d3-timeline

A d3 v4 version of d3-timeline

Get something that looks like

for a dataset that looks like

var testData = [
	{label: "person a", times: [
		{"starting_time": 1355752800000, "ending_time": 1355759900000},
		{"starting_time": 1355767900000, "ending_time": 1355774400000}]},
	{label: "person b", times: [
		{"starting_time": 1355759910000, "ending_time": 1355761900000}]},
	{label: "person c", times: [
		{"starting_time": 1355761910000, "ending_time": 1355763910000}]}
	];

with a call that looks like

var chart = d3.timeline();

var svg = d3.select("#timeline1").append("svg").attr("width", 500)
	.datum(testData).call(chart);

Works with circles. In case the rectangular edges are too pointy.

Combine rectangles and circles to your liking

by adding a display key to the data:

var rectAndCircleTestData = [
		{times: [{"starting_time": 1355752800000, "display": "circle"},
						 {"starting_time": 1355767900000, "ending_time": 1355774400000}]},
		{times: [{"starting_time": 1355759910000, "display":"circle"}, ]},
		{times: [{"starting_time": 1355761910000, "ending_time": 1355763910000}]}
	];

Make a pseudo-gantt chart thingy

with icons

For your really long charts, it supports scrolling. It can even do things on hover, click, and scroll for when someone accidentally interacts with your chart.

You can also specify an optional class key in the data dictionary. This will label each timeline rectangle item within the visualization with the following id property: "timelineItem_"+class. For example, this data

var testData = [
	{class: "pA", label: "person a", times: [
		{"starting_time": 1355752800000, "ending_time": 1355759900000},
		{"starting_time": 1355767900000, "ending_time": 1355774400000}]},
	{class: "pB", label: "person b", times: [
		{"starting_time": 1355759910000, "ending_time": 1355761900000}]},
	{class: "pC", label: "person c", times: [
		{"starting_time": 1355761910000, "ending_time": 1355763910000}]}
	];

would generate <rect> with the following classes: timelineItem_pA,timelineItem_pB,timelineItem_pC. This means that you can dynamically change the visual properties of each timeline item using JQuery like so: $(".timelineSeries_pA").css("fill","blue");. If no custom class is provided, the class attribute will be generated sequentially in the order they have been provided in. e.g.: timelineSeries_0.

Also optional is an id field per data element.

var testData = [
	{label: "person a", times: [
		{"starting_time": 1355752800000, "ending_time": 1355759900000, "id": "A1"},
		{"starting_time": 1355767900000, "ending_time": 1355774400000, "id": "A2"}]}
	];

This generates <rect>s with A1 and A2 as ids. If no id is provided, the id attribute will be generated sequentially in the order they have been provided in. e.g.: timelineItem_0_0.

Look at the examples for more details.

Data formats

The simplest data format only requires starting_time and ending_time for each series of data.

[
	{times: [
		{"starting_time": 1355752800000, "ending_time": 1355759900000},
		{"starting_time": 1355767900000, "ending_time": 1355774400000}]},
	{times: [
		{"starting_time": 1355759910000, "ending_time": 1355761900000}]}
];

label can be added if you want names by each series of data. In order for this to properly show up, the timeline needs to be called with .stack()

[
	{label: "person a", times: [
		{"starting_time": 1355752800000, "ending_time": 1355759900000},
		{"starting_time": 1355767900000, "ending_time": 1355774400000}]},
	{label: "person b", times: [
		{"starting_time": 1355759910000, "ending_time": 1355761900000}]}
];

icon can be added if you want icons by each series of data. In order for this to properly show up, the timeline needs to be called with .stack(). Icons and labels can also be mixed in together.

[
	{icon: "path/to/img.png", times: [
		{"starting_time": 1355752800000, "ending_time": 1355759900000},
		{"starting_time": 1355767900000, "ending_time": 1355774400000}]},
	{label: "person b", times: [
		{"starting_time": 1355759910000, "ending_time": 1355761900000}]}
];

'times' elements array

Each item in the times array must have starting_time and ending_time. You could also specify optional color or label elements within a times item, as well as a property mapped to a color.

[
	 {label: "person a", times: [{"color":"green", "label":"Weeee", "starting_time": 1355752800000, "ending_time": 1355759900000}, {"color":"blue", "label":"Weeee", "starting_time": 1355767900000, "ending_time": 1355774400000}]},
	 {label: "person b", times: [{"color":"pink", "label":"Weeee", "starting_time": 1355759910000, "ending_time": 1355761900000}, ]},
	 {label: "person c", times: [{"color":"yellow", "label":"Weeee", "starting_time": 1355761910000, "ending_time": 1355763910000}]}
];

Method Calls

All methods that take in arguments return the current settings if no argument is passed.

.width(width)

sets the width of the timeline. If the width of the timeline is longer than the width of the svg object, the timeline will automatically scroll. The width of the timeline will default to the width of the svg if width is not set.

.height(height)

sets the height of the timeline. The height of the timeline will be automatically calculated from the height of each item if height is not set on the timeline or the svg.

.itemHeight(height)

sets the height of the data series in the timeline. Defaults to 20px.

.itemMargin(height)

sets the margin between the data series in the timeline. Defaults to 5px.

.margin({left: , right: , top: , bottom: })

sets the margin of the entire timeline inside of the svg. Defaults to 30px all around.

.display("circle" | "rect")

Displays the data series as either circles or rectangles. Defaults to "rect".

.labelFormat(callback)

registers a function to be called when the text for the label needs to be generated. Useful if your label looks like this:

{
	en: "my label",
	fr: "mon étiquette"
}

The callback function is passed the whatever the datum.label returns, so in this case it would be the object above. So the labelFormat might look something like this:

.labelFormat(function(label){ return label[currentLocale];})

.tickFormat({format: , tickTime: , tickInterval: , tickSize: , numTicks: , tickValues})

sets the formatting of the ticks in the timeline. Defaults to

{
	format: d3.time.format("%I %p"),
	tickTime: d3.time.hours,
	tickInterval: 1,
	tickSize: 6
}

Tick interval/values can be set with:

• tickTime and tickInterval
• numTicks and tickInterval
• tickValues

.rotateTicks(degrees)

sets the degree of rotation of the tickmarks. Defaults to no rotation (0 degrees).

.orient("bottom" | "top")

sets the placement of the axis. Defaults to bottom.

.colors(callback)

sets the d3 color scale the data series in the timeline. Defaults to d3.scale.category20().

.colorProperty(propertyName)

sets the data item property name that maps your data items to your color scale. For example if you set your chart's colors() and colorsProperty() as follows:

var colorScale = d3.scale.ordinal().range(['#6b0000','#ef9b0f','#ffee00'])
						.domain(['apple','orange','lemon']);

var chart = d3.timeline()
						.colors( colorScale )
						.colorProperty('fruit');

And pass this dataset:

var testData = [
	{label: "fruit 1", fruit: "orange", times: [
		{"starting_time": 1355759910000, "ending_time": 1355761900000}]},
	{label: "fruit 2", fruit: "apple", times: [
		{"starting_time": 1355752800000, "ending_time": 1355759900000},
		{"starting_time": 1355767900000, "ending_time": 1355774400000}]},
	{label: "fruit3", fruit: "lemon", times: [
		{"starting_time": 1355761910000, "ending_time": 1355763910000}]}
	];

Your chart's bar colors will be determined based on the value of the fruit property:

You can also set the color property for a specific time object:

var testData = [
	{label: "fruit 2", fruit: "apple", times: [
		{fruit: "orange", "starting_time": 1355752800000, "ending_time": 1355759900000},
		{"starting_time": 1355767900000, "ending_time": 1355774400000},
		{fruit: "lemon", "starting_time": 1355774400000, "ending_time": 1355775500000}]}
	];

Properties set in the time object will override the property set for the series:

.beginning(date)

sets the time that the timeline should start. If beginning and ending are not set, the timeline will calculate it based off of the smallest and largest times.

.ending(date)

sets the time that the timeline should end. If beginning and ending are not set, the timeline will calculate it based off of the smallest and largest times.

.stack()

Takes in no arguments. Toggles the stacking/unstacking of data series in the timeline. Needs to be true in order for icons and labels to show up properly.

.relativeTime()

Takes in no arguments. Toggles the calculation and use of relative timestamps. The origin of the timeline will be set to 0 and the starting_time of the first data dictionary in the data array will be subtracted from every subsequent timestamp.

.showToday()

Takes in no arguments. Toggles a vertical line showing the current Date.now() time. Uses showTodayFormat for the line formatting.

.showTodayFormat({marginTop: , marginBottom: , width: , color: })

Sets the formatting of the showToday line. Color cycle can also be of the format rgb(x, y, z).

.showBorderLine()

Takes in no arguments. Toggles a vertical line showing the borders of one specific timeline. Uses showBorderFormat for the line formatting.

.showBorderFormat({marginTop: , marginBottom:, width: , color: })

Sets the formatting of the showBorder line. Color cycle can also be of the format rgb(x, y, z).

.showTimeAxis()

Takes in no arguments. Toggles the visibility of the time axis.

.showTimeAxisTick()

Takes in no arguments. Shows tick marks along the X axis according to the arguments for showTimeAxisTickFormat. Useful for datasets with a lot of stacked elements.

.showTimeAxisTickFormat(format)

Format for showTimeAxisTick. Defaults to {stroke: "stroke-dasharray", spacing: "4 10"}.

Defaults to

{
	marginTop: 25,
	marginBottom: 0,
	width: 1,
	color: colorCycle
}

.rowSeparators(color)

Sets the display of horizontal lines betweens rows.

.background(color)

Sets the background of the rows. Useful for creating a continuous effect when there are gaps in your data.

.hover(callback)

takes in a callback called on mousemove of the timeline data. Example

d3.timeline()
	.hover(function (d, i, datum) {
		// d is the current rendering object
		// i is the index during d3 rendering
		// datum is the data object
	});

.xAxisClass

Default is timeline-xAxis, but can be used to set any class.

.axisZoom

Default is true. Takes a boolean. If set to false, does not allow zooming

.mouseover(callback)

takes in a callback called on mouseover of the timeline data. Example

d3.timeline()
	.mouseover(function (d, i, datum) {
		// d is the current rendering object
		// i is the index during d3 rendering
		// datum is the data object
	});

.mouseout(callback)

takes in a callback called on mouseout of the timeline data. Example

d3.timeline()
	.mouseout(function (d, i, datum) {
		// d is the current rendering object
		// i is the index during d3 rendering
		// datum is the data object
	});

.click(callback)

takes in a callback called on click of the timeline data. Example

d3.timeline()
	.click(function (d, i, datum) {
		// d is the current rendering object
		// i is the index during d3 rendering
		// datum is the data object
	});

.scroll(callback)

takes in a callback called on scroll of the timeline data. Example

d3.timeline()
	.scroll(function (x, scale) {
		// x is the current position of the scroll
		// scale is the scale of the axis used
	});

Build

To build:

• npm install
• npm test
• copy the build/d3-timelines.js to dist/d3-timelines.js

Previous work

Inspired by jiahuang - https://github.com/jiahuang/d3-timeline

License

MIT