Beispiel #1
0
class PenLogRegTests(unittest.TestCase):
    def setUp(self):
        self.stat_calc = Identity(degree=1, cross=False)
        self.distancefunc = PenLogReg(self.stat_calc)
        self.rng = np.random.RandomState(1)

    def test_distance(self):
        d1 = 0.5 * self.rng.randn(100, 2) - 10
        d2 = 0.5 * self.rng.randn(100, 2) + 10
        d3 = 0.5 * self.rng.randn(95, 2) + 10

        d1 = d1.tolist()
        d2 = d2.tolist()
        d3 = d3.tolist()
        # Checks whether wrong input type produces error message
        self.assertRaises(TypeError, self.distancefunc.distance, 3.4, d2)
        self.assertRaises(TypeError, self.distancefunc.distance, d1, 3.4)

        # completely separable datasets should have a distance of 1.0
        self.assertEqual(self.distancefunc.distance(d1, d2), 1.0)

        # equal data sets should have a distance of 0.0
        self.assertEqual(self.distancefunc.distance(d1, d1), 0.0)

        # equal data sets should have a distance of 0.0; check that in case where n_samples is not a multiple of n_folds
        # in cross validation (10)
        self.assertEqual(self.distancefunc.distance(d3, d3), 0.0)

        # check if it returns the correct error when the number of datasets:
        self.assertRaises(RuntimeError, self.distancefunc.distance, d1, d3)

    def test_dist_max(self):
        self.assertTrue(self.distancefunc.dist_max() == 1.0)
Beispiel #2
0
class PenLogRegTests(unittest.TestCase):
    def setUp(self):
        self.stat_calc = Identity(degree = 1, cross = 0)
        self.distancefunc = PenLogReg(self.stat_calc)
    
    def test_distance(self):
        d1 = 0.5 * np.random.randn(100,2) - 10
        d2 = 0.5 * np.random.randn(100,2) + 10
        #Checks whether wrong input type produces error message
        self.assertRaises(TypeError, self.distancefunc.distance, 3.4, d2)
        self.assertRaises(TypeError, self.distancefunc.distance, d1, 3.4)

        # completely separable datasets should have a distance of 1.0
        self.assertEqual(self.distancefunc.distance(list(d1),list(d2)), 1.0)

        # equal data sets should have a distance of 0.0
        self.assertEqual(self.distancefunc.distance(list(d1),list(d1)), 0.0)
        
    def test_dist_max(self):
        self.assertTrue(self.distancefunc.dist_max() == 1.0)        
Beispiel #3
0
 def setUp(self):
     self.stat_calc = Identity(degree=1, cross=0)
     self.distancefunc = PenLogReg(self.stat_calc)
Beispiel #4
0
 def setUp(self):
     self.stat_calc = Identity(degree=1, cross=False)
     self.distancefunc = PenLogReg(self.stat_calc)
     self.rng = np.random.RandomState(1)