In [132]:

    

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from matplotlib.ticker import FuncFormatter
%matplotlib inline
pd.set_option('display.max_rows',8)

from pandas import *
from pandas.io.parsers import read_csv


    

In [133]:

    

ral = read_csv('cern30min.csv', index_col=False, header=None,
               names=['timestamp','host','size','duration'],
               usecols=['timestamp', 'duration'])
ral['timestamp'] = pd.to_datetime(ral['timestamp'],unit='s')
ral['time'] = ral['timestamp'] - min(ral['timestamp'])
ral.set_index('time', inplace=True)


    

In [134]:

    

ral.describe()


    

    
Out[134]:







  

    

      

      
duration
    
  
  

    

      
count
      
255.000000
    
    

      
mean
      
23.380024
    
    

      
std
      
2.256339
    
    

      
min
      
18.405646
    
    

      
25%
      
21.832625
    
    

      
50%
      
23.031074
    
    

      
75%
      
24.549370
    
    

      
max
      
38.850552
    
  

In [135]:

    

ral.plot(x=ral.index.astype('timedelta64[s]'), y='duration')


    

    
Out[135]:





<matplotlib.axes._subplots.AxesSubplot at 0x118e5ed68>