Add args for initial ranges to RangeToolLink

[droumis]

fixes #5437

This PR allows users to specify a start and end to the axes ranges in the call to RangeToolLink. Previously, this required a hook (#5437).

Why? Users often need to adjust the initial selection range to have the target plot focused on a subregion, rather than having the entire data range on view at the start. Often, the view of the entire dataset is less useful for initial exploration than a particular focused region, while the entire range still needs to be available for easy pan and zoom navigation. The improvement in this PR will also be helpful for future developments for particularly large amounts of data, where we could set an absolute limit on the viewport range and limit the out-of-viewport data sent to the browser, to prevent crashes or slowdowns. This PR therefore addresses the immediate utility of focused navigation and sets the stage for further developments with large data.

update: changed args and logic to boundsx/y rather than axes_start/end

import pandas as pd
import holoviews as hv
from holoviews import opts
from holoviews.plotting.links import RangeToolLink

hv.extension('bokeh')
from bokeh.sampledata.stocks import AAPL

aapl_df = pd.DataFrame(AAPL['close'], columns=['close'], index=pd.to_datetime(AAPL['date']))
aapl_df.index.name = 'Date'

aapl_curve = hv.Curve(aapl_df, 'Date', ('close', 'Price ($)'))

tgt = aapl_curve.relabel('AAPL close price').opts(width=800, labelled=['y'], toolbar='disable')

src = aapl_curve.opts(width=800, height=100, yaxis=None, default_tools=[])

RangeToolLink(src, tgt, axes=['x', 'y'], boundsx= (pd.to_datetime('2006'), pd.to_datetime('2010')), boundsy=(None, 400))

layout = (tgt + src).cols(1)
layout.opts(opts.Layout(shared_axes=False, merge_tools=False))

[codecov-commenter]

Codecov Report

Merging #5800 (d9847e3) into main (8e85674) will decrease coverage by 0.02%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main    #5800      +/-   ##
==========================================
- Coverage   88.14%   88.13%   -0.02%     
==========================================
  Files         307      307              
  Lines       62863    62897      +34     
==========================================
+ Hits        55413    55434      +21     
- Misses       7450     7463      +13     

Flag	Coverage Δ
ui-tests	?

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
holoviews/plotting/bokeh/links.py	73.29% <100.00%> (+1.27%)	⬆️
holoviews/plotting/links.py	94.91% <100.00%> (+0.17%)	⬆️
holoviews/tests/plotting/bokeh/test_links.py	98.52% <100.00%> (+0.31%)	⬆️

... and 2 files with indirect coverage changes

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[ahuang11]

Looks good! Just some suggestions for refactoring.

[droumis]

one quirk - when you press the 'reset' button in the Bokeh toolbar, users would probably expect the plot to go back to the initial state that was displayed with constrained rangetool bounds... However, upon 'reset', the rangetool currently fills the whole data space.

I don't think this should block this PR.. just to note for a future enhancement

[Hoxbro]

Some small comments. That definitely are not blockers.

[Hoxbro]

Thank you for the PR. Great job!