def atest_probabilistic_resampling(self):
self.set_positions_resampling_set()
setNpRandomSeed()
indices = [1,3,0,0,0,0]
solutions = self.resample(indices)
dautils.probabilisticResampling(self.resamplingDrifterSet, self.resamplingVar)
assert2DListAlmostEqual(self, self.resamplingDrifterSet.getDrifterPositions().tolist(), solutions, 2, "probabilistic resampling, probabilistic duplicates")
def hhhtest_default_constructor(self):
self.create_resampling_particle_set()
defaultParticleSet = self.resamplingParticleSet
self.assertEqual(defaultParticleSet.getNumParticles(),
self.resampleNumParticles)
self.assertEqual(defaultParticleSet.getObservationVariance(), 0.1)
positions = defaultParticleSet.observeParticles()
defaultPosition = [0.0, 0.0]
self.assertEqual(positions.shape, ((self.resampleNumParticles, 2)))
for i in range(self.resampleNumParticles):
self.assertEqual(positions[i, :].tolist(), defaultPosition)
observation = defaultParticleSet.observeTrueState()
self.assertEqual(observation.shape, ((2, )))
self.assertEqual(observation.tolist(), defaultPosition)
self.assertEqual(defaultParticleSet.getDomainSizeX(), 1.0)
self.assertEqual(defaultParticleSet.getDomainSizeY(), 1.0)
weight = 1.0 / self.resampleNumParticles
weights = [weight] * self.resampleNumParticles
self.assertEqual(
dautils.getGaussianWeight(
defaultParticleSet.getDistances(),
defaultParticleSet.getObservationVariance()).tolist(), weights)
self.assertEqual(
dautils.getCauchyWeight(
defaultParticleSet.getDistances(),
defaultParticleSet.getObservationVariance()).tolist(), weights)
# Check boundary condition
self.assertEqual(defaultParticleSet.getBoundaryConditions().get(),
[1, 1, 1, 1])
def atest_probabilistic_resampling(self):
self.set_positions_resampling_set()
setNpRandomSeed()
indices = [1,3,0,0,0,0]
solutions = self.resample(indices)
dautils.probabilisticResampling(self.resamplingDrifterSet, self.resamplingVar)
assert2DListAlmostEqual(self, self.resamplingDrifterSet.getDrifterPositions().tolist(), solutions, 2, "probabilistic resampling, probabilistic duplicates")
def test_residual_sampling(self):
self.set_positions_resampling_set()
setNpRandomSeed()
indices = [0,0,3,3,1,4]
solutions = self.resample(indices)
dautils.residualSampling(self.resamplingParticleSet, self.resamplingVar)
assert2DListAlmostEqual(self, self.resamplingParticleSet.observeParticles().tolist(), solutions, 2, "residual sampling, probabilistic duplicates")
def atest_stochastic_universal_sampling(self):
self.set_positions_resampling_set()
setNpRandomSeed()
indices = [0,0,1,3,3,4]
solutions = self.resample(indices)
dautils.stochasticUniversalSampling(self.resamplingDrifterSet, self.resamplingVar)
assert2DListAlmostEqual(self, self.resamplingDrifterSet.getDrifterPositions().tolist(), solutions, 2, "stochastic universal sampling, probabilistic duplicates")
def atest_stochastic_universal_sampling(self):
self.set_positions_resampling_set()
setNpRandomSeed()
indices = [0,0,1,3,3,4]
solutions = self.resample(indices)
dautils.stochasticUniversalSampling(self.resamplingDrifterSet, self.resamplingVar)
assert2DListAlmostEqual(self, self.resamplingDrifterSet.getDrifterPositions().tolist(), solutions, 2, "stochastic universal sampling, probabilistic duplicates")
def atest_monte_carlo_metropolis_hasting_sampling_with_duplicates(self):
self.set_positions_resampling_set()
setNpRandomSeed()
indices = [0,0,0,3,4,4]
solutions = self.resample(indices)
dautils.metropolisHastingSampling(self.resamplingDrifterSet)
self.assertEqual(self.resamplingDrifterSet.getDrifterPositions().tolist(), solutions)
def atest_monte_carlo_metropolis_hasting_sampling_with_duplicates(self):
self.set_positions_resampling_set()
setNpRandomSeed()
indices = [0,0,0,3,4,4]
solutions = self.resample(indices)
dautils.metropolisHastingSampling(self.resamplingDrifterSet)
self.assertEqual(self.resamplingDrifterSet.getDrifterPositions().tolist(), solutions)
def atest_stochastic_universal_sampling_with_duplicates(self):
self.set_positions_resampling_set()
setNpRandomSeed()
indices = [0,0,1,3,3,4]
solutions = self.resample(indices)
dautils.stochasticUniversalSampling(self.resamplingDrifterSet)
self.assertEqual(self.resamplingDrifterSet.getDrifterPositions().tolist(), \
solutions)
def atest_probabilistic_resampling_with_duplicates(self):
self.set_positions_resampling_set()
setNpRandomSeed()
indices = [1,3,0,0,0,0]
solutions = self.resample(indices)
dautils.probabilisticResampling(self.resamplingDrifterSet)
self.assertEqual(self.resamplingDrifterSet.getDrifterPositions().tolist(), \
solutions)
def test_residual_sampling_with_duplicates(self):
self.set_positions_resampling_set()
setNpRandomSeed()
indices = [0, 0, 3, 3, 1, 4]
solutions = self.resample(indices)
dautils.residualSampling(self.resamplingParticleSet)
self.assertEqual(self.resamplingParticleSet.observeParticles().tolist(), \
solutions)
def test_residual_sampling_with_duplicates(self):
self.set_positions_resampling_set()
setNpRandomSeed()
indices = [0,0,3,3,1,4]
solutions = self.resample(indices)
dautils.residualSampling(self.resamplingParticleSet)
self.assertEqual(self.resamplingParticleSet.observeParticles().tolist(), \
solutions)
def atest_monte_carlo_metropolis_hasting_sampling(self):
self.set_positions_resampling_set()
setNpRandomSeed()
indices = [0,0,0,3,4,4]
solutions = self.resample(indices)
dautils.metropolisHastingSampling(self.resamplingDrifterSet, self.resamplingVar)
#print self.resamplingDrifterSet.getDrifterPositions().tolist()
assert2DListAlmostEqual(self, self.resamplingDrifterSet.getDrifterPositions().tolist(), solutions, 2, "metropolis hasting sampling, probabilistic duplicates")
def atest_stochastic_universal_sampling_with_duplicates(self):
self.set_positions_resampling_set()
setNpRandomSeed()
indices = [0, 0, 1, 3, 3, 4]
solutions = self.resample(indices)
dautils.stochasticUniversalSampling(self.resamplingDrifterSet)
self.assertEqual(self.resamplingDrifterSet.getDrifterPositions().tolist(), \
solutions)
def atest_probabilistic_resampling_with_duplicates(self):
self.set_positions_resampling_set()
setNpRandomSeed()
indices = [1, 3, 0, 0, 0, 0]
solutions = self.resample(indices)
dautils.probabilisticResampling(self.resamplingDrifterSet)
self.assertEqual(self.resamplingDrifterSet.getDrifterPositions().tolist(), \
solutions)
def atest_monte_carlo_metropolis_hasting_sampling(self):
self.set_positions_resampling_set()
setNpRandomSeed()
indices = [0,0,0,3,4,4]
solutions = self.resample(indices)
dautils.metropolisHastingSampling(self.resamplingDrifterSet, self.resamplingVar)
#print self.resamplingDrifterSet.getDrifterPositions().tolist()
assert2DListAlmostEqual(self, self.resamplingDrifterSet.getDrifterPositions().tolist(), solutions, 2, "metropolis hasting sampling, probabilistic duplicates")
def test_residual_sampling(self):
self.set_positions_resampling_set()
setNpRandomSeed()
indices = [0, 0, 3, 3, 1, 4]
solutions = self.resample(indices)
dautils.residualSampling(self.resamplingParticleSet,
self.resamplingVar)
assert2DListAlmostEqual(
self,
self.resamplingParticleSet.observeParticles().tolist(), solutions,
2, "residual sampling, probabilistic duplicates")
def test_gaussian_weights(self):
self.set_positions_resampling_set()
obtainedWeights = dautils.getGaussianWeight(
self.resamplingParticleSet.getDistances(),
self.resamplingParticleSet.getObservationVariance())
referenceWeights = [
3.77361928e-01, 1.22511481e-01, 1.26590824e-04, 3.77361928e-01,
1.22511481e-01, 1.26590824e-04
]
assertListAlmostEqual(self, obtainedWeights.tolist(), referenceWeights,
6, 'gaussian weights')
def test_cauchy_weights(self):
self.set_positions_resampling_set()
obtainedWeights = dautils.getCauchyWeight(
self.resamplingParticleSet.getDistances(),
self.resamplingParticleSet.getObservationVariance())
referenceWeights = [
0.28413284, 0.16789668, 0.04797048, 0.28413284, 0.16789668,
0.04797048
]
assertListAlmostEqual(self, obtainedWeights.tolist(), referenceWeights,
6, 'cauchy weights')
