def reference_func():
func.__globals__['np'] = ag_np
if preserve_result:
val, gradval = ag_value_and_grad(func)(*deepcopy(args))
return gradval, val
else:
return ag_grad(func)(*deepcopy(args))
def reference_func():
func.__globals__['np'] = ag_np
if preserve_result:
# Note: ag_value_and_grad returns (val, grad) but we need (grad, val)
val, gradval = ag_value_and_grad(func)(*deepcopy(args))
return gradval, val
else:
return ag_grad(func)(*deepcopy(args))
def test_nested_dict(motion, optimized):
p = dict(i=dict(j=3.0, k=4.0))
func = nested_dict
df = tangent.grad(func, motion=motion, optimized=optimized, verbose=True)
dx = df(p)
df_ag = ag_grad(func)
dx_ag = df_ag(p)
for k in p['i']:
assert np.allclose(dx['i'][k], dx_ag['i'][k])
def test_logistic_regression(motion, optimized):
func = logistic_regression
w = np.random.randn(3, 5)
b = np.random.randn(5)
input_ = np.random.rand(3)
label = np.zeros(5)
label[1] = 1
func.__globals__['np'] = np
df = tangent.grad(func,
wrt=(2, 3),
motion=motion,
optimized=optimized,
verbose=True)
dw, db = df(input_, label, w, b)
func.__globals__['np'] = ag_np
ag_dw = ag_grad(func, argnum=2)(input_, label, w, b)
ag_db = ag_grad(func, argnum=3)(input_, label, w, b)
assert np.allclose(ag_dw, dw)
assert np.allclose(ag_db, db)
def test_inlining_contextmanager(motion, optimized, a):
func = inlining_contextmanager
func = tangent.tangent(func)
func.__globals__['np'] = np
df = tangent.grad(func, motion=motion, optimized=optimized, verbose=True)
dx = df(a)
func.__globals__['np'] = ag_np
df_ag = ag_grad(func)
df_ag(a)
assert np.allclose(dx, 2.9 * a**2)
def test_nested_dict(motion, optimized):
p = dict(i=dict(j=3.0, k=4.0))
func = nested_dict
df = tangent.autodiff(
func,
motion=motion,
optimized=optimized,
verbose=True,
input_derivative=INPUT_DERIVATIVE.DefaultOne)
dx = df(p)
df_ag = ag_grad(func)
dx_ag = df_ag(p)
for k in p['i']:
assert np.allclose(dx['i'][k], dx_ag['i'][k])
def test_inlining_contextmanager(motion, optimized, a):
func = inlining_contextmanager
func = tangent.tangent(func)
func.__globals__['np'] = np
df = tangent.autodiff(
func,
motion=motion,
optimized=optimized,
verbose=True,
input_derivative=INPUT_DERIVATIVE.DefaultOne)
dx = df(a)
func.__globals__['np'] = ag_np
df_ag = ag_grad(func)
df_ag(a)
assert np.allclose(dx, 2.9 * a**2)
def _setdiff(self):
self._subgradient = ag_grad(self.eval)
self._jacobian = ag_jacobian(self.eval)
self._hessian = ag_hessian(self.eval)
def _setbatchdiff(self):
self._batch_subgradient = ag_grad(self._batch_eval)
self._batch_jacobian = ag_jacobian(self._batch_eval)
self._batch_hessian = ag_hessian(self._batch_eval)
def reference_func():
func.__globals__['np'] = ag_np
return ag_grad(ag_grad(func))(*args)