Group or hide hover tools in Bokeh toolbar

[droumis]

Is your feature request related to a problem? Please describe.

Too many hover tools when overlaying many elements with custom hover.

Describe the solution you'd like

Group or hide the hover tooltips

Additional context

Code

import numpy as np
import pandas as pd
from datetime import datetime, timedelta

import holoviews as hv; hv.extension('bokeh')
import panel as pn; pn.extension()

amp_dim = hv.Dimension("amplitude", unit="µV")
time_dim = hv.Dimension("time", unit="ms")

n_channels = 10
n_seconds = 5
total_samples = 256*n_seconds
start_datetime = datetime(2024, 1, 1)
time = np.array([start_datetime + timedelta(seconds=t) for t in np.linspace(0, n_seconds, total_samples)])

data = np.random.randn(n_channels, total_samples).cumsum(axis=1)
channels = [f"EEG {i}" for i in range(n_channels)]

df = pd.DataFrame(data.T, index=time, columns=channels)
df.index.name = 'time'

hover_tooltips=[
    ("type", "$group"),
    ("channel", "$label"),
    ("time", '@time{%H:%M:%S.%3N}'),
    ("amplitude"),
]

curves = {}
for channel_name, channel_data in df.items():
    ds = hv.Dataset((channel_data.index, channel_data, channel), [time_dim, amp_dim, "channel"])
    curve = hv.Curve(ds, time_dim, [amp_dim, "channel"], label=channel_name, group='EEG')
    curve.opts(color="black", line_width=1, subcoordinate_y=True, subcoordinate_scale=3,
              hover_tooltips = hover_tooltips)
    curves[channel_name] = curve

curves_overlay = hv.Overlay(curves, kdims="channel").opts(padding=0, aspect=2, responsive=True,show_legend=False)

curves_overlay

[droumis]

@mattpap , this could be solved with the ability to group tools which is sounds like you are working on.

[mattpap]

Yes. The relevant bokeh issue is bokeh/bokeh#5497 (a very old one).

[droumis]

observation: if I create a custom bokeh HoverTool, and pass that in like .opts(tools=[HoverTool()] then the hover tools are combined in the toolbar.

[droumis]

observation: adding a plot with many hover tool icons into a layout container seems to group the tools...

[droumis]

even just hv.Layout((overlay, hv.Empty())

[philippjfr]

Yeah, we should probably run the group_tools function even on a singular plot.

[droumis]

e.g. as we do here for a Layout toolbar