I have been using R, igraph, and gephi to create semantic networks. I generally export gml file from R through igraph, then import the gml file into gephi for further adjustment. However, I am now analyzing a large corpus containing more than 20million online posts, and the gml file it created is so large that it overwhelmed my computer. Are there better formats to transfer network data from R to gephi? Or are there better ways to visualize data? (I feel ggplot2 too complicated and unintuitive.
Thanks
Let me know how igraph goes about drawing a graph. Which rules does it use?
Well, it does not even reach the point of drawing a graph. I stuck in the phase of importing data. the data just too large
Hi,
I would recommend you to use NetworkX library (Python). I have observed is better to manage large datasets.
Now, regarding visualization, 20 million nodes is not possible to observe. I would recommend you to prune your network to 10,000 - 100,000 links. Then, Cytoscape is the best tool to visualize large networks.
- Badges
- Science topic
- Similar topics
- Filing