Unpacking Data Selection Techniques in NumPy Arrays

Hello, Data Explorer! Today's lesson will focus on "selecting data" in NumPy arrays—our goal is to master integer and Boolean indexing and slicing. So, let's jump in!

In this section, we'll discuss "Indexing". It's akin to item numbering in lists, and Python implements it with a zero-based index system. Let's see this principle in action.

import numpy as np

# Let's form a 1D array and select, say, the 4th element
array_1d = np.array([2, 3, 5, 7, 11, 13, 17, 19, 23, 29])
fourth_element = array_1d[3]  # Selected element: 7

# To grab the last element, use a negative index
last_element = array_1d[-1]  # Selected element: 29

Essentially, indexing is a numeric system for items in an array—relatively straightforward!

Are you mindful of higher dimensions? NumPy arrays can range from 1D to N-dimensions. To access specific elements, we use the index pair (i,j) for a 2D array, (i,j,k) for a 3D array, and so forth.

# Form a 2D array and get the element at row 2 (indexed 1) and column 1 (indexed 0)
array_2d = np.array([[1, 2, 3], [4, 5, 6], [7, 8, 9]])
element = array_2d[1,0]  # Selected element: 4

In this example, we selected the first element of the second row in a 2D-array—it's pretty simple!

Are you ready for some magic? Enter "Boolean Indexing", which functions like a 'Yes/No' filter, with 'Yes' representing True and 'No' for False.

# A boolean index for even numbers
array_1d = np.array([8, 4, 7, 3, 4, 11])
even_index = array_1d % 2 == 0
even_numbers = array_1d[even_index]
print(even_numbers)  # Output: [8 4 4]

Or we can put the condition directly into [] brackets:

# A boolean index for even numbers
array_1d = np.array([8, 4, 7, 3, 4, 11])
even_numbers = array_1d[array_1d % 2 == 0]
print(even_numbers)  # Output: [8 4 4]

Voila! Now, we can filter data based on custom conditions.