def test_cdf(self):
sim_model = EconDensity()
x = np.ones(shape=(2000, 1))
y = np.random.uniform(0.01, 5, size=(2000, 1))
p_sim = sim_model.cdf(x, y)
p_true = stats.norm.cdf(y, loc=1, scale=2)
diff = np.sum(np.abs(p_sim - p_true))
self.assertAlmostEquals(diff, 0.0, places=2)
def generate_report(): econ_density = EconDensity() X, Y = econ_density.simulate(n_samples=1000) nke = NeighborKernelDensityEstimation() nke.fit_by_cv(X, Y) n_samples = 500 X_test = np.asarray([1 for _ in range(n_samples)]) Y_test = np.linspace(0, 8, num=n_samples) Z = nke.pdf(X_test, Y_test)
def test_cdf_sample_consistency(self):
from statsmodels.distributions.empirical_distribution import ECDF
model = EconDensity()
x_cond = np.asarray([0.1 for _ in range(200000)])
_, y_sample = model.simulate_conditional(x_cond)
emp_cdf = ECDF(y_sample.flatten())
cdf = lambda y: model.cdf(x_cond, y)
mean_cdf_diff = np.mean(
np.abs(emp_cdf(y_sample).flatten() - cdf(y_sample).flatten()))
self.assertLessEqual(mean_cdf_diff, 0.01)
def test_conditional_value_at_risk(self):
sim_model = EconDensity()
x_cond = np.array([[0], [1]])
CVaR = sim_model.conditional_value_at_risk(x_cond, alpha=0.03)
CVaR_mc = super(EconDensity,
sim_model).conditional_value_at_risk(x_cond,
alpha=0.03,
n_samples=10**7)
print("CVaR Analytic:", CVaR)
print("CVaR MC:", CVaR_mc)
print("VaR", sim_model.value_at_risk(x_cond, alpha=0.03))
diff = np.mean(np.abs(CVaR_mc - CVaR))
self.assertAlmostEqual(diff, 0, places=2)
def test_value_at_risk(self):
sim_model = EconDensity()
x_cond = np.array([[0], [1]])
VaR = sim_model.value_at_risk(x_cond, alpha=0.05)
VaR_cdf = super(EconDensity, sim_model).value_at_risk(x_cond,
alpha=0.05)
diff = np.sum(np.abs(VaR_cdf - VaR))
self.assertAlmostEqual(VaR[0],
stats.norm.ppf(0.05, loc=0, scale=1),
places=4)
self.assertAlmostEqual(VaR[1],
stats.norm.ppf(0.05, loc=1, scale=2),
places=4)
self.assertAlmostEqual(diff, 0, places=4)
def test_random_seed(self):
sim_model1 = EconDensity(random_seed=22)
X1, Y1 = sim_model1.simulate(n_samples=100)
sim_model2 = EconDensity(random_seed=22)
X2, Y2 = sim_model2.simulate(n_samples=100)
diff_x = np.sum(np.abs(X1[:100] - X2[:]))
diff_y = np.sum(np.abs(Y1[:100] - Y2[:]))
self.assertAlmostEquals(diff_x, 0, places=2)
self.assertAlmostEquals(diff_y, 0, places=2)
def eval_econ_data():
gmm = GaussianMixture(ndim_x=1, ndim_y=1)
econ_density = EconDensity()
# print("ECON DATA --------------")
# print("KMN")
# for n_centers in [50, 100, 200]:
# kmn = KernelMixtureNetwork(n_centers=n_centers)
# gof = GoodnessOfFit(kmn, econ_density, n_observations=2000, print_fit_result=False, repeat_kolmogorov=1)
# gof_results = gof.compute_results()
# print("N_Centers:", n_centers)
# print(gof_results)
print("LAZY-Learner:")
nkde = KernelMixtureNetwork(n_training_epochs=10)
gof = GoodnessOfFit(nkde, gmm, n_observations=100, print_fit_result=False)
gof_results = gof.compute_results()
print(gof_results)
print(gof_results.report_dict())