sqrt

Overview

sqrt is a continuous non-linear scale that is essentially a pow scale with a fixed exponent of 0.5. The mapping function of the sqrt scale is y = x ^ 0.5 + b, where x is the input data and b is the offset.

The sqrt scale is a type of continuous scale that is suitable for:

Data with large numerical range differences that need compression
Visualization scenarios that emphasize smaller numerical differences
Visual channels that require square root transformation, such as area mapping

Compared to linear scales, sqrt scales have a higher compression degree for large values and are suitable for handling data with large variation ranges.

Mapping Effect Examples

The sqrt scale can make data more evenly distributed visually, especially for datasets with large numerical ranges.

Using sqrt scale effect

import { Chart } from '@antv/g2';

const chart = new Chart({
  container: 'container',
});

chart.options({
  type: 'view',
  autoFit: true,
  data: [
    { year: '1991', value: 1 },
    { year: '1992', value: 4 },
    { year: '1993', value: 9 },
    { year: '1994', value: 16 },
    { year: '1995', value: 25 },
  ],
  encode: { x: 'year', y: 'value' },
  scale: { y: { type: 'sqrt' } },
  children: [
    { type: 'line', labels: [{ text: 'value', style: { dx: -10, dy: -12 } }] },
    { type: 'point', style: { fill: 'white' }, tooltip: false },
  ],
});

chart.render();

Without sqrt scale effect

import { Chart } from '@antv/g2';

const chart = new Chart({
  container: 'container',
});

chart.options({
  type: 'view',
  autoFit: true,
  data: [
    { year: '1991', value: 1 },
    { year: '1992', value: 4 },
    { year: '1993', value: 9 },
    { year: '1994', value: 16 },
    { year: '1995', value: 25 },
  ],
  encode: { x: 'year', y: 'value' },
  children: [
    { type: 'line', labels: [{ text: 'value', style: { dx: -10, dy: -12 } }] },
    { type: 'point', style: { fill: 'white' }, tooltip: false },
  ],
});

chart.render();

Options

Property	Description	Type	Default Value
domain	Set the domain range of the data	`number[]`	Min-max range of input data
domainMin	Set the minimum value of the data domain	`number`	Minimum value of input data
domainMax	Set the maximum value of the data domain	`number`	Maximum value of input data
range	Set the value range for data mapping	`number[]` \| `string[]`	`[0, 1]`
rangeMin	Set the minimum value of the data mapping range	`number \| string`	`0`
rangeMax	Set the maximum value of the data mapping range	`number \| string`	`1`
unknown	The returned data for null values like `undefined`, `NaN`, `null`	`any`	`undefined`
tickCount	Set the recommended number of ticks to generate, tickCount is only a suggested value, not absolute	`number`	`5`
tickMethod	Set the method for generating ticks, commonly used for custom ticks	`(min: number, max: number, count: number) => number[]`	`d3-ticks`
round	Round the output values	`boolean`	`false`
clamp	Limit the mapped values to the range	`boolean`	`false`
nice	Extend the domain range to make the output ticks display more nicely	`boolean`	`false`
interpolate	Custom interpolation function	`(a: number, b: number) => (t: number) => T`	`(a, b) => (t) => a * (1 - t) + b * t`
exponent	Set the exponent, for sqrt scale, this value is fixed at `0.5`	`number`	`0.5`

Examples

Basic Usage

Use sqrt scale to map point sizes in a scatter plot to make numerical differences more apparent.

import { Chart } from '@antv/g2';

const chart = new Chart({
  container: 'container',
});

chart.options({
  type: 'point',
  autoFit: true,
  data: {
    type: 'fetch',
    value: 'https://gw.alipayobjects.com/os/antvdemo/assets/data/bubble.json',
  },
  encode: {
    x: 'GDP',
    y: 'LifeExpectancy',
    size: 'Population',
    color: 'continent',
    shape: 'point',
  },
  scale: { size: { type: 'sqrt', range: [4, 50] } },
  style: { fillOpacity: 0.3, lineWidth: 1 },
  legend: { size: false },
});

chart.render();

Code Explanation

This is a bubble chart (point chart), where each point represents a country.
Data is fetched remotely and includes fields like GDP, life expectancy, population, etc.
The encode section maps GDP to the x-axis, life expectancy to the y-axis, population to point size, and continent to color.
scale.size uses sqrt to map population through power transformation to point radius, with range: [4, 20] controlling the minimum and maximum point sizes.
Visual styles set transparency and stroke width, with size legend disabled.
The chart is rendered with chart.render().

Combined with Color Gradient

Using sqrt scale for color gradient mapping

import { Chart } from '@antv/g2';

const chart = new Chart({
  container: 'container',
});

chart.options({
  type: 'view',
  autoFit: true,
  children: [
    {
      type: 'point',
      data: {
        type: 'fetch',
        value:
          'https://gw.alipayobjects.com/os/bmw-prod/56b6b137-e04e-4757-8af5-d75bafaef886.csv',
      },
      encode: { x: 'date', y: 'value', color: 'value', shape: 'point' },
      scale: {
        color: { type: 'sqrt', domain: [0, 1], range: ['#1689F1', '#1AC07D'] },
      },
      style: { stroke: '#000', strokeOpacity: 0.2 },
      tooltip: {
        items: [
          {
            channel: 'x',
            name: 'year',
            valueFormatter: (d) => d.getFullYear(),
          },
          { channel: 'y' },
        ],
      },
    },
    { type: 'lineY', data: [0], style: { stroke: '#000', strokeOpacity: 0.2 } },
  ],
});
chart.render();

Code Explanation

This is a point chart (scatter plot) with auxiliary lines.
The main component is a point chart with data also fetched remotely, including date and value fields.
The encode section maps date to the x-axis, value to the y-axis, and value is also mapped to color.
scale.color uses sqrt to map value through power transformation to color gradient (blue to green), with domain: [0, 1] specifying the input data range.
Points have black strokes with some transparency.
Tooltip is configured with custom year formatting.
A y=0 auxiliary line (lineY) is also added.
The chart is rendered with chart.render().

In this example, data is mapped to color through the sqrt scale, which can better display regional differences compared to linear mapping.