Interactive Twitter Visualization: Election 2012 (made with d3.js)
1–10 of 12 posts
Re: Interactive Twitter Visualization: Election 2012 (made with d3.js)
#2This looks great. One thing that's a little annoying is how the vertical axis automatically scales to the range of the data. It makes it hard to compare peaks that don't occur within the same view range.
Re: Interactive Twitter Visualization: Election 2012 (made with d3.js)
#3Really cool, any plans to add other data sources? FB might be harder, but I think G+ has a public search API. I see it's just a demo for now, but a dashboard view would be great as that "go-to source" for top breaking news.
Re: Interactive Twitter Visualization: Election 2012 (made with d3.js)
#4This looks great. One thing that's a little annoying is how the vertical axis automatically scales to the range of the data. It makes it hard to compare peaks that don't occur within the same view range.
I noticed that too, although I'm not sure how they'd deal with an overly large peak distorting the rest of the data, even years later.
Re: Interactive Twitter Visualization: Election 2012 (made with d3.js)
#5Very slick UX and very informative as well! I would love to be able to create custom topics, such as 'D3.JS' and generate timelines with top tweets.
Quick question: how did you obtain the real-time metric "# of tweets per hour"? Are you sampling to determine an estimate using the Search API, or do you have access to a proverbial "firehose"?
Re: Interactive Twitter Visualization: Election 2012 (made with d3.js)
#6This looks great. One thing that's a little annoying is how the vertical axis automatically scales to the range of the data. It makes it hard to compare peaks that don't occur within the same view range.
The aspect ratio is also problematic with taller windows; best practice is to bank to 45º:
http://vis.berkeley.edu/papers/banking/
It's hard to pick the right aspect ratio without knowing the data beforehand, but you could make some reasonable assumptions about how spiky the peaks are. Given those assumptions, you could choose a fixed-height (say, 400px), and then adjust the visible time range if you still want variable-width. This would allow the chart to increase in size without compromising the aspect ratio.
Alternatively, you could use a responsive layout and just choose between two or three fixed sizes.
Re: Interactive Twitter Visualization: Election 2012 (made with d3.js)
#7This looks great. One thing that's a little annoying is how the vertical axis automatically scales to the range of the data. It makes it hard to compare peaks that don't occur within the same view range.
The aspect ratio is also problematic with taller windows; best practice is to bank to 45º: http://vis.berkeley.edu/papers/banking/ It's hard to pick the right aspect ratio without knowing the data beforehand, but you could make some reasonable assumptions about how spiky the peaks are. Given those assumptions, you could choose a fixed-height (say, 400px), and then adjust the visible time range if you still want varia…
Thanks for the suggestions - it's certainly an issue we're thinking about and looking to improve.
Also, d3.js is a pleasure to work with. Really appreciate all the work you've put into it.
Re: Interactive Twitter Visualization: Election 2012 (made with d3.js)
#8Very slick UX and very informative as well! I would love to be able to create custom topics, such as 'D3.JS' and generate timelines with top tweets. Quick question: how did you obtain the real-time metric "# of tweets per hour"? Are you sampling to determine an estimate using the Search API, or do you have access to a proverbial "firehose"?
We're actually using the Twitter Streaming API, which has pretty high limits even with access levels lower than the full firehose.
Re: Interactive Twitter Visualization: Election 2012 (made with d3.js)
#9Looks real nice, but the information is pretty useless.
* When events unfold, twitter users predictably react.
* Some events are bigger than others.
Re: Interactive Twitter Visualization: Election 2012 (made with d3.js)
#10Looks real nice, but the information is pretty useless. * When events unfold, twitter users predictably react. * Some events are bigger than others.
Yes, but how much bigger? whose getting the attention? that is what's interesting about this visualization.