Template driven data visualisation for ambitious applications.

Design

Maximum Plaid is designed to be a high level visualisation library built exclusively for Ember. It uses Ember's templating language to efficiently and declaratively produce easily maintained and tested visualisations.

We're building this because traditional charting libraries often times have poor separation of concerns from data and presentation, which leads to poor maintainability and code reuse, as well as exhibiting poor performance with medium to large datasets (10K to 100K data points).

We desire something which exposes abstractions for efficiently expressing a wide variety of visualisations, while balancing expressiveness such that you're not locked in to one particular style of visualisation which doesn't quite express the discoveries you've made in your data.

Proposed API

Visualisations are composed in layers, starting with data transformation down to visualisation and interaction.

Data Transformation

In order to provide a consistent API for each visualisation layer, we do the data setup ahead of time. After you've loaded your data source, you should transform it to the necessary shape for the type of visualisation you're rendering.

We provide a few helpers to make this easy (more soon).

Series Transforms

# pair-by(attr, [attr2,] series)

Basically a multi-property map-by.

Takes an array of objects containing discrete properties for each axis of your visualisations, and returns an array of tuples containing only the values for each attribute.

Given the series:

let timeSeriesData = [
	{ timestamp: 1450345920000, value: 1, projectId: 200 },
	{ timestamp: 1450345930000, value: 2, projectId: 200 },
	{ timestamp: 1450345940000, value: 3, projectId: 200 },
	{ timestamp: 1450345950000, value: 4, projectId: 200 },
	{ timestamp: 1450345960000, value: 5, projectId: 200 }
];

Using pair-by in your template:

{{pair-by "timestamp" "value" timeSeriesData}}

Would produce:

[
	[1450345920000, 1],
	[1450345930000, 2],
	[1450345940000, 3],
	[1450345950000, 4],
	[1450345960000, 5]
]

Which is ideal for using as the values argument to plaid.line and plaid.area.

Coordinates

SVG doesn't use the same coordinate space as CSS, so we have to calculate margins manually. To help out with this we've included an area helper, which returns an object containing the positioning data for your chart.

# area(width height [margin="TOP RIGHT BOTTOM LEFT"])

The margin string follows the typical box model margin convention of TOP, RIGHT, BOTTOM, LEFT. All values are unitless and inherit their units from the coordinate space with the SVG context.

Presentation

On their own, components for even the simplest elements in a visualisation can quickly require complicated APIs. To solve this, Ember contextual components can provide lower level primitive components with the necessary inputs for scaling and positioning. This reduces the API surface for the user to quickly produce visualisations in very few lines of code.

{{#plaid-plot xScale yScale plotArea as |plot|}}
  {{plot.right-axis}}
  {{plot.line values}}
{{/plaid-plot}}

A more complete example, using scales from ember-d3-scale and helpers from ember-composable-helpers

{{#with (area 720 220 margin="0 50 72 50") as |plotArea|}}
  {{#plaid-plot
    (time-scale (extent (map-by "timestamp" responseTimeMean)) (array 0 plotArea.width))
    (linear-scale (extent (map-by "value" responseTimeMean) toZero=true) (array plotArea.height 0))
    plotArea as |plot|}}

    {{#with (pair-by "timestamp" "value" responseTimeMean) as |values|}}

      {{plot.line values stroke="#673AB7" strokeWidth="2"  curve=(curve "basis")}}
      {{plot.area values fill="#D1C4E9" curve=(curve "basis")}}
      {{plot.bottom-axis values ticks=10}}
      {{plot.left-axis values tickFormat=(format-fn "0.1s" suffix="ms") ticks=2}}
    {{/with}}
  {{/plaid-plot}}
{{/with}}

This will produce a pretty simple line + area chart:

Components

plaid-symbol

Provides an easy to use and straight forward interface to d3-shape's symbol generators. primitive-symbol can be used in the same way as any Ember component and will render as a <path> tag containing the path data for the specified symbol type.

Options

• type: Symbol to render, can be any of circle, diamond, cross, square, star, triangle, wye.
• size: Specifies the symbol render size. This is the area of the symbol, which typicall equates to 1/4th of the actual width or height, depending on the shape.
• fill: SVG path fill property.
• stroke: SVG path stroke property.
• strokeWidth: SVG path stroke-width property.
• top: Top offset (applied using transform).
• left: Left offset (applied using transform).

Mixins

PlotArea

Provides simple calculations for specifying the position of the main graphic in a visualisation.

Example:

import { PlotArea } from 'maximum-plaid/mixins/plot-area';

export default Component.extend(PlotArea, {
  margin: '16 24',

  didRender() {
    const { top, left } = this.get('plotArea');
    select(this.element).select('g.plot').attr('transform', `translate(${left},${top})`);
  }
});

Utils

computedExtent

Creates a computed property which calculates the extent of all inputs provided using extent from d3-array.

Installation

NOTE: maximum-plaid requires Ember v2.3+

ember install maximum-plaid

Running

• ember server
• Visit your app at http://localhost:4200.

Running Tests

• npm test (Runs ember try:testall to test your addon against multiple Ember versions)
• ember test
• ember test --server

Building

• ember build

For more information on using ember-cli, visit http://www.ember-cli.com/.

FAQ

Why maximum-plaid?

Ivan: Initially I was searching name which represented both a type of fabric for the visualisation metaphore, but also I've been told that friends have lost me in crowds walking around Williamsburg because I wear so much plaid. Also, the goal of this project is to make data visualisation easy and efficient, so in keeping with the promised performance mode of the next Tesla Roadster, I found "maximum plaid" to be fitting.

You spelt visualization wrong

That's not a question.

As the main author of this project is Australian — a country which speaks a variation of British English, words such as visualisation are spelt with an s instead of a z. Another word you may find incorrectly spelt is colour. Please don't issue Pull Requests to fix this.