VEL TECH HIGH TECH

Dr. RANGARAJAN Dr. SAKUNTHALA ENGINEERING

COLLEGE
An Autonomous Institution
Approved by AICTE-New Delhi, Affiliated to Anna University, Chennai
Accredited by NBA, New Delhi & Accredited by NAAC with “A” Grade & CGPA of 3.27

DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE

Course code Semester

Category OPEN ELECTIVE(OE) L T P C
Course Title PYTHON FOR DATA SCIENCE 3 0 0 3

UNIT II TOWARDS DATA SCIENCE USING NUMPY

Understanding Data Types in Python - The Basics of NumPy Arrays - Computation on NumPy Arrays:
Universal Functions - Aggregations: Min, Max, and Everything in Between Computation on Arrays:
Broadcasting-Comparisons, Masks, and Boolean Logic Fancy Indexing-Sorting Arrays.

COURSEOBJECTIVES:
· To describe the fundamentals for exploring and managing data with Python.

· To examine the various data analytics techniques for labelled/columnar data using Python.

· To demonstrate a flexible range of data visualizations techniques in Python.

· To describe the various Machine learning algorithms for data modelling with Python.

COURSEOUTCOMES:

Blooms
CO.No. CourseOutcomes
level

OnsuccessfulcompletionofthisCourse,studentswillbeableto

C305. 2 Make use of knowledge on NumPy to write programs on K2

array operations.

Understanding Data Types in Python

In Python, data types define the kind of data a variable can hold. Understanding these is essential because different
operations are valid for different data types, and correct usage ensures efficient coding and data processing.

. Basic Built-in Data Types

Type Description Example
int Integer numbers x = 10

float Decimal numbers y = 3.14

bool Boolean values (True or flag = True

False)
str Text (string of characters) name = "Alice"

2. Type Conversion (Casting)

You can convert between types using built-in functions:

python
CopyEdit
int("10") # Converts string to integer 10
float("3.5") # Converts string to float 3.5
str(100) # Converts integer to string '100'
bool(0) # Converts to boolean False

3. Collection Data Types

Type Description Example

lis Ordered, mutable collection fruits = ["apple",

t "banana"]

tup Ordered, immutable collection coords = (10, 20)

le

set Unordered collection of unique elements ids = {101, 102, 103}

dic Key-value pairs student = {"name": "John",

t "age": 21}

4. Type Checking
Use the type() function to check the data type of a variable:

python
CopyEdit
x = 25
print(type(x)) # Output: <class 'int'>

5. NumPy Data Types (For Data Science)

In scientific computing with NumPy, special data types are used for efficient computation:

NumPy Data Type Description

np.int32, np.int64 Fixed-size integers

np.float32, Floating-point numbers

np.float64

np.bool_ Boolean

np.str_ Unicode string

Example:
python
CopyEdit
import numpy as np
arr = np.array([1, 2, 3], dtype=np.int32)
print(arr.dtype) # Output: int32

Why Understanding Data Types is Important in Data Science?

● Ensures correct operations (e.g., numeric vs. text processing)

● Saves memory and improves performance

● Helps prevent errors in data transformation and modeling

● Required when using libraries like Pandas, NumPy, and scikit-learn

The Basics of NumPy Arrays

What is NumPy?
NumPy (Numerical Python) is a powerful Python library used for:

● Efficient numerical computation,

● Creating and manipulating n-dimensional arrays,

● Performing mathematical operations on large datasets.

1. What is a NumPy Array?

A NumPy array is a grid of values, all of the same type, and is indexed by a tuple of non-negative integers.

● NumPy arrays are more efficient than Python lists.

● Arrays support element-wise operations and broadcasting.

import numpy as np

arr = np.array([1, 2, 3]) # 1D array

2. Creating NumPy Arrays

🔹 From Python lists:
python
CopyEdit
a = np.array([1, 2, 3]) # 1D array
b = np.array([[1, 2], [3, 4]]) # 2D array

Using built-in functions:

python
CopyEdit
np.zeros((2, 3)) # 2x3 array of zeros
np.ones((2, 2)) # 2x2 array of ones
np.arange(0, 10, 2) # [0 2 4 6 8]
np.linspace(0, 1, 5) # 5 evenly spaced numbers from 0 to 1
np.eye(3) # 3x3 identity matrix

3. Array Attributes
Attribute Description Example

ndim Number of dimensions arr.ndi

m

shape Tuple representing array dimensions arr.sha

pe

size Total number of elements arr.siz

e

dtype Data type of array elements arr.dty

pe

arr = np.array([[1, 2, 3], [4, 5, 6]])

print(arr.ndim) # 2
print(arr.shape) # (2, 3)
print(arr.size) # 6

4. Indexing and Slicing

Just like Python lists, NumPy arrays support indexing and slicing.

python
CopyEdit
arr = np.array([10, 20, 30, 40])
print(arr[1]) # 20
print(arr[1:3]) # [20 30]

matrix = np.array([[1, 2], [3, 4]])

print(matrix[0, 1]) # 2

5. Array Operations
NumPy supports vectorized operations (element-wise), which are faster than loops.

python
CopyEdit
a = np.array([1, 2, 3])
b = np.array([4, 5, 6])

print(a + b) # [5 7 9]
print(a * 2) # [2 4 6]
print(a ** 2) # [1 4 9]
print(np.sin(a)) # Applies sin to each element

6. Reshaping Arrays
You can change the shape of arrays without changing data.

python
CopyEdit
arr = np.array([[1, 2], [3, 4], [5, 6]])
reshaped = arr.reshape((2, 3))

7. Array Copying vs. View

● arr.copy() creates a new array.

● arr.view() or slicing returns a view (reference) to the same data.

Why NumPy Arrays Are Important in Data Science:

● Memory-efficient and fast computations

● Support for vectorized operations (no explicit loops)

● Foundation for Pandas, Scikit-learn, TensorFlow, etc.

Computation on NumPy Arrays

NumPy is optimized for performing fast, element-wise operations on large arrays. It includes mathematical, statistical, and logical operations that are
both vectorized and highly efficient.
1. Arithmetic Operations
NumPy allows arithmetic operations to be applied element-wise on arrays of the same shape.

python

import numpy as np

a = np.array([1, 2, 3])

b = np.array([4, 5, 6])

print(a + b) # [5 7 9]

print(a - b) # [-3 -3 -3]

print(a * b) # [4 10 18]

print(a / b) # [0.25 0.4 0.5 ]

print(a ** 2) # [1 4 9]

2. Universal Functions (ufuncs)

NumPy provides many universal functions, which are fast, vectorized operations implemented in C.

🔹 Examples:

Function Description

np.add Element-wise addition

(a, b)

np.sub Element-wise subtraction

tract(
a, b)

np.mul Element-wise multiplication

tiply(
a, b)

np.div Element-wise division

ide(a,
b)

Trigonometric Functions:
x = np.array([0, np.pi/2, np.pi])

print(np.sin(x)) # [0.0, 1.0, 0.0]

Exponential and Logarithmic:
python

CopyEdit

np.exp([1, 2, 3]) # [e^1,

e^2, e^3]

np.log([1, np.e, np.e**2]) # [0.,

1., 2.]

Aggregation Functions
Aggregations compute summary statistics over the entire
array or along an axis.

Common Aggregations:

Function Description

np.sum() Sum of all

elements

np.mean() Mean
(average)

np.std() Standard
deviation

np.min() Minimum
value

np.max() Maximum
value

np.median( Median
)

np.percent Nth percentile

ile()

Example:

arr = np.array([[1, 2, 3], [4, 5, 6]])

print(arr.sum()) # 21

print(arr.mean()) # 3.5
print(arr.sum(axis=0)) # [5 7 9] - column-wise sum

print(arr.sum(axis=1)) # [6 15] - row-wise sum

4. Broadcasting in NumPy
Broadcasting allows NumPy to work with arrays of different shapes during
arithmetic operations.

Rules:
1. If arrays have different dimensions, the smaller one is
stretched to match.

2. Dimensions must be compatible (or one must be 1).

Example:
python

CopyEdit

a = np.array([1, 2, 3]) # shape (3,)

b = np.array([[10], [20]]) # shape (2, 1)

print(a + b)

# Output: [[11 12 13]

# [21 22 23]]

5. Comparison and Boolean

Logic
You can use NumPy to compare elements or apply Boolean conditions.

python

CopyEdit

a = np.array([1, 2, 3, 4])

print(a > 2) # [False False True True]

print(a == 3) # [False False True False]

Logical Operators:

Operator Description

np.logical_ Element-wise
and() AND

np.logical_ Element-wise
or()
OR

np.logical_ Element-wise
not()
NOT

a = np.array([1, 2, 3])

b = np.array([3, 2, 1])

np.logical_and(a < 3, b < 3) # [ True False False]

6. Fancy Indexing
Allows retrieving multiple elements using an array of indices.

python

CopyEdit

arr = np.array([10, 20, 30, 40])

idx = [0, 3]

print(arr[idx]) # [10 40]

7. Sorting Arrays
python

CopyEdit

a = np.array([3, 1, 2])

print(np.sort(a)) # [1 2 3]

For multi-dimensional arrays:

python

CopyEdit

arr = np.array([[3, 1], [4, 2]])

np.sort(arr, axis=1) # Sort rows

Summary Table
Topic Examples / Functions

Arithmetic
Ops +, -, *, /, **

Universal
Funcs np.sin(),
np.exp(),
np.power()

Aggregations
np.sum(),
np.mean(),
np.max()

Broadcasting Add arrays of different shapes

Comparisons
a > b, a == b

Boolean Logic
np.logical_and(
),
np.logical_not(
)

Fancy
Indexing arr[[0, 2]]

Sorting
np.sort()