Add scalebar support

[hoxbro]

Fixes #5948

I'm not entirely sure what plot classes need to have a scalebar option: all of them or a selective few. For now, I have added it to RGB to start with.

Also, I am not sure how many of the Scalebar options we should have available.

import holoviews as hv
hv.extension("bokeh")

hv.RGB.load_image("5948_pollen.png").opts(scalebar=True, scale_opts={"background_fill_alpha": 1})

image_download

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 97.84946% with 2 lines in your changes are missing coverage. Please review.

Project coverage is 88.49%. Comparing base (f39e0a0) to head (d6bb610).
Report is 1 commits behind head on main.

Files	Patch %	Lines
holoviews/plotting/bokeh/element.py	94.11%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #6002      +/-   ##
==========================================
+ Coverage   88.48%   88.49%   +0.01%     
==========================================
  Files         323      323              
  Lines       67640    67728      +88     
==========================================
+ Hits        59848    59933      +85     
- Misses       7792     7795       +3     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[hoxbro]

Maybe not the best way, but I couldn't figure out other ways to get the kdims.

[hoxbro]

I added a scalebar to the base ElementPlot class, which means that most elements should support having a scalebar.

Example of scalebar with subcordinates_y:

import holoviews as hv
import numpy as np
hv.extension("bokeh")

c1 = hv.Curve(np.random.rand(1000), label='c1').opts(subcoordinate_y=True, scalebar=True, scalebar_location="bottom_left", scalebar_unit="cm")
c2 = hv.Curve(np.random.rand(1000), label='c2').opts(subcoordinate_y=True, scalebar=True, scalebar_location="top_left")

curves = hv.Overlay(c1 * c2)
curves.opts(width=800, height=800)

simplescreenrecorder-2023-12-22_13.18.15.mp4

I'm not sure where to put the documentation for this feature; honestly, I would rather wait with it until custom units are supported in Bokeh.

[droumis]

Linking to the Bokeh PR for custom units

[droumis]

What is left to do for this PR?

[hoxbro]

Implement the custom unit introduced in bokeh/bokeh#13625

[droumis]

Just to connect this PR with the original use-case requirements spec: holoviz-topics/neuro#34

[hoxbro]

I have added a tool to hide scale bars.

import holoviews as hv
import numpy as np

hv.extension("bokeh")

c1 = hv.Curve(np.random.rand(1000) + 2, label="c1").opts(
    scalebar=True, scalebar_location="bottom_left", scalebar_unit="cm"
)
c2 = hv.Curve(np.random.rand(1000), label="c2").opts(scalebar=True, scalebar_location="top_left")

curves = hv.Overlay(c1 * c2)
curves.opts(width=800, height=800)

screenrecord-2024-05-28_19.37.06.mp4

[droumis]

What's with the html? I see other reference notebooks just using markdown

<div class="contentcontainer med left" style="margin-left: -50px;">
<dl class="dl-horizontal">
  <dt>Title</dt> <dd> Scalebar</dd>
  <dt>Dependencies</dt> <dd>Bokeh</dd>
  <dt>Backends</dt> 
    <dd><a href='./Scalebar.ipynb'>Bokeh</a></dd>
</dl>
</div>

[droumis]

I don't quite understand how to use the kdims to apply an inferred unit.

From the description ('The unit of the scalebar tries to infer it for the first kdims.'), it sounds like this should change the unit to Voltage:

And this to Time:

Code

from datetime import datetime, timedelta

times = [datetime.now() + timedelta(seconds=i) for i in range(1000)]
hv.Curve((times, np.random.rand(1000)), kdims=['Time']).opts(scalebar=True)

[droumis]

hmm.. ok I kind of see how to set it via hv.dim

But I guess there's no way to set both the base-dimension and the dimension with hv.Dimension. It might be a pretty narrow use case where the 'inferred' (semi-explicit but partial via hv.Dimension) approach would be sufficient. I think we should consider moving the inferred approach to a future PR, unless I'm really missing something.

[hoxbro]

What's with the html? I see other reference notebooks just using markdown

Honest to god, I just clicked on a random reference guide, and that was how it was done in that 🤷‍♂️

hmm.. ok I kind of see how to set it via hv.dim

You are right. That is the way to do it. Just if you are unaware, hv.dim is something else than hv.Dimension.

Maybe we should update the Example to show how to use hv.Dimension.

[droumis]

I only see success when using a fully specified scalebar_unit arg, so I will only document that approach until future PRs address other approaches.

If it is going to take a while to address the identified issues, I think we should consider delaying other methods for specifying the units until future PRs (including only supplying a single string to scalebar_unit), so we can proceed with the release.

Unsuccessful approaches

[droumis]

I also don't yet see support for a scalebar_icon arg so I'll remove this from the documentation for now

[hoxbro]

I also don't yet see support for a scalebar_icon arg so I'll remove this from the documentation for now

Was an earlier name for scalebar_tool 🙃

[droumis]

OK I think the docs are ready.

[philippjfr]

Suggested change

	scalebar_range =param.Selector(default="x", objects=["x", "y"], doc="""
	scalebar_range = param.Selector(default="x", objects=["x", "y"], doc="""

[philippjfr]

I'm okay with merging this but ideally we'd handle updates to the scalebar options.

[github-actions]

This pull request has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.