def ComputeJacobianMatrix(self,
vec_x,
vec_xi=None,
rep_loc=None): #need validation
"""
Calcul l'inverse du Jacobien aux points de gauss dans le cas d'un élément isoparamétrique (c'est à dire que les mêmes fonctions de forme sont utilisées)
vec_xi est un tableau dont les lignes donnent les coordonnées dans le repère de référence où on souhaite avoir le jacobien (en général pg)
"""
if vec_xi is None: vec_xi = self.xi_pg
self.ComputeDetJacobian(vec_x, vec_xi)
if rep_loc != None:
rep_pg = self.interpolateLocalFrame(
rep_loc,
vec_xi) #interpolation du repère local aux points de gauss
self.JacobianMatrix = sp.matmul(
self.JacobianMatrix, sp.swapaxes(rep_pg, 2, 3)
) #to verify - sp.swapaxes(rep_pg,2,3) is equivalent to a transpose over the axis 2 and 3
# for k,J in enumerate(self.JacobianMatrix): self.JacobianMatrix[k] = sp.dot(J, rep_pg[k].T)
if self.JacobianMatrix.shape[-2] == self.JacobianMatrix.shape[-1]:
self.inverseJacobian = linalg.inv(self.JacobianMatrix)
# self.inverseJacobian = [linalg.inv(J) for J in self.JacobianMatrix]
else: #l'espace réel est dans une dimension plus grande que l'espace de l'élément de référence
J = self.JacobianMatrix
JT = sp.swapaxes(self.JacobianMatrix, 2, 3)
self.inverseJacobian = sp.matmul(
JT, linalg.inv(sp.matmul(J, JT))
) #inverseJacobian.shape = (Nel,len(vec_xi)=nb_pg, dim:vec_x.shape[-1], dim:vec_xi.shape[-1])
def rho_dot(Hamiltonian, Gamma, rho, closed):
"""
Calculate the time derivative of the density matrix
This function uses the Lindblad form of the quantum master equation (Steck eq. 5.177 & 5.178 pg 181)
.. math::
\\dot{\\rho} = -\\frac{i}{\\hbar} [H, \\rho] + \\Gamma D[c] \\rho
where
.. math::
D[c] \\rho = c\\rho c^{\\dagger} - \\frac{1}{2} \\{c^{\\dagger} c, \\rho\\}.
:math:`\Gamma` is the decay rate and :math:`c` along with its Hermitian conjugate :math:`c^{\dagger}` are the transition matrices
of the n level system.
The documentation also has some extra information.
For now only radiative decay is considered (no dephasing).
:param Hamiltonian: Complex ndarray. Hamiltonian.
:param Gamma: Ndarray. Decay matrix.
:param rho: Complex ndarray. density matrix.
:return: Complex ndarray. Time derivative of the density matrix.
"""
return -1j / hbar * commutator(Hamiltonian, rho) \
- 1 / 2 * anticommutator(Gamma, rho) \
+ sum(sp.matmul(closed[0], sp.matmul(rho, closed[1])))
def ComputeJacobianMatrix(self, vec_x, vec_xi=None, rep_loc=None):
if vec_xi is None: vec_xi == self.xi_pg
self.ComputeDetJacobian(vec_x, vec_xi)
if self.JacobianMatrix.shape[-2] == self.JacobianMatrix.shape[-1]:
if rep_loc != None:
rep_pg = self.interpolateLocalFrame(
rep_loc, vec_xi
) #interpolation du repère local aux points de gauss -> shape = (len(vec_x),len(vec_xi),dim,dim)
self.JacobianMatrix = sp.matmul(
self.JacobianMatrix, sp.swapaxes(rep_pg, 2, 3)
) #to verify - sp.swapaxes(rep_pg,2,3) is equivalent to a transpose over the axis 2 and 3
# rep_pg = self.interpolateLocalFrame(rep_loc, vec_xi) #interpolation du repère local aux points de gauss
# for k,J in enumerate(self.JacobianMatrix): self.JacobianMatrix[k] = sp.dot(J, rep_pg[k].T)
else: #l'espace réel est dans une dimension plus grande que l'espace de l'élément de référence
if rep_loc is None:
J = self.JacobianMatrix
JT = sp.swapaxes(self.JacobianMatrix, 2, 3)
self.inverseJacobian = sp.matmul(
JT, linalg.inv(sp.matmul(J, JT))
) #inverseJacobian.shape = (Nel,len(vec_xi)=nb_pg, dim:vec_x.shape[-1], dim:vec_xi.shape[-1])
# self.inverseJacobian = [sp.dot(J.T , linalg.inv(sp.dot(J,J.T))) for J in self.JacobianMatrix] #this line may have a high computational cost
return
else:
rep_pg = self.repLocFromJac(rep_loc, vec_xi)[:, 0:2, :]
for k, J in enumerate(self.JacobianMatrix):
self.JacobianMatrix[k] = sp.dot(J, rep_pg[k].T)
self.inverseJacobian = linalg.inv(self.JacobianMatrix)
def F1(params, flag=False):
if m0:
_m, _w = m0, params[0]
elif w0:
_m, _w = params[0], w0
else:
_m, _w = params
tmp = np.power(np.abs(tc - t), _m)
X = [
np.ones(N), tmp, tmp * np.cos(_w * np.log(np.abs(tc - t))),
tmp * np.sin(_w * np.log(np.abs(tc - t)))
]
X = np.vstack(X).T
try:
beta = sp.matmul(sp.linalg.inv(sp.matmul(X.T, X)),
sp.matmul(X.T, y))
except:
# print(_m, _w)
raise ValueError("X is singular")
SSE = np.sum(np.power(y - np.matmul(beta, X.T), 2))
if flag:
return SSE, beta
return SSE
def predict(self):
"""
The predict step in Kalman filtering.
:return:
"""
# Predict the state and covariance
self.x = matmul(self.A, self.x) + matmul(self.B, self.u)
self.P = matmul(matmul(self.A, self.P), self.A.T) + self.Q
def least_squares_interpolant(x, y):
xmat = sp.zeros((len(x), 2))
for i in range(len(x)):
for j in range(2):
xmat[i][j] = x[i] ** j
a = sp.matmul(sp.transpose(xmat), xmat)
b = sp.matmul(sp.transpose(xmat), y)
ans = numpy.linalg.solve(a, b)
return ans, xmat
def update_reservoir(self, u, n, Y):
# u is input at specific time
# u has shape (N_u (3 for L63))
# See page 16 eqtn 18 of Lukosevicius PracticalESN for feedback info.
x_n_tilde = cp.tanh(
cp.matmul(self.W, self.x[n]) +
cp.array(sp.matmul(self.W_in, sp.hstack((sp.array([1]), u)))) +
cp.array(sp.matmul(self.W_fb, Y)))
self.x[n+1] = cp.multiply((1-cp.array(self.alpha_matrix)), cp.array(self.x[n])) \
+ cp.multiply(cp.array(self.alpha_matrix), x_n_tilde)
def update_reservoir(self, u, n, Y):
# u is input at specific time
# u has shape (N_u (3 for L63))
# See page 16 eqtn 18 of Lukosevicius PracticalESN for feedback info.
x_n_tilde = sp.tanh(
sp.matmul(self.W, self.x[n]) +
sp.matmul(self.W_in, sp.hstack((sp.array([1]), u))) +
sp.matmul(self.W_fb, Y)) # TODO: Add derivative term?
self.x[n+1] = sp.multiply((1-self.alpha_matrix), self.x[n]) \
+ sp.multiply(self.alpha_matrix, x_n_tilde)
def chi_formal(Ham, psi_i, psi_0, t):
extract = Matrix_diag(Ham(t))
H_d = extract[0]
M = extract[1]
M_dag = extract[2]
exponential = sp.zeros([len(H_d), len(H_d)], dtype=complex)
for i in range(len(H_d)):
exponential[i][i] = sp.exp(-1j * t * H_d[i][i])
psi_i_bra = sp.matrix.conjugate(sp.transpose(psi_i))
psi_t = sp.matmul(M, sp.matmul(exponential, sp.matmul(M_dag, psi_0)))
chi_rooted = sp.matmul(psi_i_bra, psi_t)
return abs(chi_rooted)**2
def cart2frac(self):
"""Convert Cartesian coordinates to fractional"""
if self.cart:
lat_inv = inv(self.lat)
for i in range(self.nat):
self.r[i,:] = sp.matmul(lat_inv, self.r[i,:])
self.cart = False
def output_Y(self, u, n):
one_by_one = sp.array([1])
# Eqtn 4
concatinated_matrix = sp.hstack((sp.hstack(
(one_by_one, u)), cp.asnumpy(self.x[n])))
output_Y_return = sp.matmul(self.W_out, concatinated_matrix)
return output_Y_return
def WikiCG(A, b, *, e=1e-8, iter_max=100):
x = sp.zeros(b.shape, dtype=float)
r = b - sp.matmul(A, x)
p = sp.copy(r)
rr = sp.dot(r, r)
for _ in range(iter_max):
Ap = sp.matmul(A, p)
alpha = rr / sp.matmul(p, Ap)
x += alpha * p
r -= alpha * Ap
if linalg.norm(r) < e:
break
beta = rr
rr = sp.dot(r, r)
beta = rr / beta
p = r + beta * p
return x
def fun(t, z):
i = 1
Fr = sp.zeros(40)
while i < 20:
Fr[i +
20] = -(CAP[i]) * invM[i, i] * sp.tanh((z[20 + i]) - z[20 + i - 1])
i += 1
return sp.matmul(A, z) + Fr
def calculate_W_out(self, Y_target, N_x, beta, train_start_timestep,
train_end_timestep):
# see Lukosevicius Practical ESN eqtn 11
# Using ridge regression
N_u = sp.shape(Y_target)[0]
X = sp.vstack(
(sp.ones((1, train_end_timestep - train_start_timestep)),
Y_target[:, train_start_timestep:train_end_timestep],
self.x[train_start_timestep:train_end_timestep].transpose()))
# Ridge Regression
W_out = sp.matmul(
sp.array(Y_target[:, train_start_timestep + 1:train_end_timestep +
1]),
sp.matmul(
X.transpose(),
sp.linalg.inv(
sp.matmul(X, X.transpose()) +
beta * sp.identity(1 + N_x + N_u))))
self.W_out = W_out
def update(self, z):
"""
The update step in Kalman filtering.
:param z: The measurements.
:return:
"""
# Calculate the Kalman gain
S = matmul(matmul(self.H, self.P), self.H.T) + self.R
K = matmul(matmul(self.P, self.H.T), inv(S))
# Correction based on the measurement
residual = z - matmul(self.H, self.x)
self.x = self.x + matmul(K, residual)
self.P = self.P - matmul(matmul(K, self.H), self.P)
self.state.append(self.x)
self.residual.append(norm(residual))
self.epsilon.append(matmul(residual.T, matmul(inv(S), residual)))
self.sigma.append(sqrt(S[0, 0]**2 + S[1, 1]**2 + S[2, 2]**2))
self.process_noise.append(self.Q[0, 0])
def fun(t,z):
# --- Reporte de paso de tiempo
if t > fun.tnextreport:
print " {} at t = {}".format(fun.solver, fun.tnextreport)
fun.tnextreport += 1
# --- Cálculo
Famort = sp.zeros((40)) # vector de fuerza friccional de amortiguamiento
Famort[0] = -(Cap[0] * (1./M[0,0]) * sp.tanh((z[20]/vr)))
Ft=sp.zeros(40)
if t<226.81:
Ft[20:]=f0(t)*9.8
return sp.matmul(A,z) + Famort + Ft
def BookCG(A, b, e=1e-8, k_max=100):
k = 0
d_old = 0
x = sp.zeros((A.shape[0]))
r = b - sp.matmul(A, x)
d = sp.dot(r, r)
while sp.sqrt(d) > e * sp.sqrt(sp.dot(b, b)) and k < k_max:
k = k + 1
if k == 1:
p = r
else:
beta = d / d_old
p = r + beta * p
w = sp.matmul(A, p)
a = d / sp.dot(p, w)
x = x + a * p
r = r - a * w
d_old = d
d = sp.dot(r, r)
return x
def rho_dot(Hamiltonian, Gamma, rho, closed):
"""
Calculate the time derivative of the density matrix using the quantum master equation (Steck eq. 5.177 & 5.178 pg 181)
rho_dot = -i / hbar * [H, rho] + Gamma * D[c] * rho
where
D[c] * rho = 1 / 2 * {c_dagger * c, rho}.
Gamma is the decay rate and c are the transition matrices of the n level system. In practice c will be a sum of
multiple transition matrices.
NOTE: 1.) For now only radiative decay is considered (no dephasing) and all levels have the same decay rate.
2.) The Lindblad operator does not include the term that adds population back in. Therefore this function
simulates open quantum systems
:param Hamiltonian: Hamiltonian.
:param Gamma: Decay matrix.
:param rho: density matrix.
:return: Time derivative of the density matrix as a 2-D ndarray (dtype=complex).
"""
return -1j / hbar * commutator(Hamiltonian, rho) \
- 1 / 2 * anticommutator(Gamma, rho) \
+ sum(sp.matmul(closed[0], sp.matmul(rho, closed[1])))
def main(n = 3):
A = [[20*random() for j in range(n)] for i in range(n)]
b = [20*random() for i in range(n)]
L, U = descompLU(A, b)
printMatrix(A, 'A')
print()
printMatrix(L, 'L')
print()
printMatrix(U, 'U')
print()
printMatrix(matmul(L, U), 'LU')
def rotation_matrix_from_vectors(a, b):
a = scipy.asarray(a)
b = scipy.asarray(b)
a /= scipy.linalg.norm(a)
b /= scipy.linalg.norm(b)
v = scipy.cross(a, b)
c = scipy.dot(a, b)
s = scipy.linalg.norm(v)
I = scipy.identity(3, dtype='f8')
k = scipy.array([[0., -v[2], v[1]], [v[2], 0., -v[0]], [-v[1], v[0], 0.]])
if s == 0.: return I
return I + k + scipy.matmul(k, k) * ((1. - c) / (s**2))
def reflect_by_householder(v, u, c):
print("Reflect by Householder:")
print("chord:", v)
print("Unit normal vector:", u)
center_ = center(len(v))
tensor_ = scipy.outer(u, u)
print("tensor_:\n", tensor_)
product_ = scipy.multiply(tensor_, 2)
print("product_:", product_)
identity_ = scipy.eye(len(v))
print("identity_:\n", identity_)
householder = scipy.subtract(identity_, product_)
print("householder:\n", householder)
reflected_voices = scipy.matmul(householder, v)
print("reflected_voices:", reflected_voices)
translated_voices = scipy.subtract(v, center_)
print("moved to origin:", translated_voices)
reflected_translated_voices = scipy.matmul(householder, translated_voices)
print("reflected_translated_voices:", reflected_translated_voices)
reflection = scipy.add(reflected_translated_voices, center_)
print("moved from origin:", reflection)
print("reflection by householder:", reflection)
return reflection
def oht_alg(d2d_to_d2d_gains_diag, uav_to_d2d_gains, d2d_to_d2d_gains_diff, eta, power_UAV, power_cir_UAV):
theta_ini = Parameter(value=1 / 0.5)
iter = 0
epsilon = 1
theta_sol = 0
iter_phi = []
while epsilon >= 1e-2 and iter <= 20:
iter += 1
if iter == 1:
theta_ref = theta_ini.value
else:
theta_ref = theta_sol
term_x = sp.divide(1,
sp.multiply(sp.subtract(theta_ref, 1), sp.matmul(d2d_to_d2d_gains_diag, uav_to_d2d_gains)))
term_y = sp.add(
sp.multiply(sp.subtract(theta_ref, 1), sp.matmul(sp.transpose(d2d_to_d2d_gains_diff), uav_to_d2d_gains)),
sp.divide(1, eta * power_UAV))
a_1 = sp.add(sp.divide(sp.multiply(2, sp.log(sp.add(1, sp.divide(1, sp.multiply(term_x, term_y))))), theta_ref),
sp.divide(2, sp.multiply(theta_ref, sp.add(sp.multiply(term_x, term_y), 1))))
b_1 = sp.divide(1, sp.multiply(theta_ref, sp.multiply(term_x, sp.add(sp.multiply(term_x, term_y), 1))))
c_1 = sp.divide(1, sp.multiply(theta_ref, sp.multiply(term_y, sp.add(sp.multiply(term_x, term_y), 1))))
d_1 = sp.divide(sp.log(sp.add(1, sp.divide(1, sp.multiply(term_x, term_y)))), sp.square(theta_ref))
theta = NonNegative(1)
t_max = NonNegative(1)
obj_opt = Maximize(t_max)
constraints = [theta >= 1]
constraints.append(
t_max <= a_1 - sp.divide(b_1, sp.matmul(d2d_to_d2d_gains_diag, uav_to_d2d_gains)) * inv_pos(theta - 1)
- mul_elemwise(c_1,
sp.matmul(sp.transpose(d2d_to_d2d_gains_diff), uav_to_d2d_gains) * (theta - 1)
+ sp.divide(1, eta * power_UAV))
- d_1 * theta)
t1 = time.time()
prob = Problem(obj_opt, constraints)
prob.solve(solver=ECOS_BB)
theta_sol = theta.value
phi_n_sol = sp.multiply((theta_sol - 1) * eta * power_UAV, uav_to_d2d_gains)
x_rate = sp.matmul(d2d_to_d2d_gains_diag, phi_n_sol)
term_rate = sp.matmul(sp.transpose(d2d_to_d2d_gains_diff), phi_n_sol) + 1
rate_sol_ue = sp.divide(sp.log(sp.add(1, sp.divide(x_rate, term_rate))), theta_sol)
iter_maximin_rate = min(rate_sol_ue)
term_pow_iter = sp.subtract(1, sp.divide(1, theta_sol)) * eta * power_UAV * sp.add(1, sp.sum(
uav_to_d2d_gains)) + power_cir_UAV
iter_phi.append(t_max.value)
if iter >= 2:
epsilon = sp.divide(sp.absolute(sp.subtract(iter_phi[iter - 1], iter_phi[iter - 2])),
sp.absolute(iter_phi[iter - 2]))
iter_EE = sp.divide(sp.multiply(1e3, sp.divide(sp.sum(rate_sol_ue), term_pow_iter)), sp.log(2))
return iter_EE, theta_sol, iter_maximin_rate
def RK4(t_tot, N, Ham, psi):
dt = t_tot / N #N is number of steps
t = sp.linspace(0, t_tot, N + 1)
y = [psi]
for i in range(len(t) - 1):
k1 = dt * TDSE_T(Ham, t[i], y[len(y) - 1])
k2 = dt * TDSE_T(Ham, t[i] + 0.5 * dt, y[len(y) - 1] + k1 / 2)
k3 = dt * TDSE_T(Ham, t[i] + 0.5 * dt, y[len(y) - 1] + k2 / 2)
k4 = dt * TDSE_T(Ham, t[i] + dt, y[len(y) - 1] + k3)
ynew = y[len(y) - 1] + (k1 + 2 * k2 + 2 * k3 + k4) / 6
y.append(ynew)
chi = []
for i in range(len(y)):
inneri = sp.matmul(sp.conjugate(psi), y[i])
inner = abs(inneri)**2
chi.append(inner)
return t, chi
def _operator(self, op):
"""
This is the function that acts an operator on the full hierarchy.
PARAMETERS
----------
1. op : an operator
RETURNS
-------
None
"""
phi_mat = np.transpose(
np.reshape(self.phi,
[int(len(self.phi) / self.n_state), self.n_state],
order="C"))
self.storage.phi = np.reshape(np.transpose(sp.matmul(op, phi_mat)),
len(self.phi))
def fun(t, z):
#---- Esto sirve solo para reportar en que tiempo vamos
if t > fun.tnextreport:
print " {} at t = {}".format(fun.solver, fun.tnextreport)
fun.tnextreport += 1
if fun.solver != "Euler":
z = np.squeeze(z)
Famort = sp.zeros(40) #vector de fuerza friccional de amortiguamiento
Famort[0] = -(CAP[0] * (1. / M[0, 0]) * sp.tanh((z[20] / vr)))
for i in range(1, 20):
Famort[i] = -(1. / M[i, i]) * CAP[i] * sp.tanh(
(z[i + 20] - z[i - 1 + 20]) / vr)
Ft = sp.zeros(40)
if t < 226.81:
Ft[20:] = inte(t) * 9.8
return sp.matmul(A, z) + Famort + Ft
def write_spectrum(self,
row,
col,
states,
meas,
geom,
flush_immediately=False):
"""Write data from a single inversion to all output buffers."""
self.writes = self.writes + 1
if len(states) == 0:
# Write a bad data flag
atm_bad = s.zeros(len(self.fm.statevec)) * -9999.0
state_bad = s.zeros(len(self.fm.statevec)) * -9999.0
data_bad = s.zeros(self.fm.instrument.n_chan) * -9999.0
to_write = {
'estimated_state_file': state_bad,
'estimated_reflectance_file': data_bad,
'estimated_emission_file': data_bad,
'modeled_radiance_file': data_bad,
'apparent_reflectance_file': data_bad,
'path_radiance_file': data_bad,
'simulated_measurement_file': data_bad,
'algebraic_inverse_file': data_bad,
'atmospheric_coefficients_file': atm_bad,
'radiometry_correction_file': data_bad,
'spectral_calibration_file': data_bad,
'posterior_uncertainty_file': state_bad
}
else:
# The inversion returns a list of states, which are
# intepreted either as samples from the posterior (MCMC case)
# or as a gradient descent trajectory (standard case). For
# gradient descent the last spectrum is the converged solution.
if self.iv.method == 'MCMC':
state_est = states.mean(axis=0)
else:
state_est = states[-1, :]
# Spectral calibration
wl, fwhm = self.fm.calibration(state_est)
cal = s.column_stack(
[s.arange(0, len(wl)), wl / 1000.0, fwhm / 1000.0])
# If there is no actual measurement, we use the simulated version
# in subsequent calculations. Naturally in these cases we're
# mostly just interested in the simulation result.
if meas is None:
meas = self.fm.calc_rdn(state_est, geom)
# Rodgers diagnostics
lamb_est, meas_est, path_est, S_hat, K, G = \
self.iv.forward_uncertainty(state_est, meas, geom)
# Simulation with noise
meas_sim = self.fm.instrument.simulate_measurement(meas_est, geom)
# Algebraic inverse and atmospheric optical coefficients
x_surface, x_RT, x_instrument = self.fm.unpack(state_est)
rfl_alg_opt, Ls, coeffs = invert_algebraic(
self.fm.surface, self.fm.RT, self.fm.instrument, x_surface,
x_RT, x_instrument, meas, geom)
rhoatm, sphalb, transm, solar_irr, coszen, transup = coeffs
L_atm = self.fm.RT.get_L_atm(x_RT, geom)
L_down_transmitted = self.fm.RT.get_L_down_transmitted(x_RT, geom)
L_up = self.fm.RT.get_L_up(x_RT, geom)
atm = s.column_stack(
list(coeffs[:4]) + [s.ones((len(wl), 1)) * coszen])
# Upward emission & glint and apparent reflectance
Ls_est = self.fm.calc_Ls(state_est, geom)
apparent_rfl_est = lamb_est + Ls_est
# Radiometric calibration
factors = s.ones(len(wl))
if 'radiometry_correction_file' in self.outfiles:
if 'reference_reflectance_file' in self.infiles:
reference_file = self.infiles['reference_reflectance_file']
self.rfl_ref = reference_file.read_spectrum(row, col)
self.wl_ref = reference_file.wl
w, fw = self.fm.instrument.calibration(x_instrument)
resamp = resample_spectrum(self.rfl_ref,
self.wl_ref,
w,
fw,
fill=True)
meas_est = self.fm.calc_meas(state_est, geom, rfl=resamp)
factors = meas_est / meas
else:
logging.warning('No reflectance reference')
# Assemble all output products
to_write = {
'estimated_state_file':
state_est,
'estimated_reflectance_file':
s.column_stack((self.fm.surface.wl, lamb_est)),
'estimated_emission_file':
s.column_stack((self.fm.surface.wl, Ls_est)),
'estimated_reflectance_file':
s.column_stack((self.fm.surface.wl, lamb_est)),
'modeled_radiance_file':
s.column_stack((wl, meas_est)),
'apparent_reflectance_file':
s.column_stack((self.fm.surface.wl, apparent_rfl_est)),
'path_radiance_file':
s.column_stack((wl, path_est)),
'simulated_measurement_file':
s.column_stack((wl, meas_sim)),
'algebraic_inverse_file':
s.column_stack((self.fm.surface.wl, rfl_alg_opt)),
'atmospheric_coefficients_file':
atm,
'radiometry_correction_file':
factors,
'spectral_calibration_file':
cal,
'posterior_uncertainty_file':
s.sqrt(s.diag(S_hat))
}
for product in self.outfiles:
logging.debug('IO: Writing ' + product)
self.outfiles[product].write_spectrum(row, col, to_write[product])
if (self.writes % flush_rate) == 0 or flush_immediately:
self.outfiles[product].flush_buffers()
# Special case! samples file is matlab format.
if 'mcmc_samples_file' in self.output:
logging.debug('IO: Writing mcmc_samples_file')
mdict = {'samples': states}
s.io.savemat(self.output['mcmc_samples_file'], mdict)
# Special case! Data dump file is matlab format.
if 'data_dump_file' in self.output:
logging.debug('IO: Writing data_dump_file')
x = state_est
xall = states
Seps_inv, Seps_inv_sqrt = self.iv.calc_Seps(x, meas, geom)
meas_est_window = meas_est[self.iv.winidx]
meas_window = meas[self.iv.winidx]
xa, Sa, Sa_inv, Sa_inv_sqrt = self.iv.calc_prior(x, geom)
prior_resid = (x - xa).dot(Sa_inv_sqrt)
rdn_est = self.fm.calc_rdn(x, geom)
rdn_est_all = s.array(
[self.fm.calc_rdn(xtemp, geom) for xtemp in states])
x_surface, x_RT, x_instrument = self.fm.unpack(x)
Kb = self.fm.Kb(x, geom)
xinit = invert_simple(self.fm, meas, geom)
Sy = self.fm.instrument.Sy(meas, geom)
cost_jac_prior = s.diagflat(x - xa).dot(Sa_inv_sqrt)
cost_jac_meas = Seps_inv_sqrt.dot(K[self.iv.winidx, :])
meas_Cov = self.fm.Seps(x, meas, geom)
lamb_est, meas_est, path_est, S_hat, K, G = \
self.iv.forward_uncertainty(state_est, meas, geom)
A = s.matmul(K, G)
# Form the MATLAB dictionary object and write to file
mdict = {
'K': K,
'G': G,
'S_hat': S_hat,
'prior_mu': xa,
'Ls': Ls,
'prior_Cov': Sa,
'meas': meas,
'rdn_est': rdn_est,
'rdn_est_all': rdn_est_all,
'x': x,
'xall': xall,
'x_surface': x_surface,
'x_RT': x_RT,
'x_instrument': x_instrument,
'meas_Cov': meas_Cov,
'wl': wl,
'fwhm': fwhm,
'lamb_est': lamb_est,
'coszen': coszen,
'cost_jac_prior': cost_jac_prior,
'Kb': Kb,
'A': A,
'cost_jac_meas': cost_jac_meas,
'winidx': self.iv.winidx,
'windows': self.iv.windows,
'prior_resid': prior_resid,
'noise_Cov': Sy,
'xinit': xinit,
'rhoatm': rhoatm,
'sphalb': sphalb,
'transm': transm,
'transup': transup,
'solar_irr': solar_irr,
'L_atm': L_atm,
'L_down_transmitted': L_down_transmitted,
'L_up': L_up
}
s.io.savemat(self.output['data_dump_file'], mdict)
# Write plots, if needed
if len(states) > 0 and 'plot_directory' in self.output:
if 'reference_reflectance_file' in self.infiles:
reference_file = self.infiles['reference_reflectance_file']
self.rfl_ref = reference_file.read_spectrum(row, col)
self.wl_ref = reference_file.wl
for i, x in enumerate(states):
# Calculate intermediate solutions
lamb_est, meas_est, path_est, S_hat, K, G = \
self.iv.forward_uncertainty(state_est, meas, geom)
plt.cla()
red = [0.7, 0.2, 0.2]
wl, fwhm = self.fm.calibration(x)
xmin, xmax = min(wl), max(wl)
fig = plt.subplots(1, 2, figsize=(10, 5))
plt.subplot(1, 2, 1)
meas_est = self.fm.calc_meas(x, geom)
for lo, hi in self.iv.windows:
idx = s.where(s.logical_and(wl > lo, wl < hi))[0]
p1 = plt.plot(wl[idx], meas[idx], color=red, linewidth=2)
p2 = plt.plot(wl, meas_est, color='k', linewidth=1)
plt.title("Radiance")
plt.title("Measurement (Scaled DN)")
ymax = max(meas) * 1.25
ymax = max(meas) + 0.01
ymin = min(meas) - 0.01
plt.text(500, ymax * 0.92, "Measured", color=red)
plt.text(500, ymax * 0.86, "Model", color='k')
plt.ylabel(r"$\mu$W nm$^{-1}$ sr$^{-1}$ cm$^{-2}$")
plt.ylabel("Intensity")
plt.xlabel("Wavelength (nm)")
plt.ylim([-0.001, ymax])
plt.ylim([ymin, ymax])
plt.xlim([xmin, xmax])
plt.subplot(1, 2, 2)
lamb_est = self.fm.calc_lamb(x, geom)
ymax = min(max(lamb_est) * 1.25, 0.10)
for lo, hi in self.iv.windows:
# black line
idx = s.where(s.logical_and(wl > lo, wl < hi))[0]
p2 = plt.plot(wl[idx], lamb_est[idx], 'k', linewidth=2)
ymax = max(max(lamb_est[idx] * 1.2), ymax)
# red line
if 'reference_reflectance_file' in self.infiles:
idx = s.where(
s.logical_and(self.wl_ref > lo,
self.wl_ref < hi))[0]
p1 = plt.plot(self.wl_ref[idx],
self.rfl_ref[idx],
color=red,
linewidth=2)
ymax = max(max(self.rfl_ref[idx] * 1.2), ymax)
# green and blue lines - surface components
if hasattr(self.fm.surface, 'components') and \
self.output['plot_surface_components']:
idx = s.where(
s.logical_and(self.fm.surface.wl > lo,
self.fm.surface.wl < hi))[0]
p3 = plt.plot(self.fm.surface.wl[idx],
self.fm.xa(x, geom)[idx],
'b',
linewidth=2)
for j in range(len(self.fm.surface.components)):
z = self.fm.surface.norm(
lamb_est[self.fm.surface.idx_ref])
mu = self.fm.surface.components[j][0] * z
plt.plot(self.fm.surface.wl[idx],
mu[idx],
'g:',
linewidth=1)
plt.text(500, ymax * 0.86, "Remote estimate", color='k')
if 'reference_reflectance_file' in self.infiles:
plt.text(500, ymax * 0.92, "In situ reference", color=red)
if hasattr(self.fm.surface, 'components') and \
self.output['plot_surface_components']:
plt.text(500, ymax * 0.80, "Prior mean state ", color='b')
plt.text(500,
ymax * 0.74,
"Surface components ",
color='g')
plt.ylim([-0.0010, ymax])
plt.xlim([xmin, xmax])
plt.title("Reflectance")
plt.title("Source Model")
plt.xlabel("Wavelength (nm)")
fn = self.output['plot_directory'] + ('/frame_%i.png' % i)
plt.savefig(fn)
plt.close()
def Matrix_diag(H):
w1, v1 = sp.linalg.eig(H)
M_dag = sp.matrix.conjugate(sp.transpose(v1))
M = v1 ## conjugate(transpose(M))
H_diag = sp.matmul(sp.matmul(M_dag, H), M)
return H_diag, M, M_dag, w1
d2d_to_d2d_gains = max_d2d_to_d2d_gains[:num_d2d_pairs, :
num_d2d_pairs, Mon]
d2d_to_d2d_gains_diff = max_d2d_to_d2d_gains_diff[:num_d2d_pairs, :
num_d2d_pairs]
d2d_to_d2d_gains_diag = sp.subtract(d2d_to_d2d_gains,
d2d_to_d2d_gains_diff)
# ############################################################
# This code is used to find the initial point for EEmax algorithm
# ############################################################
theta_ini = Parameter(value=1 / (1 - 0.5))
phi_n_ini = sp.multiply((theta_ini.value - 1) * eta *
sp.divide(power_UAV, num_d2d_pairs),
uav_to_d2d_gains)
x_rate = sp.matmul(d2d_to_d2d_gains_diag, phi_n_ini)
term_rate = sp.matmul(sp.transpose(d2d_to_d2d_gains_diff),
phi_n_ini) + 1
rate_sol_ue = sp.divide(
sp.log(sp.add(1, sp.divide(x_rate, term_rate))),
theta_ini.value)
# print rate_sol_ue
rmin_ref = min(rate_sol_ue)
if rmin_ref <= 0.2 * sp.log(2):
rmin = rmin_ref
else:
rmin = 0.2 * sp.log(2)
pow_ = NonNegative(num_d2d_pairs)
objective = Minimize(sum_entries(pow_) / theta_ini)
def cavity(k, n_f, w_c):
return w_c * sp.kron(sp.identity(k),
(sp.matmul(alphadag(n_f), alpha(n_f))))
def TDSE_T(Hami, t, psi):
diff = -1j * sp.matmul(Hami(t), psi)
return diff