Creating a Hilbert matrix using SciPy

A Hilbert matrix is a square matrix whose elements are all unit fractions of integer sequence. The unit fraction at a specific location of a Hilbert matrix is given by eij = 1/i+j-1.

Linear Algebra Updated 28 May 2025 1 min read

Overview

  • A Hilbert matrix is an nxn matrix whose elements are given by eij = 1/i+j-1.
  • The elements of the Hilbert matrix are Unit fractions. For example, if n=3, the Hilbert matrix is

    (1 1/2 1/3

    1/2 1/3 1/4

    1/3 1/4 1/5)

  • Unit fractions have one as their numerator. Unit fractions are of the form 1/n where n is an integer.
  • Hilbert matrices have several interesting properties
    • They are square matrices and symmetrical.
    • Their inverses contain all integer elements
    • The anti-diagonal of any Hilbert matrix are constants. i.e, The elements of the anti-diagonal will have the same values.
    • Hilbert matrices are ill-conditioned matrices whose condition numbers are very large. A large condition number means a small change in the input to the system of linear equations as represented by a matrix will result in larger changes in the output.
  • Hilbert matrices are used in image processing to obtain gray scale images. They are used in cryptography and security applications.

Example

python
# Example Python program that creates a Hilbert matrix

            # using the function hilbert() from the scipy.linalg module

import scipy.linalg as sclg

# A Hilbert matrix of size 6x6

            hilbertMatrix = sclg.hilbert(6)

            print("Hilbert matrix:")

            print(hilbertMatrix)

Output

Output
Hilbert matrix:

            [[1.         0.5        0.33333333 0.25       0.2        0.16666667]

             [0.5        0.33333333 0.25       0.2        0.16666667 0.14285714]

             [0.33333333 0.25       0.2        0.16666667 0.14285714 0.125     ]

             [0.25       0.2        0.16666667 0.14285714 0.125      0.11111111]

             [0.2        0.16666667 0.14285714 0.125      0.11111111 0.1       ]

             [0.16666667 0.14285714 0.125      0.11111111 0.1        0.09090909]]