Problem 12.2. Heat Maps

Create a heat map that displays the number count of delayed flights at the O'Hare international airport (ORD) as a function of month and Day. We define a flight to be delayed if its departure delay is greater than or equal to 15 minutes.



In [ ]:

    
%matplotlib inline
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

I hope you actually tried the examples in lesson 1 because you are required to use the sflights.h5 file you have created as you read along the lesson 1 notebook. When we use the sflights.h5 file, the data DataFrame looks as follows:

print(data.head())

month  Day  dTime  aDelay  dDelay depart arrive  distance
  0      1    3   1806      -3      -4    BWI    CLT       361
  1      1    4   1805       4      -5    BWI    CLT       361
  2      1    5   1821      23      11    BWI    CLT       361
  3      1    6   1807      10      -3    BWI    CLT       361
  4      1    7   1810      20       0    BWI    CLT       361

  [5 rows x 8 columns]



In [ ]:

    
data = pd.read_hdf('/data/database/sflights.h5', 'table')
print(data.head())

Function: pivot_table()

Write a function named pivot_table() that takes a DataFrame. It should return a DataFrame that has the *number count of flights that

departed from ORD
had departure delays greater than or equal to 15 minutes as a function of month and Day.

You should also pivot the data to make the pivot table so that we can use the resulting DataFrame in sns.heatmap(). Week 12 lesson 2 shows you how to do this. When I ran

pivoted = pivot_table(data)
  print(pivoted)

I got

Day       1     2     3     4     5     6     7
  month                                          
  1      1112   984  1157   599  1205   368   800
  2       436   550   978  1097  1071   774   950
  3       874   500   481  1317  1332   727   807
  4      1147   678   903  1080   948   881  1077
  5      1433   852   840  1126  1184   349   596
  6       679  1410  1031   911  1110   853   651
  7      1018  1296  1138   751   457   752  1683
  8       673   733  1001  1866  1240  1199   822
  9       588   331   373   447   772   655  1150
  10      597   498   822   811   537   544   646
  11      474   338   318   592   644   535   404
  12      420   304   909   640   611   613  1035

  [12 rows x 7 columns]



In [ ]:

    
def pivot_table(df):
    '''
    Takes a DataFrame and returns a DataFrame.
    
    Parameters
    ----------
    df: A Pandas DataFrame, e.g. pd.read_hdf('/data/database/sflights.h5', 'table')
    
    Returns
    -------
    A Pandas DataFrame. It's a pivot table of delayed flihts
    as a function of month and Day. See assignment 12.2 instructions.
    '''
    
    #### your code goes here ###
    
    return result

Run the following code cell to test your function.



In [ ]:

    
pivoted = pivot_table(data)
print(pivoted)

Plot: sns.heatmap()

Finally, use the pivoted DataFrame in sns.heatmap() to create heat map.



In [ ]:

    
#### your code goes here ####