Skip to content

Latest commit

 

History

History
102 lines (90 loc) · 4.61 KB

README.md

File metadata and controls

102 lines (90 loc) · 4.61 KB
Raw

Basin Hopping

Requirements

  • ase >= ?.?.?
  • NumPy >= 1.21
  • PyYaml >= 5.1

Usage

Serial

python basin_hopping.py -f ./config/config.yaml -v

Parallel

mpirun -n 4 python basin_hopping.py -f config/config.yaml -v

Examples

Arguments Description
-f config/config.yaml Run the basin hopping algorithm with settings defined in config.yaml
-f config/config.json Run the basin hopping algorithm with settings defined in config.json
-f config/config.yaml -v Turn on log messages
-n 10 Run the basin hopping algorithm for a cluster with 10 atoms
-n 10 -m 500 Run the basin hopping algorithm for 500 steps

Full usage

usage: basin_hopping.py (-f CONFIG | -n CLUSTER_SIZE) [options]

Run the basin hopping algorithm.

required arguments:
  One of these arguments must be present

  -f CONFIG, --config CONFIG
                        The location of the config file (default: None)
  -n CLUSTER_SIZE, --cluster-size CLUSTER_SIZE
                        The size of the cluster (default: None)

optional arguments:
  -r RADIUS, --radius RADIUS
                        Radius of the sphere the initial atoms configuration
                        is uniformly distributed in (default: 1)
  -c MAX_RADIUS, --max-radius MAX_RADIUS
                        The maximum radius of the cube to constrain the atoms
                        in. If not set, no constraint is placed on the atoms
                        (default: None)
  --temperature TEMPERATURE
                        The temperature parameter for the Metropolis
                        acceptance criterion (default: 0.008617330337217212)
  --step-size STEP_SIZE
                        The initial value of the step size (default: 0.5)
  --accept-rate ACCEPT_RATE
                        The desired step acceptance rate (default: 0.5)
  --step-size-factor STEP_SIZE_FACTOR
                        The factor to multiply and divide the step size by
                        (default: 0.9)
  --step-size-interval STEP_SIZE_INTERVAL
                        The interval for how often to update the step size
                        (default: 50)

stop conditions:
  -m MAX_STEPS, --max-steps MAX_STEPS
                        The maximum number of steps the algorithm will take
                        (default: 500)
  -ss STOP_STEPS, --stop-steps STOP_STEPS
                        The number of steps, without there being a new
                        minimum, the algorithm will take before stopping. If
                        not set, the algorithm will run for the maximum number
                        of steps (default: None)
  -st STOP_TIME, --stop-time STOP_TIME
                        The maximum amount of time, in seconds, the algorithm
                        will run for before stopping. If not set, the
                        algorithm will run for the maximum number of steps
                        (default: None)

filter arguments:
  -fn, --no-filter      Use no filter (default: None)
  -fs, --filter-significant-figures
                        Use significant figures filter (default: None)
  -fd, --filter-difference
                        Use difference filter (default: None)
  -sf SIGNIFICANT_FIGURES, --significant-figures SIGNIFICANT_FIGURES
                        Significant figures to round the potential energy to
                        to check for uniqueness (default: 2)
  -d DIFFERENCE, --difference DIFFERENCE
                        Minimum potential energy difference between unique
                        local minima to check for uniqueness (default: 0.1)

logging & help:
  -h, --help            show this help message and exit
  -v, --verbose         Print information about each step (default: False)
  -vr, --view-results   Open ase gui after completion (default: False)
  -db DATABASE, --database DATABASE
                        Database file for storing global minima (default:
                        None)
  -tr TRAJECTORY, --trajectory TRAJECTORY
                        Trajectory file for storing local minima (default:
                        None)
  -trf FILTERED_TRAJECTORY, --filtered-trajectory FILTERED_TRAJECTORY
                        Trajectory file for storing filtered local minima. If
                        None, filtered local minima are stored in the original
                        trajectory file (default: None)