Nanocluster Workflow

This workflow simply relaxes the structure of a set of nanoclusters

If atomic energies (atomic_energies) are given, the cohesive energies of the relaxed nanoclusters are computed. Otherwise, the total energy is used to compare nanoclusters of the same composition.

Currently, stable nanoclusters have to be picked manually once the workflow has finished, but this can be automated in the future.

This is an example how to use the nanocluster workflow:

from fireworks import LaunchPad, Workflow
import pathlib
import os,time, sys
import logging
import ase
from scipy.spatial.distance import pdist
import getpass

# internal modules
from critcatworks.workflows import get_nanoclusters_workflow
from critcatworks.database import mylaunchpad

def read_structures_locally(path):
    structures = []
    path = pathlib.Path(path).resolve()
    for idx, p in enumerate(pathlib.Path(path).iterdir()):
        if p.is_file():
            logging.debug("nanocluster path " + str(p) + " stem " + str(p.stem))
            try:
                atoms = ase.io.read(str(p))
                # set cell to 2.5 the diameter
                pos = atoms.get_positions()
                pdist(pos)
                diameter = pdist(pos).max()
                mpl = 2.5
                
                atoms.set_cell([diameter * mpl, diameter * mpl, diameter * mpl])
                structures.append(atoms)
                logging.debug(atoms)
            except ValueError:
                logging.warning("WARNING: file type not understood" + str(p) )
                continue
            except:
                logging.error("Unexpected error:", sys.exc_info()[0])
    return structures

if __name__ == "__main__":
    IS_QUEUE = True
    USERNAME = "myusername"
    PASSWORD = getpass.getpass()
    if IS_QUEUE:
        logging.basicConfig(format='%(name)s:%(levelname)s:%(message)s', level=logging.INFO)
    else:
        logdir = str(pathlib.Path(".").resolve())
        logging.basicConfig(filename = logdir + "/nanocluster_workflow.log", level=logging.INFO)

    # set up the LaunchPad and reset it
    launchpad = mylaunchpad.create_launchpad(USERNAME, PASSWORD, lpadname = "mjfireworkstriton")
    #launchpad.reset('', require_password=False)

    structures = read_structures_locally("./ptx55")
    wf = get_nanoclusters_workflow(username = "myusername", password = PASSWORD,
        source_path = None,
        template_path = str(pathlib.Path("templates/triton_gopt.inp").resolve()), 
        #worker_target_path = "/wrk/jagermar/DONOTREMOVE/workflow_runs/nanoclusters/production/ptcu_selected_clusters",
        worker_target_path = "/scratch/work/jagerm1/workflow_runs/nanoclusters/production/selected_ptni_clusters",
        structures = structures,
        extdb_ids = None,
        skip_dft = False,
        extdb_connect = {"db_name": "ncdb"},
        )

    # store workflow 
    launchpad.add_wf(wf)

As in most workflows, the initial structures can be read in three ways:

structures

list of ase.Atoms objects

extdb_ids

list of unique identifiers of the simulations collection. The simulations in the database need to have the correct form

source_path

absolute path on the computing resource to the directory where to read the structures from

There are a few workflow-specific arguments:

atomic_energies (dict)

used for computing cohesive energies, not required

The other arguments are common to all workflows, such as the username and password for the database or the path to the DFT template.