NumPy – Single Detailed Notes

1. Introduction to NumPy
NumPy (Numerical Python) is a core Python library used for numerical and scientific computing. It
provides fast, memory‑efficient array operations and mathematical functions.

Why NumPy? - Faster than Python lists - Less memory usage - Built‑in mathematical operations -
Foundation for libraries like Pandas, SciPy, TensorFlow

import numpy as np

2. NumPy Array (ndarray)

The main object in NumPy is the ndarray (n‑dimensional array).

Creating an Array

a = [Link]([1, 2, 3])

- Converts a list into a NumPy array - All elements must be of the same data type

Types of Arrays

• 1D Array: [1, 2, 3]
• 2D Array (Matrix): [[1, 2], [3, 4]]
• 3D Array: Multiple matrices

b = [Link]([[1,2],[3,4]])

3. Difference: Python List vs NumPy Array

Feature Python List NumPy Array

Speed Slow Fast

Data Type Mixed Same

Math Ops Manual Automatic

Memory More Less

1
4. Array Attributes (Detailed Notes)

a = [Link]([[1,2,3],[4,5,6]])

[Link] → Number of Dimensions

• ndim stands for number of dimensions of the array

• It tells whether the array is 1D, 2D, 3D, etc.
• For the given array, there are rows and columns, so it is 2D

[Link] # 2

[Link] → Structure of the Array

• shape shows how the data is arranged

• It is written as (rows, columns) for 2D arrays

[Link] # (2, 3)

Meaning: - 2 → number of rows - 3 → number of columns

[Link] → Total Number of Elements

• size gives the total count of elements in the array

• Formula:

size = rows × columns

[Link] # 6

Because:

2 × 3 = 6

[Link] → Data Type of Elements

• dtype tells the type of data stored in the array

• NumPy stores all elements using the same data type
• This makes NumPy fast and memory efficient

2
 [Link] # int32 or int64

Examples:

[Link]([1,2,3]).dtype # int
[Link]([1.5,2.5]).dtype # float
[Link]([1,2.5]).dtype # float (automatic conversion)

Quick Summary Table

Attribute Meaning Example Output

ndim Number of dimensions 2

shape Rows, Columns (2, 3)

size Total elements 6

dtype Data type int32 / int64

5. Creating Special Arrays (Detailed Notes)

NumPy provides built-in functions to quickly create arrays with specific values. These are called special
arrays.

1. [Link]() → Array Filled with Zeros

[Link]((2,3))

Output:

[[0. 0. 0.]
[0. 0. 0.]]

Explanation: - Creates an array where all elements are 0 - (2,3) means 2 rows and 3 columns -
Default data type is float

Use: - Initializing arrays - Mathematical calculations

2. [Link]() → Array Filled with Ones

3
 [Link]((3,3))

Output:

[[1. 1. 1.]
[1. 1. 1.]
[1. 1. 1.]]

Explanation: - Creates an array where all elements are 1 - (3,3) means 3 rows and 3 columns

Use: - Creating test data - Mathematical operations

3. [Link]() → Identity Matrix

[Link](3)

Output:

[[1. 0. 0.]
[0. 1. 0.]
[0. 0. 1.]]

Explanation: - Creates a square matrix - Diagonal elements are 1 - All other elements are 0

Use: - Linear algebra - Matrix calculations

4. [Link]() → Sequence of Numbers

[Link](1, 10, 2)

Output:

[1 3 5 7 9]

Explanation: - Works like Python range() - Syntax:

[Link](start, stop, step)

- start = 1 - stop = 10 (not included) - step = 2

Use: - Creating number sequences

4
5. [Link]() → Evenly Spaced Numbers

[Link](1, 10, 5)

Output:

[ 1. 3.25 5.5 7.75 10. ]

Explanation: - Generates 5 evenly spaced numbers between 1 and 10 - Both start and end values are
included

Use: - Graph plotting - Scientific calculations

Summary Table

Function Purpose Example Output

zeros() All zeros 0 values

ones() All ones 1 values

eye() Identity matrix Diagonal = 1

arange() Number sequence Step-based

linspace() Even spacing Fixed count

6. Indexing and Slicing (Detailed Notes)

Indexing and slicing are used to access elements or parts of an array.

1D Array Indexing and Slicing

a = [Link]([10, 20, 30, 40])

• a[0] → First element → 10

• a[-1] → Last element → 40
• a[1:3] → Elements from index 1 to 2 → [20 30]

Note: - Indexing starts from 0 - Last index is excluded in slicing

5
2D Array Indexing

b = [Link]([[1,2,3],[4,5,6]])

• b[1,2] → Element in 2nd row, 3rd column → 6

Syntax:

array[row_index, column_index]

7. Mathematical Operations (Element-wise)

NumPy performs element-wise operations by default.

a = [Link]([1,2,3])
b = [Link]([4,5,6])

• a + b → [5 7 9] (Addition)
• a - b → [-3 -3 -3] (Subtraction)
• a * b → [4 10 18] (Multiplication)
• a / b → [0.25 0.4 0.5] (Division)

Scalar Operation

a * 2

Output:

[2 4 6]

8. Universal Functions (ufuncs)

Universal functions perform fast mathematical operations on arrays.

[Link](a) # Square root

[Link](a) # Natural logarithm
[Link](a) # Exponential
[Link](a) # Sine
[Link](a) # Cosine

6
9. Aggregation Functions
Aggregation functions return a single value from an array.

a = [Link]([1,2,3,4,5])

• [Link](a) → Sum of elements

• [Link](a) → Average
• [Link](a) → Maximum value
• [Link](a) → Minimum value
• [Link](a) → Standard deviation

10. Reshaping Arrays

Reshaping changes the shape of an array without changing data.

a = [Link]([1,2,3,4,5,6])
[Link](2,3)

Output:

[[1 2 3]
[4 5 6]]

11. Matrix Operations

A = [Link]([[1,2],[3,4]])
B = [Link]([[5,6],[7,8]])

Matrix Multiplication

A @ B

Transpose of a Matrix

A.T

7
12. Copy vs View

a = [Link]([1,2,3])
b = [Link]() # Independent copy
c = [Link]() # Shared memory

Explanation: - copy() → Changes do not affect original array - view() → Changes affect original
array

End of Notes