d3-milestones

A d3 based timeline visualization.

The Viking Timeline

NPM: https://www.npmjs.com/package/d3-milestones
Github: https://github.com/walterra/d3-milestones
Storybook demos: https://walterra.github.io/d3-milestones

If you’re using d3-milestones out in the wild I’d love to see what you came up with, just ping me on twitter.com/walterra.

Installing

yarn add d3-milestones.

The most quick way to get going is to use unpkg.com as a CDN to include the library directly into your HTML file.

<link rel="stylesheet" href="https://unpkg.com/d3-milestones/build/d3-milestones.css">
<script src="https://unpkg.com/d3-milestones/build/d3-milestones.min.js"></script>
<div id="timeline"></div>
<script>
  milestones('#timeline')
    .mapping({
      'timestamp': 'year',
      'text': 'title'
    })
    .parseTime('%Y')
    .aggregateBy('year')
    .render([
      { year: 789, title: 'Vikings begin attacks on England.' },
      { year: 840, title: 'Vikings found Dublin in Ireland.' }
      ...
      { year: 1050, title: 'The city of Oslo is founded in Norway.' },
      { year: 1066, title: 'Battle of Hastings.' }
    ]);
</script>

Head over here to see this example in action: @walterra/vikings-timeline"">https://beta.observablehq.com/@walterra/vikings-timeline.

Examples

Examples are included using storybook:

Vikings Timeline
Windows/macOS Timeline
Ordinal Scale Example - Demonstrates using an ordinal scale instead of time scale
Ordinal Scale with Categories - Shows how to use ordinal scales with multiple categories

API Reference

General Usage

To initialize d3-milestones, pass a DOM Id to its main factory:

  const vis = milestones('#wrapper');

The returned object exposes the following API:

# vis.aggregateBy(interval)

Sets the aggregation interval for the event data, where interval can be one of second, minute, hour, day, week, month, quarter or year.

# vis.distribution(string)

Sets the label distribution, can be top-bottom, top or bottom. Defaults to top-bottom. The options don’t change for vertical layouts. top refers to labels on the left and bottom to labels on the right for that layout.

# vis.mapping(configObject)

Sets overrides for the default attributes for the expected data structure of an event. This defaults to:

  {
    category: undefined,
    entries: undefined,
    timestamp: 'timestamp',  // Used only for time based scales
    value: 'value',          // Used only for ordinal scale values
    text: 'text',
    url: 'url',
    id: 'id'
  };

The method allows you to override single or multiple attributes to map them to fields in your original data with a single call like:

    vis.mapping({
      'timestamp': 'year',
      'text': 'title'
    })

# vis.optimize(boolean)

Enables/Disables the label optimizer. When enabled, the optimizer attempts to avoid label overlap by vertically displacing labels.

# vis.autoResize(boolean)

Enables/Disables auto resizing. Enabled by default, this adds listeners to resizing events of the browser window.

# vis.orientation(string)

Sets the orientation of the timeline, can be either horizontal or vertical. Defaults to horizontal.

# vis.scaleType(string)

Sets the scale type of the timeline, can be either time or ordinal. Defaults to time.

time: Uses a time scale for chronological data with timestamps
ordinal: Uses an ordinal scale for categorical data without timestamps

# vis.parseTime(specifier)

Specifies the formatter for the timestamp field. The specifier string is expected to resemble a format described here: https://github.com/d3/d3-time-format#locale_format

# vis.labelFormat(specifier)

The labelFormat for the time label for each milestones defaults to '%Y-%m-%d %H:%M'. Using aggregateBy, labelFormat will be set automatically to a reasonable format corresponding to the aggregation level. Still, this method is available to override this behavior with a custom labelFormat.

# vis.urlTarget(string)

Customizes the target attribute when labels are provided with a URL. Can be _blank, _self, _parent or _top.

# vis.useLabels(boolean)

Enables/Disables the display of labels.

# vis.render([data])

When called without data this triggers re-rendering the existing visualization.

data is expected to be an array of event objects with fields matching either the expected defaults (timestamp and text attribute) or the provided mapping via .mapping().

# vis.renderCallback(function)

Sets a callback function that is executed after the visualization is fully rendered, allowing you to apply custom styling or modifications to the rendered elements. The callback is called after both initial renders and automatic re-renders due to window resizing.

vis.renderCallback(function() {
  // Apply additional customizations after rendering is complete
  d3.select(".milestones").style("margin-left", "10%");
}).render(data);

# vis.onEventClick(function)

Set a callback which is executed when the text or image is clicked.

  vis.onEventClick((d) => {
    console.log('click', d);
    alert(`
      ${d.text} | ${d.timestamp}
      ${JSON.stringify(d.attributes)}
    `);
  })

# vis.onEventMouseOver(function)

Set a callback which is executed when the mouse cursor is over text or image.

  vis.onEventMouseOver((d) => {
    console.log('mouseover', d);
  })

# vis.onEventMouseLeave(function)

Set a callback which is executed when the mouse cursor is leaving text or image.

  vis.onEventMouseLeave((d) => {
    console.log('mouseleave', d);
  })

Disclaimer on DevOps/CI Observability

This repository uses Elastic APM to track performance of functional tests. It uses @elastic/apm-rum as a devDependency to be run only as part of the tests. No telemetry library is part of the published package.

d3-milestones is also available as a visualization plugin for Kibana here: https://github.com/walterra/kibana-milestones-vis