Skip to content

A library for visualizing data trees with multiple parents, such as family trees. Built on top of D3.

License

Notifications You must be signed in to change notification settings

ErikGartner/dTree

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

dTree

A library for visualizing data trees with multiple parents built on top of D3.

npm Bower

Using dTree? Send me a message with a link to your website to be listed below.

The Online Viewer

There exists an online viewer for dTree graphs called Treehouse, similar to https://bl.ocks.org/ for D3. Treehouse allows anybody to host a dTree graph without having to create a website or interact directly with the library. It fetches data from Github's gists and displays it in a nice format. All graphs are unlisted so without your Gist ID nobody else can view them. Checkout the demo graph for dTree:

https://treehouse.gartner.io/ErikGartner/58e58be650453b6d49d7

The same demo is also available on JSFiddle.

Installation

There are several ways to use dTree. One way is to simply include the compiled file dTree.js that then exposes a dTree variable. dTree is available on both NPM and Bower as d3-dtree.

npm install d3-dtree
bower install d3-dtree
yarn add d3-dtree

Lastly dTree is also available through several CDNs such as jsDelivr:

https://cdn.jsdelivr.net/npm/d3-dtree@2.4.1/dist/dTree.min.js

Requirements

To use the library the follow dependencies must be loaded:

Usage

To create a graph from data use the following command:

tree = dTree.init(data, options);

The data object should have the following structure:

[
{
    "name": "Father", // The name of the node
    "class": "node", // The CSS class of the node
    "textClass": "nodeText", // The CSS class of the text in the node
    "depthOffset": 1, // Generational height offset
    "marriages": [
    { // Marriages is a list of nodes
        "spouse":
        { // Each marriage has one spouse
            "name": "Mother",
        },
        "children": [
        { // List of children nodes
            "name": "Child",
        }]
    }],
    "extra":
    {} // Custom data passed to renderers
}]

The following CSS sets some good defaults:

.linage {
    fill: none;
    stroke: black;
}
.marriage {
    fill: none;
    stroke: black;
}
.node {
    background-color: lightblue;
    border-style: solid;
    border-width: 1px;
}
.nodeText{
    font: 10px sans-serif;
}
.marriageNode {
    background-color: black;
    border-radius: 50%;
}

The options object has the following default values:

{
  target: '#graph',
  debug: false,
  width: 600,
  height: 600,
  hideMarriageNodes: true,
  marriageNodeSize: 10,
  callbacks: {
    /*
      Callbacks should only be overwritten on a need to basis.
      See the section about callbacks below.
    */
  },
  margin: {
    top: 0,
    right: 0,
    bottom: 0,
    left: 0
  },
  nodeWidth: 100,
  styles: {
    node: 'node',
    linage: 'linage',
    marriage: 'marriage',
    text: 'nodeText'
  }
}

Zooming

The returned object, tree = dTree.init(data, options), contains functions to control the viewport.

  • tree.resetZoom(duration = 500) - Reset zoom and position to initial state
  • zoomTo(x, y, zoom = 1, duration = 500) - Zoom to a specific position
  • zoomToNode(nodeId, zoom = 2, duration = 500) - Zoom to a specific node
  • zoomToFit(duration = 500) - Zoom to fit the entire tree into the viewport

Callbacks

Below follows a short descriptions of the available callback functions that may be passed to dTree. See dtree.js for the default implementations. Information about e.g. mouse cursor position can retrieved by interacting with the this object, i.e. d3.mouse(this).

nodeClick

function(name, extra, id)

The nodeClick function is called by dTree when the node or text is clicked by the user. It shouldn't return any value.

nodeRightClick

function(name, extra, id)

The nodeRightClick function is called by dTree when the node or text is right-clicked by the user. It shouldn't return any value.

nodeRenderer

function(name, x, y, height, width, extra, id, nodeClass, textClass, textRenderer)

The nodeRenderer is called once for each node and is expected to return a string containing the node. By default the node is rendered using a div containing the text returned from the default textRendeder. See the JSFiddle above for an example on how to set the callback.

nodeHeightSeperation

function(nodeWidth, nodeMaxHeight)

The nodeHeightSeperation is called during intial layout calculation. It shall return one number representing the distance between the levels in the graph.

nodeSize

function(nodes, width, textRenderer)

This nodeSize function takes all nodes and a preferred width set by the user. It is then expected to return an array containing the width and height for all nodes (they all share the same width and height during layout though nodes may be rendered as smaller by the nodeRenderer).

nodeSorter

function(aName, aExtra, bName, bExtra)

The nodeSorterer takes two nodes names and extra data, it then expected to return -1, 0 or 1 depending if A is less, equal or greater than B. This is used for sorting the nodes in the tree during layout.

textRenderer

function(name, extra, textClass)

The textRenderer function returns the formatted text to the nodeRenderer. This way the user may chose to overwrite only what text is shown but may opt to keep the default nodeRenderer.

marriageClick

function(extra, id)

Same as nodeClick but for the marriage nodes (connector).

marriageRightClick

function(extra, id)

Same as nodeRightClick but for the marriage nodes (connector).

marriageRenderer

function(x, y, height, width, extra, id, nodeClass)

Same as nodeRenderer but for the marriage nodes (connector).

marriageSize

function(nodes, size)

Same as nodeSize but for the marriage nodes (connector).

Built with dTree

Development

dTree has the following development environment:

  • node v11.x (use Docker image node:11)
  • gulp 3.x
  • Yarn instead of npm.

To setup and build the library from scratch follow these steps:

  1. yarn install
  2. yarn run build

A demo is available by running:

yarn run demo

It hosts a demo on localhost:3000/ by serving test/demo and using the latest compiled local version of the library.

Contributing

Contributions are very welcomed! Checkout the CONTRIBUTING document for style information. A good place to start is to make a pull request to solve an open issue. Feel free to ask questions regarding the issue since most have a sparse description.

License

The MIT License (MIT)

Copyright (c) 2015-2024 Erik Gärtner