gexf4j
Posted: 15 Aug 2012 11:09
Hi
I'm unsure where to post my questions, ignore, move or delete this thread if it's at a completely wrong place.
I'm combining crawler4j and gexf4j to create graphs of websites where nodes are pages and links are the edges.
Now to my questions:
- Why isn't it possible to create parallel edges? (Is this because it's not in the standard of gexf1.2, or because gephi doesn't support it? I couldn't find out)
- As I'm crawling a page, I need to connect the pages(nodes) with links(edges), the only way to add a new edge to the graph is by using Node.connectTo(String, Node), am I right? But sometimes it could be necessary to connect a node which doesn't exist by now (In my case when the target page isn't crawled yet, So I create dummy nodes which are later on filled with infos. Again that leads to another problem as it doesn't seem possible to me to easily get a node from the graph. Shouldn't there be something like Graph.getNodeById(id) ?)
I got around those problems by creating the according methods, and am willing to commit those to the repository, but would appreciate if the maintainer would have a look at it first. Or maybe it's like that by design and I don't understand it.
Regards
I'm unsure where to post my questions, ignore, move or delete this thread if it's at a completely wrong place.
I'm combining crawler4j and gexf4j to create graphs of websites where nodes are pages and links are the edges.
Now to my questions:
- Why isn't it possible to create parallel edges? (Is this because it's not in the standard of gexf1.2, or because gephi doesn't support it? I couldn't find out)
- As I'm crawling a page, I need to connect the pages(nodes) with links(edges), the only way to add a new edge to the graph is by using Node.connectTo(String, Node), am I right? But sometimes it could be necessary to connect a node which doesn't exist by now (In my case when the target page isn't crawled yet, So I create dummy nodes which are later on filled with infos. Again that leads to another problem as it doesn't seem possible to me to easily get a node from the graph. Shouldn't there be something like Graph.getNodeById(id) ?)
I got around those problems by creating the according methods, and am willing to commit those to the repository, but would appreciate if the maintainer would have a look at it first. Or maybe it's like that by design and I don't understand it.
Regards