Source code for toqito.states.gisin

"""Gisin states are the mixtures of separable mixed states and some purely entangled states."""

import numpy as np




def gisin(lambda_var: float, theta: float) -> np.ndarray:
    r"""Produce a Gisin state [@Gisin_1996_Hidden].

    Returns the Gisin state described in [@Gisin_1996_Hidden]. Specifically, the Gisin state can be defined as:

    \[
        \begin{equation}
            \rho_{\lambda, \theta} = \lambda
                                    \begin{pmatrix}
                                        0 & 0 & 0 & 0 \\
                                        0 & \sin^2(\theta) &
                                        -\sin(\theta)\cos(\theta) & 0 \\
                                        0 & -\sin(\theta)\cos(\theta) &
                                        \cos^2(\theta) & 0 \\
                                        0 & 0 & 0 & 0
                                    \end{pmatrix} +
                                    \frac{1 - \lambda}{2}
                                    \begin{pmatrix}
                                        1 & 0 & 0 & 0 \\
                                        0 & 0 & 0 & 0 \\
                                        0 & 0 & 0 & 0 \\
                                        0 & 0 & 0 & 1
                                    \end{pmatrix}.
        \end{equation}
    \]

    Examples:
        The following code generates the Gisin state \(\rho_{0.5, 1}\).

        ```python exec="1" source="above"
        from toqito.states import gisin
        print(gisin(0.5, 1))
        ```

    Raises:
        ValueError: If `lambda_var` is not a real number.

    Args:
        lambda_var: A real parameter in [0, 1].
        theta: A real parameter.

    Returns:
        Gisin state.

    """
    if lambda_var < 0 or lambda_var > 1:
        raise ValueError("InvalidLambda: Variable lambda must be between 0 and 1.")

    rho_theta = np.array(
        [
            [0, 0, 0, 0],
            [0, np.sin(theta) ** 2, -np.sin(2 * theta) / 2, 0],
            [0, -np.sin(2 * theta) / 2, np.cos(theta) ** 2, 0],
            [0, 0, 0, 0],
        ]
    )

    rho_uu_dd = np.array([[1, 0, 0, 0], [0, 0, 0, 0], [0, 0, 0, 0], [0, 0, 0, 1]])

    return lambda_var * rho_theta + (1 - lambda_var) * rho_uu_dd / 2