In [5]:

    

import random
from datetime import datetime
import numpy as np

#year = random.randint(1950, 2000)
#month = random.randint(1, 12)
year = 2017
month = 12

# Generate 300 random timestamps and users
n_events = 300

events=[]
for i in range(n_events): 
    day = random.randint(1, 28)
    # add a few null values
    #if i & 3 == 2 :
    #    user = np.nan
    #else:
    user = int(random.randint(1,10))
    timestamp = datetime(year, month, day).timestamp()
    events.append([timestamp,user])

    

#>>> from datetime import datetime
#>>> datetime.fromtimestamp(1172969203.1)
#datetime.datetime(2007, 3, 4, 0, 46, 43, 100000)


    

In [6]:

    

import numpy as np
import pandas as pd
# numpy array to df

events = np.array(events)
columns = ['timestamp','user']
df = pd.DataFrame(events,columns=columns)
df.head(10)


    

    
Out[6]:







  

    

      

      
timestamp
      
user
    
  
  

    

      
0
      
1.513066e+09
      
6.0
    
    

      
1
      
1.513930e+09
      
2.0
    
    

      
2
      
1.513152e+09
      
NaN
    
    

      
3
      
1.513152e+09
      
1.0
    
    

      
4
      
1.512720e+09
      
8.0
    
    

      
5
      
1.512720e+09
      
10.0
    
    

      
6
      
1.513238e+09
      
NaN
    
    

      
7
      
1.513325e+09
      
4.0
    
    

      
8
      
1.512979e+09
      
6.0
    
    

      
9
      
1.512893e+09
      
6.0
    
  

In [7]:

    

df.to_csv('random_timestamps.csv')


    

In [ ]: