Python 3 compatibility improved

README.md

@@ -13,24 +13,27 @@ for binary sequences and even kernels for labeled graphs.
 
 ## Sample usage
 
-    from sklearn.svm import SVC
-    from sklearn.metrics import accuracy_score
-    import numpy as np
+```python
+from __future__ import print_function
+from sklearn.svm import SVC
+from sklearn.metrics import accuracy_score
+import numpy as np
 
-    from pykernels.basic import RBF
+from pykernels.basic import RBF
 
-    X = np.array([[1,1], [0,0], [1,0], [0,1]])
-    y = np.array([1, 1, 0, 0])
+X = np.array([[1,1], [0,0], [1,0], [0,1]])
+y = np.array([1, 1, 0, 0])
 
-    print 'Testing XOR'
+print('Testing XOR')
 
-    for clf, name in [(SVC(kernel=RBF(), C=1000), 'pykernel'), (SVC(kernel='rbf', C=1000), 'sklearn')]:
-        clf.fit(X, y)
-        print name
-        print clf
-        print 'Predictions:', clf.predict(X)
-        print 'Accuracy:', accuracy_score(clf.predict(X), y)
-        print
+for clf, name in [(SVC(kernel=RBF(), C=1000), 'pykernel'), (SVC(kernel='rbf', C=1000), 'sklearn')]:
+    clf.fit(X, y)
+    print(name)
+    print(clf)
+    print('Predictions:', clf.predict(X))
+    print('Accuracy:', accuracy_score(clf.predict(X), y))
+    print()
+```
 
 ## implemented Kernels
 

example.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 from sklearn.svm import SVC
 from sklearn.metrics import accuracy_score
 import numpy as np
@@ -7,12 +8,11 @@
 X = np.array([[1,1], [0,0], [1,0], [0,1]])
 y = np.array([1, 1, 0, 0])
 
-print 'Testing XOR'
+print('Testing XOR')
 
 for clf, name in [(SVC(kernel=RBF(), C=1000), 'pykernel'), (SVC(kernel='rbf', C=1000), 'sklearn')]:
     clf.fit(X, y)
-    print name
-    print clf
-    print 'Predictions:', clf.predict(X)
-    print 'Accuracy:', accuracy_score(clf.predict(X), y)
-    print
+    print(name)
+    print(clf)
+    print('Predictions:', clf.predict(X))
+    print('Accuracy:', accuracy_score(clf.predict(X), y))

pykernels/__init__.py

@@ -1,6 +1,6 @@
-from basic import Linear, Polynomial, RBF
-from regular import Exponential, Laplacian, RationalQuadratic, InverseMultiquadratic, Cauchy, TStudent,\
+from .basic import Linear, Polynomial, RBF
+from .regular import Exponential, Laplacian, RationalQuadratic, InverseMultiquadratic, Cauchy, TStudent,\
                     ANOVA, Spline, Min, Log, Power, Chi2, AdditiveChi2, GeneralizedHistogramIntersection,\
                     Tanimoto, Sorensen, MinMax, Wavelet, Fourier
-from graph.randomwalk import RandomWalk
-from graph.allgraphlets import All34Graphlets
+from .graph.randomwalk import RandomWalk
+from .graph.allgraphlets import All34Graphlets

pykernels/basic.py

@@ -7,7 +7,7 @@
 
 from pykernels.base import Kernel
 import numpy as np
-from utils import euclidean_dist_matrix
+from .utils import euclidean_dist_matrix
 
 class Linear(Kernel):
     """

pykernels/graph/__init__.py

@@ -1,3 +1,3 @@
-from randomwalk import RandomWalk
-from allgraphlets import All34Graphlets
-from shortestpath import ShortestPath
+from .randomwalk import RandomWalk
+from .allgraphlets import All34Graphlets
+from .shortestpath import ShortestPath

pykernels/graph/allgraphlets.py

@@ -7,7 +7,8 @@
 import itertools
 import numpy as np
 from pykernels.base import Kernel, GraphKernel
-import basic
+from .basic import graphs_to_adjacency_lists
+
 
 def dec2bin(k, bitlength=0):
     """Decimal to binary"""
@@ -114,8 +115,8 @@ def __init__(self, k=3):
         self.graphlet_array = _generate_graphlets(k)
 
     def _compute(self, data_1, data_2):
-        data_1 = basic.graphs_to_adjacency_lists(data_1)
-        data_2 = basic.graphs_to_adjacency_lists(data_2)
+        data_1 = graphs_to_adjacency_lists(data_1)
+        data_2 = graphs_to_adjacency_lists(data_2)
         d1 = np.zeros((data_1.shape[0], _number_of_graphlets(self.k)))
         d2 = np.zeros((data_2.shape[0], _number_of_graphlets(self.k)))
         for i, g in enumerate(data_1):

pykernels/graph/basic.py

@@ -21,10 +21,10 @@ def graphs_to_adjacency_lists(data):
     try:
         if data.ndim == 3:
             return np.array(data)
-    except Exception, exc:
+    except Exception:
         try:
             return np.array([G.adjacency_matix for G in data])
-        except Exception, exc:
+        except Exception:
             return np.array(data)
 
 def relabel(data, data_2):

pykernels/graph/randomwalk.py

@@ -8,7 +8,7 @@
 from pykernels.base import Kernel, GraphKernel
 from scipy.sparse import lil_matrix, kron,identity
 from scipy.sparse.linalg import lsqr
-import basic
+from .basic import graphs_to_adjacency_lists
 
 def _norm(adj_mat):
     """Normalize adjacency matrix"""
@@ -29,8 +29,8 @@ def __init__(self, lmb=0.5, tolerance=1e-8, maxiter=20):
 
     # either tensor of dimention 3 (list of adjacency matrices)
     def _compute(self, data_1, data_2):
-        data_1 = basic.graphs_to_adjacency_lists(data_1)
-        data_2 = basic.graphs_to_adjacency_lists(data_2)
+        data_1 = graphs_to_adjacency_lists(data_1)
+        data_2 = graphs_to_adjacency_lists(data_2)
         res = np.zeros((len(data_1), len(data_2)))
         N = len(data_1) * len(data_2)
         for i, graph1 in enumerate(data_1):

pykernels/graph/shortestpath.py

@@ -5,7 +5,7 @@
 
 import numpy as np
 import numpy.matlib as matlib
-import basic
+from .basic import graphs_to_adjacency_lists, relabel
 from pykernels.base import Kernel, GraphKernel
 from scipy.sparse import lil_matrix
 
@@ -18,9 +18,9 @@ def floyd_warshall(adj_mat, weights):
     res = res + ((adj_mat != 0) * weights)
     res[res == 0] = np.inf
     np.fill_diagonal(res, 0)
-    for i in xrange(N):
-        for j in xrange(N):
-            for k in xrange(N):
+    for i in range(N):
+        for j in range(N):
+            for k in range(N):
                 if res[i, j] + res[j, k] < res[i, k]:
                     res[i, k] = res[i, j] + res[j, k]
     return res
@@ -85,16 +85,16 @@ def _create_accum_list(self, shortest_paths, maxpath):
         return res
 
     def _compute(self, data_1, data_2):
-        ams_1 = basic.graphs_to_adjacency_lists(data_1)
-        ams_2 = basic.graphs_to_adjacency_lists(data_2)
+        ams_1 = graphs_to_adjacency_lists(data_1)
+        ams_2 = graphs_to_adjacency_lists(data_2)
         sp_1, max1 = _apply_floyd_warshall(np.array(ams_1))
         sp_2, max2 = _apply_floyd_warshall(np.array(ams_2))
         maxpath = max(max1, max2)
         if not self.labeled:
             accum_list_1 = self._create_accum_list(sp_1, maxpath)
             accum_list_2 = self._create_accum_list(sp_2, maxpath)
         else:
-            labels_1, labels_2, numlabels = basic.relabel(
+            labels_1, labels_2, numlabels = relabel(
                 [G.node_labels for G in data_1], [G.node_labels for G in data_2])
             accum_list_1 = self._create_accum_list_labeled(sp_1, maxpath,
                                                            labels_1, numlabels)

pykernels/regular.py

@@ -7,7 +7,7 @@
 
 from pykernels.base import Kernel
 import numpy as np
-from utils import euclidean_dist_matrix
+from .utils import euclidean_dist_matrix
 import warnings
 
 class Cossim(Kernel):