{"id":4224,"date":"2022-08-08T04:37:30","date_gmt":"2022-08-08T04:37:30","guid":{"rendered":"https:\/\/www.pythontutorial.net\/?page_id=4224"},"modified":"2022-08-16T03:10:15","modified_gmt":"2022-08-16T03:10:15","slug":"create-numpy-array","status":"publish","type":"page","link":"https:\/\/www.pythontutorial.net\/python-numpy\/create-numpy-array\/","title":{"rendered":"Create NumPy Array"},"content":{"rendered":"\n

Summary<\/strong>: in this tutorial, you’ll learn how to create NumPy arrays including one-dimensional, two-dimensional, and three-dimensional arrays.<\/p>\n\n\n\n

The array is the core data structure of the NumPy library. A NumPy array is a grid of values with the same type and indexed by a tuple of non-negative integers. <\/p>\n\n\n\n

All arrays are instances of the ndarray<\/code> class. To create a new NumPy array, you use the array()<\/code> function of the NumPy library.<\/p>\n\n\n\n

Creating one-dimensional arrays #<\/a><\/h2>\n\n\n\n

The following example uses the array()<\/code> function to create a one-dimensional (1-D) array:<\/p>\n\n\n

import<\/span> numpy as<\/span> np\n\na = np.array([1<\/span>, 2<\/span>, 3<\/span>])\n\nprint(type(a))\nprint(a)<\/code><\/span>Code language:<\/span> Python<\/span> (<\/span>python<\/span>)<\/span><\/small><\/pre>\n\n\n
Output:<\/p>\n\n\n
<class<\/span> 'numpy<\/span>.ndarray<\/span>'>\n[1 2 3]<\/span><\/code><\/span>Code language:<\/span> Python<\/span> (<\/span>python<\/span>)<\/span><\/small><\/pre>\n\n\n
How it works.<\/p>\n\n\n\n
First, import the numpy<\/code> library as np<\/code>:<\/p>\n\n\n
import<\/span> numpy as<\/span> np<\/code><\/span>Code language:<\/span> Python<\/span> (<\/span>python<\/span>)<\/span><\/small><\/pre>\n\n\n
a = np.array([1<\/span>, 2<\/span>, 3<\/span>])<\/code><\/span>Code language:<\/span> Python<\/span> (<\/span>python<\/span>)<\/span><\/small><\/pre>\n\n\n
The array()<\/code> function returns a new instance of the ndarray<\/code> type. Therefore, the type(a)<\/code> returns <class 'numpy.ndarray'><\/code>.<\/p>\n\n\n\n
A 1-D array is known as a vector<\/em>.<\/p>\n\n\n\n
Getting the dimension of an array #<\/a><\/h3>\n\n\n\n
To get the number of dimensions of an array, you use the ndim<\/code> property. In NumPy, dimensions are called axes. For example:<\/p>\n\n\n
import<\/span> numpy as<\/span> np\n\na = np.array([1<\/span>, 2<\/span>, 3<\/span>])\n\nprint(a.ndim)<\/code><\/span>Code language:<\/span> Python<\/span> (<\/span>python<\/span>)<\/span><\/small><\/pre>\n\n\n
Output:<\/p>\n\n\n
1<\/span><\/code><\/span>Code language:<\/span> Python<\/span> (<\/span>python<\/span>)<\/span><\/small><\/pre>\n\n\n
In this example, The ndim<\/code> property returns one as expected.<\/p>\n\n\n\n
Getting the data type of array elements #<\/a><\/h3>\n\n\n\n
To get the data type of the elements of an array, you use the dtype<\/code> property. For example:<\/p>\n\n\n
import<\/span> numpy as<\/span> np\n\na = np.array([1<\/span>, 2<\/span>, 3<\/span>])\n\nprint(a.dtype)<\/code><\/span>Code language:<\/span> Python<\/span> (<\/span>python<\/span>)<\/span><\/small><\/pre>\n\n\n
Output:<\/p>\n\n\n
int32<\/code><\/span>Code language:<\/span> Python<\/span> (<\/span>python<\/span>)<\/span><\/small><\/pre>\n\n\n
In this example, the type of the elements is int32<\/code>. If you want to set the type of the array’s elements, you can use the dtype<\/code> argument of the array()<\/code> function. For example:<\/p>\n\n\n
import<\/span> numpy as<\/span> np\n\na = np.array([1<\/span>, 2<\/span>, 3<\/span>], dtype=np.float64)\n\nprint(a)\nprint(a.dtype)<\/code><\/span>Code language:<\/span> Python<\/span> (<\/span>python<\/span>)<\/span><\/small><\/pre>\n\n\n
Output:<\/p>\n\n\n
[1.<\/span> 2.<\/span> 3.<\/span>]\nfloat64<\/code><\/span>Code language:<\/span> Python<\/span> (<\/span>python<\/span>)<\/span><\/small><\/pre>\n\n\n
In this example, the numbers of the array have the decimal point (.<\/code>) and the data type of its elements is float64<\/code>.<\/p>\n\n\n\n
Creating two-dimensional arrays #<\/a><\/h2>\n\n\n\n
The following example uses the array()<\/code> function to create a two-dimensional (2-D) array:<\/p>\n\n\n
import<\/span> numpy as<\/span> np\n\nb = np.array(\n    [\n        [1<\/span>, 2<\/span>, 3<\/span>],\n        [4<\/span>, 5<\/span>, 6<\/span>]\n    ]\n)\n\nprint(b)\nprint(b.ndim)<\/code><\/span>Code language:<\/span> Python<\/span> (<\/span>python<\/span>)<\/span><\/small><\/pre>\n\n\n
Output:<\/p>\n\n\n
[[1<\/span> 2<\/span> 3<\/span>] \n [4<\/span> 5<\/span> 6<\/span>]]\n\n2<\/span><\/code><\/span>Code language:<\/span> Python<\/span> (<\/span>python<\/span>)<\/span><\/small><\/pre>\n\n\n
In this example, we pass a list of a list of integers to the array()<\/code> function. The ndim<\/code> property returns 2 as expected.<\/p>\n\n\n\n
A good tip to get the number of dimensions of an array is that you count the square brackets ([<\/code>) until you encounter the first number. The number of square brackets is the number of dimensions or axes.<\/p>\n\n\n\n
A two-dimensional array is also called a matrix<\/em>.<\/p>\n\n\n\n
Creating three-dimensional array #<\/a><\/h2>\n\n\n\n
The following example uses the array()<\/code> function to create a three-dimensional (3-D) array:<\/p>\n\n\n
import<\/span> numpy as<\/span> np\n\nc = np.array(\n    [\n        [\n            [1<\/span>, 2<\/span>, 3<\/span>],\n            [4<\/span>, 5<\/span>, 6<\/span>]\n        ],\n        [\n            [7<\/span>, 8<\/span>, 9<\/span>],\n            [10<\/span>, 11<\/span>, 12<\/span>]\n        ],\n    ]\n)\n\nprint(c.ndim)<\/code><\/span>Code language:<\/span> Python<\/span> (<\/span>python<\/span>)<\/span><\/small><\/pre>\n\n\n
Output:<\/p>\n\n\n
3<\/span><\/code><\/span>Code language:<\/span> Python<\/span> (<\/span>python<\/span>)<\/span><\/small><\/pre>\n\n\n
Note that a 3-D array is also called a tensor.<\/p>\n\n\n\n
Getting shapes of arrays #<\/a><\/h2>\n\n\n\n
To find the number of axes and the number of elements on each axis of an array, you use the shape<\/code> property. For example:<\/p>\n\n\n
import<\/span> numpy as<\/span> np\n\n\na = np.array([1<\/span>, 2<\/span>, 3<\/span>])\nprint(a.shape)  # (3,)<\/span>\n\nb = np.array(\n    [\n        [1<\/span>, 2<\/span>, 3<\/span>],\n        [4<\/span>, 5<\/span>, 6<\/span>]\n    ]\n)\nprint(b.shape)  # (2, 3)<\/span>\n\nc = np.array(\n    [\n        [\n            [1<\/span>, 2<\/span>, 3<\/span>],\n            [4<\/span>, 5<\/span>, 6<\/span>]\n        ],\n        [\n            [7<\/span>, 8<\/span>, 9<\/span>],\n            [10<\/span>, 11<\/span>, 12<\/span>]\n        ],\n    ]\n)\nprint(c.shape)  # (2, 2, 3)<\/span><\/code><\/span>Code language:<\/span> Python<\/span> (<\/span>python<\/span>)<\/span><\/small><\/pre>\n\n\n
Output:<\/p>\n\n\n
(3<\/span>,)\n(2<\/span>, 3<\/span>)   \n(2<\/span>, 2<\/span>, 3<\/span>)<\/code><\/span>Code language:<\/span> Python<\/span> (<\/span>python<\/span>)<\/span><\/small><\/pre>\n\n\n
The following picture explains the shape of each array a, b, and c:<\/p>\n\n\n\n
\"\"<\/figure>\n\n\n\n
The shape<\/code> property returns a tuple:<\/p>\n\n\n\n
Second, create a 1D array by passing a list<\/a> of three integers:<\/p>\n\n\n