Update pages/4_Simple_EDA.py

pages/4_Simple_EDA.py

@@ -99,3 +99,41 @@ st.markdown('''
 - Whether the collected text data contains any data /time
 ''')
 
+st.code('''
+            import pandas as pd
+            import numpy as np
+            import re
+            import emoji
+            
+            def simple_eda(data,column):
+                lower_upper = data[column].apply(lambda x:True if (x.lower()) or (x.upper()) else False).sum()
+                tags = data[column].apply(lambda x:True if re.search("<.*?>",x) else False).sum()
+                urls = data[column].apply(lambda x:True if re.search("https://\S+",x) else False).sum()
+                mails = data[column].apply(lambda x:True if re.search("\S+@\S+",x) else False).sum()
+                mentions = data[column].apply(lambda x:True if re.search("\B[@#]\S+",x) else False).sum()
+                emojis = data[column].apply(lambda x:True if emoji.emoji_count(x) else False).sum()
+                digit = data[column].apply(lambda x:True if re.search("\d",x) else False).sum()
+                punc = data[column].apply(lambda x:True if re.search('[!"#$%&\'()*+,-./:;<=>?@[\\]^_`{|}~]',x) else False).sum()
+                dates = data[column].apply(lambda x:True if re.search(r"^[0-9]{1,2}\/[0-9]{1,2}\/[0-9]{4}$",x) else False).sum()
+
+    
+                if lower_upper >0:
+                    print("text have combination")
+                if tags > 0:
+                    print("text have tags")
+                if urls >0:
+                    print("text have urls")
+                if mails > 0:
+                    print("text have mails")
+                if mentions >0:
+                    print("text have mentions")
+                if emojis > 0:
+                    print("text have emojis")
+                if digit >0:
+                    print("text have digit")
+                if punc > 0:
+                    print("text have punctuations")
+                if dates >0:
+                    print("text have dates")
+        
+''')