Filtering fix

clarin_datasets/kpwr_ner_datasets.py

@@ -104,7 +104,7 @@ class KpwrNerDataset(DatasetToShow):
             on="class",
         )
         with class_distribution:
-            st.header("Class distribution in each subset (without '0' and 'I-*')")
+            st.header("Class distribution in each subset (without 'O' and 'I-*')")
             st.dataframe(class_distribution_df)
             st.text_area(
                 label="LaTeX code", value=class_distribution_df.style.to_latex()
@@ -117,10 +117,10 @@ class KpwrNerDataset(DatasetToShow):
                     "ner": labels_all,
                 }
             )
-            full_df_unzipped = full_df_unzipped.loc[full_df_unzipped["ner"] != 0]
+            full_df_unzipped = full_df_unzipped.loc[full_df_unzipped["ner"] != "O" and not full_df_unzipped["ner"].str.starstwith("I-")]
             possible_options = sorted(full_df_unzipped["ner"].unique())
             with most_common_tokens:
-                st.header("10 most common tokens from selected class (without '0')")
+                st.header("10 most common tokens from selected class (without 'O')")
                 selected_class = st.selectbox(
                     label="Select class to show", options=possible_options
                 )