Why Data Libraries: Lists to Arrays

Plain Python lists are flexible, but for numbers they are slow and verbose. Doubling every value in a list means writing a loop:

prices = [10, 20, 30]
doubled = []
for p in prices:
    doubled.append(p * 2)
print(doubled)   # [20, 40, 60]

numpy (Numerical Python) gives you the ndarray: a grid of numbers stored in one tight block of memory. Math on a whole array happens in fast compiled C, with no Python loop. This is called vectorization.

import numpy as np
prices = np.array([10, 20, 30])
print(prices * 2)        # [20 40 60]   no loop needed
print(prices.sum())      # 60
print(prices.mean())     # 20.0

The whole data and machine-learning stack sits on top of numpy, so this is the natural first step. We import it as np by convention.

Build an array with np.array(a_list), then call methods like .sum() and .mean() straight on it. They return single numbers (a Python/numpy scalar), not lists.