Ejemplo n.º 1
0
    def test_inverse_gamma_prior(self):

        # Test input parameters
        self.assertRaises(ValueError, pints.InverseGammaLogPrior, 0, 0)
        self.assertRaises(ValueError, pints.InverseGammaLogPrior, 2, -2)
        self.assertRaises(ValueError, pints.InverseGammaLogPrior, -2, 2)

        a1 = 0.123
        a2 = 4.567

        b1 = 2.345
        b2 = 0.356

        p1 = pints.InverseGammaLogPrior(a1, b1)
        p2 = pints.InverseGammaLogPrior(a2, b2)

        points = [-2., 0.0, 0.001, 0.1, 1.0, 2.45, 6.789]

        # Test means
        self.assertTrue(np.isnan(p1.mean()))
        self.assertAlmostEqual(p2.mean(), 0.0998037566582562)

        # Test n_parameters
        self.assertEqual(p1.n_parameters(), 1)

        # Test specific points
        for point in points:
            to_test = [point]
            self.assertAlmostEqual(
                scipy.stats.invgamma.logpdf(to_test[0], a=a1, scale=b1),
                p1(to_test), places=9)
            self.assertAlmostEqual(
                scipy.stats.invgamma.logpdf(to_test[0], a=a2, scale=b2),
                p2(to_test), places=9)

        # Test derivatives
        p1_derivs = [0., np.inf, 2.343877e6, 223.27, 1.2220000000000002,
                     -0.06769679300291548, -0.1145365009000441]

        p2_derivs = [0., np.inf, 350433.00000000006, -20.07,
                     -5.211000000000001, -2.2129362765514373,
                     -0.8122790150278407]

        for point, deriv in zip(points, p1_derivs):
            calc_val, calc_deriv = p1.evaluateS1([point])
            self.assertAlmostEqual(calc_deriv[0], deriv)
            self.assertAlmostEqual(calc_val,
                                   scipy.stats.invgamma.logpdf(point, a=a1,
                                                               scale=b1))

        for point, deriv in zip(points, p2_derivs):
            calc_val, calc_deriv = p2.evaluateS1([point])
            self.assertAlmostEqual(calc_deriv[0], deriv)
            self.assertAlmostEqual(calc_val,
                                   scipy.stats.invgamma.logpdf(point, a=a2,
                                                               scale=b2))
Ejemplo n.º 2
0
    def test_inverse_gamma_prior_sampling(self):
        p1 = pints.InverseGammaLogPrior(5.0, 40.)
        self.assertEqual(len(p1.sample()), 1)

        n = 1000
        samples1 = p1.sample(n)
        self.assertEqual(len(samples1), n)

        # Mean should be b/(a-1) = 40/4 = 10, so we check that this is very
        # roughly the case, to ensure the parameterisation is correct
        mean = np.mean(samples1).item()
        self.assertTrue(9. < mean < 11.)
Ejemplo n.º 3
0
 def test_inverse_gamma_prior_cdf_icdf(self):
     p1 = pints.InverseGammaLogPrior(5.0, 4.0)
     self.assertAlmostEqual(p1.cdf(3.5), 0.9936442962684809)
     self.assertAlmostEqual(p1.icdf(0.55), 0.9078166853539807)