def evalerr_l(prm):
dimN=1
lmbda_l= prm[0]
b_l = comcbpdn.ComplexConvBPDN(D1_d, s_test, lmbda_l, opt=opt_par, dimK=None, dimN=dimN)
x_0_l = b_l.solve()
x_1_l = x_0_l.squeeze()
return (np.sum(np.abs(x_1_l)))
def evalerr(prm):
dimN=1
lmbda = prm[0]
b = comcbpdn.ComplexConvBPDN(D0_d, s_test, lmbda, opt=opt_par, dimK=None, dimN=dimN)
x_0 = b.solve()
x_1 = x_0.squeeze()
return (np.sum(np.abs(x_1)))
def test_05(self):
N = 16
Nd = 5
K = 2
M = 4
D = np.random.randn(Nd, Nd, M)
s = np.random.randn(N, N, K)
lmbda = 1e-1
b = comcbpdn.ComplexConvBPDN(D, s, lmbda)
assert b.cri.dimC == 0
assert b.cri.dimK == 1
def test_03(self):
N = 16
Nd = 5
Cd = 3
M = 4
D = np.random.randn(Nd, Nd, Cd, M)
s = np.random.randn(N, N, Cd)
lmbda = 1e-1
b = comcbpdn.ComplexConvBPDN(D, s, lmbda)
assert b.cri.dimC == 1
assert b.cri.dimK == 0
def test_02(self):
N = 16
Nd = 5
Cs = 3
K = 5
M = 4
D = np.random.randn(Nd, Nd, M)
s = np.random.randn(N, N, Cs, K)
lmbda = 1e-1
b = comcbpdn.ComplexConvBPDN(D, s, lmbda)
assert b.cri.dimC == 1
assert b.cri.dimK == 1
def test_06(self):
N = 16
Nd = 5
K = 2
M = 4
D = np.random.randn(Nd, Nd, M)
s = np.random.randn(N, N, K)
dt = np.complex
opt = comcbpdn.ComplexConvBPDN.Options({
'Verbose': False,
'MaxMainIter': 20,
'AutoRho': {
'Enabled': True
},
'DataType': dt
})
lmbda = 1e-1
b = comcbpdn.ComplexConvBPDN(D, s, lmbda, opt=opt)
b.solve()
assert b.X.dtype == dt
assert b.Y.dtype == dt
assert b.U.dtype == dt
optd = comccmod.ComConvCnstrMODOptions({'Verbose': False, 'MaxMainIter': 1,
'rho': 10, 'ZeroMean': True},
method='cns')
#
#Dictionary support projection and normalisation (cropped).
#Normalise dictionary according to dictionary Y update options.
D0n = cnvrep.Pcn(D0, D0.shape, cri.Nv, dimN=1, dimC=0, crp=True,
zm=optd['ZeroMean'])
#
# Update D update options to include initial values for Y and U.
optd.update({'Y0': cnvrep.zpad(cnvrep.stdformD(D0n, cri.Cd, cri.M), cri.Nv),
'U0': np.zeros(cri.shpD + (cri.K,))})
# # #
#Create X update object.
xstep = comcbpdn.ComplexConvBPDN(D0n, s_noise, lmbda, optx)
# # the first one is coefficient map
#Create D update object.
dstep = comccmod.ComConvCnstrMOD(None, s_noise, D0.shape, optd, method='cns')
#
opt = dictlrn.DictLearn.Options({'Verbose': True, 'MaxMainIter': 1000})
d = dictlrn.DictLearn(xstep, dstep, opt)
D1 = d.solve()
x = d.getcoef()
rec = reconstruct(D1,x,cri)
rec = rec.squeeze()
print("DictLearn solve time: %.2fs" % d.timer.elapsed('solve'), "\n")
itsx = xstep.getitstat()
fig, ax = plot.subplots(nrows=2, ncols=3, sharex=True, sharey=True, figsize=(18, 8))
rec_r = b_r.reconstruct().squeeze()
fig.suptitle('real value solver ' + str(np.count_nonzero(x_r) * 100 / x_r.size) + '% non zero elements', fontsize=14)
plot.plot(s_clean.real, title='clean(real)', fig=fig, ax=ax[0, 0])
plot.plot(s_clean.imag, title='clean(imag)', fig=fig, ax=ax[1, 0])
plot.plot(s_noise.real, title='corrupted(real)', fig=fig, ax=ax[0, 1])
plot.plot(s_noise.imag, title='corrupted(imag)', fig=fig, ax=ax[1, 1])
plot.plot(rec_r.real, title='reconstructed(real)', fig=fig, ax=ax[0, 2])
plot.plot(rec_r.imag, title='reconstructed(imag)', fig=fig, ax=ax[1, 2])
fig.show()
# solve with a complex valued signal with complex valued dictionary
b_c = comcbpdn.ComplexConvBPDN(D0, s_noise, lmbda_0, opt_par, None, dimN)
x_c = b_c.solve()
rec_c = b_c.reconstruct().squeeze()
fig, ax = plot.subplots(nrows=2, ncols=3, sharex=True, sharey=True, figsize=(18, 8))
fig.suptitle('complex value solver ' + str(np.count_nonzero(x_c) * 100 / x_c.size) + '% non zero elements', fontsize=14)
plot.plot(s_clean.real, title='clean(real)', fig=fig, ax=ax[0, 0])
plot.plot(s_clean.imag, title='clean(imag)', fig=fig, ax=ax[1, 0])
plot.plot(s_noise.real, title='corrupted(real)', fig=fig, ax=ax[0, 1])
plot.plot(s_noise.imag, title='corrupted(imag)', fig=fig, ax=ax[1, 1])
plot.plot(rec_c.real, title='reconstructed(real)', fig=fig, ax=ax[0, 2])
plot.plot(rec_c.imag, title='reconstructed(imag)', fig=fig, ax=ax[1, 2])
fig.show()