Lesson 4

In this lesson were going to go back to the basics. We will be working with a small data set so that you can easily understand what I am trying to explain. We will be adding columns, deleting columns, and slicing the data many different ways. Enjoy!



In [1]:

    
# Import libraries
import pandas as pd
import sys



In [2]:

    
print 'Python version ' + sys.version
print 'Pandas version: ' + pd.__version__









    



Python version 2.7.5 |Anaconda 2.1.0 (64-bit)| (default, Jul  1 2013, 12:37:52) [MSC v.1500 64 bit (AMD64)]
Pandas version: 0.15.2



In [3]:

    
# Our small data set
d = [0,1,2,3,4,5,6,7,8,9]

# Create dataframe
df = pd.DataFrame(d)
df



In [4]:

    
# Lets change the name of the column
df.columns = ['Rev']
df



In [5]:

    
# Lets add a column
df['NewCol'] = 5
df



In [6]:

    
# Lets modify our new column
df['NewCol'] = df['NewCol'] + 1
df



In [7]:

    
# We can delete columns
del df['NewCol']
df



In [8]:

    
# Lets add a couple of columns
df['test'] = 3
df['col'] = df['Rev']
df



In [9]:

    
# If we wanted, we could change the name of the index
i = ['a','b','c','d','e','f','g','h','i','j']
df.index = i
df

We can now start to select pieces of the dataframe using loc.



In [10]:

    
df.loc['a']









    Out[10]:





Rev     0
test    3
col     0
Name: a, dtype: int64



In [11]:

    
# df.loc[inclusive:inclusive]
df.loc['a':'d']



In [12]:

    
# df.iloc[inclusive:exclusive]
# Note: .iloc is strictly integer position based. It is available from [version 0.11.0] (http://pandas.pydata.org/pandas-docs/stable/whatsnew.html#v0-11-0-april-22-2013) 
df.iloc[0:3]

We can also select using the column name.



In [13]:

    
df['Rev']









    Out[13]:





a    0
b    1
c    2
d    3
e    4
f    5
g    6
h    7
i    8
j    9
Name: Rev, dtype: int64



In [14]:

    
df[['Rev', 'test']]



In [15]:

    
# df['ColumnName'][inclusive:exclusive]
df['Rev'][0:3]









    Out[15]:





a    0
b    1
c    2
Name: Rev, dtype: int64



In [16]:

    
df['col'][5:]









    Out[16]:





f    5
g    6
h    7
i    8
j    9
Name: col, dtype: int64



In [17]:

    
df[['col', 'test']][:3]

There is also some handy function to select the top and bottom records of a dataframe.



In [18]:

    
# Select top N number of records (default = 5)
df.head()



In [19]:

    
# Select bottom N number of records (default = 5)
df.tail()

Author: David Rojas