def test_function_overloading():
a = pe.pseudo_Obs(17, 2.9, 'e1')
b = pe.pseudo_Obs(4, 0.8, 'e1')
fs = [
lambda x: x[0] + x[1], lambda x: x[1] + x[0], lambda x: x[0] - x[1],
lambda x: x[1] - x[0], lambda x: x[0] * x[1], lambda x: x[1] * x[0],
lambda x: x[0] / x[1], lambda x: x[1] / x[0], lambda x: np.exp(x[0]),
lambda x: np.sin(x[0]), lambda x: np.cos(x[0]), lambda x: np.tan(x[0]),
lambda x: np.log(x[0]), lambda x: np.sqrt(np.abs(x[0])),
lambda x: np.sinh(x[0]), lambda x: np.cosh(x[0]),
lambda x: np.tanh(x[0])
]
for i, f in enumerate(fs):
t1 = f([a, b])
t2 = pe.derived_observable(f, [a, b])
c = t2 - t1
assert c.is_zero()
assert np.log(np.exp(b)) == b
assert np.exp(np.log(b)) == b
assert np.sqrt(b**2) == b
assert np.sqrt(b)**2 == b
np.arcsin(1 / b)
np.arccos(1 / b)
np.arctan(1 / b)
np.arctanh(1 / b)
np.sinc(1 / b)
def add_to_funcs(low, upp, i):
if (low is None) and (upp is None):
funcs.append(lambda x: pass_through(x))
inv_f.append(lambda x: pass_through(x))
elif (low == 0) and (upp == 1):
D = 10
funcs.append(lambda x: D * np.arctanh((2 * x) - 1))
inv_f.append(lambda x: (np.tanh(x / D) + 1) / 2)
elif (upp is None):
funcs.append(lambda x: (inv_adj_relu(x - low)))
inv_f.append(lambda x: (adj_relu(x) + low))
elif (low is None):
funcs.append(lambda x: inv_rev_adj_relu(x - np.copy(upp)))
inv_f.append(lambda x: np.copy(upp) + rev_adj_relu(x))
else:
funcs.append(lambda x: pass_through(x))
inv_f.append(lambda x: pass_through(x))
def test_arctanh():
fun = lambda x : 3.0 * np.arctanh(x)
d_fun = grad(fun)
check_grads(fun, 0.2)
check_grads(d_fun, 0.3)
def test_arctanh():
fun = lambda x : 3.0 * np.arctanh(x)
d_fun = grad(fun)
check_grads(fun, 0.2)
check_grads(d_fun, 0.3)
def test_arctanh():
fun = lambda x: 3.0 * np.arctanh(x)
check_grads(fun)(0.2)
def arctanh(a: Numeric):
return anp.arctanh(a)
def inverse(self, p):
return np.arctanh((np.clip(p, 0.4, 0.6)))
def test_arctanh():
fun = lambda x : 3.0 * np.arctanh(x)
check_grads(fun)(0.2)
if model_type is "scalar":
filename_prex = "figure/initial_scalar_" + state_name + "_"
savefilename = "data/initial_scalar_solution_" + state_name
opt_data = np.load(savefilename + "_full.npz", allow_pickle=True)
configurations = opt_data["configurations"]
controls = opt_data["controls_opt"]
simulation_first_confirmed = opt_data["simulation_first_confirmed"]
parameters = opt_data["parameters_opt"]
# y0, t_total, N_total, number_group, population_proportion, \
# t_control, number_days_per_control_change, number_control_change_times, number_time_dependent_controls = configurations
alpha = np.arctanh(2 * controls[0] - 1)
controls = (alpha, )
misfit = Misfit(configurations, parameters, controls,
simulation_first_confirmed)
# sigma = 10 * np.ones_like(x)
prior = Laplacian(misfit.dimension, gamma=10, mean=alpha)
# prior = Laplacian(misfit.dimension, gamma=10, regularization=False)
model = Model(prior, misfit)
if __name__ == "__main__":
print(misfit.t_total)
sum(temporalKL))
if __name__ == '__main__':
#with open('powerData.pkl') as f:
# X = pickle.load(f)
with open('pendulous.pkl') as f:
X = pickle.load(f)
inputDim = 121
seqLen = 200
numSeq = 1
step_size = 0.0001
fakeData = np.random.randn(seqLen, numSeq, inputDim)
print(fakeData.shape)
frame_to_vect = lambda frame: np.reshape(np.arctanh(2.0 * frame - 1.0), 121
)
vect_to_frame = lambda vect: np.reshape(0.5 * np.tanh(vect) + 0.5,
(11, 11))
dataDims = {'x': 80, 'u': 1, 'a': 20}
dataDims = {'x': 121, 'u': 0, 'a': 0}
X = map(frame_to_vect, X)
X = np.concatenate(map(lambda x: np.expand_dims(x, axis=0), X), axis=0)
inputs = {'x': np.expand_dims(X, axis=0)}
outputs = {'x': inputs['x'][:, 1:, :]}
inputs['x'] = inputs['x'][:, :-1, :]
#inputs = {'x': X}
#inputs = {'x':fakeData[:,:,:dataDims['x']], 'u':fakeData[:,:,dataDims['x']:dataDims['x']+dataDims['u']],
# 'a':fakeData[:,:,:dataDims['a']]}
hiddenDims = {
def inv_sigmoid(x):
return np.arctanh(2.0 * x - 1.0)
