)
parser.add_argument("output_dir", default="PDBs")
args = parser.parse_args()
if -1 in args.states:
print "Ripping PDBs for all states"
args.states = "all"
if args.conformations_per_state == -1:
print "Getting all PDBs for each state"
args.conformations_per_state = "all"
atom_indices = np.loadtxt(args.lprmsd_atom_indices, np.int)
assignments = Serializer.LoadData(args.assignments)
project = Project.load_from_hdf(args.project)
if args.lprmsd_permute_atoms == "None":
permute_indices = None
else:
permute_indices = ReadPermFile(args.lprmsd_permute_atoms)
if args.lprmsd_alt_indices == "None":
alt_indices = None
else:
alt_indices = np.loadtxt(args.lprmsd_alt_indices, np.int)
run(
project,
assignments,
args.conformations_per_state,
default='Data/Gens.lh5')
parser.add_argument('output_dir', default='PDBs')
args = parser.parse_args()
if -1 in args.states:
print "Ripping PDBs for all states"
args.states = 'all'
if args.conformations_per_state == -1:
print "Getting all PDBs for each state"
args.conformations_per_state = 'all'
atom_indices = np.loadtxt(args.lprmsd_atom_indices, np.int)
assignments = Serializer.LoadData(args.assignments)
project = Project.load_from_hdf(args.project)
if args.lprmsd_permute_atoms == 'None':
permute_indices = None
else:
permute_indices = ReadPermFile(args.lprmsd_permute_atoms)
if args.lprmsd_alt_indices == 'None':
alt_indices = None
else:
alt_indices = np.loadtxt(args.lprmsd_alt_indices, np.int)
run(project, assignments, args.conformations_per_state, args.states,
args.output_dir, args.generators, atom_indices, permute_indices,
alt_indices, args.total_memory_gb)
def main(args, metric, logger):
project = Project.load_from_hdf(args.project)
if not os.path.exists(args.generators):
raise IOError('Could not open generators')
generators = os.path.abspath(args.generators)
output_dir = os.path.abspath(args.output_dir)
# connect to the workers
try:
json_file = client_json_file(args.profile, args.cluster_id)
client = parallel.Client(json_file, timeout=2)
except parallel.error.TimeoutError as exception:
msg = '\nparallel.error.TimeoutError: ' + str(exception)
msg += "\n\nPerhaps you didn't start a controller?\n"
msg += "(hint, use ipcluster start)"
print >> sys.stderr, msg
sys.exit(1)
lview = client.load_balanced_view()
# partition the frames into a bunch of vtrajs
all_vtrajs = local.partition(project, args.chunk_size)
# initialze the containers to save to disk
f_assignments, f_distances = local.setup_containers(
output_dir, project, all_vtrajs)
# get the chunks that have not been computed yet
valid_indices = np.where(
f_assignments.root.completed_vtrajs[:] == False)[0]
remaining_vtrajs = np.array(all_vtrajs)[valid_indices].tolist()
logger.info('%d/%d jobs remaining', len(remaining_vtrajs), len(all_vtrajs))
# send the workers the files they need to get started
# dview.apply_sync(remote.load_gens, generators, project['ConfFilename'],
# metric)
# get the workers going
n_jobs = len(remaining_vtrajs)
amr = lview.map(remote.assign,
remaining_vtrajs, [generators] * n_jobs, [metric] * n_jobs,
chunksize=1)
pending = set(amr.msg_ids)
while pending:
client.wait(pending, 1e-3)
# finished is the set of msg_ids that are complete
finished = pending.difference(client.outstanding)
# update pending to exclude those that just finished
pending = pending.difference(finished)
for msg_id in finished:
# we know these are done, so don't worry about blocking
async = client.get_result(msg_id)
assignments, distances, chunk = async .result[0]
vtraj_id = local.save(f_assignments, f_distances, assignments,
distances, chunk)
log_status(logger, len(pending), n_jobs, vtraj_id, async)
f_assignments.close()
f_distances.close()
logger.info('All done, exiting.')
