Update classification.py
Browse files- classification.py +5 -5
classification.py
CHANGED
@@ -181,7 +181,7 @@ def match_categories(df, category_df, treshold=0.45):
|
|
181 |
categories_list.append("@~@".join([category_df.loc[index, 'description'] for index in high_score_indices]))
|
182 |
experts_list.append("@~@".join(["@~@".join(list(set(category_df.loc[index, 'experts']))) for index in high_score_indices]))
|
183 |
topic_list.append("@~@".join([category_df.loc[index, 'topic'] for index in high_score_indices]))
|
184 |
-
scores_list.append("@~@".join([
|
185 |
for j in high_score_indices:
|
186 |
df.loc[index, category_df.loc[j, 'topic']] = float(cos_scores[j])
|
187 |
else:
|
@@ -189,10 +189,10 @@ def match_categories(df, category_df, treshold=0.45):
|
|
189 |
experts_list.append(np.nan)
|
190 |
topic_list.append(np.nan)
|
191 |
scores_list.append(np.nan)
|
192 |
-
|
193 |
-
|
194 |
-
|
195 |
-
|
196 |
return df
|
197 |
|
198 |
|
|
|
181 |
categories_list.append("@~@".join([category_df.loc[index, 'description'] for index in high_score_indices]))
|
182 |
experts_list.append("@~@".join(["@~@".join(list(set(category_df.loc[index, 'experts']))) for index in high_score_indices]))
|
183 |
topic_list.append("@~@".join([category_df.loc[index, 'topic'] for index in high_score_indices]))
|
184 |
+
scores_list.append("@~@".join([str(float(cos_scores[index])) for index in high_score_indices]))
|
185 |
for j in high_score_indices:
|
186 |
df.loc[index, category_df.loc[j, 'topic']] = float(cos_scores[j])
|
187 |
else:
|
|
|
189 |
experts_list.append(np.nan)
|
190 |
topic_list.append(np.nan)
|
191 |
scores_list.append(np.nan)
|
192 |
+
df["Description"] = categories_list
|
193 |
+
df["Expert"] = experts_list
|
194 |
+
df["Topic"] = topic_list
|
195 |
+
df["Score"] = scores_list
|
196 |
return df
|
197 |
|
198 |
|