Trajectories module

Low-level tools for handling trajectory files.

Python library usage

Functions to help generate and manipulate trajectories

esteem.trajectories.generate_md_trajectory(model, seed, target, traj_label, trajname_suffix, md_func, count_snaps, count_equil, md_steps, md_timestep, temp, calc_params, constraints=None, dynamics=None, singlepoint_func=None)

Runs equilibration followed by statistics generation MD for a given model.

This generic function gets called by both AIMD and by MLMD.

model: ASE Atoms

Initial geometry for the MD trajectory

seed: str

String containing name of molecule and excited state index

target: int or None

Excited state index

traj_label: character or str

String labelling trajectory (usually A,B,C..)

seed: str

String containing name of molecule and excited state index

trajname_suffix: str

String appended to seed_state_traj to give full trajectory filename

md_func: function

Wrapper function that runs md_steps steps of MD on model

count_snaps: int

Number of snapshot runs

count_equil: int

Number of equilibration runs

md_steps: int

Number of molecular dynamics timesteps in each run above

md_timestep: float

Molecular dynamics timestep within the runs

temp: float

Thermostat temperature (NVT ensemble)

calc_params:

Control-parameters for the wrapper (for QM this is the basis, functional, and target; for ML this is the calculator seed, suffix, prefix and target)

constraints: wrapper-dependent type

Variables controlling the constraints

dynamics: wrapper-dependent type

Variables controlling the dynamics

esteem.trajectories.cycle_step_labels_and_restarts(seed, traj_label, prevdir, currdir, nextdir, prevtarg, currtarg, nexttarg, prevstep, currstep, nextstep, count, prevcount, rst_ext=None)

Move restart files around so that each step continues from the previous one

esteem.trajectories.find_initial_geometry(seed, geom_opt_func, calc_params, which_traj=None)

Obtains a suitable initial geometry for the current seed and state. Optimises it if not present.

seed: str

String indicating name of molecule, used to find xyz file

geom_opt_func: function

Wrapper function that runs a geometry optimisation

calc_params:

Control-parameters for the wrapper (for QM this is the basis, functional, and target; for ML this is the calculator seed, suffix, prefix and target)

esteem.trajectories.recalculate_trajectory(seed, target, traj_label, trajname_suffix, input_target, input_suffix, singlepoint_func, calc_params, input_traj_range=None, output_traj_offset=0, solvent=None)

Loads snapshots from a trajectory and recalculates the energy and forces with the current settings seed: str

String containing name of molecule and excited state index

target: int or None

Excited state index

traj_label: character or str

String labelling trajectory (usually A,B,C..)

esteem.trajectories.cycle_restarts(seed, traj_label, trajname_suffix, prevtarg, currtarg, prevstep, currstep)

Move db files around so that each step continues from the previous one

esteem.trajectories.get_trajectory_list(ntraj)

Returns a list of characters to be used as trajectory labels Currently ABCDEDFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz (52)

esteem.trajectories.merge_traj(trajnames, trajfile)

Merges a list of trajectories supplied as a list of filenames, and writes the result to another trajectory supplied as a filename

esteem.trajectories.split_traj(input_traj_file, output_trajs=None, ntrajs=None, randomise=False, start=0, end=-1)

Takes in a trajectory filename, and splits it into sections, randomly or sequentially

esteem.trajectories.fit_atom_energies_to_traj(traj, atom_traj_file)

Uses linear regression to best fit atom energies to trajectory with varying numbers of atoms of each species

esteem.trajectories.subtract_atom_energies_from_traj(traj, atom_traj, trajout)

Subtracts the energies associated with isolated atoms from the total energies in a trajectory

esteem.trajectories.compare_traj_to_traj(trajx, trajy, plot_file=None, xlabel=None, ylabel=None)

Compares two trajectories to each other and calculates statistics for how much they differ in energy and force

trajx: ASE Trajectory

Trajectory whose energy is plotted along x-axis

trajy: ASE Trajectory

Trajectory whose energy is plotted along y-axis

plot_file:

Filename to write plot image to.

xlabel, ylabel: str

Axis labels for plots.