def test_train_rdp_classifier_and_assign_taxonomy_no_model_output(self):
obs = train_rdp_classifier_and_assign_taxonomy(
self.reference_file, self.taxonomy_file, self.test_trained_input)
exp = {'X67228': (
'Bacteria;Proteobacteria;Alphaproteobacteria;Rhizobiales;'
'Rhizobiaceae;Rhizobium', 1.0
)}
self.assertEqual(obs, exp)
def __call__(self, seq_path, result_path=None, log_path=None):
"""Returns dict mapping {seq_id:(taxonomy, confidence)} for
each seq.
Parameters:
seq_path: path to file of sequences
result_path: path to file of results. If specified, dumps the
result to the desired path instead of returning it.
log_path: path to log, which should include dump of params.
"""
tmp_dir = get_qiime_temp_dir()
min_conf = self.Params['Confidence']
training_data_properties_fp = self.Params[
'training_data_properties_fp']
reference_sequences_fp = self.Params['reference_sequences_fp']
id_to_taxonomy_fp = self.Params['id_to_taxonomy_fp']
max_memory = self.Params['max_memory']
seq_file = open(seq_path, 'U')
if reference_sequences_fp and id_to_taxonomy_fp:
# Train and assign taxonomy
taxonomy_file, training_seqs_file = self._generate_training_files()
results = rdp_classifier.train_rdp_classifier_and_assign_taxonomy(
training_seqs_file,
taxonomy_file,
seq_file,
min_confidence=min_conf,
classification_output_fp=result_path,
max_memory=max_memory,
tmp_dir=tmp_dir)
if result_path is None:
results = self._training_set.fix_results(results)
else:
self._training_set.fix_output_file(result_path)
else:
# Just assign taxonomy, using properties file if passed
if training_data_properties_fp:
fix_ranks = False
else:
fix_ranks = True
results = rdp_classifier.assign_taxonomy(
seq_file,
min_confidence=min_conf,
output_fp=result_path,
training_data_fp=training_data_properties_fp,
max_memory=max_memory,
fixrank=fix_ranks,
tmp_dir=tmp_dir)
if log_path:
self.writeLog(log_path)
return results
def __call__(self, seq_path, result_path=None, log_path=None):
"""Returns dict mapping {seq_id:(taxonomy, confidence)} for
each seq.
Parameters:
seq_path: path to file of sequences
result_path: path to file of results. If specified, dumps the
result to the desired path instead of returning it.
log_path: path to log, which should include dump of params.
"""
tmp_dir = get_qiime_temp_dir()
min_conf = self.Params['Confidence']
training_data_properties_fp = self.Params[
'training_data_properties_fp']
reference_sequences_fp = self.Params['reference_sequences_fp']
id_to_taxonomy_fp = self.Params['id_to_taxonomy_fp']
max_memory = self.Params['max_memory']
seq_file = open(seq_path, 'U')
if reference_sequences_fp and id_to_taxonomy_fp:
# Train and assign taxonomy
taxonomy_file, training_seqs_file = self._generate_training_files()
results = rdp_classifier.train_rdp_classifier_and_assign_taxonomy(
training_seqs_file, taxonomy_file, seq_file,
min_confidence=min_conf,
classification_output_fp=result_path,
max_memory=max_memory, tmp_dir=tmp_dir)
if result_path is None:
results = self._training_set.fix_results(results)
else:
self._training_set.fix_output_file(result_path)
else:
# Just assign taxonomy, using properties file if passed
if training_data_properties_fp:
fix_ranks = False
else:
fix_ranks = True
results = rdp_classifier.assign_taxonomy(
seq_file, min_confidence=min_conf, output_fp=result_path,
training_data_fp=training_data_properties_fp,
max_memory=max_memory, fixrank=fix_ranks, tmp_dir=tmp_dir)
if log_path:
self.writeLog(log_path)
return results
