CHANGES.md 82.8 KB
Newer Older
Mike Bostock's avatar
Mike Bostock committed
1
# Changes in D3 4.0
Mike Bostock's avatar
Mike Bostock committed
2

Mike Bostock's avatar
Mike Bostock committed
3 4 5 6 7 8 9 10
* [Modules](#modules)
* [Arrays](#arrays-d3-array)
* [Axes](#axes-d3-axis)
* [Brushes](#brushes-d3-brush)
* [Collections](#collections-d3-collection)
* [Colors](#colors-d3-color)
* [Dispatches](#dispatches-d3-dispatch)
* [Dragging](#dragging-d3-drag)
Mike Bostock's avatar
Mike Bostock committed
11
* [Delimiter-Separated Values](#delimiter-separated-values-d3-dsv)
Mike Bostock's avatar
Mike Bostock committed
12 13
* [Easings](#easings-d3-ease)
* [Forces](#forces-d3-force)
Mike Bostock's avatar
Mike Bostock committed
14
* [Number Formats](#number-formats-d3-format)
Mike Bostock's avatar
Mike Bostock committed
15
* [Geographies](#geographies-d3-geo)
Mike Bostock's avatar
Mike Bostock committed
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
* [Hierarchies](#hierarchies-d3-hierarchy)
* [Interpolators](#interpolators-d3-interpolate)
* [Paths](#paths-d3-path)
* [Polygons](#polygons-d3-polygon)
* [Quadtrees](#quadtrees-d3-quadtree)
* [Queues](#queues-d3-queue)
* [Random Numbers](#random-numbers-d3-random)
* [Requests](#requests-d3-request)
* [Scales](#scales-d3-scale)
* [Selections](#selections-d3-selection)
* [Shapes](#shapes-d3-shape)
* [Time Formats](#time-formats-d3-time-format)
* [Time Intervals](#time-intervals-d3-time)
* [Timers](#timers-d3-timer)
* [Transitions](#transitions-d3-transition)
* [Voronoi Diagrams](#voronoi-diagrams-d3-voronoi)
* [Zooming](#zooming-d3-zoom)

Mike Bostock's avatar
Mike Bostock committed
34
N.B.: This document is a [work-in-progress](https://github.com/d3/d3/issues/2841). It does not yet cover all API changes.
Mike Bostock's avatar
Mike Bostock committed
35

Mike Bostock's avatar
Mike Bostock committed
36
## Modules
Mike Bostock's avatar
Mike Bostock committed
37

Mike Bostock's avatar
Mike Bostock committed
38
D3 3.x was a monolithic library: the core functionality resided in a single [repository](https://github.com/d3/d3) and was published in a [single file](https://d3js.org/d3.v3.js). It was possible to create a custom build using a [nonstandard tool](https://github.com/mbostock/smash), but not easy and few did. (There were also plugins, but these could only add features and had their own [monolithic repository](https://github.com/d3/d3-plugins).)
Mike Bostock's avatar
Mike Bostock committed
39

Mike Bostock's avatar
Mike Bostock committed
40
D3 4.0 is modular. Instead of one library, D3 is now [many small libraries](https://github.com/d3) that are designed to work together. You can pick and choose which parts to use as you see fit. Each library is maintained in a separate repository, allowing decentralized ownership and independent release cycles. Want to own a new repository in the [D3 organization](https://github.com/d3)? [Let me know!](https://twitter.com/mbostock)
Mike Bostock's avatar
Mike Bostock committed
41

Mike Bostock's avatar
Mike Bostock committed
42
The [default bundle](https://d3js.org/d3.v4.0.0-alpha.50.js) of D3 4.0 conveniently aggregates [about thirty](https://github.com/d3/d3/blob/master/index.js) of these microlibraries.
Mike Bostock's avatar
Mike Bostock committed
43 44

```html
Mike Bostock's avatar
Mike Bostock committed
45
<script src="https://d3js.org/d3.v4.0.0-alpha.50.min.js"></script>
Mike Bostock's avatar
Mike Bostock committed
46
```
Mike Bostock's avatar
Mike Bostock committed
47

Mike Bostock's avatar
Mike Bostock committed
48
As before, you can load optional plugins on top of the default bundle, such as [ColorBrewer scales](https://github.com/d3/d3-scale-chromatic):
Mike Bostock's avatar
Mike Bostock committed
49

Mike Bostock's avatar
Mike Bostock committed
50
```html
Mike Bostock's avatar
Mike Bostock committed
51
<script src="https://d3js.org/d3.v4.0.0-alpha.50.min.js"></script>
Mike Bostock's avatar
Mike Bostock committed
52
<script src="https://d3js.org/d3-scale-chromatic.v0.3.min.js"></script>
Mike Bostock's avatar
Mike Bostock committed
53 54
```

Mike Bostock's avatar
Mike Bostock committed
55
But now in 4.0, you don’t have to use the default bundle! Custom bundles are useful for applications that use a subset of D3’s features; for example, a React charting library might use D3’s scales and shapes, but use React instead of selections to manipulate the DOM. Or if you’re just using [d3-selection](https://github.com/d3/d3-selection), it’s only 5KB instead of 64KB for the default bundle. You can load D3 microlibraries using vanilla script tags or RequireJS (great for HTTP/2!):
Mike Bostock's avatar
Mike Bostock committed
56 57

```html
Mike Bostock's avatar
Mike Bostock committed
58
<script src="https://d3js.org/d3-selection.v0.9.min.js"></script>
Mike Bostock's avatar
Mike Bostock committed
59 60
```

Mike Bostock's avatar
Mike Bostock committed
61
You can `cat` D3 microlibraries into a custom bundle, or use tools such as [Webpack](https://webpack.github.io/) or [Rollup](http://rollupjs.org/) to create [optimized bundles](https://bl.ocks.org/mbostock/bb09af4c39c79cffcde4). The D3 microlibraries are written as [ES6 modules](http://www.2ality.com/2014/09/es6-modules-final.html), and Rollup lets you pick at the symbol level to produce the smallest bundles!
Mike Bostock's avatar
Mike Bostock committed
62

Mike Bostock's avatar
Mike Bostock committed
63
Small files are nice, but modularity is also about making D3 *fun* again. Microlibraries are easier to understand, develop and test. They make it easier for new people to get involved and contribute. They reduce the distinction between a “core module” and a “plugin”, and increase the pace of development in D3 features.
Mike Bostock's avatar
Mike Bostock committed
64

Mike Bostock's avatar
Mike Bostock committed
65
If you don’t care about modularity, you can mostly ignore this change and keep using the default bundle. However, there’s an unavoidable consequence of adopting ES6 modules: every symbol in D3 4.0 now shares a flat namespace rather than the nested one of D3 3.x. For example, d3.scale.linear is now d3.scaleLinear, and d3.layout.treemap is now d3.treemap. The adoption of ES6 modules also means that D3 is now written exclusively in [strict mode](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Strict_mode) and has better readability. And there have been many other significant improvements to D3’s features! (Nearly all of the code from D3 3.x has been rewritten.) These changes are covered in the sections below.
Mike Bostock's avatar
Mike Bostock committed
66

Mike Bostock's avatar
Mike Bostock committed
67
The default D3 [UMD bundle](https://github.com/umdjs/umd) is now [anonymous](https://github.com/requirejs/requirejs/wiki/Updating-existing-libraries#register-as-an-anonymous-module-), rather being named “d3”. No `d3` global is exported if AMD or CommonJS is detected. In a vanilla environment, the D3 microlibraries share the `d3` global, meaning the code you write for the default D3 bundle works identically if you load the modules separately. (See [Let’s Make a (D3) Plugin](https://bost.ocks.org/mike/d3-plugin/) for more.) The generated UMD bundles are no longer stored in the Git repository; Bower has been repointed to [d3-bower](https://github.com/mbostock-bower/d3-bower), and you can find the generated files on [npmcdn](https://npmcdn.com/d3@next/) or attached to the [latest release](https://github.com/d3/d3/releases/latest). The non-minified default bundle is no longer mangled, making it more readable and preserving inline comments.
Mike Bostock's avatar
Mike Bostock committed
68

Mike Bostock's avatar
Mike Bostock committed
69
To the consternation of some users, D3 3.x employed Unicode variable names such as τ and π for a concise representation of mathematical operations. A downside of this approach was that a SyntaxError would occur if you loaded the non-minified D3 using ISO-8859-1 instead of UTF-8. D3 3.x also used Unicode string literals, such as the SI-prefix µ for 1e-6. D3 4.0 uses only ASCII variable names and ASCII string literals (see [rollup-plugin-ascii](https://github.com/mbostock/rollup-plugin-ascii)), avoiding these encoding problems.
Mike Bostock's avatar
Mike Bostock committed
70

Mike Bostock's avatar
Mike Bostock committed
71
## [Arrays (d3-array)](https://github.com/d3/d3-array/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
72

Mike Bostock's avatar
Mike Bostock committed
73
The new [d3.scan](https://github.com/d3/d3-array#scan) method performs a linear scan of an array, returning the index of the least element according to the specified comparator. This is similar to [d3.min](https://github.com/d3/d3-array#min) and [d3.max](https://github.com/d3/d3-array#max), except you can use it to find the position of an extreme element, rather than just calculate an extreme value.
Mike Bostock's avatar
Mike Bostock committed
74 75 76 77 78 79 80 81 82 83 84 85 86

```js
var data = [
  {name: "Alice", value: 2},
  {name: "Bob", value: 3},
  {name: "Carol", value: 1},
  {name: "Dwayne", value: 5}
];

var i = d3.scan(data, function(a, b) { return a.value - b.value; }); // 2
data[i]; // {name: "Carol", value: 1}
```

Mike Bostock's avatar
Mike Bostock committed
87
The new [d3.ticks](https://github.com/d3/d3-array#ticks) and [d3.tickStep](https://github.com/d3/d3-array#tickStep) methods are useful for generating human-readable numeric ticks. These methods are a low-level alternative to [*continuous*.ticks](https://github.com/d3/d3-scale#continuous_ticks) from [d3-scale](https://github.com/d3/d3-scale). The new implementation is also more accurate, returning the optimal number of ticks as measured by relative error.
Mike Bostock's avatar
Mike Bostock committed
88

Mike Bostock's avatar
Mike Bostock committed
89 90 91
```js
var ticks = d3.ticks(0, 10, 5); // [0, 2, 4, 6, 8, 10]
```
Mike Bostock's avatar
Mike Bostock committed
92

Mike Bostock's avatar
Mike Bostock committed
93 94
The [d3.range](https://github.com/d3/d3-array#range) method no longer makes an elaborate attempt to avoid floating-point error when *step* is not an integer. The returned values are strictly defined as *start* + *i* \* *step*, where *i* is an integer. (Learn more about [floating point math](http://0.30000000000000004.com/).) d3.range returns the empty array for infinite ranges, rather than throwing an error.

Mike Bostock's avatar
Mike Bostock committed
95
The method signature for optional accessors has been changed to be more consistent with array methods such as [*array*.forEach](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Array/forEach): the accessor is passed the current element (*d*), the index (*i*), and the array (*data*), with *this* as undefined. This affects [d3.min](https://github.com/d3/d3-array#min), [d3.max](https://github.com/d3/d3-array#max), [d3.extent](https://github.com/d3/d3-array#extent), [d3.sum](https://github.com/d3/d3-array#sum), [d3.mean](https://github.com/d3/d3-array#mean), [d3.median](https://github.com/d3/d3-array#median), [d3.quantile](https://github.com/d3/d3-array#quantile), [d3.variance](https://github.com/d3/d3-array#variance) and [d3.deviation](https://github.com/d3/d3-array#deviation). The [d3.quantile](https://github.com/d3/d3-array#quantile) method previously did not take an accessor. Some methods with optional arguments now treat those arguments as missing if they are null or undefined, rather than strictly checking arguments.length.
Mike Bostock's avatar
Mike Bostock committed
96

Mike Bostock's avatar
Mike Bostock committed
97
The new [d3.histogram](https://github.com/d3/d3-array#histograms) API replaces d3.layout.histogram. Rather than exposing *bin*.x and *bin*.dx on each returned bin, the histogram exposes *bin*.x0 and *bin*.x1, guaranteeing that *bin*.x0 is exactly equal to *bin*.x1 on the preceeding bin. The “frequency” and “probability” modes are no longer supported; each bin is simply an array of elements from the input data, so *bin*.length is equal to D3 3.x’s *bin*.y in frequency mode. To compute a probability distribution, divide the number of elements in each bin by the total number of elements.
Mike Bostock's avatar
Mike Bostock committed
98

Mike Bostock's avatar
Mike Bostock committed
99
The *histogram*.range method has been renamed [*histogram*.domain](https://github.com/d3/d3-array#histogram_domain) for consistency with scales. The *histogram*.bins method has been renamed [*histogram*.thresholds](https://github.com/d3/d3-array#histogram_thresholds), and no longer accepts an upper value: *n* thresholds will produce *n* + 1 bins. If you specify a desired number of bins rather than thresholds, d3.histogram now uses [d3.ticks](https://github.com/d3/d3-array#ticks) to compute nice bin thresholds. In addition to the default Sturges’ formula, D3 now implements the [Freedman-Diaconis rule](https://github.com/d3/d3-array#thresholdFreedmanDiaconis) and [Scott’s normal reference rule](https://github.com/d3/d3-array#thresholdScott).
Mike Bostock's avatar
Mike Bostock committed
100

Mike Bostock's avatar
Mike Bostock committed
101
## [Axes (d3-axis)](https://github.com/d3/d3-axis/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
102

Mike Bostock's avatar
Mike Bostock committed
103
To render axes properly in D3 3.x, you needed to style them:
Mike Bostock's avatar
Mike Bostock committed
104 105 106

```html
<style>
Mike Bostock's avatar
Mike Bostock committed
107

Mike Bostock's avatar
Mike Bostock committed
108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129
.axis path,
.axis line {
  fill: none;
  stroke: #000;
  shape-rendering: crispEdges;
}

.axis text {
  font: 10px sans-serif;
}

</style>
<script>

d3.select(".axis")
    .call(d3.svg.axis()
        .scale(x)
        .orient("bottom"));

</script>
```

Mike Bostock's avatar
Mike Bostock committed
130
If you didn’t, you saw this:
Mike Bostock's avatar
Mike Bostock committed
131 132 133

<img src="https://raw.githubusercontent.com/d3/d3/master/img/axis-v3.png" width="100%" height="105">

Mike Bostock's avatar
Mike Bostock committed
134
D3 4.0 provides default styles and shorter syntax. In place of d3.svg.axis and *axis*.orient, D3 4.0 now provides four constructors for each orientation: [d3.axisTop](https://github.com/d3/d3-axis#axisTop), [d3.axisRight](https://github.com/d3/d3-axis#axisRight), [d3.axisBottom](https://github.com/d3/d3-axis#axisBottom), [d3.axisLeft](https://github.com/d3/d3-axis#axisLeft). These constructors accept a scale, so you can reduce all of the above to:
Mike Bostock's avatar
Mike Bostock committed
135 136 137 138 139 140 141 142 143 144

```html
<script>

d3.select(".axis")
    .call(d3.axisBottom(x));

</script>
```

Mike Bostock's avatar
Mike Bostock committed
145
And get this:
Mike Bostock's avatar
Mike Bostock committed
146 147 148

<img src="https://raw.githubusercontent.com/d3/d3/master/img/axis-v4.png" width="100%" height="105">

Mike Bostock's avatar
Mike Bostock committed
149
As before, you can customize the axis appearance either by applying stylesheets or by modifying the axis elements. The default appearance has been changed slightly to offset the axis by a half-pixel;  this fixes a crisp-edges rendering issue on Safari where the axis would be drawn two-pixels thick.
Mike Bostock's avatar
Mike Bostock committed
150

151
There’s now an [*axis*.tickArguments](https://github.com/d3/d3-axis#axis_tickArguments) method, as an alternative to [*axis*.ticks](https://github.com/d3/d3-axis#axis_ticks) that also allows the axis tick arguments to be inspected. The [*axis*.tickSize](https://github.com/d3/d3-axis#axis_tickSize) method has been changed to only allow a single argument when setting the tick size; use [*axis*.tickSizeInner](https://github.com/d3/d3-axis#axis_tickSizeInner) or [*axis*.tickSizeOuter](https://github.com/d3/d3-axis#axis_tickSizeOuter) to set the inner and outer tick size separately.
Mike Bostock's avatar
Mike Bostock committed
152

Mike Bostock's avatar
Mike Bostock committed
153
## [Brushes (d3-brush)](https://github.com/d3/d3-brush/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
154

Mike Bostock's avatar
Mike Bostock committed
155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173
There are now three classes of brush, replacing d3.svg.brush, for brushing along the *x*-dimension, the *y*-dimension, or both: [d3.brushX](https://github.com/d3/d3-brush#brushX), [d3.brushY](https://github.com/d3/d3-brush#brushY), [d3.brush](https://github.com/d3/d3-brush#brush). Brushes are no longer dependent on [scales](#scales-d3-scale); instead, each brush defines a selection in screen coordinates. This selection can be [inverted](https://github.com/d3/d3-scale#continuous_invert) if you want to compute the corresponding data domain.

Rather than relying on the scales’ ranges to determine the brushable area, there is now a [*brush*.extent](https://github.com/d3/d3-brush#brush_extent) method for setting it. If you do not set the brush extent, the default behavior is to allow brushing over the full extent of the owner SVG element. The *brush*.clamp method has also been eliminated; brushing is always restricted to the brushable area defined by the brush extent.

Brushes no longer store the active brush selection internally. The selection is now stored on any elements to which the brush has been applied. The selection is available as *event*.selection within a brush event, or by calling [d3.brushSelection](https://github.com/d3/d3-brush#brushSelection) on a given *element*. To move the brush to a new position, use [*brush*.move](https://github.com/d3/d3-brush#brush_move) with a given [selection](#selections-d3-selection) or [transition](#transitions-d3-transition); see the [brush snapping example](http://bl.ocks.org/mbostock/6232537).

Brush interaction has been improved. By default, the brush now ignores right-clicks intended for the context menu; you can change this behavior using [*brush*.filter](https://github.com/d3/d3-brush#brush_filter). Holding down SHIFT (⇧) while brushing locks the *x*- or *y*-position of the brush. Holding down META (⌘) while clicking and dragging always starts a new brush selection, rather than dragging the existing brush selection.

The default appearance of the brush has also been improved and slightly simplified. Previously it was necessary to apply styles to the brush to give it a reasonable appearance, such as:

```css
.brush .extent {
  stroke: #fff;
  fill-opacity: .125;
  shape-rendering: crispEdges;
}
```

These styles are now applied by default as attributes; if you want to customize the brush appearance, you can still apply external styles or modify the brush elements. (D3 4.0 features a similar improvement to [axes](#axes-d3-axis).) A new [*brush*.handleSize](https://github.com/d3/d3-brush#brush_handleSize) method lets you override the brush handle size; it defaults to six pixels.
Mike Bostock's avatar
Mike Bostock committed
174

Mike Bostock's avatar
Mike Bostock committed
175
The brush now consumes handled events, making it easier to combine the brush with other interactive behaviors such as [drag](#dragging-d3-drag) and [zoom](#zooming-d3-zoom). The *brushstart* and *brushend* events have been renamed to *start* and *end*, respectively. The brush event no longer reports a *event*.mode to distinguish between resizing and dragging the brush.
Mike Bostock's avatar
Mike Bostock committed
176

Mike Bostock's avatar
Mike Bostock committed
177
## [Chords (d3-chord)](https://github.com/d3/d3-chord/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
178

Mike Bostock's avatar
Mike Bostock committed
179 180
TODO

Mike Bostock's avatar
Mike Bostock committed
181 182
This module is not yet implemented in D3 4.0, but I’m working on it.

Mike Bostock's avatar
Mike Bostock committed
183 184
* d3.svg.chord ↦ d3.ribbon
* d3.layout.chord ↦ d3.chord
Mike Bostock's avatar
Mike Bostock committed
185

Mike Bostock's avatar
Mike Bostock committed
186
## [Collections (d3-collection)](https://github.com/d3/d3-collection/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
187

Mike Bostock's avatar
Mike Bostock committed
188
The [d3.set](https://github.com/d3/d3-collection#set) constructor now accepts an existing set for making a copy. If you pass an array to d3.set, you can also pass a value accessor. This accessor takes the standard arguments: the current element (*d*), the index (*i*), and the array (*data*), with *this* undefined. For example:
Mike Bostock's avatar
Mike Bostock committed
189 190 191 192 193 194 195 196 197 198 199 200

```js
var yields = [
  {yield: 22.13333, variety: "Manchuria",        year: 1932, site: "Grand Rapids"},
  {yield: 26.76667, variety: "Peatland",         year: 1932, site: "Grand Rapids"},
  {yield: 28.10000, variety: "No. 462",          year: 1931, site: "Duluth"},
  {yield: 38.50000, variety: "Svansota",         year: 1932, site: "Waseca"},
  {yield: 40.46667, variety: "Svansota",         year: 1931, site: "Crookston"},
  {yield: 36.03333, variety: "Peatland",         year: 1932, site: "Waseca"},
  {yield: 34.46667, variety: "Wisconsin No. 38", year: 1931, site: "Grand Rapids"}
];

Mike Bostock's avatar
Mike Bostock committed
201
var sites = d3.set(yields, function(d) { return d.site; }); // Grand Rapids, Duluth, Waseca, Crookston
Mike Bostock's avatar
Mike Bostock committed
202 203
```

204
The [d3.map](https://github.com/d3/d3-collection#map) constructor also follows the standard array accessor argument pattern.
Mike Bostock's avatar
Mike Bostock committed
205

Mike Bostock's avatar
Mike Bostock committed
206
The *map*.forEach and *set*.forEach methods have been renamed to [*map*.each](https://github.com/d3/d3-collection#map_each) and [*set*.each](https://github.com/d3/d3-collection#set_each) respectively. The order of arguments for *map*.each has also been changed to *value*, *key* and *map*, while the order of arguments for *set*.each is now *value*, *value* and *set*. This is closer to ES6 [*map*.forEach](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Map/forEach) and [*set*.forEach](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Set/forEach). Also like ES6 Map and Set, *map*.set and *set*.add now return the current collection (rather than the added value) to facilitate method chaining. New [*map*.clear](https://github.com/d3/d3-collection#map_clear) and [*set*.clear](https://github.com/d3/d3-collection#set_clear) methods can be used to empty collections.
Mike Bostock's avatar
Mike Bostock committed
207

Mike Bostock's avatar
Mike Bostock committed
208
The [*nest*.map](https://github.com/d3/d3-collection#nest_map) method now always returns a d3.map instance. For a plain object, use [*nest*.object](https://github.com/d3/d3-collection#nest_object) instead. When used in conjunction with [*nest*.rollup](https://github.com/d3/d3-collection#nest_rollup), [*nest*.entries](https://github.com/d3/d3-collection#nest_entries) now returns {key, value} objects for the leaf entries, instead of {key, values}. This makes *nest*.rollup easier to use in conjunction with [hierarchies](#hierarchies-d3-hierarchy), as in this [Nest Treemap example](http://bl.ocks.org/mbostock/2838bf53e0e65f369f476afd653663a2).
Mike Bostock's avatar
Mike Bostock committed
209

Mike Bostock's avatar
Mike Bostock committed
210
## [Colors (d3-color)](https://github.com/d3/d3-color/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
211

Mike Bostock's avatar
Mike Bostock committed
212
All colors now have opacity exposed as *color*.opacity, which is a number in [0, 1]. You can pass an optional opacity argument to the color space constructors [d3.rgb](https://github.com/d3/d3-color#rgb), [d3.hsl](https://github.com/d3/d3-color#hsl), [d3.lab](https://github.com/d3/d3-color#lab), [d3.hcl](https://github.com/d3/d3-color#hcl) or [d3.cubehelix](https://github.com/d3/d3-color#cubehelix).
Mike Bostock's avatar
Mike Bostock committed
213

Mike Bostock's avatar
Mike Bostock committed
214
You can now parse rgba(…) and hsla(…) CSS color specifiers or the string “transparent” using [d3.color](https://github.com/d3/d3-color#color). The “transparent” color is defined as an RGB color with zero opacity and undefined red, green and blue channels; this differs slightly from CSS which defines it as transparent black, but is useful for simplifying color interpolation logic where either the starting or ending color has undefined channels. The [*color*.toString](https://github.com/d3/d3-color#color_toString) method now likewise returns an rgb(…) or rgba(…) string with integer channel values, not the hexadecimal RGB format, consistent with CSS computed values. This improves performance by short-circuiting transitions when the element’s starting style matches its ending style.
Mike Bostock's avatar
Mike Bostock committed
215

Mike Bostock's avatar
Mike Bostock committed
216 217 218 219 220 221 222
The new [d3.color](https://github.com/d3/d3-color#color) method is the primary method for parsing colors: it returns a d3.color instance in the appropriate color space, or null if the CSS color specifier is invalid. For example:

```js
var red = d3.color("hsl(0, 80%, 50%)"); // {h: 0, l: 0.5, s: 0.8, opacity: 1}
```

The parsing implementation is now more robust. For example, you can no longer mix integers and percentages in rgb(…), and it correctly handles whitespace, decimal points, number signs, and other edge cases. The color space constructors d3.rgb, d3.hsl, d3.lab, d3.hcl and d3.cubehelix now always return a copy of the input color, converted to the corresponding color space. While [*color*.rgb](https://github.com/d3/d3-color#color_rgb) remains, *rgb*.hsl has been removed; use d3.hsl to convert a color to the RGB color space.
Mike Bostock's avatar
Mike Bostock committed
223

Mike Bostock's avatar
Mike Bostock committed
224
The RGB color space no longer greedily quantizes and clamps channel values when creating colors, improving accuracy in color space conversion. Quantization and clamping now occurs in *color*.toString when formatting a color for display. You can use the new [*color*.displayable](https://github.com/d3/d3-color#color_displayable) to test whether a color is [out-of-gamut](https://en.wikipedia.org/wiki/Gamut).
Mike Bostock's avatar
Mike Bostock committed
225

Mike Bostock's avatar
Mike Bostock committed
226
The [*rgb*.brighter](https://github.com/d3/d3-color#rgb_brighter) method no longer special-cases black. This is a multiplicative operator, defining a new color *r*′, *g*′, *b*′ where *r*′ = *r* × *pow*(0.7, *k*), *g*′ = *g* × *pow*(0.7, *k*) and *b*′ = *b* × *pow*(0.7, *k*); a brighter black is still black.
Mike Bostock's avatar
Mike Bostock committed
227

Mike Bostock's avatar
Mike Bostock committed
228
There’s a new [d3.cubehelix](https://github.com/d3/d3-color#cubehelix) color space, generalizing Dave Green’s color scheme! (See also [d3.interpolateCubehelixDefault](https://github.com/d3/d3-scale#interpolateCubehelixDefault) from [d3-scale](#scales-d3-scale).) You can continue to define your own custom color spaces, too; see [d3-hsv](https://github.com/d3/d3-hsv) for an example.
Mike Bostock's avatar
Mike Bostock committed
229

Mike Bostock's avatar
Mike Bostock committed
230
## [Dispatches (d3-dispatch)](https://github.com/d3/d3-dispatch/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
231

Mike Bostock's avatar
Mike Bostock committed
232
Rather than decorating the *dispatch* object with each event type, the dispatch object now exposes generic [*dispatch*.call](https://github.com/d3/d3-dispatch#dispatch_call) and [*dispatch*.apply](https://github.com/d3/d3-dispatch#dispatch_apply) methods which take the *type* string as the first argument. For example, in D3 3.x, you might say:
Mike Bostock's avatar
Mike Bostock committed
233 234 235 236 237

```js
dispatcher.foo.call(that, "Hello, Foo!");
```

Mike Bostock's avatar
Mike Bostock committed
238
To dispatch a *foo* event in D3 4.0, you’d say:
Mike Bostock's avatar
Mike Bostock committed
239 240 241 242 243 244 245 246 247 248 249 250 251 252

```js
dispatcher.call("foo", that, "Hello, Foo!");
```

The [*dispatch*.on](https://github.com/d3/d3-dispatch#dispatch_on) method now accepts multiple typenames, allowing you to add or remove listeners for multiple events simultaneously. For example, to send both *foo* and *bar* events to the same listener:

```js
dispatcher.on("foo bar", function(message) {
  console.log(message);
});
```

This matches the new behavior of [*selection*.on](https://github.com/d3/d3-selection#selection_on) in [d3-selection](#selections-d3-selection). The *dispatch*.on method now validates that the specifier *listener* is a function, rather than throwing an error in the future.
Mike Bostock's avatar
Mike Bostock committed
253

Mike Bostock's avatar
Mike Bostock committed
254
The new implementation d3.dispatch is faster, using fewer closures to improve performance. There’s also a new [*dispatch*.copy](https://github.com/d3/d3-dispatch#dispatch_copy) method for making a copy of a dispatcher; this is used by [d3-transition](#transitions-d3-transition) to improve the performance of transitions in the common case where all elements in a transition have the same transition event listeners.
Mike Bostock's avatar
Mike Bostock committed
255

Mike Bostock's avatar
Mike Bostock committed
256
## [Dragging (d3-drag)](https://github.com/d3/d3-drag/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
257

Mike Bostock's avatar
Mike Bostock committed
258 259
TODO

Mike Bostock's avatar
Mike Bostock committed
260
* d3.behavior.drag ↦ d3.drag
Mike Bostock's avatar
Mike Bostock committed
261 262 263 264 265 266 267 268 269
* add *drag*.filter
* add *drag*.subject - for Canvas-based dragging
* add *drag*.container - for Canvas-based dragging, or avoiding feedback loop
* ignore emulated mouse events on iOS
* *dragstart* event ↦ *start* event
* *dragend* event ↦ *end* event
* add d3.dragEnable, d3.dragDisable - dealing with browser quirks
* new *event*.active property makes it easier to tell if any active gesture
* new *event*.on lets you register temporary listeners for the current gesture
Mike Bostock's avatar
Mike Bostock committed
270

Mike Bostock's avatar
Mike Bostock committed
271 272
## [Delimiter-Separated Values (d3-dsv)](https://github.com/d3/d3-dsv/blob/master/README.md)

Mike Bostock's avatar
Mike Bostock committed
273
Pursuant to the great namespace flattening, various CSV and TSV methods have new names:
Mike Bostock's avatar
Mike Bostock committed
274

Mike Bostock's avatar
Mike Bostock committed
275 276 277 278 279 280 281 282
* d3.csv.parse ↦ [d3.csvParse](https://github.com/d3/d3-dsv#csvParse)
* d3.csv.parseRows ↦ [d3.csvParseRows](https://github.com/d3/d3-dsv#csvParseRows)
* d3.csv.format ↦ [d3.csvFormat](https://github.com/d3/d3-dsv#csvFormat)
* d3.csv.formatRows ↦ [d3.csvFormatRows](https://github.com/d3/d3-dsv#csvFormatRows)
* d3.tsv.parse ↦ [d3.tsvParse](https://github.com/d3/d3-dsv#tsvParse)
* d3.tsv.parseRows ↦ [d3.tsvParseRows](https://github.com/d3/d3-dsv#tsvParseRows)
* d3.tsv.format ↦ [d3.tsvFormat](https://github.com/d3/d3-dsv#tsvFormat)
* d3.tsv.formatRows ↦ [d3.tsvFormatRows](https://github.com/d3/d3-dsv#tsvFormatRows)
Mike Bostock's avatar
Mike Bostock committed
283

Mike Bostock's avatar
Mike Bostock committed
284
The [d3.csv](https://github.com/d3/d3-request#csv) and [d3.tsv](https://github.com/d3/d3-request#tsv) methods for loading files of the corresponding formats have not been renamed, however! Those are defined in [d3-request](#requests-d3-request).There’s no longer a d3.dsv method, which served the triple purpose of defining a DSV formatter, a DSV parser and a DSV requestor; instead, there’s just [d3.dsvFormat](https://github.com/d3/d3-dsv#dsvFormat) which you can use to define a DSV formatter and parser. You can use [*request*.response](https://github.com/d3/d3-request#request_response) to make a request and then parse the response body, or just use [d3.text](https://github.com/d3/d3-request#text).
Mike Bostock's avatar
Mike Bostock committed
285 286 287 288

The [*dsv*.parse](https://github.com/d3/d3-dsv#dsv_parse) method now exposes the column names and their input order as *data*.columns. For example:

```js
Mike Bostock's avatar
Mike Bostock committed
289
d3.csv("cars.csv", function(error, data) {
Mike Bostock's avatar
Mike Bostock committed
290
  if (error) throw error;
Mike Bostock's avatar
Mike Bostock committed
291
  console.log(data.columns); // ["Year", "Make", "Model", "Length"]
Mike Bostock's avatar
Mike Bostock committed
292 293 294 295 296 297 298 299 300
});
```

You can likewise pass an optional array of column names to [*dsv*.format](https://github.com/d3/d3-dsv#dsv_format) to format only a subset of columns, or to specify the column order explicitly:

```js
var string = d3.csvFormat(data, ["Year", "Model", "Length"]);
```

Mike Bostock's avatar
Mike Bostock committed
301
The parser is a bit faster and the formatter is a bit more robust: inputs are coerced to strings before formatting, fixing an obscure crash, and deprecated support for falling back to [*dsv*.formatRows](https://github.com/d3/d3-dsv#dsv_formatRows) when the input *data* is an array of arrays has been removed.
Mike Bostock's avatar
Mike Bostock committed
302

Mike Bostock's avatar
Mike Bostock committed
303
## [Easings (d3-ease)](https://github.com/d3/d3-ease/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
304

Mike Bostock's avatar
Mike Bostock committed
305
D3 3.x used strings, such as “cubic-in-out”, to identify easing methods; these strings could be passed to d3.ease or *transition*.ease. D3 4.0 uses symbols instead, such as [d3.easeCubicInOut](https://github.com/d3/d3-ease#easeCubicInOut). Symbols are simpler and cleaner. They work well with Rollup to produce smaller custom bundles. You can still define your own custom easing function, too, if desired. Here’s the full list of equivalents:
Mike Bostock's avatar
Mike Bostock committed
306

Mike Bostock's avatar
Mike Bostock committed
307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347
* linear ↦ [d3.easeLinear](https://github.com/d3/d3-ease#easeLinear)¹
* linear-in ↦ [d3.easeLinear](https://github.com/d3/d3-ease#easeLinear)¹
* linear-out ↦ [d3.easeLinear](https://github.com/d3/d3-ease#easeLinear)¹
* linear-in-out ↦ [d3.easeLinear](https://github.com/d3/d3-ease#easeLinear)¹
* linear-out-in ↦ [d3.easeLinear](https://github.com/d3/d3-ease#easeLinear)¹
* poly-in ↦ [d3.easePolyIn](https://github.com/d3/d3-ease#easePolyIn)
* poly-out ↦ [d3.easePolyOut](https://github.com/d3/d3-ease#easePolyOut)
* poly-in-out ↦ [d3.easePolyInOut](https://github.com/d3/d3-ease#easePolyInOut)
* poly-out-in ↦ REMOVED²
* quad-in ↦ [d3.easeQuadIn](https://github.com/d3/d3-ease#easeQuadIn)
* quad-out ↦ [d3.easeQuadOut](https://github.com/d3/d3-ease#easeQuadOut)
* quad-in-out ↦ [d3.easeQuadInOut](https://github.com/d3/d3-ease#easeQuadInOut)
* quad-out-in ↦ REMOVED²
* cubic-in ↦ [d3.easeCubicIn](https://github.com/d3/d3-ease#easeCubicIn)
* cubic-out ↦ [d3.easeCubicOut](https://github.com/d3/d3-ease#easeCubicOut)
* cubic-in-out ↦ [d3.easeCubicInOut](https://github.com/d3/d3-ease#easeCubicInOut)
* cubic-out-in ↦ REMOVED²
* sin-in ↦ [d3.easeSinIn](https://github.com/d3/d3-ease#easeSinIn)
* sin-out ↦ [d3.easeSinOut](https://github.com/d3/d3-ease#easeSinOut)
* sin-in-out ↦ [d3.easeSinInOut](https://github.com/d3/d3-ease#easeSinInOut)
* sin-out-in ↦ REMOVED²
* exp-in ↦ [d3.easeExpIn](https://github.com/d3/d3-ease#easeExpIn)
* exp-out ↦ [d3.easeExpOut](https://github.com/d3/d3-ease#easeExpOut)
* exp-in-out ↦ [d3.easeExpInOut](https://github.com/d3/d3-ease#easeExpInOut)
* exp-out-in ↦ REMOVED²
* circle-in ↦ [d3.easeCircleIn](https://github.com/d3/d3-ease#easeCircleIn)
* circle-out ↦ [d3.easeCircleOut](https://github.com/d3/d3-ease#easeCircleOut)
* circle-in-out ↦ [d3.easeCircleInOut](https://github.com/d3/d3-ease#easeCircleInOut)
* circle-out-in ↦ REMOVED²
* elastic-in ↦ [d3.easeElasticOut](https://github.com/d3/d3-ease#easeElasticOut)²
* elastic-out ↦ [d3.easeElasticIn](https://github.com/d3/d3-ease#easeElasticIn)²
* elastic-in-out ↦ REMOVED²
* elastic-out-in ↦ [d3.easeElasticInOut](https://github.com/d3/d3-ease#easeElasticInOut)²
* back-in ↦ [d3.easeBackIn](https://github.com/d3/d3-ease#easeBackIn)
* back-out ↦ [d3.easeBackOut](https://github.com/d3/d3-ease#easeBackOut)
* back-in-out ↦ [d3.easeBackInOut](https://github.com/d3/d3-ease#easeBackInOut)
* back-out-in ↦ REMOVED²
* bounce-in ↦ [d3.easeBounceOut](https://github.com/d3/d3-ease#easeBounceOut)²
* bounce-out ↦ [d3.easeBounceIn](https://github.com/d3/d3-ease#easeBounceIn)²
* bounce-in-out ↦ REMOVED²
* bounce-out-in ↦ [d3.easeBounceInOut](https://github.com/d3/d3-ease#easeBounceInOut)²
Mike Bostock's avatar
Mike Bostock committed
348

Mike Bostock's avatar
Mike Bostock committed
349 350
¹ The -in, -out and -in-out variants of linear easing are identical, so there’s just d3.easeLinear.
<br>² Elastic and bounce easing were inadvertently reversed in 3.x, so 4.0 eliminates -out-in easing!
Mike Bostock's avatar
Mike Bostock committed
351

Mike Bostock's avatar
Mike Bostock committed
352
For convenience, there are also default aliases for each easing method. For example, [d3.easeCubic](https://github.com/d3/d3-ease#easeCubic) is an alias for [d3.easeCubicInOut](https://github.com/d3/d3-ease#easeCubicInOut). Most default to -in-out; the exceptions are [d3.easeBounce](https://github.com/d3/d3-ease#easeBounce) and [d3.easeElastic](https://github.com/d3/d3-ease#easeElastic), which default to -out.
Mike Bostock's avatar
Mike Bostock committed
353

Mike Bostock's avatar
Mike Bostock committed
354
Rather than pass optional arguments to d3.ease or *transition*.ease, parameterizable easing functions now have named parameters: [*poly*.exponent](https://github.com/d3/d3-ease#poly_exponent), [*elastic*.amplitude](https://github.com/d3/d3-ease#elastic_amplitude), [*elastic*.period](https://github.com/d3/d3-ease#elastic_period) and [*back*.overshoot](https://github.com/d3/d3-ease#back_overshoot). For example, in D3 3.x you might say:
Mike Bostock's avatar
Mike Bostock committed
355

Mike Bostock's avatar
Mike Bostock committed
356 357 358 359 360 361 362 363 364
```js
var e = d3.ease("elastic-out-in", 1.2);
```

The equivalent in D3 4.0 is:

```js
var e = d3.easeElastic.amplitude(1.2);
```
Mike Bostock's avatar
Mike Bostock committed
365

Mike Bostock's avatar
Mike Bostock committed
366
Many of the easing functions have been optimized for performance and accuracy. Several bugs have been fixed, as well, such as the interpretation of the overshoot parameter for back easing, and the period parameter for elastic easing. Also, [d3-transition](#transitions-d3-transition) now explicitly guarantees that the last tick of the transition happens at exactly *t* = 1, avoiding floating point errors in some easing functions.
Mike Bostock's avatar
Mike Bostock committed
367

Mike Bostock's avatar
Mike Bostock committed
368
There’s now a nice [visual reference](https://github.com/d3/d3-ease/blob/master/README.md) and an [animated reference](http://bl.ocks.org/mbostock/248bac3b8e354a9103c4) to the new easing functions, too!
Mike Bostock's avatar
Mike Bostock committed
369

Mike Bostock's avatar
Mike Bostock committed
370
## [Forces (d3-force)](https://github.com/d3/d3-force/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
371

Mike Bostock's avatar
Mike Bostock committed
372 373
TODO

Mike Bostock's avatar
Mike Bostock committed
374 375
* velocity verlet instead of position verlet
* deterministic initialization and forces; D3 does not play dice!
Mike Bostock's avatar
Mike Bostock committed
376
* d3.layout.force ↦ d3.forceSimulation
Mike Bostock's avatar
Mike Bostock committed
377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412
* *force*.friction ↦ *force*.drag

the simulation is extensible rather than hard-coding several forces:

* *force*.gravity ↦ d3.forceX, d3.forceY
* *force*.charge ↦ d3.forceManyBody
* *force*.link ↦ d3.forceLink
* new d3.forceCenter
* new d3.forceCollision - more stable than prior examples, too
* new *forceManyBody*.distanceMin

the new forces are more flexible, and better:

* force strengths can typically be configured per-node or per-link
* separate positioning forces for *x* and *y*
* better default link strength and bias heuristics to improve stability

easier controls for heating and cooling the layout:

* *simulation*.alpha
* *simulation*.alphaMin - control when the internal timer stops
* *simulation*.alphaDecay - control how quickly the simulation cools
* *simulation*.alphaTarget - smooth reheating during interaction!

better controls for starting and stopping the internal timer, independent of heat:

* *simulation*.restart
* *simulation*.stop
* *simulation*.tick

the dependency on the drag behavior is removed. instead:

* *simulation*.fix
* *simulation*.unfix
* *simulation*.unfixAll
* *simulation*.find
Mike Bostock's avatar
Mike Bostock committed
413

Mike Bostock's avatar
Mike Bostock committed
414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437
## [Number Formats (d3-format)](https://github.com/d3/d3-format/blob/master/README.md)

TODO

* treat negative zero (-0) and very small numbers that round to zero as unsigned zero
* the `c` directive is now for character data (i.e., literals), not for character codes
* the `b` and `d` directives now round to the nearest integer rather than returning the empty string
* new `(` sign option uses parentheses for negative values
* new `=` align option places any sign and symbol to the left of any padding
* improve accuracy by relying on *number*.toExponential to extract the mantissa and exponent
* locales are now published as JSON data; can load from npmcdn.com if desired

changed the behavior for default precision. now 6 for all directives except none, which defaults to 12. *none* is a new directive type that is like `g` except it trims insignificant trailing zeros. d3.round and d3.requote are now removed.

new methods for computing the suggested decimal precision for formatting values (used by d3-scale for tick formatting)

* d3.precisionFixed
* d3.precisionPrefix
* d3.precisionRound

new d3.formatSpecifier method for parsing, validating and debugging format specifiers. also good for deriving related format specifiers, such as when you want to set the precision automatically.

quite a few more changes… TODO describe them

Mike Bostock's avatar
Mike Bostock committed
438
## [Geographies (d3-geo)](https://github.com/d3/d3-geo/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
439

Mike Bostock's avatar
Mike Bostock committed
440 441
TODO

Mike Bostock's avatar
Mike Bostock committed
442
* d3.geo.graticule ↦ d3.geoGraticule
Mike Bostock's avatar
Mike Bostock committed
443 444 445 446
* *graticule*.majorExtent ↦ *graticule*.extentMajor
* *graticule*.minorExtent ↦ *graticule*.extentMinor
* *graticule*.majorStep ↦ *graticule*.stepMajor
* *graticule*.minorStep ↦ *graticule*.stepMinor
Mike Bostock's avatar
Mike Bostock committed
447
* d3.geo.circle ↦ d3.geoCircle
Mike Bostock's avatar
Mike Bostock committed
448 449
* *circle*.origin ↦ *circle*.center
* *circle*.angle ↦ *circle*.radius
Mike Bostock's avatar
Mike Bostock committed
450 451 452 453 454 455 456 457
* d3.geo.area ↦ d3.geoArea
* d3.geo.bounds ↦ d3.geoBounds
* d3.geo.centroid ↦ d3.geoCentroid
* d3.geo.distance ↦ d3.geoDistance
* d3.geo.interpolate ↦ d3.geoInterpolate
* d3.geo.length ↦ d3.geoLength
* d3.geo.rotation ↦ d3.geoRotation
* d3.geo.stream ↦ d3.geoStream
Mike Bostock's avatar
Mike Bostock committed
458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477
* d3.geo.path ↦ d3.geoPath
* d3.geo.projection ↦ d3.geoProjection
* d3.geo.projectionMutator ↦ d3.geoProjectionMutator?

Projections:

* d3.geo.albers ↦ d3.geoAlbers
* d3.geo.albersUsa ↦ d3.geoAlbersUsa
* d3.geo.azimuthalEqualArea ↦ d3.geoAzimuthalEqualArea
* d3.geo.azimuthalEquidistant ↦ d3.geoAzimuthalEquidistant
* d3.geo.conicConformal ↦ d3.geoConicConformal
* d3.geo.conicEqualArea ↦ d3.geoConicEqualArea
* d3.geo.conicEquidistant ↦ d3.geoConicEquidistant
* d3.geo.equirectangular ↦ d3.geoEquirectangular
* d3.geo.gnomonic ↦ d3.geoGnomonic
* d3.geo.mercator ↦ d3.geoMercator
* d3.geo.orthographic ↦ d3.geoOrthographic
* d3.geo.stereographic ↦ d3.geoStereographic
* d3.geo.transverseMercator ↦ d3.geoTransverseMercator

Mike Bostock's avatar
Mike Bostock committed
478
TODO Raw projections?
Mike Bostock's avatar
Mike Bostock committed
479 480 481 482 483 484 485 486 487 488 489 490 491 492 493

* d3.geo.azimuthalEqualArea.raw ↦ d3.geoRawAzimuthalEqualArea
* d3.geo.azimuthalEquidistant.raw ↦ d3.geoRawAzimuthalEquidistant
* d3.geo.conicConformal.raw ↦ d3.geoRawConicConformal
* d3.geo.conicEqualArea.raw ↦ d3.geoRawConicEqualArea
* d3.geo.conicEquidistant.raw ↦ d3.geoRawConicEquidistant
* d3.geo.equirectangular.raw ↦ d3.geoRawEquirectangular
* d3.geo.gnomonic.raw ↦ d3.geoRawGnomonic
* d3.geo.mercator.raw ↦ d3.geoRawMercator
* d3.geo.orthographic.raw ↦ d3.geoRawOrthographic
* d3.geo.stereographic.raw ↦ d3.geoRawStereographic
* d3.geo.transverseMercator.raw ↦ d3.geoRawTransverseMercator

A new d3.geoPipeline API is in development for D3 5.0.

Mike Bostock's avatar
Mike Bostock committed
494
## [Hierarchies (d3-hierarchy)](https://github.com/d3/d3-hierarchy/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
495

Mike Bostock's avatar
Mike Bostock committed
496 497
TODO

Mike Bostock's avatar
Mike Bostock committed
498 499 500 501 502 503
* d3.layout.cluster ↦ d3.cluster
* d3.layout.hierarchy ↦ d3.hierarchy
* d3.layout.pack ↦ d3.pack
* d3.layout.partition ↦ d3.partition
* d3.layout.tree ↦ d3.tree
* d3.layout.treemap ↦ d3.treemap
Mike Bostock's avatar
Mike Bostock committed
504 505 506 507 508 509 510 511 512 513
* new d3.stratify API for converting tabular data to hierarchies!
* new d3.hierarchy API for working with hierarchical data!
* new *treemap*.tile - the treemap tiling algorithms are now extensible
* reimplemented squarified treemaps, fixing bugs with padding and rounding
* reimplemented circle-packing layout, fixing major bugs and improving results
* new d3.treemapBinary for binary treemaps
* new d3.packSiblings for circle-packing (non-hierarchical circles)
* new d3.packEnclose uses Welzl’s algorithm to compute the exact enclosing circle
* *treemap*.sticky ↦ d3.treemapResquarify
* new treemap padding parameters, distinguishing parent and sibling padding
Mike Bostock's avatar
Mike Bostock committed
514 515
* new nested treemap example
* new treemap + d3.nest example
Mike Bostock's avatar
Mike Bostock committed
516 517
* new partition padding parameter
* space-filling layouts now output *x0*, *x1*, *y0*, *y1* instead of *x0*, *dx*, *y0*, *dy*; better accuracy
Mike Bostock's avatar
Mike Bostock committed
518
* d3.layout.bundle ↦ *node*.path
Mike Bostock's avatar
Mike Bostock committed
519
* see d3.curveBundle in d3-shape for hierarchical edge bundling
Mike Bostock's avatar
Mike Bostock committed
520

Mike Bostock's avatar
Mike Bostock committed
521
## [Interpolators (d3-interpolate)](https://github.com/d3/d3-interpolate/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
522

Mike Bostock's avatar
Mike Bostock committed
523
The [d3.interpolate](https://github.com/d3/d3-interpolate#interpolate) method no longer delegates to d3.interpolators, which has been removed; its behavior is now defined by the library. It is now slightly faster in the common case that *b* is a number. It only uses [d3.interpolateRgb](https://github.com/d3/d3-interpolate#interpolateRgb) if *b* is a valid CSS color specifier (and not approximately one). And if the end value *b* is null, undefined, true or false, d3.interpolate now returns a constant function which always returns *b*.
Mike Bostock's avatar
Mike Bostock committed
524 525

The behavior of [d3.interpolateObject](https://github.com/d3/d3-interpolate#interpolateObject) and [d3.interpolateArray](https://github.com/d3/d3-interpolate#interpolateArray) has changed slightly with respect to properties or elements in the start value *a* that do not exist in the end value *b*: these properties and elements are now ignored, such that the ending value of the interpolator at *t* = 1 is now precisely equal to *b*. So, in 3.x:
Mike Bostock's avatar
Mike Bostock committed
526

Mike Bostock's avatar
Mike Bostock committed
527
```js
Mike Bostock's avatar
Mike Bostock committed
528
d3.interpolateObject({foo: 2, bar: 1}, {foo: 3})(0.5); // {bar: 1, foo: 2.5} in 3.x
Mike Bostock's avatar
Mike Bostock committed
529
```
Mike Bostock's avatar
Mike Bostock committed
530

Mike Bostock's avatar
Mike Bostock committed
531
Whereas in 4.0, *a*.bar is ignored:
Mike Bostock's avatar
Mike Bostock committed
532

Mike Bostock's avatar
Mike Bostock committed
533
```js
Mike Bostock's avatar
Mike Bostock committed
534
d3.interpolateObject({foo: 2, bar: 1}, {foo: 3})(0.5); // {foo: 2.5} in 4.0
Mike Bostock's avatar
Mike Bostock committed
535
```
Mike Bostock's avatar
Mike Bostock committed
536

Mike Bostock's avatar
Mike Bostock committed
537
If *a* or *b* are undefined or not an object, they are now implicitly converted to the empty object or empty array as appropriate, rather than throwing a TypeError.
Mike Bostock's avatar
Mike Bostock committed
538

Mike Bostock's avatar
Mike Bostock committed
539
The d3.interpolateTransform interpolator has been renamed to [d3.interpolateTransformSvg](https://github.com/d3/d3-interpolate#interpolateTransformSvg), and there is a new [d3.interpolateTransformCss](https://github.com/d3/d3-interpolate#interpolateTransformCss) to interpolate CSS transforms! This allows [d3-transition](#transitions-d3-transition) to automatically interpolate both the SVG [transform attribute](https://www.w3.org/TR/SVG/coords.html#TransformAttribute) and the CSS [transform style property](https://www.w3.org/TR/css-transforms-1/#transform-property). (Note, however, that only 2D CSS transforms are supported.) The d3.transform method has been removed.
Mike Bostock's avatar
Mike Bostock committed
540

Mike Bostock's avatar
Mike Bostock committed
541
Color space interpolators now interpolate opacity (see [d3-color](#colors-d3-color)) and return rgb(…) or rgba(…) CSS color specifier strings rather than using the RGB hexadecimal format. This is necessary to support opacity interpolation, but is also beneficial because it matches CSS computed values. When a channel in the start color *a* is undefined, color interpolators now use the corresponding channel value from the end color *b*, or *vice versa*. This logic previously applied to some channels (such as saturation in HSL), but now applies to all channels in all color spaces, and is especially useful when interpolating to or from transparent.
Mike Bostock's avatar
Mike Bostock committed
542

Mike Bostock's avatar
Mike Bostock committed
543
There are now “long” versions of cylindrical color space interpolators: [d3.interpolateHslLong](https://github.com/d3/d3-interpolate#interpolateHslLong), [d3.interpolateHclLong](https://github.com/d3/d3-interpolate#interpolateHclLong) and [d3.interpolateCubehelixLong](https://github.com/d3/d3-interpolate#interpolateCubehelixLong). These interpolators use linear interpolation of hue, rather than using the shortest path around the 360° hue circle. See [d3.interpolateRainbow](https://github.com/d3/d3-scale#interpolateRainbow) for an example. The Cubehelix color space is now supported by [d3-color](#colors-d3-color), and so there are now [d3.interpolateCubehelix](https://github.com/d3/d3-interpolate#interpolateCubehelix) and [d3.interpolateCubehelixLong](https://github.com/d3/d3-interpolate#interpolateCubehelixLong) interpolators.
Mike Bostock's avatar
Mike Bostock committed
544

Mike Bostock's avatar
Mike Bostock committed
545
[Gamma-corrected color interpolation](https://web.archive.org/web/20160112115812/http://www.4p8.com/eric.brasseur/gamma.html) is now supported for both RGB and Cubehelix color spaces as [*interpolate*.gamma](https://github.com/d3/d3-interpolate#interpolate_gamma). For example, to interpolate from purple to orange with a gamma of 2.2 in RGB space:
Mike Bostock's avatar
Mike Bostock committed
546

Mike Bostock's avatar
Mike Bostock committed
547 548 549
```js
var interpolate = d3.interpolateRgb.gamma(2.2)("purple", "orange");
```
Mike Bostock's avatar
Mike Bostock committed
550

Mike Bostock's avatar
Mike Bostock committed
551
There are new interpolators for uniform non-rational [B-splines](https://en.wikipedia.org/wiki/B-spline)! These are useful for smoothly interpolating between an arbitrary sequence of values from *t* = 0 to *t* = 1, such as to generate a smooth color gradient from a discrete set of colors. The [d3.interpolateBasis](https://github.com/d3/d3-interpolate#interpolateBasis) and [d3.interpolateBasisClosed](https://github.com/d3/d3-interpolate#interpolateBasisClosed) interpolators generate one-dimensional B-splines, while [d3.interpolateRgbBasis](https://github.com/d3/d3-interpolate#interpolateRgbBasis) and [d3.interpolateRgbBasisClosed](https://github.com/d3/d3-interpolate#interpolateRgbBasisClosed) generate three-dimensional B-splines through RGB color space. These are used by [d3-scale-chromatic](https://github.com/d3/d3-scale-chromatic) to generate continuous color scales from ColorBrewer’s discrete color schemes, such as [PiYG](http://bl.ocks.org/mbostock/048d21cf747371b11884f75ad896e5a5).
Mike Bostock's avatar
Mike Bostock committed
552

Mike Bostock's avatar
Mike Bostock committed
553
There’s also now a [d3.quantize](https://github.com/d3/d3-interpolate#quantize) method for generating uniformly-spaced discrete samples from a continuous interpolator. This is useful for taking one of the built-in color scales (such as [d3.interpolateViridis](https://github.com/d3/d3-scale#interpolateViridis)) and quantizing it for use with [d3.scaleQuantize](https://github.com/d3/d3-scale#scaleQuantize), [d3.scaleQuantile](https://github.com/d3/d3-scale#scaleQuantile) or [d3.scaleThreshold](https://github.com/d3/d3-scale#scaleThreshold).
Mike Bostock's avatar
Mike Bostock committed
554

Mike Bostock's avatar
Mike Bostock committed
555
## [Paths (d3-path)](https://github.com/d3/d3-path/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
556

Mike Bostock's avatar
Mike Bostock committed
557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572
The [d3.path](https://github.com/d3/d3-path#path) serializer implements the [CanvasPathMethods API](https://www.w3.org/TR/2dcontext/#canvaspathmethods), allowing you to write code that can render to either Canvas or SVG. For example, given some code that draws to a canvas:

```js
function drawCircle(context, radius) {
  context.moveTo(radius, 0);
  context.arc(0, 0, radius, 0, 2 * Math.PI);
}
```

You can render to SVG as follows:

```js
var context = d3.path();
drawCircle(context, 40);
pathElement.setAttribute("d", context.toString());
```
Mike Bostock's avatar
Mike Bostock committed
573

Mike Bostock's avatar
Mike Bostock committed
574
The path serializer enables [d3-shape](#shapes-d3-shape) to support both Canvas and SVG; see [*line*.context](https://github.com/d3/d3-shape#line_context) and [*area*.context](https://github.com/d3/d3-shape#area_context), for example.
Mike Bostock's avatar
Mike Bostock committed
575 576

## [Polygons (d3-polygon)](https://github.com/d3/d3-polygon/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
577

Mike Bostock's avatar
Mike Bostock committed
578
There’s no longer a d3.geom.polygon constructor; instead you just pass an array of vertices to the polygon methods. So instead of *polygon*.area and *polygon*.centroid, there’s [d3.polygonArea](https://github.com/d3/d3-polygon#polygonArea) and [d3.polygonCentroid](https://github.com/d3/d3-polygon#polygonCentroid). There are also new [d3.polygonContains](https://github.com/d3/d3-polygon#polygonContains) and [d3.polygonLength](https://github.com/d3/d3-polygon#polygonLength) methods. There’s no longer an equivalent to *polygon*.clip, but it [Sutherland–Hodgman clipping](https://en.wikipedia.org/wiki/Sutherland–Hodgman_algorithm) is needed, please [file a feature request](https://github.com/d3/d3-polygon/issues).
Mike Bostock's avatar
Mike Bostock committed
579

Mike Bostock's avatar
Mike Bostock committed
580
The d3.geom.hull operator has been simplified: instead of an operator with *hull*.x and *hull*.y accessors, there’s just the [d3.polygonHull](https://github.com/d3/d3-polygon#polygonHull) method which takes an array of points and returns the convex hull.
Mike Bostock's avatar
Mike Bostock committed
581

Mike Bostock's avatar
Mike Bostock committed
582
## [Quadtrees (d3-quadtree)](https://github.com/d3/d3-quadtree/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
583

Mike Bostock's avatar
Mike Bostock committed
584
The d3.geom.quadtree method has been replaced by [d3.quadtree](https://github.com/d3/d3-quadtree#quadtree). 4.0 removes the concept of quadtree “generators” (configurable functions that build a quadtree from an array of data); there are now just quadtrees, which you can create via d3.quadtree and add data to via [*quadtree*.add](https://github.com/d3/d3-quadtree#quadtree_add) and [*quadtree*.addAll](https://github.com/d3/d3-quadtree#quadtree_addAll). This code in 3.x:
Mike Bostock's avatar
Mike Bostock committed
585 586 587 588 589 590 591 592 593 594 595 596 597 598 599

```js
var quadtree = d3.geom.quadtree()
    .extent([[0, 0], [width, height]])
    (data);
```

Can be rewritten in 4.0 as:

```js
var quadtree = d3.quadtree()
    .extent([[0, 0], [width, height]])
    .addAll(data);
```

Mike Bostock's avatar
Mike Bostock committed
600
The new quadtree implementation is vastly improved! It is no longer recursive, avoiding stack overflows when there are large numbers of coincident points. The internal storage is now more efficient, and the implementation is also faster; constructing a quadtree of 1M normally-distributed points takes about one second in 4.0, as compared to three seconds in 3.x.
Mike Bostock's avatar
Mike Bostock committed
601 602 603 604 605 606 607 608 609 610 611 612 613

The change in [internal *node* structure](https://github.com/d3/d3-quadtree#nodes) affects [*quadtree*.visit](https://github.com/d3/d3-quadtree#quadtree_visit): use *node*.length to distinguish leaf nodes from internal nodes. For example, to iterate over all data in a quadtree:

```js
quadtree.visit(function(node) {
  if (!node.length) {
    do {
      console.log(node.data);
    } while (node = node.next)
  }
});
```

Mike Bostock's avatar
Mike Bostock committed
614
There’s a new [*quadtree*.visitAfter](https://github.com/d3/d3-quadtree#quadtree_visitAfter) method for visiting nodes in post-order traversal. This feature is used in [d3-force](#forces-d3-force) to implement the [Barnes–Hut approximation](https://en.wikipedia.org/wiki/Barnes–Hut_simulation).
Mike Bostock's avatar
Mike Bostock committed
615

Mike Bostock's avatar
Mike Bostock committed
616
You can now remove data from a quadtree using [*quadtree*.remove](https://github.com/d3/d3-quadtree#quadtree_remove) and [*quadtree*.removeAll](https://github.com/d3/d3-quadtree#quadtree_removeAll). When adding data to a quadtree, the quadtree will now expand its extent by repeated doubling if the new point is outside the existing extent of the quadtree. There are also [*quadtree*.extent](https://github.com/d3/d3-quadtree#quadtree_extent) and [*quadtree*.cover](https://github.com/d3/d3-quadtree#quadtree_cover) methods for explicitly expanding the extent of the quadtree after creation.
Mike Bostock's avatar
Mike Bostock committed
617

Mike Bostock's avatar
Mike Bostock committed
618
Quadtrees support several new utility methods: [*quadtree*.copy](https://github.com/d3/d3-quadtree#quadtree_copy) returns a copy of the quadtree sharing the same data; [*quadtree*.data](https://github.com/d3/d3-quadtree#quadtree_data) generates an array of all data in the quadtree; [*quadtree*.size](https://github.com/d3/d3-quadtree#quadtree_size) returns the number of data points in the quadtree; and [*quadtree*.root](https://github.com/d3/d3-quadtree#quadtree_root) returns the root node, which is useful for manual traversal of the quadtree. The [*quadtree*.find](https://github.com/d3/d3-quadtree#quadtree_find) method now takes an optional search radius, which is useful for pointer-based selection in [force-directed graphs](http://bl.ocks.org/mbostock/ad70335eeef6d167bc36fd3c04378048).
Mike Bostock's avatar
Mike Bostock committed
619

Mike Bostock's avatar
Mike Bostock committed
620 621 622 623
## [Queues (d3-queue)](https://github.com/d3/d3-queue/blob/master/README.md)

Formerly known as Queue.js and queue-async, [d3.queue](https://github.com/d3/d3-queue) is now included in the default bundle, making it easy to load data files in parallel. It has been rewritten with fewer closures to improve performance, and there are now stricter checks in place to guarantee well-defined behavior. You can now use instanceof d3.queue and inspect the queue’s internal private state.

Mike Bostock's avatar
Mike Bostock committed
624
## [Random Numbers (d3-random)](https://github.com/d3/d3-random/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
625

Mike Bostock's avatar
Mike Bostock committed
626 627 628 629 630 631
Pursuant to the great namespace flattening, the random number generators have new names:

* d3.random.normal ↦ [d3.randomNormal](https://github.com/d3/d3-random#randomNormal)
* d3.random.logNormal ↦ [d3.randomLogNormal](https://github.com/d3/d3-random#randomLogNormal)
* d3.random.bates ↦ [d3.randomBates](https://github.com/d3/d3-random#randomBates)
* d3.random.irwinHall ↦ [d3.randomIrwinHall](https://github.com/d3/d3-random#randomIrwinHall)
Mike Bostock's avatar
Mike Bostock committed
632

Mike Bostock's avatar
Mike Bostock committed
633
There are also new random number generators for [exponential](https://github.com/d3/d3-random#randomExponential) and [uniform](https://github.com/d3/d3-random#randomUniform) distributions. The [normal](https://github.com/d3/d3-random#randomNormal) and [log-normal](https://github.com/d3/d3-random#randomLogNormal) random generators have been optimized.
Mike Bostock's avatar
Mike Bostock committed
634

Mike Bostock's avatar
Mike Bostock committed
635
## [Requests (d3-request)](https://github.com/d3/d3-request/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
636

Mike Bostock's avatar
Mike Bostock committed
637 638 639
The d3.xhr method has been renamed to [d3.request](https://github.com/d3/d3-request#request). Basic authentication is now supported using [*request*.user](https://github.com/d3/d3-request#request_user) and [*request*.password](https://github.com/d3/d3-request#request_password). You can now configure a timeout using [*request*.timeout](https://github.com/d3/d3-request#request_timeout).

If an error occurs, the corresponding [ProgressEvent](https://xhr.spec.whatwg.org/#interface-progressevent) of type “error” is now passed to the error listener, rather than the [XMLHttpRequest](https://xhr.spec.whatwg.org/#interface-xmlhttprequest). Likewise, the ProgressEvent is passed to progress event listeners, rather than using [d3.event](https://github.com/d3/d3-selection#event). If [d3.xml](https://github.com/d3/d3-request#xml) encounters an error parsing XML, this error is now reported to error listeners rather than returning a null response.
Mike Bostock's avatar
Mike Bostock committed
640

Mike Bostock's avatar
Mike Bostock committed
641
With the exception of [d3.html](https://github.com/d3/d3-request#html) and [d3.xml](https://github.com/d3/d3-request#xml), Node is now supported via [node-XMLHttpRequest](https://github.com/driverdan/node-XMLHttpRequest).
Mike Bostock's avatar
Mike Bostock committed
642

Mike Bostock's avatar
Mike Bostock committed
643
## [Scales (d3-scale)](https://github.com/d3/d3-scale/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
644

Mike Bostock's avatar
Mike Bostock committed
645 646
TODO

Mike Bostock's avatar
Mike Bostock committed
647 648
[Introducing d3-scale](https://medium.com/@mbostock/introducing-d3-scale-61980c51545f#.d38uz6vuz)

Mike Bostock's avatar
Mike Bostock committed
649 650 651 652 653 654 655 656 657 658
* d3.scale.linear ↦ d3.scaleLinear
* d3.scale.sqrt ↦ d3.scaleSqrt
* d3.scale.pow ↦ d3.scalePow
* d3.scale.log ↦ d3.scaleLog
* d3.scale.quantize ↦ d3.scaleQuantize
* d3.scale.threshold ↦ d3.scaleThreshold
* d3.scale.quantile ↦ d3.scaleQuantile
* d3.scale.identity ↦ d3.scaleIdentity
* d3.scale.ordinal ↦ d3.scaleOrdinal
* d3.time.scale ↦ d3.scaleTime
Mike Bostock's avatar
Mike Bostock committed
659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676
* d3.time.scale.utc ↦ d3.scaleUtc
* quantitative scales generate ticks in the same order as the domain
* non-linear quantitative scales are more accurate
* better *log*.ticks filtering for large domains
* *time*.ticks and *time*.nice now only accept time intervals
* new d3.scaleSequential
  * new Viridis, Inferno, Magma, Plasma interpolators
  * new Warm, Cool, Rainbow interpolators
  * new default Cubehelix interpolator
* d3.scaleOrdinal constructor now takes an optional *range*
* new *ordinal*.unknown
* new d3.scaleBand, *band*.bandwidth, *band*.align
* new d3.scalePoint
* category scales now defined as arrays of colors:
  * d3.scale.category10 ↦ d3.schemeCategory10
  * d3.scale.category20 ↦ d3.schemeCategory20
  * d3.scale.category20b ↦ d3.schemeCategory20b
  * d3.scale.category20c ↦ d3.schemeCategory20c
Mike Bostock's avatar
Mike Bostock committed
677

Mike Bostock's avatar
Mike Bostock committed
678 679
Mention d3-scale-chromatic?

Mike Bostock's avatar
Mike Bostock committed
680
## [Selections (d3-selection)](https://github.com/d3/d3-selection/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
681

Mike Bostock's avatar
Mike Bostock committed
682
Selections no longer subclass Array using [prototype chain injection](http://perfectionkills.com/how-ecmascript-5-still-does-not-allow-to-subclass-an-array/#wrappers_prototype_chain_injection); they are now plain objects, improving performance. The internal fields (*selection*.\_groups, *selection*.\_parents) are private; please use the documented public API to manipulate selections. The new [*selection*.nodes](https://github.com/d3/d3-selection#selection_nodes) method generates an array of all nodes in a selection.
Mike Bostock's avatar
Mike Bostock committed
683

Mike Bostock's avatar
Mike Bostock committed
684
Selections are now immutable: the elements and parents in a selection never change. (The elements’ attributes and content will of course still be modified!) The [*selection*.sort](https://github.com/d3/d3-selection#selection_sort) and [*selection*.data](https://github.com/d3/d3-selection#selection_data) methods now return new selections rather than modifying the selection in-place. In addition, [*selection*.append](https://github.com/d3/d3-selection#selection_append) no longer merges entering nodes into the update selection; use [*selection*.merge](https://github.com/d3/d3-selection#selection_merge) to combine enter and update after a data join. For example, the following [general update pattern](http://bl.ocks.org/mbostock/a8a5baa4c4a470cda598) in 3.x:
Mike Bostock's avatar
Mike Bostock committed
685 686 687 688 689 690 691

```js
var circle = svg.selectAll("circle").data(data) // UPDATE
    .style("fill", "blue");

circle.exit().remove(); // EXIT

Mike Bostock's avatar
Mike Bostock committed
692
circle.enter().append("circle") // ENTER; modifies UPDATE! 🌶
Mike Bostock's avatar
Mike Bostock committed
693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712
    .style("fill", "green");

circle // ENTER + UPDATE
    .style("stroke", "black");
```

Would be rewritten in 4.0 as:

```js
var circle = svg.selectAll("circle").data(data) // UPDATE
    .style("fill", "blue");

circle.exit().remove(); // EXIT

circle.enter().append("circle") // ENTER
    .style("fill", "green")
  .merge(circle) // ENTER + UPDATE
    .style("stroke", "black");
```

Mike Bostock's avatar
Mike Bostock committed
713
This change is discussed further in [What Makes Software Good](https://medium.com/@mbostock/what-makes-software-good-943557f8a488#.4ukdnxqiz).
Mike Bostock's avatar
Mike Bostock committed
714

Mike Bostock's avatar
Mike Bostock committed
715
In 3.x, the [*selection*.enter](https://github.com/d3/d3-selection#selection_enter) and [*selection*.exit](https://github.com/d3/d3-selection#selection_exit) methods were undefined until you called *selection*.data, resulting in a TypeError if you attempted to access them. In 4.0, now they simply return the empty selection if the selection has not been joined to data.
Mike Bostock's avatar
Mike Bostock committed
716

Mike Bostock's avatar
Mike Bostock committed
717
In 3.x, [*selection*.append](https://github.com/d3/d3-selection#selection_append) would always append the new element as the last child of its parent. A little-known trick was to use [*selection*.insert](https://github.com/d3/d3-selection#selection_insert) without specifying a *before* selector when entering nodes, causing the entering nodes to be inserted before the following element in the update selection. In 4.0, this is now the default behavior of *selection*.append; if you do not specify a *before* selector to *selection*.insert, the inserted element is appended as the last child. This change makes the general update pattern preserve the relative order of elements and data. For example, given the following DOM:
Mike Bostock's avatar
Mike Bostock committed
718 719

```html
Mike Bostock's avatar
Mike Bostock committed
720 721 722
<div>a</div>
<div>b</div>
<div>f</div>
Mike Bostock's avatar
Mike Bostock committed
723 724 725 726 727 728
```

And the following code:

```js
var div = d3.select("body").selectAll("div")
Mike Bostock's avatar
Mike Bostock committed
729
  .data(["a", "b", "c", "d", "e", "f"], function(d) { return d || this.textContent; });
Mike Bostock's avatar
Mike Bostock committed
730 731

div.enter().append("div")
Mike Bostock's avatar
Mike Bostock committed
732
    .text(function(d) { return d; });
Mike Bostock's avatar
Mike Bostock committed
733 734 735 736 737
```

The resulting DOM will be:

```html
Mike Bostock's avatar
Mike Bostock committed
738 739 740 741 742 743
<div>a</div>
<div>b</div>
<div>c</div>
<div>d</div>
<div>e</div>
<div>f</div>
Mike Bostock's avatar
Mike Bostock committed
744 745
```

Mike Bostock's avatar
Mike Bostock committed
746 747
Thus, the entering *c*, *d* and *e* are inserted before *f*, since *f* is the following element in the update selection. Although this behavior is sufficient to preserve order if the new data’s order is stable, if the data changes order, you must still use [*selection*.order](https://github.com/d3/d3-selection#selection_order) to reorder elements.

Mike Bostock's avatar
Mike Bostock committed
748
There is now only one class of selection. 3.x implemented enter selections using a special class with different behavior for *enter*.append and *enter*.select; a consequence of this design was that enter selections in 3.x lacked [certain methods](https://github.com/d3/d3/issues/2043). In 4.0, enter selections are simply normal selections; they have the same methods and the same behavior. [Enter nodes](https://github.com/d3/d3-selection/blob/master/src/selection/enter.js) now implement [*node*.appendChild](https://developer.mozilla.org/en-US/docs/Web/API/Node/appendChild), [*node*.insertBefore](https://developer.mozilla.org/en-US/docs/Web/API/Node/insertBefore), [*node*.querySelector](https://developer.mozilla.org/en-US/docs/Web/API/Element/querySelector), and [*node*.querySelectorAll](https://developer.mozilla.org/en-US/docs/Web/API/Element/querySelectorAll).
Mike Bostock's avatar
Mike Bostock committed
749

Mike Bostock's avatar
Mike Bostock committed
750
The [*selection*.data](https://github.com/d3/d3-selection#selection_data) method has been changed slightly with respect to duplicate keys. In 3.x, if multiple data had the same key, the duplicate data would be ignored and not included in enter, update or exit; in 4.0 the duplicate data is always put in the enter selection. In both 3.x and 4.0, if multiple elements have the same key, the duplicate elements are put in the exit selection. Thus, 4.0’s behavior is now symmetric for enter and exit, and the general update pattern will now produce a DOM that matches the data even if there are duplicate keys.
Mike Bostock's avatar
Mike Bostock committed
751

Mike Bostock's avatar
Mike Bostock committed
752
Selections have several new methods! Use [*selection*.raise](https://github.com/d3/d3-selection#selection_raise) to move the selected elements to the front of their siblings, so that they are drawn on top; use [*selection*.lower](https://github.com/d3/d3-selection#selection_lower) to move them to the back. Use [*selection*.dispatch](https://github.com/d3/d3-selection#selection_dispatch) to dispatch a [custom event](https://developer.mozilla.org/en-US/docs/Web/API/CustomEvent) to event listeners.
Mike Bostock's avatar
Mike Bostock committed
753

Mike Bostock's avatar
Mike Bostock committed
754
When called in getter mode, [*selection*.data](https://github.com/d3/d3-selection#selection_data) now returns the data for all elements in the selection, rather than just the data for the first group of elements. The [*selection*.call](https://github.com/d3/d3-selection#selection_call) method no longer sets the `this` context when invoking the specified function; the *selection* is passed as the first argument to the function, so use that. The [*selection*.on](https://github.com/d3/d3-selection#selection_on) method now accepts multiple whitespace-separated typenames, so you can add or remove multiple listeners simultaneously. For example:
Mike Bostock's avatar
Mike Bostock committed
755

Mike Bostock's avatar
Mike Bostock committed
756 757 758 759 760
```js
selection.on("mousedown touchstart", function() {
  console.log(d3.event.type);
});
```
Mike Bostock's avatar
Mike Bostock committed
761

Mike Bostock's avatar
Mike Bostock committed
762
The arguments passed to callback functions has changed slightly in 4.0 to be more consistent. The standard arguments are the element’s datum (*d*), the element’s index (*i*), and the element’s group (*nodes*), with *this* as the element. The slight exception to this convention is *selection*data, which is evaluated for each group rather than each element; it is passed the group’s parent datum (*d*), the group index (*i*), and the selection’s parents (*parents*), with *this* as the group’s parent.
Mike Bostock's avatar
Mike Bostock committed
763

Mike Bostock's avatar
Mike Bostock committed
764
The new [d3.local](https://github.com/d3/d3-selection#local-variables) provides a mechanism for defining [local variables](http://bl.ocks.org/mbostock/e1192fe405703d8321a5187350910e08): state that is bound to DOM elements, and available to any descendant element. This can be a convenient alternative to using [*selection*.each](https://github.com/d3/d3-selection#selection_each) or storing local state in data.
Mike Bostock's avatar
Mike Bostock committed
765

Mike Bostock's avatar
Mike Bostock committed
766
The d3.ns.prefix namespace prefix map has been renamed to [d3.namespaces](https://github.com/d3/d3-selection#namespaces), and the d3.ns.qualify method has been renamed to [d3.namespace](https://github.com/d3/d3-selection#namespace). Several new low-level methods are now available, as well. [d3.matcher](https://github.com/d3/d3-selection#matcher) is used internally by [*selection*.filter](https://github.com/d3/d3-selection#selection_filter); [d3.selector](https://github.com/d3/d3-selection#selector) is used by [*selection*.select](https://github.com/d3/d3-selection#selection_select); [d3.selectorAll](https://github.com/d3/d3-selection#selectorAll) is used by [*selection*.selectAll](https://github.com/d3/d3-selection#selection_selectAll); [d3.creator](https://github.com/d3/d3-selection#creator) is used by [*selection*.append](https://github.com/d3/d3-selection#selection_append) and [*selection*.insert](https://github.com/d3/d3-selection#selection_insert). The new [d3.window](https://github.com/d3/d3-selection#window) returns the owner window for a given element, window or document. The new [d3.customEvent](https://github.com/d3/d3-selection#customEvent) temporarily sets [d3.event](https://github.com/d3/d3-selection#event) while invoking a function, allowing you to implement controls which dispatch custom events; this is used by [d3-drag](https://github.com/d3/d3-drag), [d3-zoom](https://github.com/d3/d3-zoom) and [d3-brush](https://github.com/d3/d3-brush).
Mike Bostock's avatar
Mike Bostock committed
767

Mike Bostock's avatar
Mike Bostock committed
768
For the sake of parsimony, the multi-value map methods have been extracted to [d3-selection-multi](https://github.com/d3/d3-selection-multi) and are no longer part of the default bundle. The multi-value map methods have also been renamed to plural form to reduce overload: [*selection*.attrs](https://github.com/d3/d3-selection-multi#selection_attrs), [*selection*.styles](https://github.com/d3/d3-selection-multi#selection_styles) and [*selection*.properties](https://github.com/d3/d3-selection-multi#selection_properties).
Mike Bostock's avatar
Mike Bostock committed
769

Mike Bostock's avatar
Mike Bostock committed
770
## [Shapes (d3-shape)](https://github.com/d3/d3-shape/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
771

Mike Bostock's avatar
Mike Bostock committed
772 773
TODO

Mike Bostock's avatar
Mike Bostock committed
774 775
[Introducing d3-shape](https://medium.com/@mbostock/introducing-d3-shape-73f8367e6d12#.78ucz4hr2)

Mike Bostock's avatar
Mike Bostock committed
776 777 778 779 780
* d3.svg.line ↦ d3.line
* d3.svg.line.radial ↦ d3.radialLine
* d3.svg.area ↦ d3.area
* d3.svg.area.radial ↦ d3.radialArea
* d3.svg.arc ↦ d3.arc
Mike Bostock's avatar
Mike Bostock committed
781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823
* more robust arc padding?

shapes can now render to canvas!

* *line*.context
* *area*.context
* *arc*.context
* see also d3-path
* fast; uses streaming geometry transforms similar to d3-geo

new curve API!

* *line*.interpolate ↦ *line*.curve
* *area*.interpolate ↦ *area*.curve
* "basis" ↦ d3.curveBasis
* "basis-closed" ↦ d3.curveBasisClosed
* "basis-open" ↦ d3.curveBasisOpen
* "bundle" ↦ d3.curveBundle, *bundle*.beta
* default bundle β to 0.85
* "cardinal" ↦ d3.curveCardinal
* "cardinal-closed" ↦ d3.curveCardinalClosed
* "cardinal-open" ↦ d3.curveCardinalOpen
* fixed interpretation of cardinal spline tension
* fixed first and last segment of basis curve
* fixed first and last segment of cardinal curve
* new d3.curveCatmullRom!
* new d3.curveCatmullRomClosed!
* new d3.curveCatmullRomOpen!
* *catmullRom*.alpha implements Yuksel et al.’s parameterization
* defaults to centripetal Catmull–Rom
* "linear" ↦ d3.curveLinear
* "linear-closed" ↦ d3.curveLinearClosed
* fixed monotone curve implementation
* "monotone" ↦ d3.curveMonotoneX
* new d3.curveMonotoneY
* new d3.curveNatural
* "step" ↦ d3.curveStep
* "step-after" ↦ d3.curveStepAfter
* "step-before" ↦ d3.curveStepBefore
* no more funky *interpolate*.reverse; curves can define different behavior for topline vs. baseline

new symbol API

Mike Bostock's avatar
Mike Bostock committed
824 825
* d3.svg.symbol ↦ d3.symbol
* d3.svg.symbolTypes ↦ d3.symbolTypes
Mike Bostock's avatar
Mike Bostock committed
826 827 828 829 830 831 832 833 834 835 836 837 838
* new d3.symbolStar
* new d3.symbolWye
* "triangle-up" ↦ d3.symbolTriangle
* removed "triangle-down"

new stack API!

* d3.layout.stack ↦ d3.stack
* no more x-accessor
* no more weird *stack*.out

removed diagonal shapes

Mike Bostock's avatar
Mike Bostock committed
839 840
* d3.svg.diagonal ↦ REMOVED
* d3.svg.diagonal.radial ↦ REMOVED
Mike Bostock's avatar
Mike Bostock committed
841

Mike Bostock's avatar
Mike Bostock committed
842
## [Time Formats (d3-time-format)](https://github.com/d3/d3-time-format/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
843

Mike Bostock's avatar
Mike Bostock committed
844 845 846 847 848 849
Pursuant to the great namespace flattening, the format constructors have new names:

* d3.time.format ↦ [d3.timeFormat](https://github.com/d3/d3-time-format#timeFormat)
* d3.time.format.utc ↦ [d3.utcFormat](https://github.com/d3/d3-time-format#utcFormat)
* d3.time.format.iso ↦ [d3.isoFormat](https://github.com/d3/d3-time-format#isoFormat)

Mike Bostock's avatar
Mike Bostock committed
850
The *format*.parse method has also been removed in favor of separate [d3.timeParse](https://github.com/d3/d3-time-format#timeParse), [d3.utcParse](https://github.com/d3/d3-time-format#utcParse) and [d3.isoParse](https://github.com/d3/d3-time-format#isoParse) parser constructors. Thus, this code in 3.x:
Mike Bostock's avatar
Mike Bostock committed
851 852 853 854 855

```js
var parseTime = d3.time.format("%c").parse;
```

Mike Bostock's avatar
Mike Bostock committed
856
Can be rewritten in 4.0 as:
Mike Bostock's avatar
Mike Bostock committed
857 858 859 860 861 862

```js
var parseTime = d3.timeParse("%c");
```

The multi-scale time format d3.time.format.multi has been replaced by [d3.scaleTime](https://github.com/d3/d3-scale#scaleTime)’s [tick format](https://github.com/d3/d3-scale#time_tickFormat). Time formats now coerce inputs to dates, and time parsers coerce inputs to strings. The `%Z` directive now allows more flexible parsing of time zone offsets, such as `-0700`, `-07:00`, `-07`, and `Z`. The `%p` directive is now parsed correctly when the locale’s period name is longer than two characters (*e.g.*, “a.m.”).
Mike Bostock's avatar
Mike Bostock committed
863

Mike Bostock's avatar
Mike Bostock committed
864
The performance of time formatting and parsing has been improved, and the UTC formatter and parser have a cleaner implementation (that avoids temporarily overriding the Date global). The locales are now published as [JSON](https://github.com/d3/d3-request#json) to [NPM](https://npmcdn.com/d3-time-format/locale/).
Mike Bostock's avatar
Mike Bostock committed
865

Mike Bostock's avatar
Mike Bostock committed
866
## [Time Intervals (d3-time)](https://github.com/d3/d3-time/blob/master/README.md)
Mike Bostock's avatar
Mike Bostock committed
867

Mike Bostock's avatar
Mike Bostock committed
868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885
Pursuant to the great namespace flattening, the local time intervals have been renamed:

* ADDED ↦ [d3.timeMillisecond](https://github.com/d3/d3-time#timeMillisecond)
* d3.time.second ↦ [d3.timeSecond](https://github.com/d3/d3-time#timeSecond)
* d3.time.minute ↦ [d3.timeMinute](https://github.com/d3/d3-time#timeMinute)
* d3.time.hour ↦ [d3.timeHour](https://github.com/d3/d3-time#timeHour)
* d3.time.day ↦ [d3.timeDay](https://github.com/d3/d3-time#timeDay)
* d3.time.sunday ↦ [d3.timeSunday](https://github.com/d3/d3-time#timeSunday)
* d3.time.monday ↦ [d3.timeMonday](https://github.com/d3/d3-time#timeMonday)
* d3.time.tuesday ↦ [d3.timeTuesday](https://github.com/d3/d3-time#timeTuesday)
* d3.time.wednesday ↦ [d3.timeWednesday](https://github.com/d3/d3-time#timeWednesday)
* d3.time.thursday ↦ [d3.timeThursday](https://github.com/d3/d3-time#timeThursday)
* d3.time.friday ↦ [d3.timeFriday](https://github.com/d3/d3-time#timeFriday)
* d3.time.saturday ↦ [d3.timeSaturday](https://github.com/d3/d3-time#timeSaturday)
* d3.time.week ↦ [d3.timeWeek](https://github.com/d3/d3-time#timeWeek)
* d3.time.month ↦ [d3.timeMonth](https://github.com/d3/d3-time#timeMonth)
* d3.time.year ↦ [d3.timeYear](https://github.com/d3/d3-time#timeYear)

Mike Bostock's avatar
Mike Bostock committed
886
The UTC time intervals have likewise been renamed:
Mike Bostock's avatar
Mike Bostock committed
887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903

* ADDED ↦ [d3.utcMillisecond](https://github.com/d3/d3-time#utcMillisecond)
* d3.time.second.utc ↦ [d3.utcSecond](https://github.com/d3/d3-time#utcSecond)
* d3.time.minute.utc ↦ [d3.utcMinute](https://github.com/d3/d3-time#utcMinute)
* d3.time.hour.utc ↦ [d3.utcHour](https://github.com/d3/d3-time#utcHour)
* d3.time.day.utc ↦ [d3.utcDay](https://github.com/d3/d3-time#utcDay)
* d3.time.sunday.utc ↦ [d3.utcSunday](https://github.com/d3/d3-time#utcSunday)
* d3.time.monday.utc ↦ [d3.utcMonday](https://github.com/d3/d3-time#utcMonday)
* d3.time.tuesday.utc ↦ [d3.utcTuesday](https://github.com/d3/d3-time#utcTuesday)
* d3.time.wednesday.utc ↦ [d3.utcWednesday](https://github.com/d3/d3-time#utcWednesday)
* d3.time.thursday.utc ↦ [d3.utcThursday](https://github.com/d3/d3-time#utcThursday)
* d3.time.friday.utc ↦ [d3.utcFriday](https://github.com/d3/d3-time#utcFriday)
* d3.time.saturday.utc ↦ [d3.utcSaturday](https://github.com/d3/d3-time#utcSaturday)
* d3.time.week.utc ↦ [d3.utcWeek](https://github.com/d3/d3-time#utcWeek)
* d3.time.month.utc ↦ [d3.utcMonth](https://github.com/d3/d3-time#utcMonth)
* d3.time.year.utc ↦ [d3.utcYear](https://github.com/d3/d3-time#utcYear)

Mike Bostock's avatar
Mike Bostock committed
904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920
The local time range aliases have been renamed:

* d3.time.seconds ↦ [d3.timeSeconds](https://github.com/d3/d3-time#timeSeconds)
* d3.time.minutes ↦ [d3.timeMinutes](https://github.com/d3/d3-time#timeMinutes)
* d3.time.hours ↦ [d3.timeHours](https://github.com/d3/d3-time#timeHours)
* d3.time.days ↦ [d3.timeDays](https://github.com/d3/d3-time#timeDays)
* d3.time.sundays ↦ [d3.timeSundays](https://github.com/d3/d3-time#timeSundays)
* d3.time.mondays ↦ [d3.timeMondays](https://github.com/d3/d3-time#timeMondays)
* d3.time.tuesdays ↦ [d3.timeTuesdays](https://github.com/d3/d3-time#timeTuesdays)
* d3.time.wednesdays ↦ [d3.timeWednesdays](https://github.com/d3/d3-time#timeWednesdays)
* d3.time.thursdays ↦ [d3.timeThursdays](https://github.com/d3/d3-time#timeThursdays)
* d3.time.fridays ↦ [d3.timeFridays](https://github.com/d3/d3-time#timeFridays)
* d3.time.saturdays ↦ [d3.timeSaturdays](https://github.com/d3/d3-time#timeSaturdays)
* d3.time.weeks ↦ [d3.timeWeeks](https://github.com/d3/d3-time#timeWeeks)
* d3.time.months ↦ [d3.timeMonths](https://github.com/d3/d3-time#timeMonths)
* d3.time.years ↦ [d3.timeYears](https://github.com/d3/d3-time#timeYears)

Mike Bostock's avatar
Mike Bostock committed
921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937
The UTC time range aliases have been renamed:

* d3.time.seconds.utc ↦ [d3.utcSeconds](https://github.com/d3/d3-time#utcSeconds)
* d3.time.minutes.utc ↦ [d3.utcMinutes](https://github.com/d3/d3-time#utcMinutes)
* d3.time.hours.utc ↦ [d3.utcHours](https://github.com/d3/d3-time#utcHours)
* d3.time.days.utc ↦ [d3.utcDays](https://github.com/d3/d3-time#utcDays)
* d3.time.sundays.utc ↦ [d3.utcSundays](https://github.com/d3/d3-time#utcSundays)
* d3.time.mondays.utc ↦ [d3.utcMondays](https://github.com/d3/d3-time#utcMondays)
* d3.time.tuesdays.utc ↦ [d3.utcTuesdays](https://github.com/d3/d3-time#utcTuesdays)
* d3.time.wednesdays.utc ↦ [d3.utcWednesdays](https://github.com/d3/d3-time#utcWednesdays)
* d3.time.thursdays.utc ↦ [d3.utcThursdays](https://github.com/d3/d3-time#utcThursdays)
* d3.time.fridays.utc ↦ [d3.utcFridays](https://github.com/d3/d3-time#utcFridays)
* d3.time.saturdays.utc ↦ [d3.utcSaturdays](https://github.com/d3/d3-time#utcSaturdays)
* d3.time.weeks.utc ↦ [d3.utcWeeks](https://github.com/d3/d3-time#utcWeeks)
* d3.time.months.utc ↦ [d3.utcMonths](https://github.com/d3/d3-time#utcMonths)
* d3.time.years.utc ↦ [d3.utcYears](https://github.com/d3/d3-time#utcYears)

Mike Bostock's avatar
Mike Bostock committed
938 939 940 941 942 943 944 945 946 947
The behavior of [*interval*.range](https://github.com/d3/d3-time#interval_range) (and the convenience aliases such as [d3.timeDays](https://github.com/d3/d3-time#timeDays)) has been changed when *step* is greater than one. Rather than filtering the returned dates using the field number, *interval*.range now behaves like [d3.range](https://github.com/d3/d3-array#range): it simply skips, returning every *step*th date. For example, the following code in 3.x returns only odd days of the month:

```js
d3.time.days(new Date(2016, 4, 28), new Date(2016, 5, 5), 2);
// [Sun May 29 2016 00:00:00 GMT-0700 (PDT),
//  Tue May 31 2016 00:00:00 GMT-0700 (PDT),
//  Wed Jun 01 2016 00:00:00 GMT-0700 (PDT),
//  Fri Jun 03 2016 00:00:00 GMT-0700 (PDT)]
```

Mike Bostock's avatar
Mike Bostock committed
948
Note the returned array of dates does not start on the *start* date because May 28 is even. Also note that May 31 and June 1 are one day apart, not two! The behavior of d3.timeDays in 4.0 is probably closer to what you expect:
Mike Bostock's avatar
Mike Bostock committed
949 950 951 952 953 954 955 956 957

```js
d3.timeDays(new Date(2016, 4, 28), new Date(2016, 5, 5), 2);
// [Sat May 28 2016 00:00:00 GMT-0700 (PDT),
//  Mon May 30 2016 00:00:00 GMT-0700 (PDT),
//  Wed Jun 01 2016 00:00:00 GMT-0700 (PDT),
//  Fri Jun 03 2016 00:00:00 GMT-0700 (PDT)]
```

Mike Bostock's avatar
Mike Bostock committed
958
If you want a filtered view of a time interval (say to guarantee that two overlapping ranges are consistent, such as when generating [time scale ticks](https://github.com/d3/d3-scale#time_ticks)), you can use the new [*interval*.every](https://github.com/d3/d3-time#interval_every) method or its more general cousin [*interval*.filter](https://github.com/d3/d3-time#interval_filter):
Mike Bostock's avatar
Mike Bostock committed
959 960 961 962 963 964 965 966 967

```js
d3.timeDay.every(2).range(new Date(2016, 4, 28), new Date(2016, 5, 5));
// [Sun May 29 2016 00:00:00 GMT-0700 (PDT),
//  Tue May 31 2016 00:00:00 GMT-0700 (PDT),
//  Wed Jun 01 2016 00:00:00 GMT-0700 (PDT),
//  Fri Jun 03 2016 00:00:00 GMT-0700 (PDT)]
```

Mike Bostock's avatar
Mike Bostock committed
968
Time intervals now expose an [*interval*.count](https://github.com/d3/d3-time#interval_count) method for counting the number of interval boundaries after a *start* date and before or equal to an *end* date. This replaces d3.time.dayOfYear and related methods in 3.x. For example, this code in 3.x:
Mike Bostock's avatar
Mike Bostock committed
969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987

```js
var now = new Date;
d3.time.dayOfYear(now); // 165
```

Can be rewritten in 4.0 as:

```js
var now = new Date;
d3.timeDay.count(d3.timeYear(now), now); // 165
```

Likewise, in place of 3.x’s d3.time.weekOfYear, in 4.0 you would say:

```js
d3.timeWeek.count(d3.timeYear(now), now); // 24
```

Mike Bostock's avatar
Mike Bostock committed
988
The new *interval*.count is of course more general. For example, you can use it to compute hour-of-week for a heatmap:
Mike Bostock's avatar
Mike Bostock committed
989 990

```js
Mike Bostock's avatar
Mike Bostock committed
991
d3.timeHour.count(d3.timeWeek(now), now); // 64
Mike Bostock's avatar
Mike Bostock committed
992
```
Mike Bostock's avatar
Mike Bostock committed
993