Python numpy.diag_indices_from方法代碼示例

# 需要導入模塊: import numpy [as 別名]
# 或者: from numpy import diag_indices_from [as 別名]
def corr_equi(k_vars, rho):
    '''create equicorrelated correlation matrix with rho on off diagonal

    Parameters
    ----------
    k_vars : int
        number of variables, correlation matrix will be (k_vars, k_vars)
    rho : float
        correlation between any two random variables

    Returns
    -------
    corr : ndarray (k_vars, k_vars)
        correlation matrix

    '''
    corr = np.empty((k_vars, k_vars))
    corr.fill(rho)
    corr[np.diag_indices_from(corr)] = 1
    return corr 

# 需要導入模塊: import numpy [as 別名]
# 或者: from numpy import diag_indices_from [as 別名]
def block_covariance(self):
        "return average covariance within block"
        if self._block_covariance is None:
            if self.ndb <= 1:  # point kriging
                self._block_covariance = self.unbias
            else:
                cov = list()
                for x1, y1, z1 in zip(self.xdb, self.ydb, self.zdb):
                    for x2, y2, z2 in zip(self.xdb, self.ydb, self.zdb):
                        # cov.append(self._cova3((x1, y1, z1), (x2, y2, z2)))
                        cov.append(cova3(
                            (x1, y1, z1), (x2, y2, z2),
                            self.rotmat, self.maxcov, self.nst,
                            self.it, self.cc, self.aa_hmax))
                cov = np.array(cov).reshape((self.ndb, self.ndb))
                cov[np.diag_indices_from(cov)] -= self.c0
                self._block_covariance = np.mean(cov)
        return self._block_covariance 

# 需要導入模塊: import numpy [as 別名]
# 或者: from numpy import diag_indices_from [as 別名]
def diag_indices_from(arr):
    """
    Return the indices to access the main diagonal of an n-dimensional array.
    See `diag_indices` for full details.

    Args:
        arr (cupy.ndarray): At least 2-D.

     .. seealso:: :func:`numpy.diag_indices_from`

    """
    if not isinstance(arr, cupy.ndarray):
        raise TypeError("Argument must be cupy.ndarray")

    if not arr.ndim >= 2:
        raise ValueError("input array must be at least 2-d")
    # For more than d=2, the strided formula is only valid for arrays with
    # all dimensions equal, so we check first.
    if not cupy.all(cupy.diff(arr.shape) == 0):
        raise ValueError("All dimensions of input must be of equal length")

    return diag_indices(arr.shape[0], arr.ndim) 

# 需要導入模塊: import numpy [as 別名]
# 或者: from numpy import diag_indices_from [as 別名]
def first_fit(self, train_x, train_y):
        """ Fit the regressor for the first time. """
        train_x, train_y = np.array(train_x), np.array(train_y)

        self._x = np.copy(train_x)
        self._y = np.copy(train_y)

        self._distance_matrix = edit_distance_matrix(self._x)
        k_matrix = bourgain_embedding_matrix(self._distance_matrix)
        k_matrix[np.diag_indices_from(k_matrix)] += self.alpha

        self._l_matrix = cholesky(k_matrix, lower=True)  # Line 2

        self._alpha_vector = cho_solve(
            (self._l_matrix, True), self._y)  # Line 3

        self._first_fitted = True
        return self 

# 需要導入模塊: import numpy [as 別名]
# 或者: from numpy import diag_indices_from [as 別名]
def do_lagrange(ma, vb, factor):
    """
    Lagrange multipliers.

    Parameters
    ----------
    ma : NumPy matrix
    vb : NumPy vector
    factor : float
    """
    mac = copy.deepcopy(ma)
    ind = np.diag_indices_from(mac)
    logger.log(5, 'A:\n{}'.format(mac))
    mac[ind] = mac[ind] + factor
    logger.log(5, 'A:\n{}'.format(mac))
    changes = solver(mac, vb)
    return [('LAGRANGE F{}'.format(factor), changes)] 

# 需要導入模塊: import numpy [as 別名]
# 或者: from numpy import diag_indices_from [as 別名]
def do_levenberg(ma, vb, factor):
    """
    Lagrange multipliers.

    Parameters
    ----------
    ma : NumPy matrix
    vb : NumPy vector
    factor : float
    """
    mac = copy.deepcopy(ma)
    ind = np.diag_indices_from(mac)
    mac[ind] = mac[ind] + factor
    logger.log(5, 'A:\n{}'.format(mac))
    changes = solver(mac, vb)
    return [('LM {}'.format(factor), changes)] 

# 需要導入模塊: import numpy [as 別名]
# 或者: from numpy import diag_indices_from [as 別名]
def _kalman_correct(x, P, z, H, R, gain_factor, gain_curve):
    PHT = np.dot(P, H.T)

    S = np.dot(H, PHT) + R
    e = z - H.dot(x)
    L = cholesky(S, lower=True)
    inn = solve_triangular(L, e, lower=True)

    if gain_curve is not None:
        q = (np.dot(inn, inn) / inn.shape[0]) ** 0.5
        f = gain_curve(q)
        if f == 0:
            return inn
        L *= (q / f) ** 0.5

    K = cho_solve((L, True), PHT.T, overwrite_b=True).T
    if gain_factor is not None:
        K *= gain_factor[:, None]

    U = -K.dot(H)
    U[np.diag_indices_from(U)] += 1
    x += K.dot(z - H.dot(x))
    P[:] = U.dot(P).dot(U.T) + K.dot(R).dot(K.T)

    return inn 

# 需要導入模塊: import numpy [as 別名]
# 或者: from numpy import diag_indices_from [as 別名]
def order_cl_pixels(x_pix,y_pix):
    dist = distance.cdist(np.array([x_pix,y_pix]).T,np.array([x_pix,y_pix]).T)
    dist[np.diag_indices_from(dist)]=100.0
    ind = np.argmin(x_pix) # select starting point on left side of image
    clinds = [ind]
    count = 0
    while count<len(x_pix):
        t = dist[ind,:].copy()
        if len(clinds)>2:
            t[clinds[-2]]=t[clinds[-2]]+100.0
            t[clinds[-3]]=t[clinds[-3]]+100.0
        ind = np.argmin(t)
        clinds.append(ind)
        count=count+1
    x_pix = x_pix[clinds]
    y_pix = y_pix[clinds]
    return x_pix,y_pix 

# 需要導入模塊: import numpy [as 別名]
# 或者: from numpy import diag_indices_from [as 別名]
def test_symmetrical_mi_nonzero():
    # test that the MI matrix for sets of uncorrelated things results
    # in zero MI

    nonzero_mi_funcs = [nonzero_mi_np, nonzero_mi_ra, nonzero_mi_list]
    for a, n_states in (f() for f in nonzero_mi_funcs):

        mi = mutual_info.mi_matrix(a, a, n_states, n_states)

        assert_almost_equal(mi[-1, -2], 0.86114, decimal=3)
        mi[-1, -2] = mi[-2, -1] = 0

        assert_almost_equal(np.diag(mi), 0.86114, decimal=2)
        mi[np.diag_indices_from(mi)] = 0

        assert_allclose(mi, 0, atol=1e-3) 

# 需要導入模塊: import numpy [as 別名]
# 或者: from numpy import diag_indices_from [as 別名]
def test_symmetrical_mi_nonzero_int_shape_spec():
    # test that when we use an integer (rather than a list of integers)
    # that we correctly assume that the integer is just repeated for all
    # of the various features.

    nonzero_mi_funcs = [nonzero_mi_np, nonzero_mi_ra, nonzero_mi_list]
    for a, n_states in (f() for f in nonzero_mi_funcs):

        mi = mutual_info.mi_matrix(a, a, 5, 5)

        assert_almost_equal(mi[-1, -2], 0.86114, decimal=3)
        mi[-1, -2] = mi[-2, -1] = 0

        assert_almost_equal(np.diag(mi), 0.86114, decimal=2)
        mi[np.diag_indices_from(mi)] = 0

        assert_allclose(mi, 0, atol=1e-3) 

# 需要導入模塊: import numpy [as 別名]
# 或者: from numpy import diag_indices_from [as 別名]
def energy_nuc(mol, charges=None, coords=None):
    '''Compute nuclear repulsion energy (AU) or static Coulomb energy

    Returns
        float
    '''
    if charges is None: charges = mol.atom_charges()
    if len(charges) <= 1:
        return 0
    #e = 0
    #for j in range(len(mol._atm)):
    #    q2 = charges[j]
    #    r2 = coords[j]
    #    for i in range(j):
    #        q1 = charges[i]
    #        r1 = coords[i]
    #        r = numpy.linalg.norm(r1-r2)
    #        e += q1 * q2 / r
    rr = inter_distance(mol, coords)
    rr[numpy.diag_indices_from(rr)] = 1e200
    if CHECK_GEOM and numpy.any(rr < 1e-5):
        for atm_idx in numpy.argwhere(rr<1e-5):
            logger.warn(mol, 'Atoms %s have the same coordinates', atm_idx)
        raise RuntimeError('Ill geometry')
    e = numpy.einsum('i,ij,j->', charges, 1./rr, charges) * .5
    return e 

# 需要導入模塊: import numpy [as 別名]
# 或者: from numpy import diag_indices_from [as 別名]
def inter_distance(mol, coords=None):
    '''
    Inter-particle distance array
    '''
    if coords is None: coords = mol.atom_coords()
    rr = numpy.linalg.norm(coords.reshape(-1,1,3) - coords, axis=2)
    rr[numpy.diag_indices_from(rr)] = 0
    return rr 

# 需要導入模塊: import numpy [as 別名]
# 或者: from numpy import diag_indices_from [as 別名]
def matrix_sign(M):
    """ The "sign" matrix of `M` """
    #Notes: sign(M) defined s.t. eigvecs of sign(M) are evecs of M
    # and evals of sign(M) are +/-1 or 0 based on sign of eigenvalues of M

    #Using the extremely numerically stable (but expensive) Schur method
    # see http://www.maths.manchester.ac.uk/~higham/fm/OT104HighamChapter5.pdf
    N = M.shape[0]; assert(M.shape == (N, N)), "M must be square!"
    T, Z = _spl.schur(M, 'complex')  # M = Z T Z^H where Z is unitary and T is upper-triangular
    U = _np.zeros(T.shape, 'complex')  # will be sign(T), which is easy to compute
    # (U is also upper triangular), and then sign(M) = Z U Z^H

    # diagonals are easy
    U[_np.diag_indices_from(U)] = _np.sign(_np.diagonal(T))

    #Off diagonals: use U^2 = I or TU = UT
    # Note: Tij = Uij = 0 when i > j and i==j easy so just consider i<j case
    # 0 = sum_k Uik Ukj =  (i!=j b/c off-diag)
    # FUTURE: speed this up by using np.dot instead of sums below
    for j in range(1, N):
        for i in range(j - 1, -1, -1):
            S = U[i, i] + U[j, j]
            if _np.isclose(S, 0):  # then use TU = UT
                if _np.isclose(T[i, i] - T[j, j], 0):  # then just set to zero
                    U[i, j] = 0.0  # TODO: check correctness of this case
                else:
                    U[i, j] = T[i, j] * (U[i, i] - U[j, j]) / (T[i, i] - T[j, j]) + \
                        sum([U[i, k] * T[k, j] - T[i, k] * U[k, j] for k in range(i + 1, j)]) \
                        / (T[i, i] - T[j, j])
            else:  # use U^2 = I
                U[i, j] = - sum([U[i, k] * U[k, j] for k in range(i + 1, j)]) / S
    return _np.dot(Z, _np.dot(U, _np.conjugate(Z.T)))

    #Quick & dirty - not always stable:
    #U,_,Vt = _np.linalg.svd(M)
    #return _np.dot(U,Vt) 

# 需要導入模塊: import numpy [as 別名]
# 或者: from numpy import diag_indices_from [as 別名]
def fun(self, x):
        dx, dy, dz = self._compute_coordinate_deltas(x)
        with np.errstate(divide='ignore'):
            dm1 = (dx**2 + dy**2 + dz**2) ** -0.5
        dm1[np.diag_indices_from(dm1)] = 0
        return 0.5 * np.sum(dm1) 

# 需要導入模塊: import numpy [as 別名]
# 或者: from numpy import diag_indices_from [as 別名]
def grad(self, x):
        dx, dy, dz = self._compute_coordinate_deltas(x)

        with np.errstate(divide='ignore'):
            dm3 = (dx**2 + dy**2 + dz**2) ** -1.5
        dm3[np.diag_indices_from(dm3)] = 0

        grad_x = -np.sum(dx * dm3, axis=1)
        grad_y = -np.sum(dy * dm3, axis=1)
        grad_z = -np.sum(dz * dm3, axis=1)

        return np.hstack((grad_x, grad_y, grad_z))