class NalaSingleModelTagger(Tagger):
def __init__(self,
class_id=MUT_CLASS_ID,
bin_model=pkg_resources.resource_filename(
'nala.data', 'all3_model'),
features_pipeline=None,
execute_pipeline=True,
execute_pp=True,
keep_silent=True,
keep_genetic_markers=True,
keep_unnumbered=True,
keep_rs_ids=True):
super().__init__([class_id])
self.class_id = class_id
self.bin_model = bin_model
self.features_pipeline = features_pipeline if features_pipeline else get_prepare_pipeline_for_best_model(
)
self.execute_pipeline = execute_pipeline
# ---
self.crf = PyCRFSuite(model_file=self.bin_model)
self.post = None
if execute_pp:
self.post = PostProcessing(
keep_silent=keep_silent,
keep_genetic_markers=keep_genetic_markers,
keep_unnumbered=keep_unnumbered,
keep_rs_ids=keep_rs_ids)
def tag(self, dataset, class_id=None):
class_id = self.class_id if class_id is None else class_id
if self.execute_pipeline:
self.features_pipeline.execute(dataset)
self.crf.annotate(dataset, class_id)
if self.post:
self.post.process(dataset, class_id=class_id)
class NalaMultipleModelTagger(Tagger):
def __init__(
self,
class_id=MUT_CLASS_ID,
st_model=pkg_resources.resource_filename('nala.data', 'st_model'),
all3_model=pkg_resources.resource_filename('nala.data',
'all3_model'),
features_pipeline=None,
execute_pp=True,
keep_silent=True,
keep_genetic_markers=True,
keep_unnumbered=True,
keep_rs_ids=True):
super().__init__([class_id])
tagger1 = NalaSingleModelTagger(class_id, st_model, features_pipeline)
tagger2 = NalaSingleModelTagger(class_id,
all3_model,
tagger1.features_pipeline,
execute_pipeline=False)
self.tagger = MultipleModelTagger(tagger1, tagger2, [class_id])
# ---
self.post = None
if execute_pp:
self.post = PostProcessing(
keep_silent=keep_silent,
keep_genetic_markers=keep_genetic_markers,
keep_unnumbered=keep_unnumbered,
keep_rs_ids=keep_rs_ids)
def tag(self, dataset, class_id=None):
class_id = self.class_id if class_id is None else class_id
self.tagger.tag(dataset, class_id)
if self.post:
self.post.process(dataset, class_id=class_id)