feat: tooltip

[flekschas]

This PR adds support for showing a tooltip upon hovering a point.

Description

What was changed in this pull request?

This PR adds a new method to the Scatter class called tooltip:

# Enable tooltip via the constructor
scatter = Scatter(data=df, x='mass', y='speed', color_by='group', size=6, tooltip=True)

# Or via the `tooltip` method
scatter.tooltip(true, contents={'color'})

Examples

Screen.Recording.2023-08-14.at.7.58.12.PM.mp4

Screen.Recording.2023-08-14.at.7.56.07.PM.mp4

Why is it necessary?

It makes it easier to interpret data points by showing the underlying data that is visually encoded by the x, y, color, opacity, or size channel.

Checklist

• Provided a concise title as a semantic commit message (e.g. "fix: correctly handle undefined properties")
• CHANGELOG.md updated
• Tests added or updated
• Documentation in API.md/README.md added or updated
• Example(s) added or updated
• Screenshot, gif, or video attached for visual changes

[flekschas]

@lmcinnes I spoke with @hamelin at SciPy and he asked about the ability of displaying a tooltip on hovering points. I wonder if this PR might be useful for you guys in the context of https://github.com/TutteInstitute/thisnotthat?

The tooltip content is currently limited to the dimensions of the data frame that are encoded visually but we could expand it to allow displaying other dimensions. I am thinking of something like this:

scatter = Scatter(data=df, x='x', y='y', color='pval')
scatter.tooltip(true, contents={'color', 'another_column'})

Imagine the data frame as four columns in this case: x, y, pval, and another_column.