15.2: NumPy

Learning Objectives

By the end of this section you should be able to

• Describe the NumPy library.
• Create a NumPy array object.
• Choose appropriate NumPy functions to process arrays of numerical data.

NumPy library

NumPy (Numerical Python) is a Python library that provides support for efficient numerical operations on large, multi-dimensional arrays and serves as a fundamental building block for data analysis in Python. The conventional alias for importing NumPy is np. In other words, NumPy is imported as import numpy as np. NumPy implements the ndarray object, which allows the creation of a multi-dimensional array of homogeneous data types (columns with the same data type) and efficient data processing. An ndarray object can have any number of dimensions and can store elements of various numeric data types. To create a NumPy ndarray object, one of the following options can be used:

• Creating an ndarray by converting a Python list or tuple using the np.array() function.
• Using built-in functions like np.zeros() and np.ones() for creating an array of all 0's or all 1's, respectively.
• Generating an array with random numbers using np.random.rand(n, m), where n and m are the number of rows and columns, respectively.
• Loading data from a file. Ex: np.genfromtxt('data.csv', delimiter=',').

Checkpoint: Creating an ndarray object

Concepts in Practice \(\PageIndex{1}\)

Which of the following creates an ndarray object with one row and two columns?

1. np.array([2, 3])
2. np.zeros(1, 2)
3. np.zeros((1, 2))

Answer

c. To create an m by n matrix of all zeros, np.zeros((m,n)) can be used.

Concepts in Practice \(\PageIndex{2}\)

Which of the following is a NumPy data type?

1. ndarray
2. list
3. array

Answer

a. ndarray is a data type supported by NumPy. When printing the type of an ndarray object, <'numpy.ndarray'> is printed.

Concepts in Practice \(\PageIndex{3}\)

What is the benefit of using anndarray object compared to a list?

1. computational efficiency
2. array-oriented computing
3. memory efficiency
4. all the above

Answer

d. NumPy array is optimized for computational and memory efficiency while also offering array-oriented computation.

NumPy operations

In addition to the ndarray data type, NumPy's operations provide optimized performance for large-scale computation. The key features of NumPy include:

• Mathematical operations: NumPy provides a range of mathematical functions and operations that can be applied to entire arrays or specific elements. These operations include arithmetic, trigonometric, exponential, and logarithmic functions.
• Array manipulation: NumPy provides various functions to manipulate the shape, size, and structure of arrays. These include reshaping, transposing, concatenating, splitting, and adding or removing elements from arrays.
• Linear algebra operations: NumPy offers a set of linear algebra functions for matrix operations, like matrix multiplication, matrix inversion, eigenvalues, and eigenvectors.

Checkpoint: NumPy array operations

Concepts in Practice \(\PageIndex{4}\): NumPy operations

What is the output of the following code?

import numpy as np

arr = np.array([[1, 2], [3, 4]])
out = 2 * arr
print(out)

1. [[2 4]
   [8 16]]

[[2 4]
[6 8]]

[[1 2]
[3 4]]

Answer

1. b. 2 * arr multiplies all elements by 2 resulting in:

[[2 4]
[6 8]]

Concepts in Practice \(\PageIndex{5}\): NumPy operations

Which of the following results in a 2 by 3 ndarray?

1. import numpy as np
   arr = np.array(2, 3)
   

2. import numpy as np
   arr = np.array([[1, 2], [1, 2], [1, 2]])
   

3. import numpy as np
   arr = np.array([[1, 2], [1, 2], [1, 2]]).

Answer

c. np.array([[1, 2], [1, 2], [1, 2]]) creates an ndarray of 3 by 2. The T operator transposes the array. The transpose operator takes an array of m by n and converts the array to an n by m array. The result of applying the transpose operator on the given array is a 2 by 3 array.

Concepts in Practice \(\PageIndex{6}\): NumPy operations

The function np.multiply(arr1, arr2) receives two ndarray objects arr1 and arr2 with the same dimensions, and performs element-wise multiplication. What is the output of the following code?

import numpy as np

arr1 = np.array([[1, 2], [3, 4]])
arr2 = np.array([[1, 0], [0, 1]])
out = np.multiply(arr1, arr2)
print(out)

1. [[1 0]
   [0 4]]
   

2. [[1 2]
   [3 4]]
   

3. [[2 2]
   [3 5]]

Answer

a. The result of element-wise multiplication between arr1 and arr2 is:

[[1 0]
[0 4]]

Exploring further

Please refer to the NumPy user guide for more information about the NumPy library.

• NumPy User Guide

Programming practice with Google

Use the Google Colaboratory document below to practice NumPy functionalities to extract statistical insights from a dataset.

Google Colaboratory document