def test_continuous_discr():
""" Ensure correct discretization in continuous state spaces """
# NOTE - if possible, test a domain with mixed discr/continuous
domain = inf_cp.InfTrackCartPole() # 2 continuous dims
rep = IndependentDiscretization(domain, discretization=20)
assert rep.features_num == 40
rep = IndependentDiscretization(domain, discretization=50)
assert rep.features_num == 100
def test_Fourier_order():
""" Ensure rep of appropriate order is created """
domain = inf_cp.InfTrackCartPole() # 2 continuous dims
order = 3
rep = Fourier(domain, order=order)
assert rep.features_num == order**domain.state_space_dims # 9
phiVec = rep.phi(np.array([0, 0]), terminal=False)
assert len(phiVec) == order**domain.state_space_dims # 9
# all phi vals should be -1 <= <= 1, since cosine
assert np.all(-1 <= phiVec) and np.all(phiVec <= 1)
def test_parametric_rep():
"""
For fixed representations: test successful kernel function use, using
varying number of features.
Ensure get expected result. Test normalization, ensure expected result.
"""
for normalization in [False, True]: # verify everything with/out norm
kernel = gaussian_kernel
domain = inf_cp.InfTrackCartPole() # 2 continuous dims
discretization = 20 # not used
num = 1 # number of basis functions to use IN EACH DIMENSION
resolution_min = 1
resolution_max = 5
rep = RandomLocalBases(
domain,
kernel,
num,
resolution_min,
resolution_max,
seed=1,
normalization=normalization,
discretization=discretization,
)
assert rep.features_num == num # in reality, theres one in each dim.
# Center lies within statespace limits
assert np.all(domain.statespace_limits[:, 0] <= rep.centers[0])
assert np.all(rep.centers[0] <= domain.statespace_limits[:, 1])
# width lies within resolution bounds
statespace_range = (domain.statespace_limits[:, 1] -
domain.statespace_limits[:, 0])
# widths[0] has `state_space_dims` cols
assert np.all(statespace_range / resolution_max <= rep.widths[0])
assert np.all(rep.widths[0] <= statespace_range / resolution_min)
phiVecOrigin = rep.phi(np.array([0, 0], dtype=np.float),
terminal=False)
assert np.all(phiVecOrigin >= 0) # nonnegative feat func values
# feature func only dependent on own dimension
phiVec2 = rep.phi(np.array([0, 1], dtype=np.float), terminal=False)
if normalization:
assert sum(phiVecOrigin) == 1
assert sum(phiVec2) == 1
def test_visual():
""" Test 2-D basis func visualization. """
kernel = gaussian_kernel
normalization = False
domain = inf_cp.InfTrackCartPole() # 2 continuous dims
discretization = 20 # not used
num = 1 # number of basis functions to use
resolution_min = 1
resolution_max = 5
rep = RandomLocalBases(
domain,
kernel,
num,
resolution_min,
resolution_max,
seed=1,
normalization=normalization,
discretization=discretization,
)
rep.plot_2d_feature_centers()
def test_nonparametric_rep():
"""
For nonparametric representations: test successful kernel function use,
ensure get expected result.
"""
for normalization in [False, True]: # verify everything with/out norm
kernel = gaussian_kernel
normalization = False
domain = inf_cp.InfTrackCartPole() # 2 continuous dims
discretization = 20 # not used
resolution = 2
# Start by making it impossible to add feats:
max_similarity = 0
rep = NonparametricLocalBases(
domain,
kernel,
max_similarity,
resolution,
normalization=normalization,
discretization=discretization,
)
assert rep.features_num == 0 # ``num`` feats in each dimension
origS = np.array([0, 0], dtype=np.float)
s2 = np.array([0, 1], dtype=np.float)
terminal = False # nonterminal states
a = 1 # arbitrary
rep.pre_discover(origS, terminal, a, s2, terminal)
# in the first call, automaticlaly add 1 feature since empty phi_s
# is always < rep.max_similarity.
# In ALL OTHER cases though, since max_similarity = 0, can never add
# any more.
assert rep.features_num == 1
# Now make it really easy to add feats:
max_similarity = np.inf
rep = NonparametricLocalBases(
domain,
kernel,
max_similarity,
resolution,
normalization=normalization,
discretization=discretization,
)
assert rep.features_num == 0 # ``num`` feats in each dimension
origS = np.array([0, 0], dtype=np.float)
s2 = np.array([0, 1], dtype=np.float)
terminal = False # nonterminal states
a = 1 # arbitrary
rep.pre_discover(origS, terminal, a, s2, terminal)
# max_similarity == inf means we definitely shouldve added feat for
# BOTH s and ns:
assert rep.features_num == 2
assert np.all(rep.centers[0, :] == origS)
assert np.all(rep.centers[1, :] == s2)
statespace_range = (domain.statespace_limits[:, 1] -
domain.statespace_limits[:, 0])
assert np.all(rep.widths == statespace_range / resolution)
phiVecOrigin = rep.phi(np.array([0, 0], dtype=np.float),
terminal=False)
assert np.all(phiVecOrigin >= 0) # nonnegative feat func values
# feature func only dependent on own dimension
phiVec2 = rep.phi(np.array([0, 1], dtype=np.float), terminal=False)
np.all(phiVec2 >= 0)
if normalization:
assert sum(phiVecOrigin) == 1
assert sum(phiVec2) == 1