def corpus_mock(tokens):
corpus = Corpus.from_numpy(
Domain([], metas=[StringVariable("texts")]),
np.empty((len(tokens), 0)),
metas=np.array([[" ".join(t)] for t in tokens]),
)
return corpus
def _send_output(self) -> None:
"""
Create corpus with scores and output it
"""
if self.corpus is None:
self.Outputs.corpus.send(None)
self.Outputs.selected_documents.send(None)
return
scores, labels = self._gather_scores()
if labels:
d = self.corpus.domain
domain = Domain(
d.attributes,
d.class_var,
metas=d.metas + tuple(
ContinuousVariable(get_unique_names(d, l))
for l in labels),
)
out_corpus = Corpus.from_numpy(
domain,
self.corpus.X,
self.corpus.Y,
np.hstack([self.corpus.metas, scores]),
)
Corpus.retain_preprocessing(self.corpus, out_corpus)
else:
out_corpus = self.corpus
self.Outputs.corpus.send(
create_annotated_table(out_corpus, self.selected_rows))
self.Outputs.selected_documents.send(
out_corpus[self.selected_rows] if self.selected_rows else None)
def _preprocess_words(corpus: Corpus, words: List[str],
callback: Callable) -> List[str]:
"""
Corpus's tokens can be preprocessed. Since they will not match correctly
with words preprocessors that change words (e.g. normalization) must
be applied to words too.
"""
# workaround to preprocess words
# TODO: currently preprocessors work only on corpus, when there will be more
# cases like this think about implementation of preprocessors for a list
# of strings
words_feature = StringVariable("words")
words_c = Corpus.from_numpy(
Domain([], metas=[words_feature]),
np.empty((len(words), 0)),
metas=np.array([[w] for w in words]),
text_features=[words_feature],
)
# only transformers and normalizers preprocess on the word level
pps = [
pp for pp in corpus.used_preprocessor.preprocessors
if isinstance(pp, (BaseTransformer, BaseNormalizer))
]
for i, pp in enumerate(pps):
words_c = pp(words_c)
callback((i + 1) / len(pps))
return [w[0] for w in words_c.tokens if len(w)]
def create_corpus(texts: List[str]) -> Corpus:
"""Create sample corpus with texts passed"""
text_var = StringVariable("Text")
domain = Domain([], metas=[text_var])
c = Corpus.from_numpy(
domain,
X=np.empty((len(texts), 0)),
metas=np.array(texts).reshape(-1, 1),
text_features=[text_var],
)
return preprocess.LowercaseTransformer()(c)
def commit(self):
"""Create a new corpus and output it"""
doc_var = StringVariable("Document")
title_var = StringVariable("Title")
domain = Domain([], metas=[title_var, doc_var])
corpus = Corpus.from_numpy(
domain,
np.empty((len(self.texts), 0)),
metas=np.array(self.texts),
text_features=[doc_var],
)
corpus.set_title_variable(title_var)
self.Outputs.corpus.send(corpus)
def _create_simple_data(self) -> None:
"""
Creat a simple dataset with 4 documents. Save it to `self.corpus`.
"""
metas = np.array([
["Lorem ipsum dolor sit amet, consectetur adipiscing elit."],
["Duis viverra elit eu mi blandit, {et} sollicitudin nisi "],
[" a porta\tleo. Duis vitae ultrices massa. Mauris ut pulvinar a"],
[
"tortor. Class (aptent) taciti\nsociosqu ad lit1ora torquent per"
],
])
text_var = StringVariable("text")
domain = Domain([], metas=[text_var])
self.corpus = Corpus.from_numpy(
domain,
X=np.empty((len(metas), 0)),
metas=metas,
text_features=[text_var],
)
def _create_corpus(self) -> Optional[Corpus]:
if len(self.tweets) == 0:
return None
def to_val(attr, val):
if isinstance(attr, DiscreteVariable):
attr.val_from_str_add(val)
return attr.to_val(val)
m = [attr for attr, _ in METAS]
domain = Domain(attributes=[], class_vars=[], metas=m)
metas = np.array(
[
[to_val(attr, t) for (attr, _), t in zip(METAS, ts)]
for ts in self.tweets.values()
],
dtype=object,
)
x = np.empty((len(metas), 0))
return Corpus.from_numpy(domain, x, metas=metas, text_features=self.text_features)
def run(
corpus: Optional[Corpus],
words: Optional[List],
cached_keywords: Dict,
scoring_methods: Set,
scoring_methods_kwargs: Dict,
agg_method: int,
state: TaskState
) -> Results:
results = Results(scores=[], labels=[], all_keywords={})
if not corpus:
return results
# passed by reference (and not copied) - to save partial results
results.all_keywords = cached_keywords
if not scoring_methods:
return results
def callback(i: float, status=""):
state.set_progress_value(i * 100)
if status:
state.set_status(status)
if state.is_interruption_requested():
raise Exception
callback(0, "Calculating...")
scores = {}
documents = corpus.documents
step = 1 / len(scoring_methods)
for method_name, func in ScoringMethods.ITEMS:
if method_name in scoring_methods:
if method_name not in results.all_keywords:
i = len(results.labels)
cb = wrap_callback(callback, start=i * step,
end=(i + 1) * step)
needs_tokens = method_name in ScoringMethods.TOKEN_METHODS
kw = {"progress_callback": cb}
kw.update(scoring_methods_kwargs.get(method_name, {}))
keywords = func(corpus if needs_tokens else documents, **kw)
results.all_keywords[method_name] = keywords
keywords = results.all_keywords[method_name]
scores[method_name] = \
dict(AggregationMethods.aggregate(keywords, agg_method))
results.labels.append(method_name)
scores = pd.DataFrame(scores)
if words:
# Normalize words
for preprocessor in corpus.used_preprocessor.preprocessors:
if isinstance(preprocessor, BaseNormalizer):
dummy = Corpus.from_numpy(
Domain((), metas=[StringVariable("Words")]),
X=np.empty((len(words), 0)),
metas=np.array(words)[:, None]
)
words = list(preprocessor(dummy).tokens.flatten())
# Filter scores using words
existing_words = [w for w in set(words) if w in scores.index]
scores = scores.loc[existing_words] if existing_words \
else scores.iloc[:0]
results.scores = scores.reset_index().sort_values(
by=[results.labels[0], "index"],
ascending=[False, True]
).values.tolist()
return results
