Toggling labels and feedback
Right after the mid-term review, I had posted on the Github page for the Sigma plugin for Gephi, asking a couple questions: Is there a way for the user to select two nodes at once? Is there a way to implement a toggle so that labels can be turned on and off? Let’s jump right into it.
That also goes for anything else interactive. A lot of the feedback I’ve received is to have more control over the visualization with sliders and selecting only two nodes to search for. I’m not sure how possible any of this is going to be.
Jason was also taking a look into my files and we dug a little deeper. We found the file that has all of my data in it (nodes and connections) which is there as a GEXF file. So the first thing I tried doing was adding an a tag within the label section of the file for a song. Naturally, since things are never meant to work out, the site broke. So I want to try seeing if Gephi supports HTML tags within the program and will the Sigma plugin export it as such.
These may offer some more functionality and interactivity than I already have. I have yet to test them out. If these become a headache, I think I’ll abandon them and continue focusing on Sigma. But, thank you for your help Jason.
On Friday, I met with Arthur once more to see what he thought about the print. I’m happy to see that his feedback was along the same lines of everyone else’s: the big graph is beautiful to look at but how functional is it? He thinks it’s a great idea to show the comparisons of each category as their own graphs. However, the layout should have a meaning. He asked me how the layout is determined on the bigger one and while it was obviously the algorithm that Gephi, but it’s also based off the amount of connections a node has.
According to Arthur, the meaning doesn’t even have to be something so complex. An idea is to have them organized, from left to right, by year. With a layout like that, it is easy to see what range of years where 120–130 BPMs were popular, for example. (screenshots for these to follow)
In order to isolate these categories on the graph, I have been making separate Gephi files in the past. But I kept reading about the filters area in Gephi. With these filters, certain data (or nodes/connections) can be filtered out to display, with the others disappearing. While I feel like I wasted some time on Friday by watching videos/reading tutorials on these filters, it wound up helping tremendously. Now I can just work with one graph, exporting different layouts. So, it’s safe to say the print(s) is/are coming along nicely.
Oh! I’ve also emailed Chris from the Gephi workshop to see if he could offer insight into incorporating a music player somewhere in the graph.
Update: Just received an email back from Chris. Here’s what he had to say:
Glad to see that the visualization has come together. I have no idea how to embed a player there. All that is pulled from the node attributes–perhaps if you included an embed code there with each song and then re-exported the data, that might be a start. I would contact the developer, as you suggested.
It seems like embedding a link may be the best course of action.