjust the forces of the system by adding PLUMED forces

adjust_positions(atoms, newpositions) → None: Method required for ASE but not used in mlp-train

adjust_potential_energy(atoms) → float: Adjust the energy of the system by adding PLUMED bias

property biasfactor_setup: str: String specifying biasfactor in PLUMED input file

property cv_sequence: str: String containing names of all collective variables separated by commas

property from_file: bool: Whether the bias is initialised using PLUMED input file

Initialise PlumedBias for metadynamics active learning by setting the required parameters.

Parameters:

pace – (int) τ_G/dt, interval at which a new gaussian is placed
height – (float) ω, initial height of placed gaussians (in eV). If not supplied will be set to 5*k_B*T, where T is the temperature at which metadynamics active learning is performed
biasfactor – (float) γ, describes how quickly gaussians shrink, larger values make gaussians to be placed less sensitive to the bias potential
cvs – (mlptrain._PlumedCV) Sequence of PLUMED collective variables which will be biased. If this variable is not set, all CVs attached to self will be biased
grid_min – (float) Lower bound of the grid
grid_max – (float) Upper bound of the grid
grid_bin – (float) Number of bins to use for each collective variable, if not specified PLUMED automatically sets bin width to 1/5 of the gaussian width (σ) value

property metad_cv_sequence: str: String containing names of collective variables used in metadynamics separated by commas

property metad_grid_setup: str: String specifying metadynamics grid parameters in PLUMED input file

property metadynamics: bool: True if any parameters required for metadynamics are set

property n_cvs: int: Number of collective variables attached to the bias

property n_metad_cvs: int: Number of collective variables attached to the bias that will be used in metadynamics

strip() → None: Change the bias such that it would only contain the definitions of collective variables and their associated upper and lower walls

property width_sequence: str: String containing width values separated by commas

write_cv_files() → None: Write files attached to the CVs to the current directory

class mlptrain.sampling.plumed.PlumedCNCV(name: str, r_ref: float, d_ref: float = 0, n: int = 6, m: int = 12, atom_groups: Sequence | None = None)

Bases: _PlumedCV

__init__(name: str, r_ref: float, d_ref: float = 0, n: int = 6, m: int = 12, atom_groups: Sequence | None = None)

PLUMED collective variable as a coordination number (CN) between two atoms or groups of atoms

e.g. [(0, 2)] gives ζ =[1-((r_02 - d_ref)/r_ref)^n]/ [1-((r_02 - d_ref)/r_ref)^m]

which corresponds to the CN between atoms 0 and 2. To ensure that the CN has continuous derivatives, we use a rational switching function consistent with PLUMED. More information: https://www.plumed.org/doc-v2.9/user-doc/html/_c_o_o_r_d_i_n_a_t_i_o_n.html

Parameters:

d_ref – (float) Parmeter d_0 in the original switching function. Default = 0.
n – (int) Parameters used to adjust the shape of the function
m – (int) Parameters used to adjust the shape of the function
atom_groups – (Sequence[Sequence[int]]) List of atom index sequences which are used to generate DOFs

class mlptrain.sampling.plumed.PlumedCalculator(calc, input, timestep, atoms=None, kT=1.0, log='', restart=False, use_charge=False, update_charge=False)

Bases: Plumed

Modified ASE calculator, instead of returning biased energies and forces, this calculator computes unbiased energies and forces, and computes PLUMED energy and force biases separately.

calculate(atoms=None, properties=['energy', 'forces', 'energy_bias', 'forces_bias'], system_changes=['positions', 'numbers', 'cell', 'pbc', 'initial_charges', 'initial_magmoms']) → None: Compute the properties and attach them to the results

compute_energy_and_forces(pos, istep) → Tuple: Compute unbiased energies and forces, and PLUMED energy and force biases separately

implemented_properties: List[str] = ['energy', 'forces', 'energy_bias', 'forces_bias']: Properties calculator can handle (energy, forces, …)

class mlptrain.sampling.plumed.PlumedCustomCV(filename: str, component: str | None = None, units: str | None = None)

Bases: _PlumedCV

Class used to initialise a PLUMED collective variable from a file

__init__(filename: str, component: str | None = None, units: str | None = None)

PLUMED collective variable from a file. The file must be written in the style of a PLUMED input file, but only contain input used in the definition of the CV, and the CV itself. There are no constraints on the CVs that can be defined.

e.g. dof0: COM ATOMS=1-5: dof1: CENTER ATOMS=3-7 cv0: CUSTOM ARG=dof0,dof1 FUNC=x^2*y^2 PERIODIC=NO

Parameters:

component – (str) Name of a component of the last CV in the supplied PLUMED input file to use as a collective variable
units – (str) Units of the collective variable, used in plots

class mlptrain.sampling.plumed.PlumedDifferenceCV(name: str, atom_groups: Sequence | None = None)

Bases: _PlumedCV

Class used to initialise a PLUMED collective variable as a difference between two degrees of freedom

__init__(name: str, atom_groups: Sequence | None = None)

PLUMED collective variable as a difference between two degrees of freedom (distances, angles, torsions),

e.g. [(0, 2), (0, 1)] gives ζ = r_02 - r_01

which are used to generate DOFs

mlptrain.sampling.plumed.get_colvar_filename(cv: _PlumedCV, **kwargs) → str: Return the name of the file where the trajectory in terms of collective variable values would be written

mlptrain.sampling.plumed.get_hills_filename(**kwargs) → str: Return the name of the file where a list of deposited gaussians would be written

mlptrain.sampling.plumed.plot_cv1_and_cv2(filenames: Sequence[str], style: str = 'scatter', cvs_units: Sequence[str] | None = None, cvs_limits: Sequence[Sequence[float]] | None = None, label: str | None = None) → None

Plot the trajectory of the system by tracking two collective variables using two colvar files. The function only works for two collective variables.

‘histogram’

Parameters:

cvs_units – Units of the CVs to be plotted
cvs_limits – Min and max limits of the CVs in the plot
label – (str) Label attached to the name of the plot, useful when multiple plots of the same CVs are generated in the same directory

mlptrain.sampling.plumed.plot_cv_versus_time(filename: str, style: str = 'trajectory', time_units: str = 'ps', cv_units: str | None = None, cv_limits: Sequence[float] | None = None, label: str | None = None) → None

Plot a collective variable as a function of time from a given colvar file. Only plot the first collective variable in the colvar file.

Parameters:

time_units – (str) Units of time
cv_units – Units of the CV to be plotted
cv_limits – Min and max limits of the CV in the plot
label – (str) Label attached to the name of the plot, useful when multiple plots of the same CVs are generated in the same directory

mlptrain.sampling.plumed.plumed_setup(bias: PlumedBias, temp: float, interval: int, **kwargs) → List[str]

Generate a list which represents the PLUMED input file

Parameters:: interval – (int) Interval between saving the geometry