nan values

b2w-reviews01.py

@@ -2,6 +2,7 @@
 
 import datasets
 import pandas as pd
+import numpy as np
 
 _CITATION = """
 @inproceedings{real2019b2w,
@@ -66,19 +67,8 @@ class Reviews(datasets.GeneratorBasedBuilder):
             except ValueError:
                 row["reviewer_birth_year"] = None
 
-            nan_list = [
-                'product_name',
-                'product_brand',
-                'site_category_lv1',
-                'site_category_lv2',
-                'review_title',
-                'recommend_to_a_friend',
-                'review_text',
-                'reviewer_gender',
-                'reviewer_state'
-            ]
             for key, value in row.items():
-                if key in nan_list and not isinstance(value, str):
+                if np.isnan(value):
                     row[key] = None
 
             return row