René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"Bar chart variants","articleBody":"A couple of bar chart variants will be supported by #80 but only with extra calculation and redundant data. We could express these semantically fairly easily, I would think.\n- [Grouped waterfall](http://wiki.analytica.com/images/b/b3/Waterfall_chart_example.png):\n  ![waterfall_chart_example](https://cloud.githubusercontent.com/assets/2678795/19140356/6794ff5e-8b5a-11e6-9baa-ae4f4c277b76.png)\n  With explicit positioning this requires the user to calculate all the subtotals... can we just make this a new barmode? One complication is how to specify the total bar at the end, which ideally has _no_ data associated with it but is a trace, so that it can be styled (and named) arbitrarily.\n  Would we need a separate mode for a single waterfall group? I don't think so, and keeping each component as a separate trace adds styling and labeling flexibility.\n- [Grouped stacks](https://www.google.com/search?q=grouped+stacked+bar+chart\u0026espv=2\u0026biw=1273\u0026bih=590\u0026source=lnms\u0026tbm=isch\u0026sa=X\u0026ved=0ahUKEwipkrCHp8XPAhXCKCYKHR0JCmoQ_AUIBigB):\n  ![example-of-cluster-stacked-chart](https://cloud.githubusercontent.com/assets/2678795/19140432/2b2eacd0-8b5b-11e6-8fc8-8f914bb87b87.png)\n  Seems like just another barmode plus per-trace attributes for group number or ID.\n","author":{"url":"https://github.com/alexcjohnson","@type":"Person","name":"alexcjohnson"},"datePublished":"2016-10-06T04:26:08.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":3},"url":"https://github.com/1015/plotly.js/issues/1015"}