Difference between Numpy array and Numpy matrix

While working with Python many times we come across the question that what exactly is the difference between a numpy array and numpy matrix, in this article we are going to read about the same.

What is np.array() in Python

The Numpy array object in Numpy is called ndarray. We can create ndarray using numpy.array() function. It is used to convert a list, tuple, etc. into a Numpy array.

Syntax: numpy.array(object, dtype=None)

Creating a NumPy array using the .array() function.

Python3

# importing numpy module
import numpy as np

# creating numpy array
res = np.array(['G', 'F', 'G'])

print("Numpy Array in python :", res)
print(type(res))

Output:

Numpy Array in python : ['G' 'F' 'G']
<class 'numpy.ndarray'>

What is numpy.matrix() in Python

A matrix in Numpy returns a matrix from a string of data or array-like object. The matrix obtained is a specialized 2D array.

Syntax: numpy.matrix(object, dtype = None) :

Creating a NumPy matrix using the .matrix() function.

Python3

import numpy as np

mat = np.matrix([[1, 2, 3],
                 [5, 6, 7],
                 [4, 6, 8]])

print("Matrix is : \n", mat)
print(type(mat))

Output:

Matrix is : 
 [[1 2 3]
 [5 6 7]
 [4 6 8]]

<class 'numpy.matrix'>

Difference between Numpy array and Numpy Matrix

Matrix is 2-dimensional while ndarray can be multi-dimensional

Example 1: 

Here, we can print all the dimensions of an array in np.array.

Python3

import numpy as np

arr1 = np.array([1, 2, 3])
print("1D array\n", arr1)
print("\n")

arr2 = np.array([[1, 2], [3, 4]])
print("2D array\n", arr2)
print("\n")

C = np.array([[[1, 2], [3, 4]],
             [[5, 6], [7, 8]],
             [[9, 10], [11, 12]]])
print("3D array\n", C)

Output:

1D array
 [1 2 3]


2D array
 [[1 2]
 [3 4]]


3D array
 [[[ 1  2]
  [ 3  4]]

 [[ 5  6]
  [ 7  8]]

 [[ 9 10]
  [11 12]]]

Example 2: 

Matrix works normally for a 2D matrix and if a 1D matrix will convert into 2D Matrix, but if we pass a 3D matrix it will through an error. 

Python3

import numpy as np

arr1 = np.matrix([1, 2, 3])
print(arr1)
print("Dimensions:", arr1.ndim)
print("\n")

arr2 = np.matrix([[[1, 2], [3, 4]],
                  [[5, 6], [7, 8]],
                  [[9, 10], [11, 12]]])
print("2D array\n", arr2)

Output:

Different functionality of  * operator in ndarray and Matrix

Example 1:

Array * operator does simple multiplication.

Python3

a = np.array([[1, 2],
             [3, 4]])
b = np.array([[1, 2],
             [3, 4]])

print("Array multiplication: \n", a*b)

Output:

Array multiplication: 
 [[ 1  4]
 [ 9 16]]

Example 2: 

While it does matrix multiplication.

Python3

a = np.matrix([[1, 2],
             [3, 4]])
b = np.matrix([[1, 2],
             [3, 4]])

print("Matrix multiplication: \n", a*b)

Output:

Matrix multiplication: 
 [[ 7 10]
 [15 22]]

Matrix has an array.I for inverse, but ndarray has linalg.inv

Example 1: 

The inverse can be done with array.I in ndarray.

Python3

arr1 = np.matrix([[1, 2],
              [3, 4]])

print('Inverse \n', arr1.I)

Output:

Inverse 
 [[-2.   1. ]
 [ 1.5 -0.5]]

Example 2:

The inverse can be done with np.linalg.inv in matrix.

Python3

b = np.array([[1, 2],
             [3, 4]])

print('Inverse \n', np.linalg.inv(b))

Output:

Inverse 
 [[-2.   1. ]
 [ 1.5 -0.5]]

Table of Differences between the Numpy Array and Numpy Matrix

Numpy np.array()	Numpy np.matrix()
Syntax:  numpy.array(object, dtype=None)	Syntax: numpy.matrix(object, dtype=None)
Numpy arrays (nd-arrays) are N-dimensional where, N=1,2,3…	Numpy matrices are strictly 2-dimensional.
nd-arrays are base classes for matrix objects	Matrix objects are a subclass of nd-array.
Matrix objects have arr.I for the inverse.	Array objects don’t.
If a and b are matrices, then a@b or np.dot(a,b) is their matrix product	If a and b are matrices, then a*b is their matrix product.