🧮 Introduction to NumPy

Your class has 40 students, and the teacher decides to give everyone 5 grace marks. With a Python list, you have to write a loop and add 5 to each student one by one. What if you could just write marks + 5 and have it done for all 40 students in one go? 🤔 That's exactly what NumPy does! 🧮 This chapter answers the BIG questions: What is NumPy and why do we need it? What is an array? How is a NumPy array different from a list? And how do we create a NumPy array from a list — both 1-D and 2-D? Let's meet NumPy! 🚀

8.1 🏁 Introduction — What is NumPy?

NumPy stands for Numerical Python. It is a library (package) for Python that is used for working with arrays and doing fast calculations on numbers.

Analogy: Python is like a smartphone 📱. It can do a lot on its own, but when you need something special — say, video editing — you install an app. NumPy is Python's "maths app": you install it once, then import it whenever you want to work with lots of numbers quickly.

8.1.1 Why Do We Need NumPy? ⭐

NumPy is widely used in data analysis, science and engineering, and machine learning — anywhere large amounts of numerical data have to be handled.

8.2 📥 Installing and Importing NumPy

NumPy does not come built into Python, so it has to be installed once. It is installed using pip (Python's package installer), typed at the command prompt, not inside Python:

pip install numpy

After that, we import it in every program that uses it:

import numpy as np
print("NumPy is ready!")

Output:

NumPy is ready!

Here np is an alias (a short nickname) for numpy. Instead of writing numpy.array(...) every time, we can simply write np.array(...). Using np is a convention followed by almost all Python programmers.

8.3 🧱 What is an Array?

An array is a collection of elements that are:

1. all of the same data type (all integers, or all floats, or all strings), and
2. stored next to each other (contiguously) in memory.

Each element can be reached by its index, which starts from 0 — just like a list.

   Index  →    0     1     2     3
             ┌─────┬─────┬─────┬─────┐
   marks  →  │ 45  │ 78  │ 92  │ 60  │    all integers, stored side by side
             └─────┴─────┴─────┴─────┘

In NumPy, an array is called an ndarray, which stands for n-dimensional array. "n-dimensional" means it can have one dimension (a single row, like above), two dimensions (rows and columns, like a table), or more.

Analogy: A list is like your school bag 🎒 — you can put anything in it: books, a tiffin, a water bottle, a cricket ball. A NumPy array is like an egg tray 🥚 — every slot is the same size and holds the same kind of thing, neatly in a row. Because everything is the same, you can deal with the whole tray at once!

8.4 ⚖️ List vs NumPy Array

A list and a NumPy array both hold many values, but they behave very differently.

See the difference in action:

import numpy as np
L = [10, 20, 30]
A = np.array([10, 20, 30])
print(L * 2)
print(A * 2)
print(A + 5)

Output:

[10, 20, 30, 10, 20, 30]
[20 40 60]
[15 25 35]

L * 2 repeats the list, but A * 2 multiplies every element by 2. And A + 5 adds 5 to every element — while L + 5 would give TypeError: can only concatenate list (not "int") to list.

Analogy: Giving grace marks with a list is like the teacher walking to every desk and changing each answer sheet one by one (a loop). With a NumPy array, it is like the teacher making one announcement — "Everyone gets +5!" — and it happens for the whole class at once. 📢

8.5 🛠️ Creating a NumPy Array from a List

The function np.array() takes a list and converts it into a NumPy array.

graph LR
    L1["📋 Simple list\n[10, 20, 30]"]
    F1["🛠️ np.array()"]
    A1["🧮 1-D array\n[10 20 30]"]
    L2["📋 Nested list\n[[1, 2], [3, 4]]"]
    F2["🛠️ np.array()"]
    A2["🔲 2-D array\n2 rows × 2 columns"]

    L1 --> F1 --> A1
    L2 --> F2 --> A2

    style F1 fill:#FF9800,color:#fff
    style F2 fill:#FF9800,color:#fff
    style A1 fill:#4CAF50,color:#fff
    style A2 fill:#2196F3,color:#fff

   arr  =  np.array( [10, 20, 30, 40] )
   ───     ────────   ────────────────
  array    function    a Python list
  name                 (inside the brackets)

8.5.1 Creating a 1-D Array 📏

A 1-D (one-dimensional) array is a single row of elements, made from a simple list.

import numpy as np
arr = np.array([10, 20, 30, 40])
print(arr)
print(type(arr))

Output:

[10 20 30 40]
<class 'numpy.ndarray'>

The list can also be stored in a variable first and then passed to np.array():

import numpy as np
marks = [45, 78, 92, 60]
arr = np.array(marks)
print("List :", marks)
print("Array:", arr)

Output:

List : [45, 78, 92, 60]
Array: [45 78 92 60]

Arrays can hold strings too:

import numpy as np
subjects = np.array(["CS", "IP", "Maths"])
print(subjects)

Output:

['CS' 'IP' 'Maths']

8.5.2 All Elements Become the Same Type 🔄

Because an array can hold only one data type, NumPy automatically converts all elements to a common type when the list has mixed values. This is called upcasting.

import numpy as np
a = np.array([1, 2.5, 3])
print(a)
b = np.array([5, -7.4, "a", 7.2])
print(b)
c = np.array([True, 2, 3])
print(c)

Output:

[1.  2.5 3. ]
['5' '-7.4' 'a' '7.2']
[1 2 3]

In a, the integers 1 and 3 became floats 1. and 3. because of 2.5. In b, all the numbers became strings (shown in quotes) because of "a". In c, True became 1.

8.5.3 Choosing the Data Type — dtype 🎛️

We can also tell np.array() which data type to use with the dtype argument. The data type of an existing array is shown by arr.dtype.

import numpy as np
a = np.array([1, 2, 3], dtype=float)
print(a)
print(a.dtype)
b = np.array([1.9, 2.7, -3.5], dtype=int)
print(b)

Output:

[1. 2. 3.]
float64
[ 1  2 -3]

With dtype=int, the decimal part is simply cut off (not rounded) — 1.9 becomes 1, and -3.5 becomes -3.

For an array of whole numbers, print(arr.dtype) shows int64 on most computers (it can be int32 on some). The number is the count of bits used to store each element.

8.5.4 Creating a 2-D Array from a Nested List 🔲

A 2-D (two-dimensional) array has rows and columns, like a table. It is made from a nested list — a list of lists, where each inner list becomes one row.

import numpy as np
table = np.array([[1, 2, 3],
                  [4, 5, 6]])
print(table)

Output:

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

              column 0   column 1   column 2
   row 0  →      1          2          3
   row 1  →      4          5          6

Another example with mixed int and float values (upcasting applies here too):

import numpy as np
b = np.array([[2.4, 3], [4.91, 7], [0, -1]])
print(b)

Output:

[[ 2.4   3.  ]
 [ 4.91  7.  ]
 [ 0.   -1.  ]]

This array has 3 rows and 2 columns, and every element is a float.

Analogy: A 2-D array is like the seating plan of your classroom 🪑 — rows of benches, and every row has the same number of seats. You can't have a row with 3 seats and another with only 1!

import numpy as np
L = [5, 10, 15, 20]
arr = np.array(L)
print(arr)

8.6 💻 Programs on Creating Arrays

Code Example: grace_marks.py — Array from a List of Marks

import numpy as np
marks = [45, 78, 92, 60]
arr = np.array(marks)
print("Original marks:", arr)
print("After grace   :", arr + 5)

Output:

Original marks: [45 78 92 60]
After grace   : [50 83 97 65]

Code Example: input_array.py — Array from Values Typed by the User

import numpy as np
L = []
n = int(input("How many numbers? "))
for i in range(n):
    L.append(int(input("Enter a number: ")))
arr = np.array(L)
print("The array is:", arr)

Sample run (the user types 3, 12, 7 and 25):

How many numbers? 3
Enter a number: 12
Enter a number: 7
Enter a number: 25
The array is: [12  7 25]

Code Example: temperatures.py — A 2-D Array of Temperatures

import numpy as np
temps = [[31, 33, 30],
         [28, 29, 27]]
arr = np.array(temps)
print("Delhi and Shimla temperatures:")
print(arr)

Output:

Delhi and Shimla temperatures:
[[31 33 30]
 [28 29 27]]

Code Example: price_array.py — Converting Prices to a Float Array

import numpy as np
price = [10, 25, 40]
arr = np.array(price, dtype=float)
print(arr)
print(arr * 1.1)

Output:

[10. 25. 40.]
[11.  27.5 44. ]

⚠️ Common Errors and Misconceptions

🔑 Quick Revision

• NumPy → Numerical Python; a library for fast work on arrays of numbers
• Install → pip install numpy (at the command prompt); Import → import numpy as np
• Array → elements of the same type, stored contiguously, reached by index from 0
• ndarray → NumPy's array object (n-dimensional array)
• List vs array → list: mixed types, more memory, needs loops; array: one type, less memory, faster, element-wise maths
• Create from a list → np.array([10, 20, 30]) or np.array(L)
• 1-D array → from a simple list: [10 20 30]
• 2-D array → from a nested list; each inner list = one row; all rows the same length
• Upcasting → bool → int → float → string; one float makes all floats, one string makes all strings
• dtype → np.array(L, dtype=float) sets the type; arr.dtype shows it; dtype=int cuts off decimals
• Printing → arrays print without commas: [1 2 3]

🎯 Sample Exam Questions

Q1: Very Short Answer [1 mark each]

a) What is the full form of NumPy? Answer: Numerical Python

b) Write the statement to import NumPy with the alias np. Answer: import numpy as np

c) Which NumPy function creates an array from a list? Answer: np.array()

d) What is the name of NumPy's array object? Answer: ndarray (n-dimensional array)

e) Can a NumPy array store elements of different data types? Answer: No — all elements of an array are of the same data type

Q2: Output Based [2 marks]

import numpy as np
L = [2, 4, 6]
A = np.array(L)
print(L * 2)
print(A * 2)

Answer:

[2, 4, 6, 2, 4, 6]
[ 4  8 12]

Q3: Output Based [3 marks]

import numpy as np
p = np.array([1, 2, 3.5])
q = np.array([7, "x", 9])
r = np.array([[1, 2], [3, 4]])
print(p)
print(q)
print(r)

Answer:

[1.  2.  3.5]
['7' 'x' '9']
[[1 2]
 [3 4]]

Q4: Short Answer [2 marks]

What is NumPy? Give two reasons why a NumPy array is preferred over a list for numerical data.

Answer: NumPy (Numerical Python) is a Python library used to create and work with arrays and to do fast numerical calculations. A NumPy array is preferred because (1) it takes less memory and is faster, as its elements are of the same type and stored contiguously, and (2) it supports element-wise operations, so an operation like arr + 5 works on all elements without a loop.

Q5: Program [3 marks]

Write a program to input 5 numbers into a list, convert the list into a NumPy array, and display both the list and the array.

Answer:

import numpy as np
L = []
for i in range(5):
    L.append(int(input("Enter a number: ")))
arr = np.array(L)
print("List :", L)
print("Array:", arr)

Sample run (the user types 4, 8, 15, 16 and 23):

Enter a number: 4
Enter a number: 8
Enter a number: 15
Enter a number: 16
Enter a number: 23
List : [4, 8, 15, 16, 23]
Array: [ 4  8 15 16 23]

✏️ Practice Problems

1. What is NumPy? Why is it called a library?
2. Write the command to install NumPy and the statement to import it with the alias np.
3. Define an array. What does "contiguous" mean for the elements of an array?
4. Write any four differences between a Python list and a NumPy array.
5. Create a NumPy array from the list [3, 6, 9, 12] and print it.
6. Write the output: print(np.array([1, 2, 3]) + 10)
7. Write the output: print(np.array([4, 5.5, 6])) and explain why 4 and 6 changed.
8. Write the output: print(np.array([1, "two", 3]))
9. Create a 2-D array from the nested list [[10, 20], [30, 40], [50, 60]]. How many rows and columns does it have?
10. Find the error: arr = np.array([[1, 2, 3], [4, 5]])
11. Write a program to store the prices of 4 items in a list, convert it into a float array using dtype, and print it.
12. What is the difference between L * 3 for a list L = [1, 2] and A * 3 for the array A = np.array([1, 2])? Write both outputs.