def test_brentsroot():
for fmt in D.available_float_fmt():
print("Set dtype to:", fmt)
D.set_float_fmt(fmt)
for _ in range(10):
ac_prod = D.array(np.random.uniform(0.9, 1.1))
a = D.array(np.random.uniform(-1, 1))
a = D.to_float(-1 * (a <= 0) + 1 * (a > 0))
c = ac_prod / a
b = D.sqrt(0.01 + 4 * ac_prod)
gt_root = -b / (2 * a) - 0.1 / (2 * a)
ub = -b / (2 * a)
lb = -b / (2 * a) - 1.0 / (2 * a)
fun = lambda x: a * x**2 + b * x + c
assert (D.to_numpy(D.to_float(D.abs(fun(gt_root)))) <=
32 * D.epsilon())
root, success = de.utilities.optimizer.brentsroot(fun, [lb, ub],
4 * D.epsilon(),
verbose=True)
assert (success)
assert (np.allclose(D.to_numpy(D.to_float(gt_root)),
D.to_numpy(D.to_float(root)), 32 * D.epsilon(),
32 * D.epsilon()))
assert (D.to_numpy(D.to_float(D.abs(fun(root)))) <=
32 * D.epsilon())
def test_brentsroot_wrong_order():
if D.backend() == 'torch':
import torch
torch.set_printoptions(precision=17)
torch.autograd.set_detect_anomaly(True)
a = D.array(1.0)
b = D.array(1.0)
gt_root = -b / a
lb, ub = -b / a - 1, -b / a + 1
fun = lambda x: a * x + b
assert (D.to_numpy(D.to_float(D.abs(fun(gt_root)))) <= 32 * D.epsilon())
root, success = de.utilities.optimizer.brentsroot(fun, [ub, lb],
4 * D.epsilon(),
verbose=True)
assert (success)
assert (np.allclose(D.to_numpy(D.to_float(gt_root)),
D.to_numpy(D.to_float(root)), 32 * D.epsilon(),
32 * D.epsilon()))
def test_addcmul_within_tolerance_out(ffmt):
D.set_float_fmt(ffmt)
pi = D.to_float(D.pi)
out = D.copy(pi)
D.addcmul(pi, D.to_float(3), D.to_float(2), value=1, out=out)
assert (pi + (1 * (3 * 2)) - 2 * D.epsilon() <= out <= pi +
(1 * (3 * 2)) + 2 * D.epsilon())
def analytic_soln(t, initial_conditions):
c1 = initial_conditions[0]
c2 = initial_conditions[1] - 1
return D.stack([
c2 * D.sin(D.to_float(D.asarray(t))) + c1 * D.cos(D.to_float(D.asarray(t))) + D.asarray(t),
c2 * D.cos(D.to_float(D.asarray(t))) - c1 * D.sin(D.to_float(D.asarray(t))) + 1
])
def test_jacobian_wrapper_calls_estimate(ffmt):
D.set_float_fmt(ffmt)
rhs = lambda x: D.exp(-x)
jac_rhs = de.utilities.JacobianWrapper(rhs, richardson_iter=0, adaptive=False, rtol=D.epsilon() ** 0.5, atol=D.epsilon() ** 0.5)
x = D.array(0.0)
assert (D.allclose(D.to_float(jac_rhs.estimate(x)), D.to_float(jac_rhs(x)), rtol=4 * D.epsilon() ** 0.5, atol=4 * D.epsilon() ** 0.5))
def test_jacobian_wrapper_exact(ffmt):
D.set_float_fmt(ffmt)
rhs = lambda x: D.exp(-x)
drhs_exact = lambda x: -D.exp(-x)
jac_rhs = de.utilities.JacobianWrapper(rhs, rtol=D.epsilon() ** 0.5, atol=D.epsilon() ** 0.5)
x = D.array(0.0)
assert (D.allclose(D.to_float(drhs_exact(x)), D.to_float(jac_rhs(x)), rtol=4 * D.epsilon() ** 0.5, atol=4 * D.epsilon() ** 0.5))
def test_newtonraphson_dims(ffmt, tol, dim):
print("Set dtype to:", ffmt)
D.set_float_fmt(ffmt)
np.random.seed(30)
if tol is not None:
tol = tol * D.epsilon()
if D.backend() == 'torch':
import torch
torch.set_printoptions(precision=17)
torch.autograd.set_detect_anomaly(False)
if ffmt == 'gdual_vdouble':
pytest.skip("Root-finding is ill-conceived with vectorised gduals")
shift = D.array(np.random.uniform(1, 10, size=(dim, )))
exponent = D.array(np.random.uniform(1, 5, size=(dim, )))
gt_root1 = shift**(1 / exponent)
gt_root2 = -shift**(1 / exponent)
def fun(x):
return x**exponent - shift
def jac(x):
return D.diag(exponent * D.reshape(x, (-1, ))**(exponent - 1))
x0 = D.array(np.random.uniform(1, 3, size=(dim, )))
print(gt_root1, gt_root2)
print(x0)
print(fun(x0))
print(jac(x0))
root, (success, num_iter,
prec) = de.utilities.optimizer.newtonraphson(fun,
x0,
jac=jac,
tol=tol,
verbose=True)
if tol is None:
tol = D.epsilon()
assert (success)
conv_root1 = D.stack([
D.array(np.allclose(D.to_numpy(D.to_float(r1)),
D.to_numpy(D.to_float(r)), 128 * tol, 32 * tol),
dtype=D.bool) for r, r1 in zip(root, gt_root1)
])
conv_root2 = D.stack([
D.array(np.allclose(D.to_numpy(D.to_float(r2)),
D.to_numpy(D.to_float(r)), 128 * tol, 32 * tol),
dtype=D.bool) for r, r2 in zip(root, gt_root2)
])
assert (D.all(conv_root1 | conv_root2))
def test_matrix_inv_bigger():
for diag_size in range(2, 101):
np.random.seed(15)
for trial in range(3):
A = np.random.normal(size=(diag_size, diag_size))
while np.abs(np.linalg.det(D.to_float(A))) <= 1e-5:
A = np.random.normal(size=(diag_size, diag_size), std=250.0)
A = D.array(D.cast_to_float_fmt(A))
Ainv = D.matrix_inv(A)
assert (
D.max(D.abs(D.to_float(Ainv @ A - D.eye(diag_size)))) <=
4 * D.epsilon()**0.5
), "Matrix inversion failed for diagonal with size: " + str(
diag_size)
def test_newtonraphson_pytorch_jacobian(ffmt, tol):
print("Set dtype to:", ffmt)
D.set_float_fmt(ffmt)
np.random.seed(21)
if tol is not None:
tol = tol * D.epsilon()
if D.backend() == 'torch':
import torch
torch.set_printoptions(precision=17)
torch.autograd.set_detect_anomaly(False)
if ffmt == 'gdual_vdouble':
pytest.skip("Root-finding is ill-conceived with vectorised gduals")
for _ in range(10):
ac_prod = D.array(np.random.uniform(0.9, 1.1))
a = D.array(np.random.uniform(-1, 1))
a = D.to_float(-1 * (a <= 0) + 1 * (a > 0))
c = ac_prod / a
b = D.sqrt(0.01 + 4 * ac_prod)
gt_root1 = -b / (2 * a) - 0.1 / (2 * a)
gt_root2 = -b / (2 * a) + 0.1 / (2 * a)
ub = -b / (2 * a) - 0.2 / (2 * a)
lb = -b / (2 * a) - 0.4 / (2 * a)
x0 = D.array(np.random.uniform(ub, lb))
fun = lambda x: a * x**2 + b * x + c
assert (D.to_numpy(D.to_float(D.abs(fun(gt_root1)))) <=
32 * D.epsilon())
assert (D.to_numpy(D.to_float(D.abs(fun(gt_root2)))) <=
32 * D.epsilon())
root, (success, num_iter,
prec) = de.utilities.optimizer.newtonraphson(fun,
x0,
tol=tol,
verbose=True)
if tol is None:
tol = D.epsilon()
conv_root1 = np.allclose(D.to_numpy(D.to_float(gt_root1)),
D.to_numpy(D.to_float(root)), 128 * tol,
32 * tol)
conv_root2 = np.allclose(D.to_numpy(D.to_float(gt_root2)),
D.to_numpy(D.to_float(root)), 128 * tol,
32 * tol)
print(conv_root1, conv_root2, root, gt_root1, gt_root2, x0,
root - gt_root1, root - gt_root2, num_iter, prec)
assert (success)
assert (conv_root1 or conv_root2)
assert (D.to_numpy(D.to_float(D.abs(fun(root)))) <= 32 * tol)
def test_softplus_within_tolerance_out(ffmt):
D.set_float_fmt(ffmt)
pi = D.to_float(D.pi)
out = D.copy(pi)
D.softplus(pi, out=out)
assert (3.18389890758499587775 - 2 * D.epsilon() <= out <=
3.18389890758499587775 + 2 * D.epsilon())
def test_square_within_tolerance_out(ffmt):
D.set_float_fmt(ffmt)
pi = D.to_float(D.pi)
out = D.copy(pi)
D.square(pi, out=out)
assert (9.8696044010893586188 - 2 * D.epsilon() <= out <=
9.8696044010893586188 + 2 * D.epsilon())
def test_matrix_inv():
A = D.array([
[-1.0, 3 / 2],
[1.0, -1.0],
], dtype=D.float64)
Ainv = D.matrix_inv(A)
assert (D.max(D.abs(D.to_float(Ainv @ A - D.eye(2)))) <= 8 * D.epsilon())
def test_brentsroot_same_sign():
if D.backend() == 'torch':
import torch
torch.set_printoptions(precision=17)
torch.autograd.set_detect_anomaly(True)
ac_prod = D.array(np.random.uniform(0.9, 1.1))
a = D.array(1.0)
b = D.array(1.0)
gt_root = -b / a
lb, ub = -b / a - 1, -b / a - 2
fun = lambda x: a * x + b
assert (D.to_numpy(D.to_float(D.abs(fun(gt_root)))) <= 32 * D.epsilon())
root, success = de.utilities.optimizer.brentsroot(fun, [lb, ub],
4 * D.epsilon(),
verbose=True)
assert (np.isinf(root))
assert (not success)
def test_frac_within_tolerance_out(ffmt):
D.set_float_fmt(ffmt)
pi = D.to_float(D.pi)
out = D.copy(pi)
D.frac(pi, out=out)
assert (0.141592653589793238 - 2 * D.epsilon() <= out <=
0.141592653589793238 + 2 * D.epsilon())
def test_brentsrootvec(ffmt, tol):
print("Set dtype to:", ffmt)
D.set_float_fmt(ffmt)
if tol is not None:
tol = tol * D.epsilon()
if D.backend() == 'torch':
import torch
torch.set_printoptions(precision=17)
torch.autograd.set_detect_anomaly(True)
if ffmt == 'gdual_vdouble':
pytest.skip("Root-finding is ill-conceived with vectorised gduals")
for _ in range(10):
slope_list = D.array(
np.copysign(np.random.uniform(0.9, 1.1, size=25),
np.random.uniform(-1, 1, size=25)))
intercept_list = slope_list
gt_root_list = -intercept_list / slope_list
fun_list = [(lambda m, b: lambda x: m * x + b)(m, b)
for m, b in zip(slope_list, intercept_list)]
assert (all(
map((lambda i: D.to_numpy(D.to_float(D.abs(i))) <= 32 * D.epsilon(
)), map((lambda x: x[0](x[1])), zip(fun_list, gt_root_list)))))
root_list, success = de.utilities.optimizer.brentsrootvec(
fun_list, [D.min(gt_root_list) - 1.,
D.max(gt_root_list) + 1.],
tol,
verbose=True)
assert (np.all(D.to_numpy(success)))
assert (np.allclose(D.to_numpy(D.to_float(gt_root_list)),
D.to_numpy(D.to_float(root_list)),
32 * D.epsilon(), 32 * D.epsilon()))
assert (all(
map((lambda i: D.to_numpy(D.to_float(D.abs(i))) <= 32 * D.epsilon(
)), map((lambda x: x[0](x[1])), zip(fun_list, root_list)))))
def test_brentsroot(ffmt, tol):
print("Set dtype to:", ffmt)
D.set_float_fmt(ffmt)
if tol is not None:
tol = tol * D.epsilon()
if D.backend() == 'torch':
import torch
torch.set_printoptions(precision=17)
torch.autograd.set_detect_anomaly(True)
for _ in range(10):
ac_prod = D.array(np.random.uniform(0.9, 1.1))
a = D.array(np.random.uniform(-1, 1))
a = D.to_float(-1 * (a <= 0) + 1 * (a > 0))
c = ac_prod / a
b = D.sqrt(0.01 + 4 * ac_prod)
gt_root = -b / (2 * a) - 0.1 / (2 * a)
ub = -b / (2 * a)
lb = -b / (2 * a) - 1.0 / (2 * a)
fun = lambda x: a * x**2 + b * x + c
assert (D.to_numpy(D.to_float(D.abs(fun(gt_root)))) <=
32 * D.epsilon())
root, success = de.utilities.optimizer.brentsroot(fun, [lb, ub],
tol,
verbose=True)
assert (success)
assert (np.allclose(D.to_numpy(D.to_float(gt_root)),
D.to_numpy(D.to_float(root)), 32 * D.epsilon(),
32 * D.epsilon()))
assert (D.to_numpy(D.to_float(D.abs(fun(root)))) <= 32 * D.epsilon())
def test_brentsrootvec():
for fmt in D.available_float_fmt():
print("Set dtype to:", fmt)
D.set_float_fmt(fmt)
if fmt == 'gdual_vdouble':
continue
for _ in range(10):
slope_list = D.array(
np.copysign(np.random.uniform(0.9, 1.1, size=25),
np.random.uniform(-1, 1, size=25)))
intercept_list = slope_list
gt_root_list = -intercept_list / slope_list
fun_list = [(lambda m, b: lambda x: m * x + b)(m, b)
for m, b in zip(slope_list, intercept_list)]
assert (all(
map((lambda i: D.to_numpy(D.to_float(D.abs(i))) <= 32 * D.
epsilon()),
map((lambda x: x[0](x[1])), zip(fun_list, gt_root_list)))))
root_list, success = de.utilities.optimizer.brentsrootvec(
fun_list, [D.min(gt_root_list) - 1.,
D.max(gt_root_list) + 1.],
4 * D.epsilon(),
verbose=True)
assert (np.all(D.to_numpy(success)))
assert (np.allclose(D.to_numpy(D.to_float(gt_root_list)),
D.to_numpy(D.to_float(root_list)),
32 * D.epsilon(), 32 * D.epsilon()))
assert (all(
map((lambda i: D.to_numpy(D.to_float(D.abs(i))) <= 32 * D.
epsilon()),
map((lambda x: x[0](x[1])), zip(fun_list, root_list)))))
def test_floor_within_tolerance(ffmt):
D.set_float_fmt(ffmt)
pi = D.to_float(D.pi)
assert (D.floor(pi) == 3)
def test_digamma_within_tolerance(ffmt):
D.set_float_fmt(ffmt)
pi = D.to_float(D.pi)
assert (0.9772133079420067 - 2 * D.epsilon() <= D.digamma(pi) <=
0.9772133079420067 + 2 * D.epsilon())
def test_fmod_within_tolerance(ffmt):
D.set_float_fmt(ffmt)
pi = D.to_float(D.pi)
assert (1.1415926535897931 - 2 * D.epsilon() <= D.fmod(pi, 2) <=
1.1415926535897931 + 2 * D.epsilon())
def test_round_within_tolerance(ffmt):
D.set_float_fmt(ffmt)
pi = D.to_float(D.pi)
assert (D.round(pi) == 3)
def test_pi_within_tolerance(ffmt):
D.set_float_fmt(ffmt)
assert (np.pi - 2 * D.epsilon() <= D.to_float(D.pi) <=
np.pi + 2 * D.epsilon())
def test_clip_within_tolerance(ffmt):
D.set_float_fmt(ffmt)
pi = D.to_float(D.pi)
assert (D.clip(pi, 1, 2) == 2)
def test_square_within_tolerance(ffmt):
D.set_float_fmt(ffmt)
pi = D.to_float(D.pi)
assert (9.8696044010893586188 - 2 * D.epsilon() <= D.square(pi) <=
9.8696044010893586188 + 2 * D.epsilon())
def test_sign_within_tolerance(ffmt):
D.set_float_fmt(ffmt)
pi = D.to_float(D.pi)
assert (D.sign(pi) == 1)
def test_trunc_within_tolerance(ffmt):
D.set_float_fmt(ffmt)
pi = D.to_float(D.pi)
assert (D.trunc(pi) == 3)
def test_softplus_within_tolerance(ffmt):
D.set_float_fmt(ffmt)
pi = D.to_float(D.pi)
assert (3.18389890758499587775 - 2 * D.epsilon() <= D.softplus(pi) <=
3.18389890758499587775 + 2 * D.epsilon())
def test_erfinv_within_tolerance(ffmt):
D.set_float_fmt(ffmt)
pi = D.to_float(D.pi)
assert (0.4769362762044699 - 2 * D.epsilon() <= D.erfinv(D.to_float(0.5))
<= 0.4769362762044699 + 2 * D.epsilon())
def test_frac_within_tolerance(ffmt):
D.set_float_fmt(ffmt)
pi = D.to_float(D.pi)
assert (0.141592653589793238 - 2 * D.epsilon() <= D.frac(pi) <=
0.141592653589793238 + 2 * D.epsilon())
def test_mvlgamma_within_tolerance(ffmt):
D.set_float_fmt(ffmt)
pi = D.to_float(D.pi)
assert (1.7891115385869942 - 2 * D.epsilon() <= D.mvlgamma(pi, 2) <=
1.7891115385869942 + 2 * D.epsilon())
