def test_standardize_and_filter(self):
model = ARGARCHModel(40.0, 0.4, 0.2, 0.3, 0.4, sc=self.sc)
n = 10000
ts = np.array(model.sample(n))
# de-heteroskedasticize
standardized = model.remove_time_dependent_effects(ts)
filtered = model.add_time_dependent_effects(standardized)
for i in range(len(filtered)):
self.assertAlmostEquals(filtered[i], ts[i], msg="%f != %f at index %d" % (filtered[i], ts[i], i))
def test_standardize_and_filter(self):
model = ARGARCHModel(40.0, 0.4, 0.2, 0.3, 0.4, sc=self.sc)
n = 10000
ts = np.array(model.sample(n))
# de-heteroskedasticize
standardized = model.remove_time_dependent_effects(ts)
filtered = model.add_time_dependent_effects(standardized)
for i in range(len(filtered)):
self.assertAlmostEquals(filtered[i],
ts[i],
msg="%f != %f at index %d" %
(filtered[i], ts[i], i))
def test_fit_model(self):
omega = 0.2
alpha = 0.3
beta = 0.5
genModel = ARGARCHModel(0.0, 0.0, alpha, beta, omega, sc=self.sc)
n = 10000
ts = np.array(genModel.sample(n))
model = GARCH.fit_model(ts, sc=self.sc)
# tolerances are a little bit larger than Scala; not certain why... -pame
self.assertAlmostEqual(model.omega, omega, delta=0.1)
self.assertAlmostEqual(model.alpha, alpha, delta=0.1)
self.assertAlmostEqual(model.beta, beta, delta=0.2)
def test_fit_model(self):
omega = 0.2
alpha = 0.3
beta = 0.5
genModel = ARGARCHModel(0.0, 0.0, alpha, beta, omega, sc=self.sc)
n = 10000
ts = np.array(genModel.sample(n))
model = GARCH.fit_model(ts, sc=self.sc)
# tolerances are a little bit larger than Scala; not certain why... -pame
self.assertAlmostEqual(model.omega, omega, delta=0.1)
self.assertAlmostEqual(model.alpha, alpha, delta=0.1)
self.assertAlmostEqual(model.beta, beta, delta=0.2)
