Function fast_cross_3d

mne/surface.py:335–367

Compute cross product between list of 3D vectors. Much faster than np.cross() when the number of cross products becomes large (>= 500). This is because np.cross() methods become less memory efficient at this stage. Parameters ---------- x : array Input array 1, shap

(x, y)

def fast_cross_3d(x, y):
    """Compute cross product between list of 3D vectors.

    Much faster than np.cross() when the number of cross products
    becomes large (>= 500). This is because np.cross() methods become
    less memory efficient at this stage.

    Parameters
    ----------
    x : array
        Input array 1, shape (..., 3).
    y : array
        Input array 2, shape (..., 3).

    Returns
    -------
    z : array, shape (..., 3)
        Cross product of x and y along the last dimension.

    Notes
    -----
    x and y must broadcast against each other.
    """
    assert x.ndim >= 1
    assert y.ndim >= 1
    assert x.shape[-1] == 3
    assert y.shape[-1] == 3
    if max(x.size, y.size) >= 500:
        out = np.empty(np.broadcast(x, y).shape)
        _jit_cross(out, x, y)
        return out
    else:
        return np.cross(x, y)


@jit()