See discussion on NetworkX group: http://groups.google.com/group/networkx ... 843de30bd1
Stephan
Statistics:Posted by unidesigner — 28 Aug 2010 14:30
]]>
i am currently in china in poor internet connection uptime and struggling with proxy to access many blocked websites like gephi.org
Well yes I thought that I could use an existing python graph lib but i hadn't time to make a clearer plan. My previous idea was just to be close to the previous libgexf.
I am developing some more graph utilities using python for others purposes (issuecrawler to navicrawler utility) and I think using a common graph lib could be a nice idea to glue those different tools.
I am glad to hear you'd like to contribute and you're more than welcome.
For the how it's up to you, but a cool way to start things up could be to draft a proposition of refactoring based on the mentioned lib ? or you start directly by setting a new branch and send me a push request ? Depends if like better writing soft architecture or code
hope to work with you on this soon,
bye
Paul
ps: I'll out of internet for about a week
Statistics:Posted by paul — 20 Aug 2010 03:38
]]>
Unidesigner
Statistics:Posted by unidesigner — 12 Aug 2010 14:58
]]>
]]>
I finally released my python library to read and write gexf file.
repository is here : http://github.com/paulgirard/pygexf
package in pypi : http://packages.python.org/pygexf/users.html
doc (to be improved...) : http://pypi.python.org/pypi/pygexf/0.2
one requirements : lxml (see doc)
Testing and doc are not mature enough but it's a good lib to play with gexf.
Feel free to extend it and share your work.
I will go on extend features following project.
hope this will help,
Paul
ps : Big thanks to gexf / gephi team and also to people from lxml who helped about installing lxml on Mac Os X.
Statistics:Posted by paul — 04 Aug 2010 15:35
]]>
I'm doing some research into data mining and network visualisation as part of my MSc. so this is relevant, but a bit too much work to do before my dissertation deadline in September. There is a possibility that I may carry this research into a Ph.D, and if that happens it would be more feasible for me to rewrite the library -- it might be a nice self-contained project to kick off the Ph.D research. I'll let you know!
Statistics:Posted by slithy — 28 Jun 2010 10:42
]]>
Thanks for your interest! Libgexf was primarily developed for static graphs. I started to work on dynamics at the end but quickly had to abandon because the data structures can't deal with this new dimension in an efficient way. The code is also under-architectured and requires a complete refactoring. I'm not very proud of it, but this first version works for what it was designed for
I won't work on it anymore because I've other priorities (a great surprise will be announced soon...), but I'd be very happy to provide a close mentoring and discuss about SWIG compilation if you want to contribute. Note that I'm not a C++/SWIG guru. This version of libgexf overcomes the main issues in SWIG compilation and XML processing, but data structures may be improved, also by supporting hierarchical graphs.
It's a work for 3 to 6 months, depending on your skills. I give you the keys of the project if you're motivated in contributing!
Statistics:Posted by admin — 24 Jun 2010 19:48
]]>
I am new to libgexf but I was just having a look at the source code to see what the status was of dynamic graphs. I thought that DynamicGraph might be implemented something like a subclass of Graph and/or a template type of Directed or Undirected Graph, but it seems to be a class all on its own. What was the intended syntax to use this class?
What was your plan to implement this? Were you going to rework some of the existing classes/code? How much work do you think it would be? I might be interested in contributing.
Statistics:Posted by slithy — 24 Jun 2010 17:16
]]>
admin wrote:Which versions of libgexf and Python do you use?
Python 2.6 and libgexf 0.1.2 from source. I see. Did not expect that, but I have no experience with SWIG...admin wrote:In Python there is always the "self" argument ...
Statistics:Posted by mhermans — 17 May 2010 22:39
]]>
Which versions of libgexf and Python do you use?First, using data.setNodeValue('1', '0' '2') does not work, it expects 4 arguments. If I use a fourth blank arguments it seems to behave. How do I call this function correctly?
In Python there is always the "self" argument taken into account. So this method has 4 arguments: data.setNodeValue(self, '1', '0', '2'). Therefore I'll check the bindings tomorrow.
They are not.Second, I can't figure out how to set the attributes in the VIZ module (e.g. size). Are these not supported?
Statistics:Posted by admin — 17 May 2010 22:17
]]>
First, using data.setNodeValue('1', '0' '2') does not work, it expects 4 arguments. If I use a fourth blank arguments it seems to behave. How do I call this function correctly?
Second, I can't figure out how to set the attributes in the VIZ module (e.g. size). Are these not supported?
Statistics:Posted by mhermans — 17 May 2010 19:34
]]>
I am testing it right now (see http://forum.gephi.org/viewtopic.php?f=4&t=111 and http://forum.gephi.org/viewtopic.php?f=4&t=112)
As soon as it's tested enough I'll release it here although it's far less complete as existing libgexf it's working for dynamic and any python addict (included me) could extend gexf features support in future.
I should have modified the existing libgexf but I could write a new one in Python in short time where C++ would have taken me ages to change and understand, sorry about that.
I'll let you know..
Paul
Statistics:Posted by paul — 02 Apr 2010 09:32
]]>
]]>
I will think about a workaround..
Paul
Statistics:Posted by paul — 31 Mar 2010 14:26
]]>
Dynamic graphs are not supported, due to major refactoring needed to implement it. As I stopped my contribution to this project, it won't be supported unless further fundings / new developer arrives.
Statistics:Posted by admin — 31 Mar 2010 12:34
]]>
I don't see any data.setLabel() function (http://gexf.net/lib/api/functions_func.html#index_s). Where did you find it?
Statistics:Posted by admin — 31 Mar 2010 10:47
]]>