Dc.js is more for coordinating interactive operations on the data from inter-connected plots, and having the shared data generate updates on the plots that rely on the data. In Bokeh, bokeh server with shared ColumnDataSources, feeding plots is performing this combined functionality. So really, any of the bokeh server examples should give you a good start for a custom dc.js-like app. Crossfilter in Bokeh is a little bit different in that it is a generalized application for viewing any column-oriented data. Dc.js is more focused on building domain specific visualizations.
The one difference is that bokeh can interact with an ajax or server-based data source, so it can support small or large data sets.
From my experience with dc.js and crossfilter.js, I think that the ColumnDataSource in Bokeh with Pandas DataFrame is kind of at the same level of what crossfilter.js functionality is provided. It is primarily focused on data manipulation.
On Friday, April 24, 2015 at 7:21:42 AM UTC-5, Fabio Annovazzi wrote:
first of all thanks for the great work.
I use dc.js (that uses crossfilter.js) to build interactive dashboards, where filtering on one charts makes the other charts filter eccetera. I have limited programming skills, but I manage to work with dc.js reasonably efficiently.
I love Bokeh Crossfilter so I thought it might be possible to use Bokeh in the place of Dc.js.Two questions:
- Is it a good idea at this stage of Bokeh’s development?
- Is there some examples or documentations for not-so-skilled people to get started on building interactive dashboards with Bokeh?