Maybe something like this (working for Bokeh 1.0.4):
import numpy as np
import pandas as pd
from bokeh.layouts import row
from bokeh.models import ColumnDataSource, CustomJS
from bokeh.plotting import figure, curdoc, show
df_main = pd.DataFrame({key: np.random.choice(30, 10) for key in [‘A’, ‘B’, ‘C’, ‘D’, ‘E’]})
df_mean = np.mean(df_main)
source_bars = ColumnDataSource({‘x’: list(range(0, df_mean.shape[0])), ‘y’: list(df_mean.values), ‘colnames’: list(df_mean.index.values)})
sources = { key: ColumnDataSource({‘x’: list(df_main.index.values), ‘y’: list(df_main.iloc[:, index].values)}) for index, key in enumerate([‘A’, ‘B’, ‘C’, ‘D’, ‘E’]) }
plot1 = figure(tools = “tap”)
plot2 = figure()
bars = plot1.vbar(x = ‘x’, top = ‘y’, source = source_bars, bottom = 0, width = 0.5)
lines = plot2.line(x = ‘x’, y = ‘y’, source = sources[‘A’])
code = ‘’’ selected_index = ds_bars.selected[“1d”][“indices”][0]
patch_name = ds_bars.data[‘colnames’][selected_index]
lines.data_source.data[‘y’] = sources[patch_name].data[‘y’]
lines.data_source.change.emit() ‘’’
plots = row(plot1, plot2)
source_bars.selected.js_on_change(‘indices’, CustomJS(args = {‘ds_bars’: source_bars, ‘lines’: lines, ‘sources’: sources}, code = code))
curdoc().add_root(plots)
show(plots)
``
···
On Friday, February 22, 2019 at 3:58:08 PM UTC+1, ARPAD ANDRASSY wrote:
Hi everyone,
I’m quite new to Python and Bokeh specifically, and while I started getting some hang of Bokeh I’ve hit the JS Callback wall, that I need to use if I want to keep having interactions embedded in independent HTML pages.
Basically my goal is to replicate something very similar to what this user did pretty neatly with Bokeh server, but wit JS Callbacks only:
Here is the link to the code he posted on Stackoverflow: Two interactive bokeh plots: select a value in one graph and change the other
Thank you in advance
