In [3]:

    

%matplotlib inline

import pandas as pd
import matplotlib.pyplot as plt


    

In [19]:

    

df = pd.read_csv('ttc-subway-delay-data-2014-01-01_2017-04-30.csv',
                 parse_dates=[['Date', 'Time']])

# The default short should be by `Date_Time`.
df.sort_values(by='Date_Time', ascending=False)

df.head()


    

    
Out[19]:






  

    

      

      
Date_Time
      
Day
      
Station
      
Code
      
Min Delay
      
Min Gap
      
Bound
      
Line
      
Vehicle
    
  
  

    

      
0
      
2014-01-01 02:06:00
      
Wednesday
      
HIGH PARK STATION
      
SUDP
      
3
      
7
      
W
      
BD
      
5001
    
    

      
1
      
2014-01-01 02:40:00
      
Wednesday
      
SHEPPARD STATION
      
MUNCA
      
0
      
0
      
NaN
      
YU
      
0
    
    

      
2
      
2014-01-01 03:10:00
      
Wednesday
      
LANSDOWNE STATION
      
SUDP
      
3
      
8
      
W
      
BD
      
5116
    
    

      
3
      
2014-01-01 03:20:00
      
Wednesday
      
BLOOR STATION
      
MUSAN
      
5
      
10
      
S
      
YU
      
5386
    
    

      
4
      
2014-01-01 03:29:00
      
Wednesday
      
DUFFERIN STATION
      
MUPAA
      
0
      
0
      
E
      
BD
      
5174
    
  

In [20]:

    

df.ix[df['Min Delay'].idxmax()]


    

    
Out[20]:





Date_Time       2015-02-17 06:56:00
Day                         Tuesday
Station      SCARBOROUGH RAPID TRAN
Code                           ERHV
Min Delay                       999
Min Gap                         999
Bound                             N
Line                            SRT
Vehicle                        3007
Name: 23171, dtype: object

In [17]:

    

delays_by_day = df.pivot(columns='Day', values='Min Delay')
delays_by_day.boxplot()


    

    
Out[17]:





<matplotlib.axes._subplots.AxesSubplot at 0x10f3b49b0>






    

In [ ]: