def test_mcmc_sampler_zonotope(self):
""" Test whether 'zonotope' MCMC sampling mode generates samples with the right 1 and 2 points inclusion probabilities when DPP defined by orthogonal projection correlation kernel K = A.T (A A.T)^-1 A
"""
A = rndm.randn(self.rank, self.N)
dpp = FiniteDPP(kernel_type='correlation',
projection=True,
**{'A_zono': A})
dpp.sample_mcmc(mode='zonotope', **{'nb_iter': 1000})
self.assertTrue(self.singleton_adequation(dpp, dpp.list_of_samples[0]))
self.assertTrue(self.doubleton_adequation(dpp, dpp.list_of_samples[0]))
def test_mcmc_sampler_basis_exchange(self):
""" Test whether 'E' (basis_exchange) MCMC sampling mode generates samples with the right 1 and 2 points inclusion probabilities when DPP defined by orthogonal projection correlation kernel K from its eigendecomposition
"""
eig_vals = np.ones(self.rank)
eig_vecs, _ = qr(rndm.randn(self.N, self.rank), mode='economic')
dpp = FiniteDPP(kernel_type='correlation',
projection=True,
**{'K_eig_dec': (eig_vals, eig_vecs)})
dpp.sample_mcmc(mode='E', **{'size': self.rank, 'nb_iter': 1000})
self.assertTrue(self.singleton_adequation(dpp, dpp.list_of_samples[0]))
self.assertTrue(self.doubleton_adequation(dpp, dpp.list_of_samples[0]))
def test_correlation_kernel_projection_A_zono(self):
A = rndm.randn(self.rank, self.N)
dpp = FiniteDPP(kernel_type='correlation',
projection=True,
**{'A_zono': A})
for mode in ('GS', 'GS_bis', 'KuTa12'):
dpp.flush_samples()
for _ in range(self.nb_samples):
dpp.sample_exact(mode)
self.check_right_cardinality(dpp, dpp.list_of_samples)
for mode in ('zonotope', 'E'):
dpp.flush_samples()
dpp.sample_mcmc(mode,
**{'size': self.rank, 'nb_iter': self.nb_samples})
self.check_right_cardinality(dpp, dpp.list_of_samples[0])
def test_likelihood_kernel_L_gram_factor(self):
phi = rndm.randn(self.rank, self.N)
dpp = FiniteDPP(kernel_type='likelihood',
projection=False,
**{'L_gram_factor': phi})
for mode in ('GS', 'GS_bis', 'KuTa12'):
dpp.flush_samples()
for _ in range(self.nb_samples):
dpp.sample_exact(mode)
self.check_right_cardinality(dpp, dpp.list_of_samples)
for mode in ('AED', 'AD'):
dpp.flush_samples()
dpp.sample_mcmc(mode,
**{'nb_iter': self.nb_samples})
self.check_right_cardinality(dpp, dpp.list_of_samples[0])
def test_correlation_kernel_projection_kernel_eig(self):
eig_vals = np.ones(self.rank)
eig_vecs, _ = qr(rndm.randn(self.N, self.rank), mode='economic')
dpp = FiniteDPP(kernel_type='correlation',
projection=True,
**{'K_eig_dec': (eig_vals, eig_vecs)})
for mode in ('GS', 'GS_bis', 'KuTa12'):
dpp.flush_samples()
for _ in range(self.nb_samples):
dpp.sample_exact(mode)
self.check_right_cardinality(dpp, dpp.list_of_samples)
for mode in ('E'):
dpp.flush_samples()
dpp.sample_mcmc(mode,
**{'size': self.rank, 'nb_iter': self.nb_samples})
self.check_right_cardinality(dpp, dpp.list_of_samples[0])
def test_likelihood_kernel(self):
eig_vals = 1 + rndm.geometric(p=0.5, size=self.rank)
eig_vecs, _ = qr(rndm.randn(self.N, self.rank), mode='economic')
dpp = FiniteDPP(kernel_type='likelihood',
projection=False,
**{'L': (eig_vecs * eig_vals).dot(eig_vecs.T)})
for mode in ('GS', 'GS_bis', 'KuTa12'):
dpp.flush_samples()
for _ in range(self.nb_samples):
dpp.sample_exact(mode)
self.check_right_cardinality(dpp, dpp.list_of_samples)
for mode in ('AED', 'AD'):
dpp.flush_samples()
dpp.sample_mcmc(mode,
**{'nb_iter': self.nb_samples})
self.check_right_cardinality(dpp, dpp.list_of_samples[0])
verbose=False)
result.append({'n': n, 'alg': 'vfx', 'time': t_vfx.secs, 'k': len(S_vfx)})
with Timer(verbose=False) as t_vfx_resample:
S_vfx_resample = vfx_dpp_sampler.sample_exact('vfx',
rls_oversample_bless=5,
rls_oversample_dppvfx=10,
random_state=r,
desired_expected_size=desired_k,
verbose=False)
result.append({'n': n, 'alg': 'vfx_resample', 'time': t_vfx_resample.secs,'k': len(S_vfx_resample)})
with Timer(verbose=False) as t_mc:
exact_mcmc_sample = FiniteDPP(kernel_type='likelihood', L=dot_func(X_train) * vfx_dpp_sampler.intermediate_sample_info.alpha_star)
S_mcmc = exact_mcmc_sample.sample_mcmc(mode='AED',
s_init=r.permutation(I_train.shape[0])[:len(S_vfx_resample)].tolist(),
nb_iter=I_train.shape[0] * len(S_vfx_resample))
result.append({'n': n, 'alg': 'mcmc', 'time': t_mc.secs, 'k': len(S_mcmc)})
if X_train.shape[0] <= 16000:
with Timer(verbose=False) as t_exact:
L = dot_func(X_train) * vfx_dpp_sampler.intermediate_sample_info.alpha_star
E,U = np.linalg.eigh(L)
exact_dpp_sample = FiniteDPP(kernel_type='likelihood', L_eig_dec=(E, U))
S_exact = exact_dpp_sample.sample_exact()
result.append({'n': n, 'alg': 'exact', 'time': t_exact.secs, 'k': len(S_exact)})
with Timer(verbose=False) as t_exact_resample:
exact_dpp_sample = FiniteDPP(kernel_type='likelihood', L_eig_dec=(E, U))
S_exact_resample = exact_dpp_sample.sample_exact()
result.append({'n': n, 'alg': 'exact_resample', 'time': t_exact_resample.secs, 'k': len(S_exact_resample)})
