All plugins

aiida-QECpWorkChain

Car-Parrinello Work Chain with Quantum Espresso. This workchain does a full CP simulation, from the choice of the electronic mass and the timestep, to the choice of the best parallelization options, and then it does the NPT equilibration and a final NVE simulation at the prescribed P and T. Automates as much as possible.

General information

Registry checks

Plugins provided

Entry points

WorkChains and work functions aiida.workflows

• qecpworkchain.cp

  class: aiida_QECpWorkChain.workflow:CpWorkChain

  No description available

  Input	Required	Valid types	Description
  cp_code	true	Code
  cp_resources_cg_list	true	List	Same as cp_resources_cp_list but when doing a CG. The CG uses a different amount of resource and can use no band or task group parallelization.
  cp_resources_cp_list	true	List	List of dictionary like the following: { 'resources' : { 'num_machines' : 2, 'num_mpiprocs_per_machine' : 48, }, 'wallclock' : 3600, 'queue' : 'queue_name', 'account': 'account_name', } c,porturrently only the first element of the list is used. 'wallclock' is the maximum time that can be requested to the scheduler. This code can decide to ask for less.
  ecutwfc	true	Float	wavefunction cutoff (Ry), like in the QE input
  pseudo_family	true	Str	pseudopotential family to use, as in usual aiida operations
  pw_code	true	Code	input pw code (used to calculate force ratio)
  pw_resources_list	true	List	Same as cp_resources_cp_list but for pw.x code.
  structure	true	StructureData, TrajectoryData	Input structure. If a trajectory is given, the workchain will use its last step to start the CG. If velocities are present, they will be used to initialize the simulation. Note that if you use a trajectory, usually kind information (like mass) are not included, so default values will be used. If you want to include kind information or override those provided with the input structure, use the input structure_kinds
  thermobarostat_points	true	List	List of dicts, each with the format [ { "temperature_K": 1000, "pressure_KBar": 10 , "equilibration_time_ps": 5.0, "thermostat_time_ps": 5.0} ]. The simulation will loop over this list of dictionaries, in the same order, equilibrating for the specified time at the given P,T point. Every point is repeated if the average T and P are not within the specified ranges
  additional_parameters_cp	false	Dict	parameters that will be included in the settings input of the QE CP plugin. These settings will be added on top of the default one. Same format as plugin input
  adjust_ionic_mass	false	Bool	Multiply the mass of the ions by the corresponding force ration between the cp forces and pw forces -- that is less than 1. Note that averages of static properties do not depend on the ionic masses.
  benchmark_emass_dt_walltime_s	false	Float	same as benchmark_parallel_walltime_s but for dermining the best electronic mass and timestep.
  benchmark_parallel_walltime_s	false	Float	time requested to the scheduler during the test for finding the best parallelization parameters.
  cmdline_cp	false	List, NoneType	additional command line parameters of the cp verlet caclulations only (for example parallelization options)
  default_nose_frequency	false	Float	default nose frequency when a frequency cannot be estimated from the vibrational spectrum
  dt	false	Float, NoneType	timestep in atomic units, if not automatically chosen.
  dt_start_stop_step	false	List	list of timesteps to try. Timesteps are changed to better integrate the equation of motion. When a new electronic mass is selected by this workchain timesteps are automatically adjusted.
  emass	false	Float, NoneType	electronic mass, atomic mass units, if not automatically chosen
  emass_list	false	List	list of electronic masses to try. The emass is selected in order to satisfy the requested CP/DFT force ratio.
  initial_atomic_velocities_A_ps	false	ArrayData, NoneType	optional input initial velocities in angstrom over picoseconds
  max_slope_const	false	Float	max slope in K/ps of the constant of motion linear fit.
  max_slope_ekinc	false	Float	max slope in K/ps of the ekinc linear fit. If not satisfied try to change emass
  max_slope_min_emass	false	Float	minimum possible value of electronic mass that can be set by the max_slope correction routine. Will not go lower than that.
  max_slope_min_ps	false	Float	minimum required lenght in ps of the last trajectory to do the linear fit on ekinc and const of motion
  metadata	false
  min_traj_steps_vdos	false	Int	minimum number of steps to consider the calculated vibrational spectrum maximum valid, to set the thermostat frequency
  minimum_nose_frequency	false	Float	minimum nose frequency: if the frequency estimated from the vibrational spectrum is lower than this value, this value is used
  nstep_initial_cg	false	Int	At the beginning of the simulation the CP algorithm is not used. This is the number of steps to do using Born-Oppenheimer molecular dynamics algorithm with a conjugate gradient minimization of the electronic ground state.
  nstep_parallel_test	false	Int	the benchmark simulations will be that long, if performed
  number_of_pw_per_trajectory	false	Int	Number of pw submitted for every trajectory during calculation of force ratio.
  nve_required_picoseconds	false	Float	The equilibrated NVE simulation will last at least this number of picoseconds. How much picoseconds do you want?
  pressure_tolerance	false	Float	Pressure tolerance in kBar used to say if the npt is equilibrated. If not setted, use the standard deviation of the P time series
  skip_emass_dt_test	false	Bool
  skip_parallel_test	false	Bool	do not run run benchmarks to discover a good internal Quantum Espresso parallelization scheme for the current system
  skip_thermobarostat	false	Bool
  structure_kinds	false	List, NoneType	These kinds will be used to override or set the masses of the various atomic types. Note that the workflow, if skip_emass_dt_test is True, will calculate the ratio between cp forces and pw forces and adjust the provided masses automatically according to this ratio. So if you provide this input, make sure to set skip_emass_dt_test to True and set also the inputs emass and dt, or "bad things can happen"
  target_force_ratio	false	Float	The forces calculated by the Car-Parrinello method are affected by two types of error: one is due to the oscillations of the electrons around the DFT energy minimum, and the second is due to the finite mass of the electronic fluid that produces a _sistematic_ error in the forces, as if the electrons add mass to the ionic core. This second kind of error is can be controlled by this parameter, that tries to adjust the electronic mass to obtain the desidered ratio between CP forces and true DFT forces. Then you may want to modify the ionic mass to correct the leading factor of this error.
  temperature_tolerance	false	Float	Temperature tolerance in K used to say if the npt is equilibrated. If not setted, use the standard deviation of the T time series
  tempw_initial_random	false	Float, NoneType	If provided, sets the initial temperature when randomly initializing the starting velocities.

  Output	Required	Valid types	Description
  dt	true	Float	The best integration timestep found
  emass	true	Float	The best electronic mass found
  full_traj	true	TrajectoryData	The full merged trajectory of all the simulations
  nve_prod_traj	true	TrajectoryData	The trajectory of the last NVE simulation
  cmdline_cp	false	List	The best command line parameters for the cp code
  kinds	false	List	The kinds for the structure with the modified ionic masses

  Exit status	Message
  1	The process has failed with an unspecified error.
  2	The process failed with legacy failure mode.
  10	The process returned an invalid output.
  11	The process did not register a required output.
  401	The initial cg steps failed. I cannot start to work.
  402	Nose-Hoover thermostat failed.
  403	Final cg after Nose-Hoover failed.
  404	Error in the NVE simulation
  405	The simulations are calculating very expensive random numbers. There is something wrong (cutoff? metal? boo?)
  406	Wrong input parameters
  407	Parallel test was not succesful, maybe there is something more wrong.
  408	Multiple errors in the simulation that cannot fix.
  409	This is a bug in the workchain.
  410	The initial simulation failed. I cannot start to work. Probably the configuration is not suitable for a molecular dynamics run or the trial integration timestep is too big