diff options
| author | Adam Janovsky | 2023-07-26 14:11:20 +0200 |
|---|---|---|
| committer | Adam Janovsky | 2023-07-26 14:11:20 +0200 |
| commit | 7ba838443a707b86194d127bccb1db7dbad61a3c (patch) | |
| tree | 6bb114199a533318d1e36d15cc9ac8f94a43cbf1 /src | |
| parent | e19d4ae8d9294d6fba001908df02cc1b19dc88b0 (diff) | |
| download | sec-certs-7ba838443a707b86194d127bccb1db7dbad61a3c.tar.gz sec-certs-7ba838443a707b86194d127bccb1db7dbad61a3c.tar.zst sec-certs-7ba838443a707b86194d127bccb1db7dbad61a3c.zip | |
notebook to measure annotator agreement
Diffstat (limited to 'src')
| -rw-r--r-- | src/sec_certs/utils/nlp.py | 2 |
1 files changed, 2 insertions, 0 deletions
diff --git a/src/sec_certs/utils/nlp.py b/src/sec_certs/utils/nlp.py index dd114662..4c136ce9 100644 --- a/src/sec_certs/utils/nlp.py +++ b/src/sec_certs/utils/nlp.py @@ -27,6 +27,8 @@ def filter_short_sentences(sentences, cert_id): def prepare_reference_annotations_df(df: pd.DataFrame): + if df.loc[(df.label != "SELF") & (df.label.notnull())].empty: + raise ValueError("No expert annotations found in the dataset of references.") df = ( df.loc[lambda df_: (df_.label != "SELF") & (df_.label.notnull())] .assign(segments=lambda df_: eval_strings_if_necessary(df_.segments)) |