def set_parameters(self, args):
"""Validate and set parameters.
Parameters
----------
args : dict
command-line arguments
Raises
------
ValueError
found invalid parameter(s)
"""
# load arguments
for key, val in vars(args).items():
setattr(self, key, val)
# check input directory and data
if self.input:
if isfile(self.input):
self.input_map = {file2id(self.input): self.input}
elif isdir(self.input):
self.input_map = {
k: join(self.input, v)
for k, v in id2file_map(self.input, ext='tsv').items()
}
else:
raise ValueError(
'Invalid input data file or directory: {}.'.format(
self.input))
if len(self.input_map) == 0:
raise ValueError('No input data are found under: {}.'.format(
self.input))
# check / create output directory
makedirs(self.output, exist_ok=True)
self.prev_map = id2file_map(self.output, 'tsv')
# load configurations
get_config(self, 'evalue', 'analyze.evalue', float)
for key in ('maxhits', 'identity', 'coverage'):
get_config(self, key, 'analyze.{}'.format(key))
for key in ('input_cov', 'self_rank', 'close_size', 'distal_top'):
get_config(self, key, 'grouping.{}'.format(key.replace('_', '')))
for key in ('weighted', 'outliers', 'orphans', 'bandwidth', 'bw_steps',
'low_part', 'noise', 'fixed', 'silhouette', 'self_low'):
get_config(self, key, 'predict.{}'.format(key.replace('_', '')))
# convert boolean values
for key in ('weighted', 'orphans', 'self_low'):
setattr(self, key, arg2bool(getattr(self, key, None)))
# convert fractions to percentages
for metric in ('input_cov', 'noise', 'fixed', 'distal_top'):
val = getattr(self, metric)
if val and val < 1:
setattr(self, metric, val * 100)
# convert distal top to a factor to save compute
self.match_th = 1 - self.distal_top / 100
# force coverage >= 50 to ensure that candidates are sequential
if (self.input_cov or 0) < 50:
raise ValueError('Taxonomy coverage for auto-interence must be at '
'least 50%.')
def set_parameters(self, args):
"""Workflow for validating and setting arguments.
Parameters
----------
args : dict
command-line arguments
"""
# load arguments
for key, val in vars(args).items():
setattr(self, key, val)
# load configurations
for key in ('capital', 'block', 'latin'):
get_config(self, key, f'taxonomy.{key}')
for key in ('retries', 'delay', 'timeout'):
get_config(self, key, f'download.{key}')
for key in ('diamond', 'makeblastdb'):
get_config(self, key, f'program.{key}')
for key in ('threads', 'tmpdir'):
get_config(self, key, f'local.{key}')
# convert boolean values
for key in ('capital', 'latin'):
setattr(self, key, arg2bool(getattr(self, key, None)))
# make temporary directory
if not self.tmpdir:
self.tmpdir = mkdtemp()
setattr(self, 'mkdtemp', True) # mark for cleanup
if not isdir(self.tmpdir):
raise ValueError(f'Invalid temporary directory: {self.tmpdir}.')
# check local executables
for key, exe in {'blast': 'makeblastdb', 'diamond': 'diamond'}.items():
if self.compile in (key, 'both'):
if getattr(self, exe) is None:
setattr(self, exe, exe)
if which(getattr(self, exe)) is None:
raise ValueError(
f'Invalid {exe} executable: {getattr(self, exe)}.')
# determine number of CPUs to use
if self.compile in ('diamond', 'both') and not self.threads:
self.threads = cpu_count()
if self.threads is None:
self.threads = 1
# default protocol
if self.default:
print('The default protocol is selected for database building.')
print('The program will download all protein sequences of NCBI '
'RefSeq genomes of bacteria, archaea, fungi and protozoa, '
'keep one genome per species, plus all NCBI-defined '
'reference and representative genomes.')
self.cats = 'microbe'
self.sample = 1
self.rank = 'species'
self.reference = True
self.representative = True
self.compile = 'diamond'
makedirs(self.output, exist_ok=True)
def test_arg2bool(self):
# true
self.assertTrue(arg2bool(True))
self.assertTrue(arg2bool('True'))
self.assertTrue(arg2bool('Yes'))
self.assertTrue(arg2bool('y'))
self.assertTrue(arg2bool('1'))
# false
self.assertFalse(arg2bool(None))
self.assertFalse(arg2bool(False))
self.assertFalse(arg2bool('False'))
self.assertFalse(arg2bool('No'))
self.assertFalse(arg2bool('n'))
self.assertFalse(arg2bool('0'))
# invalid input
with self.assertRaises(ValueError) as ctx:
arg2bool('test')
msg = 'Boolean value expected.'
self.assertEqual(str(ctx.exception), msg)