def __init__(self, config_file, values=None, logged=False):
"""
reads all options in the configuration file and stores them into a dictionary
checks values and presence of necessary information
instead of passing a configuration file, you can pass the values directly as a dictionary
"""
if not logged:
# this will write a file to cwd
logmethods.initialise("config.log")
self.log = logging.getLogger("config")
self.settings = {}
self.config_script_dir = os.path.dirname(os.path.realpath(sys.argv[0]))
self.tax_string = ""
self.ncbi_tax_db = None
self.genomes_exclude = []
if config_file is None:
if values is not None:
self.settings = values
self.tax_string = ",".join(self.settings["taxonomy_ranks"])
else:
self.log.critical("Please provide either values or a configuration file.")
sys.exit(1)
else:
self._read_file(config_file)
self._check_values()
def __init__(self, db, logged=False):
assert os.path.isfile(db)
self.tax = taxonomy_ncbi.TaxonomyNcbi(db)
self.tree = None
self.leaves = None
self.exc = None
if not logged:
# this will write a file to cwd
logmethods.initialise('Ncbi2Newick.log')
self.log = logging.getLogger('ncbi2newick')
def __init__(self, logged=False):
self.verbose = False
self.variables = {}
self.files = {} # path -> Filestatus
# only needed when a status is loaded from backup
self.failed_files = {}
if not logged:
# this will write a file to cwd
logmethods.initialise('status.log')
self.log = logging.getLogger('status')
def __init__(self, cfg, default_ans=False, backup=False, logged=False):
self.workingdir = os.path.dirname(os.path.realpath(sys.argv[0]))
# some mappings
self.tree_organism_map = {} # node --> list of organisms which were mapped to that node
self.organism_tree_map = {} # organism --> closest node on the tree
self.organism_file_map = {} # organism --> list of files including sequences
self.organisms = set() # all organisms available
self.nodes = [] # selected organisms = nodes on the taxonomic tree
self.genomes_excluded = [] # genomes that have not been considered during computation
self.organisms_invalid = set() # organisms with a lack of mapping // or that have been mapped to the root
self.sqlite_taxonomy = None # SQLite database including the NCBI taxonomy
self.config = cfg # contains all settings
# # # # # # # Advanced training options # # # # # # # # # # # # # #
# learning
self.loss_function = 1 # 0:0/1 loss, 1:path loss
self.z_standardization = 1 # 0:no standardization, 1:z standardization
self.misc_nodes = 1 # 0:no misc nodes, 1:add misc nodes
self.n_frags_per_node = 0.0 # number of fragments per node
self.yes = default_ans # say yes to every question?
self.tree_file = ""
self.backup = backup
self.stat = None
self.backupdir = None
if not logged:
# this will write a file to cwd
logmethods.initialise('train.log')
self.log = logging.getLogger('train') # better root ?
taxdb=self.config.ncbi_tax_db)
if self.yes:
tax_call = "{} -y".format(tax_call)
my_log.debug(tax_call)
success = os.system(tax_call)
if success != 0:
my_log.critical("Error in creating the SQLite database.")
sys.exit(1)
my_log.debug("The NCBI database did not exist - created one: {}".format(self.config.ncbi_tax_db))
self.sqlite_taxonomy = taxonomy_ncbi.TaxonomyNcbi(self.config.ncbi_tax_db)
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument("-c", help="configuration file", action='store', required=True)
parser.add_argument('-y', help="automatically set answers to 'yes'", action='store_true', default=False)
parser.add_argument("-fb", help="backup directory from a previous PPS run, "
"will try to restart from within the pipeline", action='store', default=None)
parser.add_argument('-b', help="write a backup to the project directory", action='store_true', default=False)
args = parser.parse_args()
logmethods.initialise('train.log')
cfg = config.Config(args.c, logged=True)
obj = Train(cfg, args.y, args.b, logged=True)
if args.fb is None:
obj.main_processing()
else:
obj.from_backup(args.fb)