def test_set_log_level(self):
cfg = Config(self.defaults, self.parsed_args)
cfg._set_log_level(20)
self.assertEqual(cfg.log_level(), 20)
cfg._set_log_level('WARNING')
self.assertEqual(cfg.log_level(), 30)
with self.assertRaises(ValueError) as ctx:
cfg._set_log_level('FOO')
self.assertEqual(str(ctx.exception),
"Invalid value for log_level: FOO.")
def main(args=None, loglevel=None):
"""
main entry point to MacSyFinder do some check before to launch :func:`main_search_systems` which is
the real function that perform a search
:param args: the arguments passed on the command line without the program name
:type args: List of string
:param loglevel: the output verbosity
:type loglevel: a positive int or a string among 'DEBUG', 'INFO', 'WARNING', 'ERROR', 'CRITICAL'
"""
args = sys.argv[1:] if args is None else args
parser, parsed_args = parse_args(args)
defaults = MacsyDefaults()
config = Config(defaults, parsed_args)
###########################
# creation of working dir
###########################
working_dir = config.working_dir()
if not os.path.exists(working_dir):
os.makedirs(working_dir)
else:
if os.path.isdir(working_dir):
if os.listdir(working_dir):
raise ValueError(
f"'{working_dir}' already exists and is not a empty")
else:
raise ValueError(
f"'{working_dir}' already exists and is not a directory")
################
# init loggers #
################
macsypy.init_logger(log_file=os.path.join(config.working_dir(),
config.log_file()),
out=not config.mute())
if not loglevel:
# logs are specify from args options
macsypy.logger_set_level(level=config.log_level())
else:
# used by unit tests to mute or unmute logs
macsypy.logger_set_level(level=loglevel)
logger = logging.getLogger('macsypy.macsyfinder')
if parsed_args.list_models:
print(list_models(parsed_args), file=sys.stdout)
sys.exit(0)
else:
if not parsed_args.previous_run and not parsed_args.models:
parser.print_help()
print()
sys.tracebacklimit = 0
raise OptionError(
"argument --models or --previous-run is required.")
elif not parsed_args.previous_run and not parsed_args.sequence_db:
parser.print_help()
print()
sys.tracebacklimit = 0
raise OptionError(
"argument --sequence-db or --previous-run is required.")
elif not parsed_args.previous_run and not parsed_args.db_type:
parser.print_help()
print()
sys.tracebacklimit = 0
raise OptionError(
"argument --db-type or --previous-run is required.")
_log.info(f"command used: {' '.join(sys.argv)}")
models = ModelBank()
genes = GeneBank()
profile_factory = ProfileFactory(config)
macsypy.hit.hit_weight = macsypy.hit.HitWeight(itself=3,
exchangeable=.75,
mandatory=2,
accessory=.25,
neutral=1.5)
logger.info("\n{:#^70}".format(" Searching systems "))
all_systems, rejected_clusters = search_systems(
config, models, genes, profile_factory, logger)
track_multi_systems_hit = HitSystemTracker(all_systems)
if config.db_type() in ('gembase', 'ordered_replicon'):
#############################
# Ordered/Gembase replicons #
#############################
###########################
# select the best systems #
###########################
logger.info("\n{:#^70}".format(" Computing best solutions "))
best_solutions = []
one_best_solution = []
# group systems found by replicon
# before to search best system combination
import time
for rep_name, syst_group in itertools.groupby(
all_systems, key=lambda s: s.replicon_name):
syst_group = list(syst_group)
logger.info(
f"Computing best solutions for {rep_name} (nb of systems {len(syst_group)})"
)
t0 = time.time()
best_sol_4_1_replicon, score = find_best_solutions(syst_group)
t1 = time.time()
logger.info(
f"It took {t1 - t0:.2f}sec to find best solution ({score:.2f}) for replicon {rep_name}"
)
# if several solutions are equivalent same number of system and score is same
# store all equivalent solution in best_solution => all_best_systems
# pick one in one_best_solution => best_systems
best_solutions.extend(best_sol_4_1_replicon)
one_best_solution.append(best_sol_4_1_replicon[0])
##############################
# Write the results in files #
##############################
logger.info("\n{:#^70}".format(" Writing down results "))
system_filename = os.path.join(config.working_dir(),
"all_systems.txt")
tsv_filename = os.path.join(config.working_dir(),
"all_systems.tsv")
with open(system_filename, "w") as sys_file:
systems_to_txt(all_systems, track_multi_systems_hit, sys_file)
with open(tsv_filename, "w") as tsv_file:
systems_to_tsv(all_systems, track_multi_systems_hit, tsv_file)
cluster_filename = os.path.join(config.working_dir(),
"rejected_clusters.txt")
with open(cluster_filename, "w") as clst_file:
rejected_clusters.sort(key=lambda clst: (
clst.replicon_name, clst.model, clst.hits))
rejected_clst_to_txt(rejected_clusters, clst_file)
if not (all_systems or rejected_clusters):
logger.info("No Systems found in this dataset.")
tsv_filename = os.path.join(config.working_dir(),
"all_best_solutions.tsv")
with open(tsv_filename, "w") as tsv_file:
solutions_to_tsv(best_solutions, track_multi_systems_hit,
tsv_file)
tsv_filename = os.path.join(config.working_dir(),
"best_solution.tsv")
with open(tsv_filename, "w") as tsv_file:
# flattern the list and sort it
one_best_solution = [
syst for sol in one_best_solution for syst in sol
]
one_best_solution.sort(
key=lambda syst: (syst.replicon_name, syst.position[0],
syst.model.fqn, -syst.score))
systems_to_tsv(one_best_solution, track_multi_systems_hit,
tsv_file)
else:
#######################
# Unordered replicons #
#######################
##############################
# Write the results in files #
##############################
logger.info("\n{:#^70}".format(" Writing down results "))
system_filename = os.path.join(config.working_dir(),
"all_systems.txt")
with open(system_filename, "w") as sys_file:
likely_systems_to_txt(all_systems, track_multi_systems_hit,
sys_file)
# forbidden = [s for s in all_systems if s.forbidden_occ]
# system_filename = os.path.join(config.working_dir(), "forbidden_components.tsv")
# with open(system_filename, "w") as sys_file:
# likely_systems_to_tsv(forbidden, track_multi_systems_hit, sys_file)
system_filename = os.path.join(config.working_dir(),
"all_systems.tsv")
with open(system_filename, "w") as sys_file:
likely_systems_to_tsv(all_systems, track_multi_systems_hit,
sys_file)
cluster_filename = os.path.join(config.working_dir(),
"uncomplete_systems.txt")
with open(cluster_filename, "w") as clst_file:
unlikely_systems_to_txt(rejected_clusters, clst_file)
if not (all_systems or rejected_clusters):
logger.info("No Systems found in this dataset.")
logger.info("END")
