apunkt/GPy
1
0
Fork 0
mirror of https://github.com/SheffieldML/GPy.git synced 2026-05-15 06:52:39 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

replicated the cubic sine example into warped_gp tests for code coverage

Browse source
This commit is contained in:
beckdaniel 2016-04-07 17:52:59 +01:00
parent 0b32bdf9a6
commit 3e27748b42
2 changed files with 26 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

4
GPy/examples/regression.py
View file

@ -564,9 +564,13 @@ def warped_gp_cubic_sine(max_iters=100):
warp_k = GPy.kern.RBF(1) warp_k = GPy.kern.RBF(1)
warp_f = GPy.util.warping_functions.TanhFunction(n_terms=2) warp_f = GPy.util.warping_functions.TanhFunction(n_terms=2)
warp_m = GPy.models.WarpedGP(X, Y, kernel=warp_k, warping_function=warp_f) warp_m = GPy.models.WarpedGP(X, Y, kernel=warp_k, warping_function=warp_f)
warp_m['.*\.d'].constrain_fixed(1.0)
m = GPy.models.GPRegression(X, Y) m = GPy.models.GPRegression(X, Y)
m.optimize_restarts(parallel=False, robust=True, num_restarts=5, max_iters=max_iters) m.optimize_restarts(parallel=False, robust=True, num_restarts=5, max_iters=max_iters)
warp_m.optimize_restarts(parallel=False, robust=True, num_restarts=5, max_iters=max_iters) warp_m.optimize_restarts(parallel=False, robust=True, num_restarts=5, max_iters=max_iters)
#m.optimize(max_iters=max_iters)
#warp_m.optimize(max_iters=max_iters)
print(warp_m) print(warp_m)
print(warp_m['.*warp.*']) print(warp_m['.*warp.*'])

22
GPy/testing/model_tests.py
View file

@ -328,6 +328,28 @@ class MiscTests(unittest.TestCase):
np.testing.assert_almost_equal(np.exp(preds), warp_preds, decimal=4) np.testing.assert_almost_equal(np.exp(preds), warp_preds, decimal=4)
def test_warped_gp_cubic_sine(self, max_iters=100):
"""
A test replicating the cubic sine regression problem from
Snelson's paper. This test doesn't have any assertions, it's
just to ensure coverage of the tanh warping function code.
"""
X = (2 * np.pi) * np.random.random(151) - np.pi
Y = np.sin(X) + np.random.normal(0,0.2,151)
Y = np.array([np.power(abs(y),float(1)/3) * (1,-1)[y<0] for y in Y])
X = X[:, None]
Y = Y[:, None]
warp_k = GPy.kern.RBF(1)
warp_f = GPy.util.warping_functions.TanhFunction(n_terms=2)
warp_m = GPy.models.WarpedGP(X, Y, kernel=warp_k, warping_function=warp_f)
warp_m['.*\.d'].constrain_fixed(1.0)
warp_m.optimize_restarts(parallel=False, robust=False, num_restarts=5,
max_iters=max_iters)
print warp_m
warp_m.predict(X)
class GradientTests(np.testing.TestCase): class GradientTests(np.testing.TestCase):
def setUp(self): def setUp(self):