Review of Python Pandas

Based on CBSE Curriculum

Informatics Practices Class-12
CHAPTER-1

By:
Neha Tyagi, PGT CS
KV no-5 2nd Shift, Jaipur
Jaipur Region
 Python Pandas ( A Review)
• Data Processing is the most important part of Data
Analysis. Because data is not avaiable every time in
desired format.
• Before analyzing the data it needs various types of
processing like - Cleaning, Restructuring or merging
etc.
• There are many tools available in python to process
the data fast Like-Numpy, Scipy, Cython and Pandas.
• Pandas are built on the top of Numpy.
• In this chapter we will learn about the basic concepts
of Python Pandas Data Series and DataFrames
which we learnt in class -11.
संजीव भदौरिया, के० वव० बािाबंकी
 Python Pandas
• Pandas is an open-source library of python providing high-performance
data manipulation and analysis tool using its powerful data structure.
• Pandas provides rich set of functions to process various types of data.
• During data analysis it is very important to make it confirm that you are
using correct data types otherwise you may face some unexpected
errors.
• Some of the pandas supporting data types are as follows -

संजीव भदौरिया, के० वव० बािाबंकी

 Pandas Series
• Series is the primary building block of Pandas.
• Series is a labeled One-Dimensional Array which can hold any type of
data.
• Data of Series is always mutable. It means, it can be changed.
• But the size of data of Series is size immutable, means can not be
changed.
• it can be seen as a data structure with two arrays: one functioning as the
index (Labels) and the other one contains the actual data.
• In Series, row labels are also called the index .
• Lets take some data which can be considered as series -

संजीव भदौरिया, के० वव० बािाबंकी

 Creation of Series Objects
– There are many ways to create series type object.
1. Using Series ( )-
<Series Object> = pandas.Series( ) it will create empty series.

2. Non-empty series creation–

Import pandas as pd
<Series Object> = pd.Series(data, index=idx) where data can be
python sequence, ndarray, python dictionary or scaler value.

Index
Index

Neha Tyagi, KV5 Jaipur II shift

 Series Objects creation
1. Creation of series with Dictionary-
Index of
Keys

2. Creation of series with Scalar value-

Neha Tyagi, KV5 Jaipur II shift

 Creation of Series Objects –Additional functionality
1. When it is needed to create a series with missing values, this
can be achieved by filling missing data with a NaN (“Not a
Number”) value.

2. Index can also be given as-

Loop is used to give Index

Neha Tyagi, KV5 Jaipur II shift

 Creation of Series Objects –Additional functionality
3. Dtype can also be passed with Data and index

Important: it is not necessary to

have unique indices but it will give
error when search will be according to
index.

4. Mathematical function/Expression can also be used-

Neha Tyagi, KV5 Jaipur II shift

 Series Object Attributes
3. Some common attributes-
<series object>.<AttributeName>

Attribute Description
Series.index Returns index of the series
Series.values Returns ndarray
Series.dtype Returns dtype object of the underlying data
Series.shape Returns tuple of the shape of underlying data
Series.nbytes Return number of bytes of underlying data
Series.ndim Returns the number of dimention
Series.size Returns number of elements
Series.intemsize Returns the size of the dtype
Series.hasnans Returns true if there are any NaN
Series.empty Returns true if series object is empty

Neha Tyagi, KV5 Jaipur II shift

Series Object Attributes

Neha Tyagi, KV5 Jaipur II shift

 Accessing Series Object
Object
slicing

Printing object value

Printing Individual value

For Object slicing, follow the following syntax-

<objectName>[<start>:<stop>:<step >]

Neha Tyagi, KV5 Jaipur II shift

 Operations on Series Object
1. Elements modification-
<series object>[index] = <new_data_value>

To change
individual value To change value in a
certain slice

Neha Tyagi, KV5 Jaipur II shift

 Operations on Series Object
1. It is possible to change indexes
<series object>.<index] = <new_index_array>

Here, indexes got

changed.

Neha Tyagi, KV5 Jaipur II shift

 head() and tail () Function
1. head(<n> ) function fetch first n rows from a pandas object. If
you do not provide any value for n, will return first 5 rows.
2. tail(<n> ) function fetch last n rows from a pandas object. If
you do not provide any value for n, will return last 5 rows.

Neha Tyagi, KV5 Jaipur II shift

 Series Objects - Series Objects -
Vector Operations Arithmetic Operations

All these are

vector operations

Arithmetic operation is
possible on objects of
same index otherwise
will result as NaN.

Neha Tyagi, KV5 Jaipur II shift

We can also store these results in other objects.
 Entries Filtering
<seriesObject> <series - boolean expression >
Other feature

To delete value of
index

Neha Tyagi, KV5 Jaipur II shift

 Difference between NumPy array Series objects

1. In case of ndarray, vector operation is possible only

when ndarray are of similar shape. Whereas in case
of series object, it will be aligned only with matching
index otherwise NaN will be returned.

2. In ndarray, index always starts from 0 and always

numeric. Whereas, in series, index can be of any
type including number and not necessary to start
from 0.
Neha Tyagi, KV5 Jaipur II shift
 DataFrame
• Pandas का मुख्य object DataFrame होता है | और यह pandas का सबसे अधिक प्रयोग
ककया जाने वाला Data Structure है |
• DataFrame एक Two -Dimensional Array होता है जो ककसी भी data type को hold
कर सकती है | और यह tabular format में data को store करता है |
• Finance, Statistics, Social Science और कई engineering branch में इसका प्रयोग
अधिकता में ककया जाता है |
• DataFrame में data और इसका size दोनों ही mutable होते हैं अर्ाात इन्हें बदला जा
सकता है |
• DataFrame में दो धवधभन्न indexes होते हैं - row index और column index |

संजीव भदौरिया, के० वव० बािाबंकी

 Creation and presentation of DataFrame
• DataFrame object can be created by passing a data in 2D
format.
import pandas as pd
<dataFrameObject> = pd.DataFrame(<a 2D Data Structure>,\ [columns=<column
sequence>],[index=<index sequence>])

• You can create a DataFrame by various methods by passing

data values. Like-
• 2D dictionaries
• 2D ndarrays
• Series type object
• Another DataFrame object

Neha Tyagi, KV5 Jaipur, II Shift

 Creation of DataFrame from 2D Dictionary
A. Creation of DataFrame from dictionary of List or ndarrays.

In the above example, index

are automatically generated
from 0 to 5 and column name
are same as keys in dictionary.

Indexes are column name are generated from

automatically keys of 2D Dictionary
generated by using
np.range(n) Neha Tyagi, KV5 Jaipur, II Shift
 Here, indexes are
specified by you.

Meaning, if you specify the sequence of index then index

wlil be the set specified by you only otherwise it will be
automatically generated from 0 to n-1.

Neha Tyagi, KV5 Jaipur, II Shift

 Creation of DataFrame from 2D Dictionary
B. Creation of DataFrame from dictionary of Dictionaries-

It is a 2D Dictionary made up of
above given dictionaries.

DataFrame object created.

Here, you can get an idea If keys of yr2015, yr2016 and yr2017 were
of how index and column different here then rows and columns of
name have assigned. dataframe would have increased and non-
matching rows and column would store
NaN.
Neha Tyagi, KV5 Jaipur, II Shift
Creation of Dataframe from 2D ndarray

column name and index have

automatically been generated here.

Here, user has given column

name .

Here, column name and index

both have given by user.

Neha Tyagi, KV5 Jaipur, II Shift

Creation of DataFarme from 2D Dictionary of same Series Object

It is a 2D Dictionary made up of series

given above.

DataFrame object created.

DataFrame object can also be

created like this.

Neha Tyagi, KV5 Jaipur, II Shift

Creation of DataFrame from object of other DataFrame

DataFrame object is created from

object of other DataFrame.

Displaying DataFrame Object

Syntax for displaying
DataFrame object.

Neha Tyagi, KV5 Jaipur, II Shift

 DataFrame Attributes
• When we create an object of a DataFrame then all information
related to it like size, datatype etc can be accessed by attributes.
<DataFrame Object>.<attribute name>
• Some attributes are -
Attribute Description
index It shows index of dataframe.
columns It shows column labels of DataFrame.
axes It return both the axes i.e. index and column.
dtypes It returns data type of data contained by dataframe.
size It returns number of elements in an object.
shape It returns tuple of dimension of dataframe.
values It return numpy form of dataframe.
empty It is an indicator to check whether dataframe is empty or not.
ndim Return an int representing the number of axes / array dimensions.
T It Transpose index and columns.

Neha Tyagi, KV5 Jaipur, II Shift

DataFrame Attributes

Neha Tyagi, KV5 Jaipur, II Shift

 Selecting and Accessing from DataFrame
• Selecting a Column-
To select a
<DataFrame Object>[<column name>] column

or <DataFrame Object>.<column name>

Selection of multiple column
<DataFrame Object>[List of column name ]

We can change the order in column.

Neha Tyagi, KV5 Jaipur, II Shift

Selection of subset from DataFrame
<DataFrameObject>.loc [<StartRow> : <EndRow>, <StartCol> : <EndCol>]

Neha Tyagi, KV5 Jaipur, II Shift

Selection of subset from DataFrame
<DataFrameObject> .iloc [<Row Index> : <RowIndex>, <ColIndex> : <ColIndex>]

Selection of an Individual Value from DataFrame

<DFObject>. <col name.[row name or row index]
or
<DFObject> . at [<row name>,<col name>]
or
<DFObject> iat[<row index>, <col index>]

Neha Tyagi, KV5 Jaipur, II Shift

 Accessing and modifying values in DataFrame
a) Syntax to add or change a column-
<DFObject>.<Col Name>[<row label>]=<new value>

A new column will be created because there is

no column with the name ‘Four’.

The values of column will get change because

there is a column with the name ‘Four’.

Neha Tyagi, KV5 Jaipur, II Shift

 Accessing and modifying values in DataFrame
b) Syntax to add or change a row-
<DFObject> at[<RowName>, : ] =<new value>
या
<DFObject> loc[<RowName>, : ] =<new value>

A new row will be created because there is no

row with the name ‘D’.

The values of row will get change because

there is a row with the name ‘D’.

Neha Tyagi, KV5 Jaipur, II Shift

 Accessing and modifying values in DataFrame
c) Syntax to change single value-
<DFObject>.<ColName>[<RowName/Lebel>]

Values can be changed like this also. Values of

row and column can be given separately.

Here, value of column ‘Three’ of row ‘D’ got

changed.

Neha Tyagi, KV5 Jaipur, II Shift

 Accessing and modifying values in DataFrame
d) Syntax for Column deletion-
del <DFObject>[<ColName>] or
df.drop([<Col1Name>,<Col2Name>, . . ], axis=1)

axis =1 specifies deletion of column.

del command does not return

value after deletion whereas
drop method returns the
value to dataframe after
deletion.

Neha Tyagi, KV5 Jaipur, II Shift

 Iteration in DataFrame
• Sometimes we need to perform iteration on complete
DataFrame. In such cases, it is difficult to write code
to access values separately. Therefore, it is
necessary to perform iteration on dataframe which is
to be done as-

• <DFObject>.iterrows( ) it represents dataframe in

row-wise subsets .

• <DFObject>.iteritems( ) it represents dataframe in

column-wise subsets.

Neha Tyagi, KV5 Jaipur, II Shift

Use of pandas.iterrows () function

These are the values

of df1 which are
processed one by
one.

Try the code given below

after creation of DataFrame.

Neha Tyagi, KV5 Jaipur, II Shift

Use of pandas.iteritems() function

These are the

values of df1
which are
processed one by
one.

Try the code given below

after creation of DataFrame.

Neha Tyagi, KV5 Jaipur, II Shift

 Program for iteration
• Write a program to iterate over a dataframe
containing names and marks, then calculates grades
as per marks (as per guideline below) and adds them
to the grade column.
Marks > =90 Grade A+
Marks 70 – 90 Grade A
Marks 60 – 70 Grade B
Marks 50 – 60 Grade C
Marks 40 – 50 Grade D
Marks < 40 Grade F

Neha Tyagi, KV5 Jaipur, II Shift

Program for iteration

Neha Tyagi, KV5 Jaipur, II Shift

 Binary Operations in a DataFrame
It is possible to perform add, subtract, multiply and devision
operations on DataFrame.
To Add - ( +, add or radd )
To Subtract - (-, sub or rsub)
To Multiply– (* or mul)
To Divide - (/ or div)

We will perform operations on following dataframes-

Neha Tyagi, KV5 Jaipur, II Shift

 Addition

DataFrame follows index matching to perform arithmetic operations. If matches, operation

takes place otherwise it shows NaN (Not a Number). It is called Data Alignment in panda
object.
This behavior of ‘data alignment’ on the basis of “matching indexes” is called MATCHING.

Neha Tyagi, KV5 Jaipur, II Shift

Subtraction

Neha Tyagi, KV5 Jaipur, II Shift

Multiplication

Neha Tyagi, KV5 Jaipur, II Shift

 Division

See the
operation of the
rdiv carefully

Neha Tyagi, KV5 Jaipur, II Shift

 Other important functions
Other important functions of DataFrame are as under-
<DF>.info ( )
<DF>.describe ( )

Neha Tyagi, KV5 Jaipur, II Shift

 Other important functions
Other important functions of DataFrame are as under-
<DF>.head ([ n=<n>] ) here, default value of n is 5.
<DF>.tail ( [n=<n>])

Neha Tyagi, KV5 Jaipur, II Shift

 Cumulative Calculations Functions
In DataFrame, for cumulative sum, function is as under-
<DF>.cumsum([axis = None]) here, axis argument is optional. |

Neha Tyagi, KV5 Jaipur, II Shift

Index of Maximum and Minimum Values
<DF>.idxmax ( )
<DF>.idxmin ( )

Neha Tyagi, KV5 Jaipur, II Shift

 Handling of Missing Data
• The values with no computational significance are called
missing values.
• Handling methods for missing values-
 Dropping missing data
 Filling missing data (Imputation)

Neha Tyagi, KV5 Jaipur, II Shift

 Comparison of Pandas Objects

equals () checks both the

objects for equality.
Neha Tyagi, KV5 Jaipur, II Shift
• कृ पया हमारे ब्लॉग को फॉलो कररए और youtube channel
को subscribe कररए | ताकक आपको और सारे chapters
धमल सकें |

www.pythontrends.wordpress.com