Python Pandas sort by Time and group by user ID

Question

I am loading a CSV file with pandas. It has three columns: a column with date and time, a column with a user id, and another 'campaignID'. Example rows:

date                 user_id              campaign_id
2018-01-10 0:21:09   151312395            GOOGLE
2018-01-10 0:21:19   151312395            GOOGLE
2018-01-10 0:21:32   151312395            GOOGLE 

I want to group the data by the user id, and then for each user id group the rows by time and the campaign ID, it should look as follows.

user_id              date                           ad_campaign
151312395            2018-01-10 0:21:09             GOOGLE
                     2018-01-10 0:21:19             GOOGLE
                     2018-01-10 0:21:32             GOOGLE 

This is what I have made until now: import pandas as pd import numpy as np import datetime

def dateparse(time_in_secs):
    return datetime.datetime.fromtimestamp(float(time_in_secs))
columnnames = ['date','user_id', 'ad_campaign']
columnnames, sep='\t' ,usecols=[0,1,3],index_col = 'date')
df=pd.read_csv(r'C:\Users\L\Desktop\Data.csv' , 
     sep='\t',names = columnnames, usecols=[0,1,3], 
    parse_dates=True,date_parser=dateparse)
df.date = pd.to_datetime(df.date)
df = df.sort_values(by = 'date')
g = df.groupby('user_id')['ad_campaign']
print(g)

This gives the following output:

<pandas.core.groupby.SeriesGroupBy object at 0x04EF26F0>
[Finished in 0.6s]

Why doesnt the print provide the sorted columns?

Answer 1

Firstly, if you are doing groupby, you don't need to sort the column explicitly.

You can do:

Method 1:

df.date = pd.to_datetime(df.date)
g = df.groupby(['user_id','date'])['ad_campaign']
print(g.first())

Method 2:

df.set_index(['user_id','date']).sort_index()