abics.mc_mpi module

class abics.mc_mpi.ParallelMC(comm, MCalgo, model, configs, kTs, subdirs=True)[source]

Bases: object

Parameters

  • comm (comm world) – MPI communicator

  • MCalgo (object for MonteCarlo algorithm) – MonteCarlo algorithm

  • model (dft_latgas object) – DFT lattice gas mapping model

  • configs (config object) – Configurations

  • kTs (list) – Temperature list

  • subdirs (boolean) – if true, working directory for this rank is made

__init__(comm, MCalgo, model, configs, kTs, subdirs=True)[source]
Parameters

  • comm (comm world) – MPI communicator

  • MCalgo (object for MonteCarlo algorithm) – MonteCarlo algorithm

  • model (dft_latgas object) – DFT lattice gas mapping model

  • configs (config object) – Configurations

  • kTs (list) – Temperature list

  • subdirs (boolean) – if true, working directory for this rank is made

run(nsteps, sample_frequency, observer=<abics.mc.observer_base object>)[source]
Parameters

  • nsteps (int) – Number of Monte Carlo steps for running.

  • sample_frequency (int) – Number of Monte Carlo steps for running.

  • observer (observer object) –

Returns

obs_buffer – Observables

Return type

numpy array

class abics.mc_mpi.RXParams[source]

Bases: object

Parameter set for replica exchange Monte Carlo

nreplicas

The number of replicas

Type

int

nprocs_per_replica

The number of processes which a replica uses

Type

int

kTstart

The lower bound of temperature range

Type

float

kTend

The upper bound of temperature range

Type

float

nsteps

The number of MC steps

Type

int

RXtrial_frequency

The number of MC steps between replica exchange operations

sample_frequency

The number of MC steps between measurements observables

print_frequency

The number of MC steps between show information

reload

Whether to restart simulation or not

Type

bool

seed

The seed of the random number generator If 0, some random number is used (e.g., system time or some random noise).

Type

int

classmethod from_dict(d)[source]

Read information from dictionary

Parameters

d (dict) – Dictionary including parameters for replica exchange Monte Carlo method

Returns

params – self

Return type

DFTParams object

classmethod from_toml(fname)[source]

Read information from toml file

Parameters

f (str) – The name of input toml File

Returns

DFTParams – self

Return type

DFTParams object

abics.mc_mpi.RX_MPI_init(rxparams)[source]
Parameters

  • rxparams (RXParams) – Parameters for replica exchange Monte Carlo method.

  • Returns –

  • ------- –

  • comm (comm world) – MPI communicator

class abics.mc_mpi.TemperatureRX_MPI(comm, MCalgo, model, configs, kTs, subdirs=True)[source]

Bases: abics.mc_mpi.ParallelMC

Parameters

  • comm (comm world) – MPI communicator

  • MCalgo (object for MonteCarlo algorithm) – MonteCarlo algorithm

  • model (dft_latgas) – DFT lattice gas mapping model

  • configs (config object) – Configuration

  • kTs (list) – Temperature list

  • subdirs (boolean) – If true, working directory for this rank is made

Xtrial(XCscheme=-1)[source]
Parameters

  • XCscheme (int) –

  • Returns –

  • ------- –

__init__(comm, MCalgo, model, configs, kTs, subdirs=True)[source]
Parameters

  • comm (comm world) – MPI communicator

  • MCalgo (object for MonteCarlo algorithm) – MonteCarlo algorithm

  • model (dft_latgas) – DFT lattice gas mapping model

  • configs (config object) – Configuration

  • kTs (list) – Temperature list

  • subdirs (boolean) – If true, working directory for this rank is made

find_procrank_from_Trank(Trank)[source]
Parameters

Trank (int) – Temperature rank

Returns

procrank

Return type

int

reload()[source]
run(nsteps, RXtrial_frequency, sample_frequency=9223372036854775807, print_frequency=9223372036854775807, observer=<abics.mc.observer_base object>, subdirs=True, save_obs=True)[source]
Parameters

  • nsteps (int) – The number of Monte Carlo steps for running.

  • RXtrial_frequency (int) – The number of Monte Carlo steps for replica exchange.

  • sample_frequency (int) – The number of Monte Carlo steps for observation of physical quantities.

  • print_frequency (int) – The number of Monte Carlo steps for saving physical quantities.

  • observer (observer object) –

  • subdirs (boolean) – If true, working directory for this rank is made

  • save_obs (boolean) –

Returns

obs_list – Observation list

Return type

list