Update appStore/prep_data.py

appStore/prep_data.py

@@ -99,9 +99,10 @@ def process_giz_worldwide():
     # Create an empty 'url' column as the new dataset has an empty URL
     giz_df['url'] = ''
     
-    # Convert CRS value: use string replacement and then integer conversion.
+    # Convert CRS value to numeric then to integer if possible.
     if 'crs_value' in giz_df.columns:
-        giz_df['crs_value'] = giz_df['crs_value'].apply(convert_crs_value)
+        # Convert to numeric, coercing errors to NaN, then fill NaN with 0 and cast to int.
+        giz_df['crs_value'] = pd.to_numeric(giz_df['crs_value'], errors='coerce').fillna(0).astype(int)
     
     # Compute text_size based on merged_text and assign full text to the 'chunks' column
     giz_df['text_size'] = giz_df['merged_text'].apply(lambda text: len(text.split()) if isinstance(text, str) else 0)