def __init__(self, A, B):

        a = not isinstance(A, Dist) or 1 and len(A)
        b = not isinstance(B, Dist) or 1 and len(B)
        length = max(a,b)
        Dist.__init__(self, A=A, B=B,
                _length=length, _advance=True)

 def __init__(self, A, B):
     if isinstance(A, Dist):
         assert not A.dependent()
     if isinstance(B, Dist):
         assert not B.dependent()
     a = not isinstance(A, Dist) or 1 and len(A)
     b = not isinstance(B, Dist) or 1 and len(B)
     length = max(a, b)
     Dist.__init__(self, A=A, B=B,
             _length=length, _advance=True)

    def __init__(self, dist, trans):
        """
Parameters
----------
dist : Dist
    Distribution to wrap the copula around
trans : Dist
    The copula wrapper. [0,1]^D \into [0,1]^D
        """
        Dist.__init__(self, dist=dist, trans=trans, _advance=True, _length=len(trans))

    def __init__(self, loc=[0,0], scale=[[1,.5],[.5,1]]):

        loc, scale = np.asfarray(loc), np.asfarray(scale)
        assert len(loc)==len(scale)

        dist = joint.Iid(normal(), len(loc))
        C = np.linalg.cholesky(scale)
        Ci = np.linalg.inv(C)
        Dist.__init__(self, dist=dist, loc=loc, C=C, Ci=Ci,
                scale=scale, _length=len(scale), _advance=True)

    def __init__(self, dist, N):
        """
Parameters
----------
dist : Dist
    Input variable. Must have `len(dist)==1`.
N : int
    Number of variable in the output.
        """
        assert len(dist)==1 and N>1
        Dist.__init__(self, dist=dist, _length=N)

 def __init__(self, dist):
     Dist.__init__(self, dist=dist, _length=len(dist),
             _advance=True)