How to link categorical x_range to a data source?

Question: Is there a way to bind a categorical x_range to a data source in a similar fashion we bind x and y to data source columns by supplying column names?

It seems to be essential to supply a list of strings as an x_range argument when initializing figure with the intention to use categorical labels. Further I want to add interaction to this plot by restricting the data source depending on the state of the widgets (NB: the fix to the MultiSelect update callback has not been released yet).

Without x_range being bound to the data source a manual update to plot.x_range.factors is required once the state of the widgets changes. There seems to an issue with visualization layer whereby circles disappear on mouse wheel action, while rects do not. Tooltips also behave strangely under this scenario.

My assertion is some of this happens because the x_range is disconnected from the data source. Hence the question above.

Hi,

Question: Is there a way to bind a categorical x_range to a data source in a similar fashion we bind x and y to data source columns by supplying column names?

If I get your question correctly, no… not at the moment.

It seems to be essential to supply a list of strings as an x_range argument when initializing figure with the intention to use categorical labels. Further I want to add interaction to this plot by restricting the data source depending on the state of the widgets (NB: the fix to the MultiSelect update callback has not been released yet).

Without x_range being bound to the data source a manual update to plot.x_range.factors is required once the state of the widgets changes. There seems to an issue with visualization layer whereby circles disappear on mouse wheel action, while rects do not. Tooltips also behave strangely under this scenario.

My assertion is some of this happens because the x_range is disconnected from the data source. Hence the question above.

Connecting ranges to DataSources could be a reasonable feature. I’d encourage you to open an issue on Github to discuss it.

Best

Fabio

···

On Wednesday, April 27, 2016 at 12:26:17 PM UTC-5, wrote: