Support networkx 3.2

In networkx 3.2 the output for cycle_basis() has changed.

After a discussion in #36486 it
seems that checking that the output is correct would not be easy. In the
spirit of supporting networkx 3.2 at the same time as networkx 3.1, the
easiest way is to mark these five tests as random output.

Author: tornaria

src/sage/graphs/generic_graph.py

@@ -5102,13 +5102,13 @@ def cycle_basis(self, output='vertex'):
         A cycle basis in Petersen's Graph ::
 
             sage: g = graphs.PetersenGraph()
-            sage: g.cycle_basis()                                                       # needs networkx
+            sage: g.cycle_basis()                                                       # needs networkx, random (changes in networkx 3.2)
             [[1, 6, 8, 5, 0], [4, 9, 6, 8, 5, 0], [7, 9, 6, 8, 5],
              [4, 3, 8, 5, 0], [1, 2, 3, 8, 5, 0], [7, 2, 3, 8, 5]]
 
         One can also get the result as a list of lists of edges::
 
-            sage: g.cycle_basis(output='edge')                                          # needs networkx
+            sage: g.cycle_basis(output='edge')                                          # needs networkx, random (changes in networkx 3.2)
             [[(1, 6, None), (6, 8, None), (8, 5, None), (5, 0, None),
              (0, 1, None)], [(4, 9, None), (9, 6, None), (6, 8, None),
              (8, 5, None), (5, 0, None), (0, 4, None)], [(7, 9, None),
@@ -5273,17 +5273,17 @@ def minimum_cycle_basis(self, algorithm=None, weight_function=None, by_weight=Fa
             [[1, 2, 3], [1, 2, 3, 4], [5, 6, 7]]
             sage: sorted(g.minimum_cycle_basis(by_weight=False))
             [[1, 2, 3], [1, 3, 4], [5, 6, 7]]
-            sage: sorted(g.minimum_cycle_basis(by_weight=True, algorithm='NetworkX'))   # needs networkx
+            sage: sorted(g.minimum_cycle_basis(by_weight=True, algorithm='NetworkX'))   # needs networkx, random (changes in networkx 3.2)
             [[1, 2, 3], [1, 2, 3, 4], [5, 6, 7]]
-            sage: g.minimum_cycle_basis(by_weight=False, algorithm='NetworkX')          # needs networkx
+            sage: g.minimum_cycle_basis(by_weight=False, algorithm='NetworkX')          # needs networkx, random (changes in networkx 3.2)
             [[1, 2, 3], [1, 3, 4], [5, 6, 7]]
 
         ::
 
             sage: g = Graph([(1, 2), (2, 3), (3, 4), (4, 5), (5, 1), (5, 3)])
             sage: sorted(g.minimum_cycle_basis(by_weight=False))
             [[1, 2, 3, 5], [3, 4, 5]]
-            sage: sorted(g.minimum_cycle_basis(by_weight=False, algorithm='NetworkX'))  # needs networkx
+            sage: sorted(g.minimum_cycle_basis(by_weight=False, algorithm='NetworkX'))  # needs networkx, random (changes in networkx 3.2)
             [[1, 2, 3, 5], [3, 4, 5]]
 
         TESTS::