def __init__(self,
debug=False,
label=None,
excluded_token_types=(Tokenizer.TYPES["WHITESPACE"],
Tokenizer.TYPES["NEWLINE"]),
case_sensitive=True,
untagged_rule_name=None,
no_rules_rule_name=None,
excluded_rule_name=None,
return_untagged_tags=False,
return_no_rules_tags=False,
return_excluded_tags=False,
return_included_tags=False,
text_name=None,
model_path=None):
super(TopicModelTagger,
self).__init__(debug=debug,
label=label,
excluded_token_types=excluded_token_types,
case_sensitive=case_sensitive,
untagged_rule_name=untagged_rule_name,
no_rules_rule_name=no_rules_rule_name,
excluded_rule_name=excluded_rule_name,
return_untagged_tags=return_untagged_tags,
return_no_rules_tags=return_no_rules_tags,
return_excluded_tags=return_excluded_tags,
return_included_tags=return_included_tags)
if text_name is None:
raise ValueError("No text_name given.")
self.text_name = text_name
if model_path is None:
raise ValueError("No path to topic model data given.")
self.model_path = model_path
# Instantiate a TopicModelDictionary to retrieve information about the topic model.
self.model = TopicModelDictionary(debug=self.debug,
model_path=model_path)
self.untagged_rule_name = untagged_rule_name
self.text_name = None
# Ramp setup
self.num_ramp_steps = None
self.min_ramped_value = -1
self.max_ramped_value = -1
# Get topic "rules".
self._rules = OrderedDict()
# Append additional info to self._full_label.
self._full_label += ".".join(
[os.path.basename(self.model_path), self.text_name])
