Allow multiple targets for RangeToolLink

[MatusGasparik]

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

Holoviews has the RangeToolLink which can be used to link the axes between the source and the target element:

from holoviews.plotting.links import RangeToolLink

data = np.random.randn(1000).cumsum()

source = hv.Curve(data).opts(width=800, height=125, axiswise=True, default_tools=[])
target = hv.Curve(data).opts(width=800, labelled=['y'], toolbar=None)

rtlink = RangeToolLink(source, target)

(target + source).opts(merge_tools=False).cols(1)

It is often the case that one needs to link multiple targets to the range tool, but in holoviews this is currently not possible:

data = np.random.randn(1000).cumsum()
x1 = np.arange(data.size)
x2 = x1 * 0.5

source = hv.Curve((x1, data)).opts(width=800, height=125, axiswise=True, default_tools=[])
target1 = hv.Curve((x1, data)).opts(width=800, height=200, labelled=['y'], toolbar=None)
target2 = hv.Area((x2, data)).opts(width=800, height=200, labelled=['y'], toolbar=None)

rtlink = RangeToolLink(source, [target1, target2])  # ERROR: 

(target1 + target2 + source).opts(merge_tools=True, shared_axes=False).cols(1)

Error:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
Cell In[68], line 11
      8 target1 = hv.Curve((x1, data)).opts(width=800, height=200, labelled=['y'], toolbar=None)
      9 target2 = hv.Area((x2, data)).opts(width=800, height=200, labelled=['y'], toolbar=None)
---> 11 rtlink = RangeToolLink(source, [target1, target2])
     13 (target1 + target2 + source).opts(merge_tools=True, shared_axes=False).cols(1)

File [~/Dropbox/FHNW/Projekte/CoGE/wt-coge-va/feature_gam/.pixi/envs/jupyter/lib/python3.11/site-packages/holoviews/plotting/links.py:44](http://localhost:8888/lab/workspaces/auto-g/tree/docs/~/Dropbox/FHNW/Projekte/CoGE/wt-coge-va/feature_gam/.pixi/envs/jupyter/lib/python3.11/site-packages/holoviews/plotting/links.py#line=43), in Link.__init__(self, source, target, **params)
     42 # Source is stored as a weakref to allow it to be garbage collected
     43 self._source = None if source is None else weakref.ref(source)
---> 44 self._target = None if target is None else weakref.ref(target)
     45 super().__init__(**params)
     46 self.link()

TypeError: cannot create weak reference to 'list' object
Selection deleted

Describe the solution you'd like

It would be nice to be able to pass a list of targets to the RangeToolLink:

rtlink = RangeToolLink(source, [target1, target2]) 

At the moment it is possbile to achive this functionality in a somewhat hacky implicit way using shared_axes=True and manipulating the kdims - see #6463. So the role of the shared_axes parameter should be clarified.

Describe alternatives you've considered

In Bokeh, this could look like so:

from bokeh.plotting import figure, show
from bokeh.models import RangeTool, Range1d, ColumnDataSource
from bokeh.layouts import column
import numpy as np


y = np.random.randn(1000).cumsum()
x = np.arange(y.size)
y_area = y - y.mean()

source = ColumnDataSource(data=dict(x=x, y=y, y_area=y_area))

overview = figure(
    height=120,
    width=800,
    tools="",
    toolbar_location=None,
    x_range=(x.min(), x.max()),
    y_axis_location="right",
)
overview.line('x', 'y', source=source, line_width=2, color="black")
overview.plot.title = "Overview"

detail1 = figure(
    height=200,
    width=800,
    tools="xpan", toolbar_location=None, 
)
detail1.varea('x', 0, 'y_area', source=source, fill_color="grey", alpha=0.4)
detail1.line(x, y_area, line_width=1, color="black")
detail1.yaxis.axis_label = "y - mean(y)"

detail2 = figure(
    height=200,
    width=800,
    tools="xpan", toolbar_location=None,
)
detail2.line('x', 'y', source=source, line_width=2, color="blue")
detail2.yaxis.axis_label = "y"

range_tool = RangeTool(x_range=Range1d(x[500], x[700]))
range_tool.overlay.fill_color = "navy"
range_tool.overlay.fill_alpha = 0.2
overview.add_tools(range_tool)

detail1.x_range = range_tool.x_range
detail2.x_range = range_tool.x_range

layout = column(overview, detail1, detail2, )
show(layout)