to_dict_of_dicts() probably should assign values of 1 rather than None

if the graph is unweighted.

In addition, add flag to request weights or not.  (I.e., do we want
the adjacency matrix or the weight matrix in sparse form.)  Default is
the old behavior of providing the weight matrix.

--HG--
extra : convert_revision : svn%3A3ed01bd8-26fb-0310-9e4c-ca1a4053419f/networkx/trunk%40442

1 files changed, 7 insertions, 4 deletions

diff --git a/networkx/convert.py b/networkx/convert.py
index 70282dde..52684115 100644
--- a/networkx/convert.py
+++ b/networkx/convert.py
@@ -152,21 +152,24 @@ def from_dict_of_lists(d,create_using=None):
     return G                         
 
 
-def to_dict_of_dicts(G,nodelist=None):
+def to_dict_of_dicts(G,nodelist=None,weighted=False):
     """Return graph G as a Python dictionary of dictionaries.
 
     If nodelist is defined return a dict of dicts with only those nodes.
-
+    
     """
     if nodelist is None:
         nodelist=G.nodes()
+    if not weighted or not hasattr(G,"get_edge"):
+        w=lambda x,y:1
+    else:
+        w=lambda x,y:G.get_edge(x,y)
 
     d = {}
     for u in nodelist:
         d[u]={}
         for v in G.neighbors(u):
-            data=G.get_edge(u,v)
-            d[u][v]=data
+            d[u][v]=w(u,v)
     return d