Using downsampling capabiliy inside a notebook

Hi everybody,

I’m pretty new to Bokeh and I’m currently to trying to use it inside an IPython Notebook.

I’m trying to plot large time series (over 50 000 points) and the interactive plot get really slow ( I often get the “script not responding” error from firefox).
From what I understood, there is a downsampling feature when using the Bokeh server which increase the performance of the interactive plotting. However I could not find any (recent) documentation on how to enable the downsampling features. Could maybe somebody explain me how this work or point me to the corresponding documentation?

So far I tried (after starting the Bokeh server on port 5006) :

import bokeh.plotting as plt
plt.output_notebook(url=“http://127.0.0.1:5006/",docname="line”)
f = plt.figure()
f.line( mySamplingTime, myTimeSerie)
plt.show( plt.vplot(f) )

It displays the plot inside the notebook through the server but it is still very slow…

Thanks in advance for any help!
Florian