nngraph: fix definition of gaussian kernel

doc/tutorials/optimization.rst

@@ -85,7 +85,7 @@ We start with the graph TV regularization. We will use the :class:`pyunlocbox.so
     >>> prob1 = pyunlocbox.solvers.solve([d, r, f], solver=solver,
     ...                                  x0=x0, rtol=0, maxit=1000)
     Solution found after 1000 iterations:
-        objective function f(sol) = 2.138668e+02
+        objective function f(sol) = 2.367630e+02
         stopping criterion: MAXIT
     >>>
     >>> fig, ax = G.plot(prob1['sol'])
@@ -107,7 +107,7 @@ This figure shows the label signal recovered by graph total variation regulariza
     >>> prob2 = pyunlocbox.solvers.solve([r, f], solver=solver,
     ...                                  x0=x0, rtol=0, maxit=1000)
     Solution found after 1000 iterations:
-        objective function f(sol) = 7.413918e+01
+        objective function f(sol) = 3.668690e+01
         stopping criterion: MAXIT
     >>>
     >>> fig, ax = G.plot(prob2['sol'])

pygsp/filters/filter.py

@@ -254,7 +254,7 @@ def filter(self, s, method='chebyshev', order=30):
         >>> _ = G.plot(s1, ax=axes[0])
         >>> _ = G.plot(s2, ax=axes[1])
         >>> print('{:.5f}'.format(np.linalg.norm(s1 - s2)))
-        0.28049
+        0.27175
 
         Perfect reconstruction with Itersine, a tight frame:
 
@@ -449,7 +449,7 @@ def estimate_frame_bounds(self, x=None):
         A=1.708, B=2.359
         >>> A, B = g.estimate_frame_bounds(G.e)
         >>> print('A={:.3f}, B={:.3f}'.format(A, B))
-        A=1.712, B=2.348
+        A=1.720, B=2.355
 
         The frame bounds can be seen in the plot of the filter bank as the
         minimum and maximum of their squared sum (the black curve):

pygsp/graphs/fourier.py

@@ -78,7 +78,7 @@ def coherence(self):
         >>> graph.compute_fourier_basis()
         >>> minimum = 1 / np.sqrt(graph.n_vertices)
         >>> print('{:.2f} in [{:.2f}, 1]'.format(graph.coherence, minimum))
-        0.93 in [0.12, 1]
+        0.91 in [0.12, 1]
         >>>
         >>> # Plot the most localized eigenvector.
         >>> import matplotlib.pyplot as plt

pygsp/graphs/nngraphs/nngraph.py

@@ -321,7 +321,7 @@ def __init__(self, features, standardize=False,
             # Users can easily do the above.
 
         def kernel(distance, width):
-            return np.exp(-distance**2 / width)
+            return np.exp(-distance**2 / width**2)
 
         W.data = kernel(W.data, kernel_width)
 

pygsp/tests/test_filters.py

@@ -149,7 +149,7 @@ def test_frame(self):
         def get_frame(freq_response):
             return self._G.U.dot(np.diag(freq_response).dot(self._G.U.T))
         gL = np.concatenate([get_frame(gl) for gl in g.evaluate(self._G.e)])
-        np.testing.assert_allclose(gL1, gL)
+        np.testing.assert_allclose(gL1, gL, atol=1e-10)
         np.testing.assert_allclose(gL2, gL, atol=1e-10)
 
     def test_complement(self, frame_bound=2.5):
@@ -228,7 +228,7 @@ def test_modulation_gabor(self):
         f2 = filters.Gabor(self._G, f)
         s1 = f1.filter(self._signal)
         s2 = f2.filter(self._signal)
-        np.testing.assert_allclose(s1, s2, atol=1e-5)
+        np.testing.assert_allclose(s1, -s2, atol=1e-5)
 
     def test_halfcosine(self):
         f = filters.HalfCosine(self._G, Nf=4)