Chapter 1

Programming in Python
Python is a simple, high level language with a clean syntax. It oers strong support for integration
with other languages and tools, comes with extensive standard libraries, and can be learned within
few days. Many Python programmers report substantial productivity gains and feel the language
encourages the development of higher quality, more maintainable code. To know more visit the
Python website.1

1.1 Getting started with Python

To start programming in Python, we have to learn how to type the source code and save it to a
le, using a text editor program. We also need to know how to open a Command Terminal and
start the Python Interpreter. The details of this process may vary from one system to another.
On MSWindows, an IDE called IDLE will be available after installing Python. IDLE screenshot
is shown in the gure1.1.
All the examples given in this document are tested on python3. In case, if you do not have any
computer with Python installed try online2

1.1.1 Two modes of using Python Interpreter

Python statements can be entered at the '>>>' prompt available on the IDLE. Use this method
only for viewing the results of single Python statements, for example to use Python as a calculator.
This method could be confusing when you start writing larger programs, having looping etc.. The
preferred method is to Open a le from the menu and enter your source code inside the text editor
window, save it to a le (with .py extension) and Run it.
In this document, we will start writing small programs showing the essential elements of the
language without going into the details. The reader is expected to run these example programs and
also to modify them in dierent ways.It is like learning to drive a car, you master it by practicing.
Let us start with a program to display the words Hello World on the computer screen. This
is the customary 'hello world' program. The Python 'hello world' program has only one line, as
shown below.
print ('Hello World')
1 http://docs.python.org/tutorial/
2 https://trinket.io/python3

1
CHAPTER 1. PROGRAMMING IN PYTHON 2

Figure 1.1: IDLE IDE

CHAPTER 1. PROGRAMMING IN PYTHON 3

You can type this at the >>> prompt and press the Enter key to view the result. After trying
this method you should open a new le from the File menu of IDLE and type the one line of code
inside it. Save it as 'hello.py' and execute it from the Run menu. Pressing the F5 key also does
the job. You need to save the code after making changes.
The lenames of python programs should not be any of the the keywords of python. For
example, naming a le as 'string.py' or 'int.py' may result in hard to track error messages.

1.2 Variables and Data Types

As mentioned earlier, any high level programming language should support several data types.
The problem to be solved is represented using variables belonging to the supported data types.
Python supports numeric data types like integers, oating point numbers and complex numbers.
To handle character strings, it uses the String data type. Python also supports other compound
data types like lists, tuples, dictionaries etc.
In languages like C, C++ and Java, we need to explicitly declare the type of a variable. This
is not required in Python. The data type of a variable is decided by the value assigned to it. This
is called dynamic data typing. The type of a particular variable can change during the execution
of the program. If required, one type of variable can be converted in to another type by explicit
type casting, like y = f loat(3). Strings are enclosed within single quotes or double quotes.
The program rst.py shows how to dene variables of dierent data types. It also shows how
to embed comments inside a program.
Example: rst.py
'
A multi-line comment, within a pair of three single quotes.
In a line, anything after a # sign is also a comment
'
x = 10
print (x, type(x)) # print x and its type
x = 10.4
print (x, type(x))
x = 3 + 4j
print (x, type(x))
x = 'I am a String '
print (x, type(x))

The output of the program is shown below. Note that the type of the variable x changes during
the execution of the program, depending on the value assigned to it.
10 <type 'int'>
10.4 <type 'float'>
(3+4j) <type 'complex'>
I am a String <type 'str'>

The program treats the variables like humans treat labelled envelopes. We can pick an envelope,
write some name on it and keep something inside it for future use. In a similar manner the program
creates a variable, gives it a name and keeps some value inside it, to be used in subsequent steps.
So far we have used four data types of Python: int, oat, complex and str. To become familiar with
them, you may write simple programs performing arithmetic and logical operations using them.
CHAPTER 1. PROGRAMMING IN PYTHON 4

Example: oper.py
x = 2
y = 4
print(x + y * 2)
s = 'Hello '
print(s + s)
print(3 * s)
print(x == y)
print(y == 2 * x)
print(5/2 , 5//2)

Running the program oper.py will generate the following output.

10
Hello Hello
Hello Hello Hello
False
True
2.5 2

Note that a String can be added to another string and it can be multiplied by an integer. Try
to understand the logic behind that and also try adding a String to an Integer to see what is the
error message you will get. We have used the logical operator == for comparing two variables.
The last line demonstrates the nature of division operator on python3. By default it does oating
point division. Integer division can be forced by the operator //.

1.3 Operators and their Precedence

Python supports a large number of arithmetic and logical operators. They are summarized in the
table 1.1. An important thing to remember is their precedence. In the expression 2+3*4, is the
addition done rst or the multiplication? According to elementary arithmetics, the multiplication
should be done rst. It means that the multiplication operator has higher precedence than the
addition operator. If you want the addition to be done rst, enforce it by using parenthesis like
(2 + 3) ∗ 4. Whenever there is ambiguity in evaluation, use parenthesis to clarify the order of
evaluation.

1.4 Python Strings

So far we have come across four data types: Integer, Float, Complex and String. Out of which,
String is somewhat dierent from the other three. It is a collection of same kind of elements,
characters. The individual elements of a String can be accessed by indexing as shown in string1.py.
String is a compound, or collection, data type.
Example: string1.py
s = 'hello world'
print(s[0]) # print first element, h
print(s[1]) # print e
print(s[-1]) # will print the last character
CHAPTER 1. PROGRAMMING IN PYTHON 5

Operator Description Expression Result

or Boolean OR 0 or 4 4
and Boolean AND 3 and 0 0
not x Boolean NOT not 0 True
in, not in Membership tests 3 in [2.2,3,12] True
<, <=, >, >=, !=, == Comparisons 2>3 False
| Bitwise OR 1|2 3
^ Bitwise XOR 1^5 4
& Bitwise AND 1&3 1
<<, >> Bitwise Shifting 1 << 3 8
+ , - Add, Subtract 6-4 2
*, / Multiply, divide 5/2 2.5
% reminder 5%2 1
// integer division 3/2 1
+x , -x Positive, Negative -5*2 -10
~ Bitwise NOT ~1 -2
** Exponentiation 2 ** 3 8
x[index] Subscription a='abcd' ; a[1] 'b'
Table 1.1: Operators in Python listed according to their precedence.

Addition and multiplication is dened for Strings, as demonstrated by string2.py.

Example: string2.py
a = 'hello'+'world'
print a
b = 'ha' * 3
print b
print a[-1] + b[0]

will give the output

helloworld
hahaha
dh
The last element of a and rst element of b are added, resulting in the string 'dh'

1.4.1 Slicing
Part of a String can be extracted using the slicing operation. It can be considered as a modied
form of indexing a single character. Indexing using s[a : b] extracts elements s[a] to s[b − 1]. We
can skip one of the indices. If the index on the left side of the colon is skipped, slicing starts from
the rst element and if the index on right side is skipped, slicing ends with the last element.
Example: slice.py
a = 'hello world'
print(a[3:5])
CHAPTER 1. PROGRAMMING IN PYTHON 6

print(a[6:])
print(a[:5])

The reader can guess the nature of slicing operation from the output of this code, shown below.
'lo'
'world'
'hello'

Please note that specifying a right side index more than the length of the string is equivalent to
skipping it. Modify slice.py to print the result of a[6 : 20] to demonstrate it.

1.5 Python Lists and Tuples

List is an important data type of Python. It is much more exible than String. The individual
elements can be of any type, even another list. Lists are dened by enclosing the elements inside
a pair of square brackets, separated by commas. The program list1.py denes a list and print its
elements.
Example: list1.py
a = [2.3, 3.5, 234] # make a list
print (a[0])
a[1] = 'haha' # Change an element
print (a)

The output is shown below.

2.3
[2.3, 'haha', 234]
Lists can be sliced in a manner similar to that if Strings. List addition and multiplication are
demonstrated by the following example. We can also have another list as an element of a list.
Example: list2.py
a = [1,2]
print (a * 2)
print (a + [3,4])
b = [10, 20, a]
print (b)

The output of this program is shown below.

[1, 2, 1, 2]
[1, 2, 3, 4]
[10, 20, [1, 2] ]

1.5.1 Tuples
Tuples are similar to lists, but they are immutable.
a = (2.3, 3.5, 234) # make a tuple
a[1] = 10 # will give error
CHAPTER 1. PROGRAMMING IN PYTHON 7

1.6 Mutable and Immutable Types

There is one major dierence between String and List types, List is mutable but String is not.
We can change the value of an element in a list, add new elements to it and remove any existing
element. This is not possible with String type. Uncomment the last line of third.py and run it to
clarify this point.
Example: third.py
s = [3, 3.5, 234] # make a list
s[2] = 'haha' # Change an element
print s
x = 'myname' # String type
#x[1] = 2 # uncomment to get ERROR

The List data type is very exible, an element of a list can be another list. We will be using lists
extensively in the coming chapters. Tuple is another data type similar to List, except that it is
immutable. List is dened inside square brackets, tuple is dened in a similar manner but inside
parenthesis, like (3, 3.5, 234).

1.7 Python Dictionary

We are familiar with Strings, Lists and Tuples, where the individual items are accessed using
indices starting from zero. In the case of a Dictionary, data is stored in key:value pairs. Instead of
the indices, we have keys. A dictionary is a collection which is ordered, changeable and does not
allow duplicates.
a = {'a':12, 'b':23, 34:3000} # A dictionary
print (a)
print (a.keys())
print (a.vaues())

We can access the values using the keys

print (a[34]) # prints 3000
for k in a.keys():
print (a[k]) # prints all the values

Dictionary does not allow duplicate keys. If you dene it with duplicte keys, the last one will
prevail.
a = {'a':12, 'b':23, 34:3000, 'a':2000}
print (a)

will print
{'a': 2000, 'b': 23, 34: 3000} , it can be seen that the key 'a' gets a value of 2000.
CHAPTER 1. PROGRAMMING IN PYTHON 8

1.8 Input from the Keyboard

Since most of the programs require some input from the user, let us introduce this feature before
proceeding further. In python3, the input() function gets String type data from the keyboard, that
can be converted to integer, oat or complex by using eval() function. A message to be displayed
can be given as an argument while calling these functions.3
Example: kin1.py
x = eval(input('Enter an integer '))
y = eval(input('Enter a complex number, (like 2+5j) ')
print ('The sum is ', x + y)
s = input('Enter a String ')
print ('You entered ', s)

It is also possible to read more than one variable using a single input() statement. String type
data read using input() may be converted into integer or oat type if they contain only the valid
characters. In order to show the eect of conversion explicitly, we multiply the variables by 2
before printing. Multiplying a String by 2 prints it twice. If the String contains invalid characters
then eval() will give error.
Example: kin2.py
x,y = eval(input('Enter x and y separated by comma '))
print 'The sum is ', x + y
s = input('Enter a decimal number ')
a = eval(s)
print s * 2 # prints the string twice
print a * 2 # prints the converted numeric value times 2

We have learned about the basic data types of Python and how to get input data from the keyboard.
This is enough to try some simple problems and algorithms to solve them.
Example: area.py
pi = 3.1416
r = input('Enter Radius ')
a = pi * r ** 2 # A = πr2
print ('Area = ', a)

The above example calculates the area of a circle. Line three calculates r2 using the exponentiation
operator ∗∗, and multiply it with π using the multiplication operator ∗. r2 is evaluated rst because
** has higher precedence than *, otherwise the result would be (πr)2 .

1.9 Python Syntax, Colon & Indentation

Python was designed to be a highly readable language. It has a relatively uncluttered visual layout,
uses English keywords frequently where other languages use punctuation, and has notably fewer
syntactic constructions than other popular structured languages.
3 Functions will be introduced later. For the time being, understand that it is an isolated piece of code, called
from the main program with some input arguments and returns some output.
CHAPTER 1. PROGRAMMING IN PYTHON 9

There are mainly two things to remember about Python syntax: indentation and colon. Python
uses indentation to delimit blocks of code. Both space characters and tab characters are
currently accepted as forms of indentation in Python. Mixing spaces and tabs can create
bugs that are hard to track, since the text editor does not show the dierence. There should
not be any extra white spaces in the beginning of any line.
The line before any indented block must end with a colon (:) character.
1.10 Iteration: while and for loops
If programs can only execute from the rst line to the last in that order, as shown in the previous
examples, it would be impossible to write any useful program. For example, we need to print the
multiplication table of eight. Using our present knowledge, it would look like the following
Example: badtable.py
print (1 * 8)
print (2 * 8)
print (3 * 8)
print (4 * 8)
print (5 * 8)

Well, we are stopping here and looking for a better way to do this job.
The solution is to use the while loop of Python. The logical expression in front of while
is evaluated, and if it is True, the body of the while loop (the indented lines below the while
statement) is executed. The process is repeated until the condition becomes false. We should
have some statement inside the body of the loop that will make this condition false after few
iterations. Otherwise the program will run in an innite loop and you will have to press Control-C
to terminate it.
The program table.py, denes a variable x and assigns it an initial value of 1. Inside the while
loop x ∗ 8 is printed and the value of x is incremented. This process will be repeated until the
value of x becomes greater than 10.
Example: table.py
x = 1
while x <= 10:
print (x * 8)
x = x + 1

As per the Python syntax, the while statement ends with a colon and the code inside the while
loop is indented. Indentation can be done using tab or few spaces. In this example, we have
demonstrated a simple algorithm.

1.11 Syntax of 'for loops'

Python for loops are slightly dierent from the for loops of other languages. Python for loop
iterates over a compound data type like a String, List or Tuple. During each iteration, one member
of the compound data is assigned to the loop variable. The exibility of this can be seen from the
examples below.
CHAPTER 1. PROGRAMMING IN PYTHON 10

Example: forloop.py
a = 'Hello'
for ch in a: # ch is the loop variable
print (ch)
b = ['haha', 3.4, 2345, 3+5j]
for item in b:
print (item)

which gives the output :

H
e
l
l
o
haha
3.4
2345
(3+5j)
For constructing for loops that executes a xed number of times, we can use the range() function
and run the for loop over that.
Example: forloop2.py
for item in mylist:
print item

The output will look like :

0
1
2
3
4
In python3 the output of range() function is an object belonging to a class called range. It used
to be list type in python2. It is possible to specify the starting point and increment as arguments in
the form range(start, end+1, step). The following example prints the table of 5 using this feature.
Example: forloop3.py
mylist = range(5,51,5)
for item in mylist:
print (item , end = ' ')

The output is shown below.

5 10 15 20 25 30 35 40 45 50
The print statement inserts a newline at the end by default. The end = ' ' argument changes
this to a space.
In some cases, we may need to traverse the list to modify some or all of the elements. This
can be done by looping over a list of indices generated by the range() function.For example, the
program forloop4.py zeros all the elements of the list.
Example: forloop4.py
CHAPTER 1. PROGRAMMING IN PYTHON 11

a = [2, 5, 3, 4, 12]
size = len(a)
for k in range(size):
a[k] = 0
print a

The output is
[0, 0, 0, 0, 0]

1.12 Conditional Execution: if, elif and else

In some cases, we may need to execute some section of the code only if certain conditions are true.
Python implements this feature using the if, elif and else keywords, as shown in the next example.
The indentation levels of if and the corresponding elif and else must be kept the same.
Example: compare.py
x = input('Enter a string ')
if x == 'hello':
print ('You typed ', x)

Example: big.py
x = eval(input('Enter a number '))
if x > 10:
print ('Bigger Number')
elif x < 10:
print ('Smaller Number')
else:
print ('Same Number')

The statement x > 10 and x < 15 can be expressed in a short form, like 10 < x < 15.
The next example uses while and if keywords in the same program. Note the level of indentation
when the if statement comes inside the while loop. Remember that, the if statement must be
aligned with the corresponding elif and else.
Example: big2.py
x = 1
while x < 11:
if x < 5:
print ('Small ', x)
else:
print ('Big ', x)
x = x + 1
print ('Done')
CHAPTER 1. PROGRAMMING IN PYTHON 12

1.13 Modify loops : break and continue

We can use the break statement to terminate a loop, if some condition is met. The continue
statement is used to skip the rest of the block and go to the beginning again. Both are demonstrated
in the program big3.py shown below.
Example: big3.py
x = 1
while x < 10:
if x < 3:
print ('skipping work', x)
x = x + 1
continue
print (x)
if x == 4:
print ('Enough of work')
break
x = x + 1
print ('Done')
The output of big3.py is listed below.
skipping work 1
skipping work 2
3
4
Enough of work
Done
Now let us write a program to nd out the largest positive number entered by the user. The
algorithm works in the following manner. To start with, we assume that the largest number is
zero. After reading a number, the program checks whether it is bigger than the current value of
the largest number. If so the value of the largest number is replaced with the current number. The
program terminates when the user enters zero. Modify max.py to work with negative numbers
also.
Example: max.py
max = 0
while True: # Infinite loop
x = eval(input('Enter a number '))
if x > max:
max = x
if x == 0:
print (max)
break

1.14 Line joining

Python interpreter processes the code line by line. A program may have a long line of code that
may not physically t in the width of the text editor. In such cases, we can split a logical line
of code into more than one physical lines, using backslash characters (\), in other words multiple
physical lines are joined to form a logical line before interpreting it.
CHAPTER 1. PROGRAMMING IN PYTHON 13

if 1900 < year < 2100 and 1 <= month <= 12 :

can be split like

if 1900 < year < 2100 \
and 1 <= month <= 12 :

Do not split in the middle of words except for Strings. A long String can be split as shown below.
longname = 'I am so long and will \
not fit in a single line'
print (longname)

1.15 Exercises
We have now covered the minimum essentials of Python; dening variables, performing arithmetic
and logical operations on them and the control ow statements. These are sucient for handling
most of the programming tasks. It would be better to get a grip of it before proceeding further,
by writing some code.
1. Modify the expression print 5+3*2 to get a result of 16
2. What will be the output of print (type(4.5))
3. Print all even numbers upto 30, suxed by a * if the number is a multiple of 6. (hint: use
% operator)
4. Write Python code to remove the last two characters of 'I am a long string' by slicing, without
counting the characters. (hint: use negative indexing)
5. s = '012345' . (a) Slice it to remove last two elements (b) remove rst two element.
6. a = [1,2,3,4,5]. Use Slicing and multiplication to generate [2,3,4,2,3,4] from it.
7. Compare the results of 5/2, 5//2 and 2.0/3.
8. Print the following pattern using a while loop
+
++
+++
++++
9. Write a program to read inputs like 8A, 10C etc. and print the integer and alphabet parts
separately.
10. Write code to print a number in the binary format (for example 5 will be printed as 101)
11. Write code to print all perfect cubes upto 2000.
12. Write a Python program to print the multiplication table of 5.
13. Write a program to nd the volume of a box with sides 3,4 and 5 inches in cm3 ( 1 inch =
2.54 cm)
CHAPTER 1. PROGRAMMING IN PYTHON 14

14. Write a program to nd the percentage of volume occupied by a sphere of diameter r tted
in a cube of side r. Read r from the keyboard.
15. Write a Python program to calculate the area of a circle.
16. Write a program to divide an integer by another without using the / operator. (hint: use -
operator)
17. Count the number of times the character 'a' appears in a String read from the keyboard.
Keep on prompting for the string until there is no 'a' in the input.
18. Create an integer division machine that will ask the user for two numbers then divide and
give the result. The program should give the result in two parts: the whole number result
and the remainder. Example: If a user enters 11 / 4, the computer should give the result 2
and remainder 3.
19. Modify the previous program to avoid division by zero error.
20. Create an adding machine that will keep on asking the user for numbers, add them together
and show the total after each step. Terminate when user enters a zero.
21. Modify the adding machine to check for errors like user entering invalid characters.
22. Create a script that will convert Celsius to Fahrenheit. The program should ask the users to
enter the temperature in Celsius and should print out the temperature in Fahrenheit, using
f = 59 c + 32.
23. Write a program to convert Fahrenheit to Celsius.
24. Create a script that uses a variable and will write 20 times "I will not talk in class." Make
each sentence on a separate line.
25. Dene 2 + 5j and 2 − 5j as complex numbers , and nd their product. Verify the result by
dening the real and imaginary parts separately and using the multiplication formula.
26. Write the multiplication table of 12 using while loop.
27. Write the multiplication table of a number, from the user, using for loop.
28. Print the powers of 2 up to 1024 using a for loop. (only two lines of code)
29. Dene the list a = [123, 12.4, 'haha', 3.4]
a) print all members using a for loop
b) print the oat type members ( use type() function)
c) insert a member after 12.4
d) append more members
30. Make a list containing 10 members using a for loop.
31. Generate multiplication table of 5 with two lines of Python code. (hint: range function)
32. Write a program to nd the sum of ve numbers read from the keyboard.
33. Write a program to read numbers from the keyboard until their sum exceeds 200. Modify
the program to ignore numbers greater than 99.
34. Write a Python function to calculate the GCD of two numbers
35. Write a Python program to nd annual compound interest. Get P,N and R from user
CHAPTER 1. PROGRAMMING IN PYTHON 15

1.16 Functions
Large programs need to be divided into small logical units. A function is generally an isolated
unit of code that has a name and performs a well dened job. A function groups several program
statements into a unit and gives it a name. This unit can be invoked from other parts of a program.
Python allows you to dene functions using the def keyword. A function may have one or more
variables as parameters, which receive their values from the calling program.
In the example shown below, function parameters (a and b) get the values 3 and 4 respectively
from the caller. One can specify more than one variables in the return statement, separated by
commas. The function will return a tuple containing those variables. Some functions may not
have any arguments, but while calling them we need to use an empty parenthesis, otherwise the
function will not be invoked. If there is no return statement, a None is returned to the caller.
Example func.py
def sum(a,b): # a trivial function
return a + b

print (sum(3, 4))

The function factorial.py calls itself recursively. The value of argument is decremented before each
call. Try to understand the working of this by inserting print statements inside the function.
Example factor.py
def factorial(n): # a recursive function
if n == 0:
return 1
else:
return n * factorial(n-1)

print (factorial(10))

Example bonacci.py
def fib(n): # print Fibonacci series up to n
a, b = 0, 1
while b < n:
print (b)
a, b = b, a+b

print fib(30)

Runing bonacci.py will print

1 1 2 3 5 8 13 21
Modify the code to replace a, b = b, a + b by two separate assignment statements, if required
introduce a third variable.
CHAPTER 1. PROGRAMMING IN PYTHON 16

1.16.1 Scope of variables

The variables dened inside a function are not known outside the function. There could be two
unrelated variables, one inside and one outside, having the same name. The program scope.py
demonstrates this feature.
Example scope.py
def change(x):
counter = x

counter = 10
change(5)
print (counter)

The program will print 10 and not 5. The two variables, both named counter, are not related
to each other. In some cases, it may be desirable to allow the function to change some external
variable. This can be achieved by using the global keyword, as shown in global.py.
Example global.py
def change(x):
global counter # use the global variable
counter = x

counter = 10
change(5)
print (counter)

The program will now print 5. Functions with global variables should be used carefully, to avoid
inadvertent side eects.

1.16.2 Optional and Named Arguments

Python allows function arguments to have default values; if the function is called without a partic-
ular argument, its default value will be taken. Due to this feature, the same function can be called
with dierent number of arguments. The arguments without default values must appear rst in
the argument list and they cannot be omitted while invoking the function. The following example
shows a function named power() that does exponentiation, but the default value of exponent is set
to 2.
Example power.py
def power(mant, exp = 2.0):
return mant ** exp

print (power(5., 3))

print (power(4.)) # prints 16
print (power()) # Gives Error

Arguments can be specied in any order by using named arguments.

Example named.py
CHAPTER 1. PROGRAMMING IN PYTHON 17

def power(mant = 10.0, exp = 2.0):

return mant ** exp

print (power(5., 3))

print (power(4.)) # prints 16
print (power(exp=3)) # mant gets 10.0, prints 1000

1.17 More on Strings and Lists

Before proceeding further, we will explore some of the functions provided for manipulating strings
and lists. Python strings can be manipulated in many ways. The following program prints the
length of a string, makes an upper case version for printing and prints a help message on the String
class.
Example: stringhelp.py
s = 'hello world'
print (len(s))
print (s.upper())
help(str) # press q to exit help

Python is an object oriented language and all variables are objects belonging to various classes.
The method upper() (a function belonging to a class is called a method) is invoked using the dot
operator. All we need to know at this stage is that there are several methods that can be used for
manipulating objects and they can be invoked like: variable_name.method_name().

1.17.1 split and join

Splitting a String will result in a list of smaller strings. If you do not specify the separator, the
space character is assumed by default. To demonstrate the working of these functions, few lines
of code and its output are listed below.
Example: split.py
s = 'I am a long string'
print (s.split())
a = 'abc.abc.abc'
aa = a.split('.')
print (aa)
mm = '+'.join(aa)
print (mm)

The result is shown below

['I', 'am', 'a', 'long', 'string']
['abc', 'abc', 'abc']
'abc+abc+abc'
The List of strings generated by split is joined using '+' character, resulting in the last line of the
output.
CHAPTER 1. PROGRAMMING IN PYTHON 18

1.17.2 Manipulating Lists

Python lists are very exible, we can append, insert, delete and modify elements of a list. The
program list3.py demonstrates some of them.
Example: list3.py
a = [] # make an empty list
a.append(3) # Add an element
a.insert(0,2.5) # insert 2.5 as first element
print (a, a[0])
print (len(a))

The output is shown below.

[2.5, 3] 2.5
2

1.17.3 Copying Lists

Lists cannot be copied like numeric data types. The statement b = a will not create a new list b
from list a, it just make a reference to a. The following example will clarify this point. To make a
duplicate copy of a list, we need to use the copy module.
Example: list_copy.py
a = [1,2,3,4]
print (a)
b = a # b refers to a
print (a == b) # True
b[0] = 5 # Modifies a[0]
print (a)
import copy
c = copy.copy(a)
c[1] = 100
print (a is c) # is False
print (a, c)

The output is shown below.

[1, 2, 3, 4]
True
[5, 2, 3, 4]
False
[5, 2, 3, 4] [5, 100, 3, 4]

1.18 Python Modules and Packages

One of the major advantages of Python is the availability of libraries for various applications like
graphics, networking and scientic computation. The standard library distributed with Python
itself has a large number of modules: time, random, sys, etc. are some of them. The site
http://docs.python.org/library/ has the complete reference.
CHAPTER 1. PROGRAMMING IN PYTHON 19

Modules are loaded by using the import keyword. Several ways of using import is explained
below, using the math (containing mathematical functions) module as an example.4

1.18.1 Dierent ways to import

simplest way to use import is shown in mathsin.py, where the function is invoked using the form
module_name.function_name(). In the next example, we use an alias for the module name.
Example mathsin.py
import math
print (math.sin(0.5)) # module_name.method_name

Example mathsin2.py
import math as m # Give another name for math
print (m.sin(0.5)) # Refer by the new name

We can also import the functions to behave like local (like the ones within our source le) function,
as shown below. The character * is a wild card for importing all the functions.
Example mathlocal.py
from math import sin # sin is imported as local
print (sin(0.5))

Example mathlocal2.py
from math import * # import everything from math
print (sin(0.5))

In the third and fourth cases, we need not type the module name every time. But there could be
trouble if two modules imported contains a function with same name. In the program conict.py,
the sin() from numpy is capable of handling a list argument. After importing math.py, line 4, the
sin function from math module replaces the one from numpy. The error occurs because the sin()
from math can accept only a numeric type argument.
Example conict.py
from numpy import *
x = [0.1, 0.2, 0.3]
print (sin(x)) # numpy's sin can handle lists
from math import * # sin of math becomes effective
print (sin(x)) # will give ERROR
4 While giving names to your Python programs, make sure that you are not directly or indirectly importing any
Python module having same name. For example, if you create a program named math.py and keep it in your working
directory, the import math statement from any other program started from that directory will try to import your
le named math.py and give error. If you ever do that by mistake, delete all the les with .pyc extension from your
directory.
CHAPTER 1. PROGRAMMING IN PYTHON 20

1.18.2 Packages
Packages are used for organizing multiple modules. The module name A.B designates a module
named B in a package named A. The concept is demonstrated using the packages Numpy5 and
Scipy.
Example submodule.py
import numpy as np
print (np.random.normal())
import scipy.special
print (scipy.special.j0(.1))

In this example random is a module inside the package NumPy. Similarly special is a module inside
the package Scipy. We use both of them in the package.module.function() format. But there is
some dierence. In the case of Numpy, the random module is loaded by default, importing scipy
does not import the module special by default. This behavior can be dened while writing the
Package and it is up to the author of the package.

1.19 File Input/Output

Disk les can be opened using the function named open() that returns a File object. Files can be
opened for reading or writing. There are several methods belonging to the File class that can be
used for reading and writing data.
Example wle.py
f = open('test.dat' , 'w')
f.write ('This is a test file')
f.close()

Above program creates a new le named 'test.dat' (any existing le with the same name will be
deleted) and writes a String to it. The following program opens this le for reading the data.
Example rle.py
f = open('test.dat' , 'r')
print (f.read())
f.close()

Note that the data written/read are character strings. read() function can also be used to read a
xed number of characters, as shown below.
Example rle2.py
f = open('test.dat' , 'r')
print (f.read(7)) # get first seven characters
print (f.read()) # get the remaining ones
f.close()
5 NumPy will be discusssed later in chapter 2.
CHAPTER 1. PROGRAMMING IN PYTHON 21

Now we will examine how to read a text data from a le and convert it into numeric type. First
we will create a le with a column of numbers.
Example wle2.py
f = open('data.dat' , 'w')
for k in range(1,4):
s = '%3d\n' %(k)
f.write(s)
f.close()

The contents of the le created will look like this.

1
2
3
Now we write a program to read this le, line by line, and convert the string type data to integer
type, and print the numbers.6
Example rle3.py
f = open('data.dat') # default is read
s = f.readlines()
for item in s:
print (eval(item))
f.close()

1.20 Formatted Printing

Formatted printing is done by using a format string followed by the % operator and the values
to be printed. If format requires a single argument, values may be a single variable. Otherwise,
values must be a tuple (just place them inside parenthesis, separated by commas) with exactly the
number of items specied by the format string.
Example: format.py
a = 2.0 /3
print (a)
print ('a = %5.3f' %(a)) # upto 3 decimal places

Data can be printed in various formats. The conversion types are summarized in the following
table. There are several ags that can be used to modify the formatting, like justication, lling
etc.
The following example shows some of the features available with formatted printing.
Example: format2.py
6 This will give error if there is a blank line in the data le. This can be corrected by changing the comparison
statement to if len(s) < 1: , so that the processing stops at a blank line. Modify the code to skip a blank line
instead of exiting (hint: use continue ).
CHAPTER 1. PROGRAMMING IN PYTHON 22

Conversion Conversion Example Result

d,i signed Integer '%6d'%(12) ' 12'
f oating point decimal '%6.4f'%(2.0/3) 0.667
e oating point exponential '%6.2e'%(2.0/3) 6.67e-01
x hexadecimal '%x'%(16) 10
o octal '%o'%(8) 10
s string '%s'%('abcd') abcd
0d modied 'd' '%05d'%(12) 00012
Table 1.2: Formatted Printing in Python

a = 'justify as you like'

print ('%30s'%a) # right justified
print ('%-30s'%a) # minus sign for left justification
for k in range(1,11): # A good looking table
print ('5 x %2d = %2d' %(k, k*5))

The output of format2.py is given below.

justify as you like
justify as you like
5 x 1 = 5
5 x 2 = 10
5 x 3 = 15
5 x 4 = 20
5 x 5 = 25
5 x 6 = 30
5 x 7 = 35
5 x 8 = 40
5 x 9 = 45
5 x 10 = 50

1.21 Exception Handling

Errors detected during execution are called exceptions, like divide by zero. If the program does
not handle exceptions, the Python Interpreter reports the exception and terminates the program.
We will demonstrate handling exceptions using try and except keywords, in the example except.py.
Example: except.py
x = input('Enter a number ')
try:
print (10.0/x)
except:
print ('Division by zero not allowed')

If any exception occurs while running the code inside the try block, the code inside the except
block is executed. The following program implements error checking on input using exceptions.
Example: except2.py
CHAPTER 1. PROGRAMMING IN PYTHON 23

def get_number():
while 1:
try:
a = input('Enter a number ')
x = atof(a)
return x
except:
print ('Enter a valid number')

print get_number()

1.22 Exercises
1. Generate multiplication table of eight and write it to a le.
2. Write a Python program to open a le and write 'hello world' to it.
3. Write a Python program to open a text le and read all lines from it.
4. Write a program to generate the multiplication table of a number from the user. The output
should be formatted as shown below
1x5= 5
2 x 5 = 10
5. Write Python code to generate the sequence of numbers
25 20 15 10 5
using the range() function. Delete 15 from the result and sort it. Print it using a for loop.
6. Dene a string s = 'mary had a little lamb'.
a) print it in reverse order
b) split it using space character as separator
7. Join the elements of the list ['I', 'am', 'in', 'pieces'] using + character. Do the same using a
for loop also.
8. Create a program that will check a sentence to see if it is a palindrome. A palindrome is a
sentence that reads the same backwards and forwards ('malayalam').
9. Read a String from the keyboard. Multiply it by an integer to make its length more than
50. How do you nd out the smallest number that does the job.
10. Write a program using for loop to reverse a string.
11. Find the syntax error in the following code and correct it.
x=1
while x <= 10:
print x * 5
12. Write a function that returns the sum of numbers in a list.
13. Write a function returns are area of a circle, default radius should be taken as 1.
14. Split 'hello world' and join back to get 'hello+world'.
CHAPTER 1. PROGRAMMING IN PYTHON 24

15. Read a 5 digit integer from keyboard and print the sum of the digits.
16. Modify the list [1,2,3,4] to make it [1,2,3,10]
Chapter 2

Arrays and Matrices

The list data type is not really suitable for representing arrays. In this chapter, we will use Numpy
arrays to represent one and two dimensional arrays.

2.1 NumPy Arrays

NumPy is a Python extension module that provides ecient operation on arrays of homogeneous
data. First thing to learn is how to create and manipulate arrays using numpy. There are several
functions that can be used for creating arrays.1 The mathematical functions like sine, cosine etc.
of numpy accepts array objects as arguments and return the results as arrays objects. NumPy
arrays can be indexed, sliced and copied like Python Lists.
The following examples shows how to make a one and two dimensional arrays.
Example numpy1.py
import numpy as np
x = np.array( [1., 2, 3] ,dtype=float) # Make array from list
print (x , type(x))
In the above example, we have created a numpy ndarray type object from a list. To make all
members oat, either you specify the datatype explicitly or make at least one member oat. In
the code, both are done, just for demonstration.

Example numpy2.py
import numpy as np
a = [ [1, 2, 3] , [4, 5, 6] ] # make a list of lists
x = np.array(a) # and convert to an array
print (x)
[[1 2 3]
[4 5 6]]
Other than than array(), there are several other functions that can be used for creating dierent
types of arrays. Some of them are described below.
1 https://docs.scipy.org/doc/numpy-1.13.0/reference/routines.array-creation.html

25
CHAPTER 2. ARRAYS AND MATRICES 26

2.1.1 arange(start, stop, step)

Creates an evenly spaced one-dimensional array. Start, stop, step-size and datatype are the argu-
ments. If datatype is not given, it is deduced from the other arguments. Note that, the values are
generated within the interval, including start but excluding stop.
np.arange(2.0, 3.0, .1)
makes the array([ 2. , 2.1, 2.2, 2.3, 2.4, 2.5, 2.6, 2.7, 2.8, 2.9])

2.1.2 linspace(start, stop, number of elements)

Similar to arange(). Start, stop and number of samples are the arguments.
np.linspace(1, 2, 11) is equivalent to array([ 1. , 1.1, 1.2, 1.3, 1.4, 1.5, 1.6, 1.7, 1.8, 1.9, 2. ])

2.1.3 zeros(shape)
Returns a new array of given shape and type, lled zeros. For example np.zeros( [2,3]) generates
a 2 x 3 array lled with zeros
[[0.,0.,0.]
[0.,0.,0.]]

2.1.4 ones(shape)
Similar to np.zeros() except that the elements are initialized to 1.

2.1.5 random.random(shape)
Similar to the functions above, but the matrix is lled with random numbers ranging from 0 to 1,
of oat type. For example, np.random.random([3,3]) will generate the 3x3 matrix;
array([[ 0.3759652 , 0.58443562, 0.41632997],
[ 0.88497654, 0.79518478, 0.60402514],
[ 0.65468458, 0.05818105, 0.55621826]])

2.1.6 reshape(array, newshape)

We can also make multi-dimensions arrays by reshaping a one-dimensional array. The function
reshape() changes dimensions of an array. The total number of elements must be preserved.
Working of reshape() can be understood by looking at reshape.py and its result.
Example reshape.py
import numpy as np
a = np.arange(20)
b = np.reshape(a, [4,5])
print (b)
The result is shown below.
array([[ 0, 1, 2, 3, 4],
[ 5, 6, 7, 8, 9],
[10, 11, 12, 13, 14],
[15, 16, 17, 18, 19]])
CHAPTER 2. ARRAYS AND MATRICES 27

The program numpy2.py demonstrates most of the functions discussed so far.

Example numpy2.py
import numpy as np
a = np.arange(1.0, 2.0, 0.1)
print (a)
b = np.linspace(1,2,11)
print (b)
c = np.ones(5,'float')
print (c)
d = zeros(5, 'int')
print (d)
e = np.random.rand(5)
print (e)

Output of this program will look like;

[ 1. 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9]
[ 1. 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2. ]
[ 1. 1. 1. 1. 1.]
[ 0. 0. 0. 0. 0.]
[ 0.89039193 0.55640332 0.38962117 0.17238343 0.01297415]

2.1.7 Copying
Numpy arrays can be copied using the copy method, as shown below.
Example array_copy.py
import numpy as np
a = np.zeros(5)
print a
b = a
c = a.copy()
c[0] = 10
print a, c
b[0] = 10
print a,c

The output of the program is shown below. The statement b = a does not make a copy of a.
Modifying b aects a, but c is a separate entity.
[ 0. 0. 0.]
[ 0. 0. 0.] [ 10. 0. 0.]
[ 10. 0. 0.] [ 10. 0. 0.]

2.1.8 Saving and Restoring

There are many ways for reading from le and writing to data les in numpy. We will discuss only
the ones saves data in the formatted text form, savetext() and loadtext() methods, as shown in
the code leio.py
Example leio.py
CHAPTER 2. ARRAYS AND MATRICES 28

import numpy as np
a = np.array([[2,3], [4,5]])
print(a)
np.savetxt('mydata.txt', a, fmt = '%2.3f', delimiter = ' ')
print('Data saved to text file')
b = np.loadtxt("mydata.txt", delimiter = ' ')
print('Data loaded from text file')
print(b)

The output of the program will look like

[[2 3]
[4 5]]
Data saved to text file
Data loaded from text file
[[ 2. 3.]
[ 4. 5.]]

The text le 'mydata.txt' will look like

2.000 3.000
4.000 5.000

2.1.9 Slicing to extract elements and columns

Since slicing is used for extracting columns from a two dimensional array, we need to understand
it a bit better.
x = np.reshape(range(12),(4,3))

creates the array having 4 rows and 3 columns.

[[ 0, 1, 2],
[ 3, 4, 5],
[ 6, 7, 8],
[ 9, 10, 11]]

and extract elements from it in dierent ways, python indexing starts from zero.
x[1] extracts the second row to give [3,4,5].
x[2,1] extracts 7, the element at third row and second column.
x[1:3, 1] extracts the elements [4,7] from second column.
x[ :, 1} gives the full second column [ 1, 4, 7, 10]
x[:2,:2] gives the array
[[0, 1],
[3, 4]]

You should remember that a statement like y = x[1] does not create a new array. It is only a
reference to the second row or x . The statement y[0] = 100 changes x[1, 0]to 100.
The program slice.py
CHAPTER 2. ARRAYS AND MATRICES 29

import numpy as np
x = np.reshape(range(12),(4,3))
print(x[1])
print(x[2,1])
print(x[1:3,1])
print(x[:,1])
print(x[:2,:2])
will generate the output
[3 4 5]
7
[4 7]
[ 1 4 7 10]
[[0 1]
[3 4]]

2.1.10 Arithmetic Operations

Arithmetic operations on array diers from that of ordinary numbers. Adding or multiplying an
array object with an ordinary number will multiply all the elements by that number. However,
operations involving more than one array could be done in dierent ways.
x = np.ones((2,2)) makes
[[ 1., 1.],
[ 1., 1.]]
x * 5 multiplies all elements by 5 to give
[[ 5., 5.],
[ 5., 5.]]

2.2 Exercises
1. Write code to make a one dimensional array with elements 5,10,15,20 and 25. make another
matrix by slicing the rst three elements from it.
2. Create a 3 × 2 array and print the sum of its elements using for loops.
3. Create a 2 × 3 array and ll it with random numbers.
4. Use linspace to make an array from 0 to 10, with step-size of 0.1
5. Use arange to make an 100 element array ranging from 0 to 10
6. Make an array a = [2,3,4,5] and copy it to b. change one element of b and print both.
7. Make a 3x3 array and multiply it by 5.
8. Create two 3x3 arrays and add them.
9. Create a 4x3 matrix using range() and reshape().
10. Extract the second row from a 4x3 matrix
11. Extract the second column from a 3x3 matrix
12. create a 2D array [ [1,2] , [3,4] ]. Save it to a le, read it back and print.
Chapter 3

Data visualization
A graph or chart is used to present numerical data in visual form. A graph is one of the easiest
ways to compare numbers. They should be used to make facts clearer and more understandable.
Results of mathematical computations are often presented in graphical format. In this chapter, we
will explore the Python modules used for generating two and three dimensional graphs of various
types.

3.1 The Matplotlib Module

Matplotlib1 is a python package that produces publication quality gures in a variety of hardcopy
formats. It also provides many functions for matrix manipulation. You can generate plots, his-
tograms, power spectra, bar charts, error-charts, scatter-plots, etc, with just a few lines of code
and have full control of line styles, font properties, axes properties, etc. The data points to the
plotting functions are supplied as Python lists or Numpy arrays. The recommended method to
import Numpy and Matplotlib is like,
import numpy as np
import matplotlib.pyplot as plt
x = np.linspace(0,10,100)
plt.plot(x, x**2)
plt.show()

3.1.1 2D plots
Let us start with some simple plots to become familiar with matplotlib.
Example plot1.py
import matplotlib.pyplot as plt
data = [1,2,5]
plt.plot(data)
plt.show()
1 https://www.tutorialspoint.com/matplotlib/matplotlib_quick_guide.htm
https://matplotlib.org/
https://matplotlib.org/tutorials/introductory/pyplot.html#sphx-glr-tutorials-introductory-pyplot-py

30
CHAPTER 3. DATA VISUALIZATION 31

Figure 3.1: Output of (a) plot4.py (b) subplot1.py (c) piechart.py

In the above example, the x-axis of the three points is taken from 0 to 2. We can specify both the
axes as shown below.
Example plot2.py
import matplotlib.pyplot as plt
x = [1,2,5]
y = [4,5,6]
plt.plot(x,y)
plt.show()
By default, the color is blue and the line style is continuous. This can be changed by an optional
argument after the coordinate data, which is the format string that indicates the color and line
type of the plot. The default format string is `b-` (blue, continuous line). Let us rewrite the above
example to plot using red circles. We will also set the ranges for x and y axes and label them, as
shown in plot3.py.
Example plot3.py
import matplotlib.pyplot as plt
x = [1,2,5]
y = [4,5,6]
plt.plot(x,y,'ro')
plt.xlabel('x-axis')
plt.ylabel('y-axis')
plt.axis([0,6,1,7])
plt.show()
The gure 3.1 shows two dierent plots in the same window, using dierent markers and colors.
It also shows how to use legends.
Example plot4.py
import numpy as np
import matplotlib.pyplot as plt
t = np.arange(0.0, 5.0, 0.2)
plt.plot(t, t**2, 'x', label='t^2') # t2
plt.plot(t, t**3, 'ro', label='t^3') # t3
plt.legend(framealpha=0.5)
plt.show()
CHAPTER 3. DATA VISUALIZATION 32

We have just learned how to draw a simple plot using the pylab interface of matplotlib. Another
way to call numpy and matplotlib is like,
import numpy as np
import matplotlib.pyplot as plt
t = np.arange(0.0, 5.0, 0.2)
plt.plot(t, t**2, color='red', marker='x')
plt.show()

The markers and colors available are,

markers = ["." , "," , "o" , "v" , "^" , "<", ">"]
colors = ['r','g','b','c','m', 'y', 'k']

3.1.2 Polar plots

Polar coordinates locate a point on a plane with one distance and one angle. The distance `r' is
measured from the origin. The angle θ is measured from some agreed starting point. Use the
positive part of the x − axis as the starting point for measuring angles. Measure positive angles
anti-clockwise from the positive x − axis and negative angles clockwise from it.
Matplotlib supports polar plots, using the polar(θ, r) function. Let us plot a circle using polar().
For every point on the circle, the value of radius is the same but the polar angle θ changes from
0to 2π . Both the coordinate arguments must be arrays of equal size. Since θ is having 100 points
, r also must have the same number. This array can be generated using the ones() function. The
axis([θmin , θmax , rmin , rmax ) function can be used for setting the scale.
Example polar.py
import numpy as np
import matplotlib.pyplot as plt
th = np.linspace(0,2*pi,100)
r = 5 * np.ones(100) # radius = 5
plt.polar(th,r)
plt.show()

3.1.3 Pie Charts

An example of a pie chart is given below. The percentage of dierent items and their names are
given as arguments. The output is shown in gure 3.1(c).
Example piechart.py
import matplotlib.pyplot as plt
labels = 'Frogs', 'Hogs', 'Dogs', 'Logs'
fracs = [25, 25, 30, 20]
plt.pie(fracs, labels=labels)
plt.show()
Chapter 4

Introduction to Pandas
We have seen that NumPy handles homogeneous numerical array data. Pandas is designed for
working with tabular or heterogeneous data, like contents of a spreadsheet. Pandas shares many
coding idioms from NumPy. To get started with pandas, you will need to get familiar with its two
main data structures; Series and DataFrame.

4.1 Series
We have seen that the elements of a one dimensional array, of N elements, is indexed using integers
ranging from 0 to N-1. A Pandas Series data type can be considered as an extension to this, where
the indices also can be specied. If not specied it takes the default, integers ranging from 0 to
N-1.
We can create a Pandas Series using;
import pandas as pd
ser = pd.Series([4, 12, 55, 100])
print (ser)

Gives the output,

0 4
1 12
2 55
3 100
dtype: int64

The rst column is the indices and the second column are the values. It looks very similar to a
column in a spreadsheet. We can specify the indices also as shown below.
import pandas as pd
ser = pd.Series([4, 12, 55, 100], index=['a', 'b', 'c', 'd'])
print (ser)

Gives the output,

a 4
b 12

33
CHAPTER 4. INTRODUCTION TO PANDAS 34

c 55
d 100
dtype: int64

Another way to think about a Series is as a xed-length, ordered dictionary, as it is a mapping of

index values to data values. . A dictionary consists of key:value pairs, and the keys are used for
indexing the coresponding values. You can create a Series from a dictionary
d = {'b':12, 'a':23, 34:3000, 'a':230} # A dictionary
s = pd.Series(d)
print (s)

The output is
b 12
a 230
34 3000
dtype: int64

You may note that the duplicate entry of index 'a' has been removed, and the latest value (230)
has prevailed. This is the behavior of Dictionary.Pandas Index can contain duplicate labels ,
as demonstrated below.
import pandas as pd
ps = pd.Series([4, 7, -5, 3], index=['d', 'a', 'a', 'c'])
print(ps)

Gives the output

d 4
a 7
a -5
c 3
dtype: int64

It can be seen that the index 'a' is appearing twice. Pandas supports non-unique index values.
If an operation that does not support duplicate index values is attempted, an exception will be
raised at that time.
Index of a Series can be altered in-place by assignment. The exact number of indices should
be given.
ser = pd.Series([10, 20])
ser.index = ['a', 'b']
print (ser)

prints the following. The default indices 0 and 1 are replaced by 'a' and 'b' respectively.
a 10
b 20
dtype: int64
CHAPTER 4. INTRODUCTION TO PANDAS 35

Reindexing
It is possible to re-index a Series. This may alter the number of indices also. The relation between
an index and the corresponding value will be preserved during this operation.
import pandas as pd
ser = pd.Series([10, 20], index =['a', 'b'])
ser = ser.reindex(['b', 1, 2])
print (ser)
The result is
b 20.0
1 NaN
2 NaN
dtype: float64
It can be seen that the value corresponding to the old index 'b' is preserved. The new indices 1 and
2 are assigned with a NaN (Not a Number) value. The function isnull() tests for the NA values.
A key dierence between Series and ndarray is that operations between Series automatically align
the data based on index.
print (pd.isnull(ser))
prints
b False
1 True
2 True
dtype: bool
Series automatically aligns by index label in arithmetic operations, as demonstrated below.
s1 = pd.Series([10, 20, 30,40], index =[0,1,2,3])
s2 = pd.Series([12, 13, 14], index =[1,2,3])
s = s1 + s2
print (s)
The output is given below.
0 NaN
1 32.0
2 43.0
3 54.0
dtype: float64
0 s2
Index and the corresponding value are available in s1. But they are missing in , so the addition
of 10 and a non-existant value results in a NaN value. Values corresponding to the other three
indices are added. Data always align with the index.
Using NumPy functions or NumPy-like operations, such as ltering with a boolean array, scalar
multiplication, or applying math functions, will also preserve the index-value link. The default
result of operations between dierently indexed objects yield the union of the indexes in order to
avoid loss of information. You have the option of dropping labels with missing data via the dropna
function.
CHAPTER 4. INTRODUCTION TO PANDAS 36

print ( s.dropna() )
1 32.0
2 43.0
3 54.0
dtype: float64

Name attribute
Both the Series object itself and its index have a name attribute.
s = pd.Series([34, 56], index = ['John', 'Ravi'], name='ages')
s.index.name = 'names'
print (s.name)
print (s.index.name)
prints
ages
names

4.2 DataFrame
A DataFrame represents a tabular, spreadsheet-like data structure containing an ordered collection
of columns, each of which can be a dierent data type. It has both a row and column index.

It has both a row and column index. ; it can be thought of as a dictionary of Series (sharing
the same index). DataFrames can be dened in several dierent ways:

From a Dictionary of Series

s1 = pd.Series([1.0, 2.0, 3.0], index=["a", "b", "c"])
s2 = pd.Series([1.0, 2.0, 3.0, 4.0], index=["a", "b", "c", "d"])
d = { "one": s1, "two": s2 }
df = pd.DataFrame(d)
print (df)
will print
one two
a 1.0 1.0
b 2.0 2.0
c 3.0 3.0
d NaN 4.0
CHAPTER 4. INTRODUCTION TO PANDAS 37

It can be seen that each column is made from a Series, where the keys become the column names.
A union of all the Series indices becomes the DataFrame index column. If any Series is missing an
index, the corresponding value is assigned a NaN.

From a Dictionary of Lists/Arrays

d = {"one": [1.0, 2.0, 3.0, 4.0], "two": [4.0, 3.0, 2.0, 1.0]}
df = pd.DataFrame(d, index = ['a', 'b', 'c', 'd'])
print (df)

will print
one two
a 1.0 4.0
b 2.0 3.0
c 3.0 2.0
d 4.0 1.0

Number of elements in all the lists/arrays in the dictionary should match with the number of
elements in the index list. The value from each dictionary item forms a column with it's key as
the label. The values can be arrays or lists.
There are several other methods to dene a DataFrame, that are not discussed here. We will
focus more on creating DataFrames by loading table data from les.

Index Properties
Pandas Index can contain duplicate labels.

df = pd.DataFrame(np.arange(20).reshape(5,4), columns=["A", "A", "C", "D"])

df.index = [1,2,2,3,4]
print(df)

will print
A A C D
1 0 1 2 3
2 4 5 6 7
2 8 9 10 11
3 12 13 14 15
4 16 17 18 19

Column selection
In the examples given below, we will be using the DataFrame created in the previous section is
assumed.
print (df['one'])

will print
CHAPTER 4. INTRODUCTION TO PANDAS 38

a 1.0
b 2.0
c 3.0
d 4.0
Name: one, dtype: float64
We can also print a selection of the column elements, by applying some condition like,
print (df['one'] > 2)
will print
a False
b False
c True
d True
Name: one, dtype: bool
df.query('one > 2')

Adding a new column

We can create new columns from existing ones or using new data values. Assigning a column that
doesn't exist will create a new column. Assigning scalar values will result in propagating it to all
the rows.
d = {'one': [1.0, 2.0, 3.0, 4.0], 'two': [4.0, 3.0, 2.0, 1.0]}
df = pd.DataFrame(d, index = ['a', 'b', 'c', 'd'])
df['three'] = df['one'] * df['two'] # new column from existing ones
df['four'] = 50 # scalar will be propagated to all rows
df["flag"] = df["one"] > 2
print (df)
The output will look like
one two three four flag
a 1.0 4.0 4.0 50 False
b 2.0 3.0 6.0 50 False
c 3.0 2.0 6.0 50 True
d 4.0 1.0 4.0 50 True
New columns can also be created using the assign() function, as shown below. This function returns
a new DataFrame without modifying the existing one.
ndf = df.assign(ratio = df['one']/ df['two'])
print(ndf)
prints the following. The variable name used in the argument becomes the new column name.
one two ratio
a 1.0 4.0 0.250000
b 2.0 3.0 0.666667
c 3.0 2.0 1.500000
d 4.0 1.0 4.000000
CHAPTER 4. INTRODUCTION TO PANDAS 39

Conditional selection or rows

We can use the query() function to conditionally select rows in to a new DataFrame
ndf = df.query('one > 2')
print(ndf)

Only the last two rows are printed, as shown below.

one two
c 3.0 2.0
d 4.0 1.0

The query() and assign() functions can be combined to create a new DataFrame as shown below.
ndf = df.query('one > 2').assign(ratio = df['one']/ df['two'])
print(ndf)

DataFrame. The output is;

one two ratio
c 3.0 2.0 1.5
d 4.0 1.0 4.0

Some rows are selected and a new column is added to it.

Deleting a column
The del keyword will delete columns as with a dict. You can also pop it out in to a new Series.
del df['two']
s1 = df.pop('one')
print (s1)

will print
a 1.0
b 2.0
c 3.0
d 4.0
Name: one, dtype: float64

Indexing and Slicing Rows

Extracting a row from a DataFrame results in a Series.
d = {'one': [1.0, 2.0, 3.0, 4.0], 'two': [4.0, 3.0, 2.0, 1.0]}
df = pd.DataFrame(d, index = ['a', 'b', 'c', 'd'])
print (df[2])
print (df.loc['b'])
print (df.iloc[2])
CHAPTER 4. INTRODUCTION TO PANDAS 40

The last two lines prints the same thing, third row of the DataFrame. Use loc when the index
is used. Use iloc if the integer representing while using the row number. The extracted row is a
Series and printed in the column format.
one 3.0
two 2.0
Name: b, dtype: float64
one 3.0
two 2.0
Name: b, dtype: float64

Slicing of a specic column can be done like;

print (df['one'][1:3])
print ( df['one'].iloc[1:3])

The output of both the lines are the same.

b 2.0
c 3.0
Name: one, dtype: float64
b 2.0
c 3.0
Name: one, dtype: float64

The entire DataFrame (all the columns) can be sliced like,

print (df[1:3])

The output is given below.

one two
b 2.0 3.0
c 3.0 2.0

Transposing a DataFrame
This operation interchanges the rows and columns.

Data alignment and arithmetic

When doing an operation between DataFrame and Series, the default behavior is to align the Series
index on the DataFrame columns. The lines given below make two DataFrames, df1 has 4 columns
and ve rows, and df2 has 3 columns and 3 rows.

df1 = pd.DataFrame(np.arange(20).reshape(5,4), columns=["A", "B", "C", "D"]) df2 = pd.DataFra

print (df1 + df2)

The result is shown below

CHAPTER 4. INTRODUCTION TO PANDAS 41

A B C D
0 0.0 2.0 4.0 NaN
1 7.0 9.0 11.0 NaN
2 14.0 16.0 18.0 NaN
3 NaN NaN NaN NaN
4 NaN NaN NaN NaN

The addition of elements takes place only where elements from both the Frames are present.
Wherever the operation is not possible, is lled with a NaN value.

Loading and Saving CSV format les

The example shown below demonstrates howto save/load a DataFrame in CSV format.

df1 = pd.DataFrame(np.arange(20).reshape(5,4), columns=["A", "B", "C", "D"])

df.index = ['a','b','c','d','e']
df.index.name = 'I'
df1.to_csv('data.csv')
df2 = pd.read_csv('data.csv')
print (df2)