Source code for toqito.matrix_props.is_unitary

"""Checks if the matrix is a unitary matrix."""

import numpy as np

from toqito.matrix_props import is_square


[docs] def is_unitary(mat: np.ndarray, rtol: float = 1e-05, atol: float = 1e-08) -> bool: r"""Check if matrix is unitary [@WikiUniMat]. A matrix is unitary if its inverse is equal to its conjugate transpose. Alternatively, a complex square matrix \(U\) is unitary if its conjugate transpose \(U^*\) is also its inverse, that is, if \[ \begin{equation} U^* U = U U^* = \mathbb{I}, \end{equation} \] where \(\mathbb{I}\) is the identity matrix. Examples: Consider the following matrix \[ X = \begin{pmatrix} 0 & 1 \\ 1 & 0 \end{pmatrix} \] our function indicates that this is indeed a unitary matrix. ```python exec="1" source="above" import numpy as np from toqito.matrix_props import is_unitary A = np.array([[0, 1], [1, 0]]) print(is_unitary(A)) ``` We may also use the `random_unitary` function from `toqito`, and can verify that a randomly generated matrix is unitary ```python exec="1" source="above" from toqito.matrix_props import is_unitary from toqito.rand import random_unitary mat = random_unitary(2) print(is_unitary(mat)) ``` Alternatively, the following example matrix \(B\) defined as \[ B = \begin{pmatrix} 1 & 0 \\ 1 & 1 \end{pmatrix} \] is not unitary. ```python exec="1" source="above" import numpy as np from toqito.matrix_props import is_unitary B = np.array([[1, 0], [1, 1]]) print(is_unitary(B)) ``` Args: mat: Matrix to check. rtol: The relative tolerance parameter (default 1e-05). atol: The absolute tolerance parameter (default 1e-08). Returns: Return `True` if matrix is unitary, and `False` otherwise. """ if not is_square(mat): return False uc_u_mat = mat.conj().T @ mat u_uc_mat = mat @ mat.conj().T id_mat = np.eye(len(mat)) # If U^* @ U = I U @ U^*, the matrix "U" is unitary. return bool( np.allclose(uc_u_mat, id_mat, rtol=rtol, atol=atol) and np.allclose(u_uc_mat, id_mat, rtol=rtol, atol=atol) )