I have a table in postgresql database. From that table I have extracted data using the sql statement mentioned below:
sql_statement = """ select a.slno, a.clientid, a.filename, a.user1_id, b.username, a.user2_id, c.username as username2, a.uploaded_ts, a.status_id
from masterdb.xmlform_joblist a
left outer join masterdb.auth_user b
on a.user1_id = b.id
left outer join masterdb.auth_user c
on a.user2_id = c.id
"""
cursor.execute(sql_statement)
result = cursor.fetchall()
From this code I accessed the database and extracted data from each fields and added the data into list using the below code:
date = []
username1 = []
username2 = []
user1_id = []
user2_id = []
status_id = []
cient_id = []
filename = []
#extracting specific data from specified fields in the database
for row in result:
date.append(row[7])
username1.append(row[4])
username2.append(row[6])
status_id.append(row[8])
cient_id.append(row[1])
filename.append(row[2])
#creating log file for the extracted fields
logger.info("Date | {} , username1 | {} , username2 | {} , status_id | {} , client_id | {} , filename | {} ".format(row[7], row[4], row[6], row[8], row[1], row[2]))
Now I want to check the frequency of usernames from the data I have collected.
username1 looks like:
username1 = ['Dean','Sarah','Sarah','Alan','Dean',......'Alan']
I want to know the count of the users in the list
Expected result:
Dean = 10
Sarah = 6
Alan = 2
Is it possible to achieve this result using python.
I have tried Pandas but it isn’t working as my database is postgresql.
2
Answers