Introduction

NumPy (Numerical Python) is a library built around one core idea: a fast, typed array that stores its data as one contiguous block of memory, and a set of operations that run on the whole array at once instead of looping over it in Python.

Why not just use a list?

A Python list can hold anything — strings, numbers, other lists — mixed together, which means Python has to check each element's type individually every time you touch it. A NumPy array holds one fixed type for every element, packed together in memory, so operations on it run in fast, compiled C loops instead of the Python interpreter's own loop. That difference is most of why NumPy exists.

import numpy as np

numbers = np.array([1, 2, 3, 4, 5])
print(numbers)
print(type(numbers))
print(numbers.dtype)

[1 2 3 4 5]
<class 'numpy.ndarray'>
int64

np.array() takes a regular Python list and converts it into an ndarray (n-dimensional array), NumPy's core data type. numbers.dtype tells you the single data type every element is stored as — here, 64-bit integers.

Operations run on the whole array at once

This is the payoff. With a plain list, converting prices to include tax means writing a loop. With a NumPy array, you write the math once and it applies to every element:

import numpy as np

prices = np.array([10, 20, 30])
with_tax = prices * 1.08
print(with_tax)

[10.8 21.6 32.4]

No loop, no index variable — prices * 1.08 multiplies every element by 1.08 and hands back a new array. This pattern, called vectorization, is the style you'll write in for the rest of this course.