Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

remove utils.document.deduplicate_annotations in favor of Document.deduplicate_annotations #437

Merged
merged 2 commits into from
Dec 17, 2024
Merged

remove utils.document.deduplicate_annotations in favor of Document.deduplicate_annotations #437

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
4134c3d
remove utils.document.deduplicate_annotations in favor of Document.de…
ArneBinder Dec 10, 2024
7c3b218
cleanup
ArneBinder Dec 17, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
26 changes: 2 additions & 24 deletions src/pytorch_ie/utils/document.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -29,27 +29,6 @@ def deduplicate_annotation_dicts(
D = TypeVar("D", bound=Document)


def deduplicate_annotations(document: D) -> D:
"""Remove duplicate annotations from a document.

Args:
document: The document to remove duplicate annotations from.

Returns:
The document with duplicate annotations removed.
"""
annotation_field_names = [field.name for field in document.annotation_fields()]
doc_dict = document.asdict()
for annotation_field_name in annotation_field_names:
doc_dict[annotation_field_name]["annotations"] = deduplicate_annotation_dicts(
doc_dict[annotation_field_name]["annotations"]
)
doc_dict[annotation_field_name]["predictions"] = deduplicate_annotation_dicts(
doc_dict[annotation_field_name]["predictions"]
)
return type(document).fromdict(doc_dict)


def save_annotation_sources_to_metadata(
document: D,
annotation_id2source: Dict[int, List[str]],
Expand Down Expand Up @@ -124,8 +103,6 @@ def merge_annotations_from_documents(
f"Document IDs do not match: {document.id} and {merged_document.id}"
)

# TODO: add_all_annotations_from_other needs to be fixed! it should return a mapping from
# original annotation *IDs* to new annotations!
# Note: this does not check for duplicates!
added_annotations = merged_document.add_all_annotations_from_other(
other=document, strict=True
Expand All @@ -135,7 +112,8 @@ def merge_annotations_from_documents(
for orig_id, new_annotation in orig_id2new_annotation.items():
added_annotation_id2source_names[new_annotation._id].append(source_name)

merged_document = deduplicate_annotations(merged_document)
# this will remove duplicates. If duplicates have different scores, the one with the highest score will be kept
merged_document = merged_document.deduplicate_annotations()

# save source names in metadata (at key metadata_key_source_annotations / metadata_key_source_predictions
# for each layer in the order of the annotations / predictions)
Expand Down
3 changes: 0 additions & 3 deletions tests/utils/test_document.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -32,11 +32,9 @@ def test_document_merge_annotations():
assert result.id == "doc1"
assert set(result.labeled_spans) == set(base_doc.labeled_spans)
assert len(result.labeled_spans) == len(base_doc.labeled_spans) == 2
assert len(result.labeled_spans.predictions) == 4
assert result.labeled_spans.predictions.resolve() == [
("label1", "This"),
("label2", "is"),
("label1", "This"),
("label3", "is"),
]
annotations_with_sources = [
Expand All @@ -59,6 +57,5 @@ def test_document_merge_annotations():
assert predictions_with_scores == [
(LabeledSpan(start=0, end=4, label="label1", score=0.9), ["doc1"]),
(LabeledSpan(start=5, end=7, label="label2", score=0.7), ["doc1", "doc2"]),
(LabeledSpan(start=0, end=4, label="label1", score=0.8), ["doc2"]),
(LabeledSpan(start=5, end=7, label="label3", score=0.6), ["doc2"]),
]
Loading