Pandas None values vs Empty dataframe

I have two queries with different results for no data in pandas dataframe.

First query:

with self.connection as conn:
            query = f"""SELECT current_client_name, client_phone_number, profile, effective_date, effective_time, 
                                          client_details.telegram_id, telegram_reference
                                          FROM `clients` JOIN `client_details` 
                                         ON clients.telegram_id=client_details.telegram_id 
                                         WHERE `status` = \'{status}\' order by effective_date"""
            df_result = pd.read_sql_query(query, conn)
            return df_result

In the result when no records found i see the empty dataframe:

Empty DataFrame
Columns: [current_client_name, client_phone_number, profile, effective_date, effective_time, telegram_id, telegram_reference]
Index: []

and can catch empty dataframe with next condition:

if df.empty is True:

Second query

with self.connection as conn:
            query = f"SELECT MAX(effective_date), effective_time FROM `client_details` WHERE `telegram_id` = {telegram_id} " \
                    f"and (`status` = \'WAIT\' or `status`= \'APPROVED\')"
            df_result = pd.read_sql_query(query, conn)
            return df_result

This query for some reasons return me not empty dataframe but dataframe with None in columns

df MAX(effective_date) effective_time
0                None           None

this result can not be handled with next code because for some reasons was returned is not empty dataframe!

if df.empty is True:

Does anybody know why its happening : why return for some cases empty dataframe and why return dataframe with None type in columns? How to avoid or handle this?

Thanks

Solution

In the first case, your query doesn't return any results, thus there are no entries in your df.

In the second case, your query returns a result, but those fields are empty in your SQL database, thus your have a non-empty DataFrame, containing None values.

Depending if you only want rows containing all non-null values (e.g. both columns in your example have to be non-null), or if some null values are ok (e.g. one of the columns is Null, while another is non-Null), you could drop rows containing these None values:

To drop rows containing any None values:

df = df.dropna(how = 'any')

To drop rows only containing None values:

df = df.dropna(how = 'all')

Then the check if df.empty should come as True.