How to plot multi charts with AjaxDataSource

[Henry]

I want to draw multi charts with one ajax call.
the ajax return a list of dictionary like this:

list(dict(
     x= series["x"].tolist(),
     y=series["y"].tolist(),
     z=series["z"].tolist(),
     date_time=series["date_time"].tolist()
 ))

I want to draw the multi charts by the return data.
each dictionary item in the list for one chart.
not sure how to use the AjaxDataSource.
I reference the

Multiple Ajaxdatasource call for a 3-glyph plot.

I already posted (a better readable post) on stackoverflow but had no answer: python - Bokeh+Flask: Multiple AjaxDataSource calls - Stack Overflow A simple 3-glyphs plot is giving me headache. This is the stripped down class: class Graph: def __init__(self): data_source = AjaxDataSource(data=dict(date_time=[], value_1=[], value_2=[], value_3…

GitHub - davidhamann/bokeh-ajax-tutorial: Bokeh ajax tutorial for blog post

but not exactly the same things.

[Bryan]

There is not any built-in / “official” way to do this, AjaxDataSource is meant to update a single columnar data source from an endpoint.

You could potentially slightly abuse the AjaxDataSource.adapter hook, by passing in other CDS objects to the args parameter of the CustomJS, and then updating those manually in the hook code. But note you would have to handle and rollover, etc. yourself, to prevent those data sources from growing without bound.