After looking at Variance I wanted to examine another graphing tool that charges for a stand-alone version of the necessary code. Tools in this vein are more likely to be sustainable and make strong candidates for use in client data visualization projects.
Plotly.js is the library that powers the web tool of the same name but can be downloaded and hosted separately. They don't publish prices but an email to email@example.com will give you a better picture (and individual service).
Plotly.js features many standard graph types but prides itself on more scientific chart types such as box plots, histograms, error bars, log scales, multiple y-axes and more. It also features a grammar of graphics similar to that used in Variance - you encode attributes of the graph in the "data" and "layout" objects, then plotly.js builds the graph from there.
What sets Plotly apart from other charting libraries (in addition to its business model) is the community that has formed around it. The Plotly feed catalogs a huge variety of charts created with the tool.
Jack Parmer, Plotly's CEO, told me:
We created the feed because we thought there should be a place where you can see what everybody else using Plotly is up to. We designed it to promote more conversation, more constructive criticism, and more collaboration.
Each item in the feed can be forked and edited with Plotly, making it incredibly easy to see how each is built. Each user has a public profile page and every graph can be commented on or improved upon by other members of the community.
All this goes a long way to making the community and the library seem friendly and easy to use, which I'm sure wins it many converts.