Detailed Explanation of Python np.array Usage

1. Introduction

In Python, NumPy (Numerical Python is a powerful library for scientific computing. It provides a high-performance multidimensional array object (ndarray) and various functions for manipulating these arrays. This article will detail the usage and features of one of NumPy’s core objects, the np.array.

2. Creating an np.array

Before using NumPy’s np.array, we must first import the NumPy library. Typically, the NumPy library is imported and renamed to np:

import numpy as np

2.1 Creating an np.array from a List or Tuple

Using NumPy’s array function, you can create an np.array object from a list or tuple. For example, we can create a one-dimensional np.array from a one-dimensional Python list:

list_data = [1, 2, 3, 4, 5]
array_data = np.array(list_data)
print(array_data)

This code will output:

[1 2 3 4 5]

Similarly, we can create a multidimensional np.array from a multidimensional Python list:

matrix = [[1, 2], [3, 4], [5, 6]]
array_matrix = np.array(matrix)
print(array_matrix)

This code will output:

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

Arrays created using the np.array function can be of any dimension. Simply provide the corresponding data in the array function.

2.2 Creating special np.array functions provided by NumPy

In addition to creating np.array using lists or tuples, NumPy also provides some functions for creating special arrays.

2.2.1 Creating an all-zero array using np.zeros

np.zeros function can create an all-zero array of a specified shape. We can specify the shape (dimensions) of an array as a parameter. For example, to create an array of all zeros with a shape of (2, 3):

zeros_array = np.zeros((2, 3))
print(zeros_array)

This code will output:

[[0. 0. 0.]
[0. 0. 0.]]

2.2.2 Creating an Array of All Ones Using np.ones

The np.ones function is similar to the np.zeros function and can create an array of all ones with a specified shape.

ones_array = np.ones((2, 3))
print(ones_array)

This code will output:

[[1. 1. 1.]
[1. 1. 1.]]

2.2.3 Creating a Continuous Integer Array Using np.arange

np.arange creates a continuous array of integers. We can specify the starting and ending values of the array, as well as the step size. For example, let’s create an array of consecutive integers from 0 to 9:

integer_array = np.arange(10)
print(integer_array)

This code will output:

[0 1 2 3 4 5 6 7 8 9]

2.2.4 Creating an Arithmetic Progression Array Using np.linspace

np.linspace creates an arithmetic progression between the specified starting and ending values. We can specify the starting and ending values, as well as the number of elements in the progression.

linspace_array = np.linspace(0, 1, 5)
print(linspace_array)

This code will output:

[0. 0.25 0.5 0.75 1. ]

2.3 Creating an np.array with Random Numbers

In addition to the methods described above, NumPy provides several methods for creating an np.array filled with random numbers.

2.3.1 Creating an Array of Random Numbers Between 0 and 1 Using np.random.rand

The np.random.rand function creates an array of random numbers between 0 and 1 of a specified dimension. We can specify the shape (dimensions) of the array.

random_array = np.random.rand(2, 3)
print(random_array)

This code will output:

[[0.4205492 0.35339967 0.15871663]
[0.2809823 0.29531438 0.68120656]]

2.3.2 Creating an Array of Random Numbers from a Standard Normal Distribution Using np.random.randn

np.random.randn creates an array of random numbers of a specified dimension that follows a standard normal distribution (mean 0, variance 1).

randn_array = np.random.randn(2, 3)
print(randn_array)

This code will output:

[[ 1.01111932 -0.23198528 0.09575871]
[ 0.72938318 0.60414215 -0.29782545]]

3. Attributes and Methods of np.array

3.1 shape Attribute

The shape attribute can be used to obtain the shape (dimensions) of an np.array.

array = np.array([[1, 2, 3, 4], [5, 6, 7, 8]])
print(array.shape)

This code will output:

(2, 4)

3.2 ndim Attribute

The ndim attribute can be used to obtain the dimensions of an np.array.

array = np.array([[1, 2, 3, 4], [5, 6, 7, 8]])
print(array.ndim)

This code will output:

2

3.3 size Attribute

The size attribute can be used to obtain the total number of elements in an np.array.

array = np.array([[1, 2, 3, 4], [5, 6, 7, 8]])
print(array.size)

This code will output:

8

3.4 dtype Attribute

The dtype attribute can be used to obtain the data type of elements in an np.array array.

array = np.array([[1, 2, 3, 4], [5, 6, 7, 8]])
print(array.dtype)

This code will output:

int64

3.5 reshape Method

The reshape method can be used to change the shape (dimensions) of an np.array.

array = np.arange(10)
reshaped_array = array.reshape((2, 5))
print(reshaped_array)

This code will output:

[[0 1 2 3 4]
[5 6 7 8 9]]

3.6 flatten Method

The flatten method can be used to flatten an np.array into a one-dimensional array.

array = np.array([[1, 2, 3], [4, 5, 6]])
flattened_array = array.flatten()
print(flattened_array)

This code will output:

[1 2 3 4 5 6]