Mosaic Plot

Introduction to Mosaic Plots

A mosaic plot (also known as a Marimekko Chart) is a chart used to display relationships between categorical data. It represents data proportions and classifications through the area and color of rectangles.

Mosaic plots are divided into two types: uniform and non-uniform. Uniform mosaic plots are more common in daily life, such as subway fare charts, while non-uniform mosaic plots are more frequently used in statistical fields.

Other Names: Mosaic Plot, Marimekko Chart

Components of a Mosaic Plot

Non-uniform Axis Mosaic Plot

Chart Type	Non-uniform Axis Mosaic Plot
Suitable Data	Multi-dimensional categorical data
Functionality	Display distribution of multi-dimensional categorical data
Data Mapping	Categorical data fields map to non-uniform coordinate axis Area and color represent data proportions and classifications
Data Size Limit	More effective with higher dimensional data

Uniform Axis Mosaic Plot

Chart Type	Uniform Axis Mosaic Plot
Suitable Data	Two-dimensional categorical data
Functionality	Display distribution of two-dimensional categorical data
Data Mapping	Categorical data fields map to uniform coordinate axis Color represents data classifications
Data Size Limit	More effective with lower dimensional data

Use Cases of Mosaic Plots

Suitable Use Cases

Example 1: Suitable for multi-dimensional categorical data analysis

Non-uniform Mosaic Plot

Example 2: Suitable for two-dimensional categorical data analysis

import * as G2 from '@antv/g2';

const chart = new G2.Chart({
  container: 'container',
  autoFit: true,
  height: 400,
});

chart.options({
  type: 'cell',
  data: [
    { product: 'Phone', region: 'North', sales: 120, category: 'Premium' },
    { product: 'Phone', region: 'East', sales: 180, category: 'Premium' },
    { product: 'Phone', region: 'South', sales: 150, category: 'Premium' },
    { product: 'Computer', region: 'North', sales: 80, category: 'Mid-range' },
    { product: 'Computer', region: 'East', sales: 110, category: 'Mid-range' },
    { product: 'Computer', region: 'South', sales: 95, category: 'Mid-range' },
    { product: 'Tablet', region: 'North', sales: 60, category: 'Mid-range' },
    { product: 'Tablet', region: 'East', sales: 85, category: 'Mid-range' },
    { product: 'Tablet', region: 'South', sales: 70, category: 'Budget' },
    { product: 'Headphone', region: 'North', sales: 40, category: 'Budget' },
    { product: 'Headphone', region: 'East', sales: 55, category: 'Budget' },
    { product: 'Headphone', region: 'South', sales: 45, category: 'Budget' },
  ],
  encode: {
    x: 'product',
    y: 'region',
    color: 'category',
    size: 'sales',
  },
  scale: {
    color: {
      palette: 'category10',
      type: 'ordinal',
    },
    size: {
      type: 'linear',
      range: [0.3, 1],
    },
  },
  style: {
    stroke: '#fff',
    strokeWidth: 2,
    inset: 2,
  },
  legend: {
    color: {
      title: 'Product Category',
      position: 'top',
    },
    size: {
      title: 'Sales Volume',
      position: 'right',
    },
  },
  axis: {
    x: {
      title: 'Product Type',
      labelAutoRotate: false,
    },
    y: {
      title: 'Sales Region',
    },
  },
  tooltip: {
    title: (d) => `${d.product} - ${d.region}`,
    items: [
      { field: 'sales', name: 'Sales', valueFormatter: (v) => `${v}K` },
      { field: 'category', name: 'Category' },
    ],
  },
});

chart.render();

Example 3: Market Segmentation Analysis (Non-uniform Mosaic Plot)

This example demonstrates how to use a non-uniform mosaic plot to display the distribution of different market segments, where the width of each rectangle represents market size, and the height represents the proportion of each market segment.

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

const chart = new Chart({
  container: 'container',
  width: 900,
  height: 800,
  paddingLeft: 0,
  paddingRight: 0,
});

chart.options({
  type: 'interval',
  data: {
    type: 'fetch',
    value:
      'https://gw.alipayobjects.com/os/bmw-prod/3041da62-1bf4-4849-aac3-01a387544bf4.csv',
  },
  transform: [
    { type: 'flexX', reducer: 'sum' }, // Flexible X-axis width
    { type: 'stackY' }, // Y-axis stacking
    { type: 'normalizeY' }, // Y-axis normalization
  ],
  encode: {
    x: 'market',
    y: 'value',
    color: 'segment',
  },
  axis: {
    y: false,
  },
  scale: {
    x: { paddingOuter: 0, paddingInner: 0.01 },
  },
  tooltip: 'value',
  labels: [
    {
      text: 'segment',
      x: 5,
      y: 5,
      textAlign: 'start',
      textBaseline: 'top',
      fontSize: 10,
      fill: '#fff',
    },
    {
      text: 'value',
      x: 5,
      y: 5,
      textAlign: 'start',
      dy: 15,
      fontSize: 10,
      fill: '#fff',
    },
  ],
});

chart.render();

Example 4: Movie Rating Distribution Analysis (Density Mosaic Plot)

This example demonstrates how to use a density mosaic plot to analyze the relationship between IMDB and Rotten Tomatoes ratings, where the color intensity represents the number of movies.

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

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

chart.options({
  type: 'rect',
  data: {
    type: 'fetch',
    value: 'https://assets.antv.antgroup.com/g2/movies.json',
  },
  encode: {
    x: 'IMDB Rating',
    y: 'Rotten Tomatoes Rating',
  },
  transform: [
    { type: 'bin', color: 'count', thresholdsX: 30, thresholdsY: 20 },
  ],
  scale: {
    color: { palette: 'ylGnBu' },
  },
  tooltip: {
    title: { channel: 'color' },
    items: [
      (d, i, data, column) => ({
        name: 'IMDB Rating',
        value: `${column.x.value[i]}, ${column.x1.value[i]}`,
      }),
      (d, i, data, column) => ({
        name: 'Rotten Tomatoes Rating',
        value: `${column.y.value[i]}, ${column.y1.value[i]}`,
      }),
    ],
    render: () => '1',
  },
});

chart.render();

Example 5: Athletes’ Physiological Data Analysis (Grouped Density Mosaic Plot)

This example shows how to use a mosaic plot to display the height and weight distribution of athletes grouped by gender, with opacity representing data point density.

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

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

chart.options({
  type: 'rect',
  data: {
    type: 'fetch',
    value: 'https://assets.antv.antgroup.com/g2/athletes.json',
  },
  encode: {
    x: 'weight',
    y: 'height',
    color: 'sex',
  },
  transform: [{ type: 'bin', opacity: 'count' }],
  legend: {
    opacity: false,
  },
  style: {
    inset: 0.5,
  },
  tooltip: {
    title: { channel: 'opacity' },
    items: [
      (d, i, data, column) => ({
        name: 'Weight',
        value: `${column.x.value[i]}, ${column.x1.value[i]}`,
      }),
      (d, i, data, column) => ({
        name: 'Height',
        value: `${column.y.value[i]}, ${column.y1.value[i]}`,
      }),
    ],
  },
});

chart.render();

Unsuitable Use Cases

Mosaic plots are not suitable for:

Displaying continuous numerical data trends
Showing precise numerical comparisons
Analyzing time series data
Displaying data with too many categories (which can make the chart cluttered and hard to read)

Comparing Mosaic Plots to Other Charts

Mosaic Plots and Heatmaps

From graphical properties perspective:
- Heatmaps represent the third dimension with linear color changes
- Mosaic plots represent the third dimension with categorical colors
- Standard heatmaps use smoothing algorithms with no clear boundaries
- Mosaic plots have clear boundaries
From data perspective:
- Uniform mosaic plots and heatmaps have very similar meanings for continuous data
From analytical needs perspective:
- Heatmaps focus on distribution and can predict data in unknown areas
- Mosaic plots have clearer boundaries and focus more on comparison

Mosaic Plots and Treemaps

Mosaic plots are better for showing relationships between multiple categorical variables
Treemaps are more suitable for displaying hierarchical data structures
Mosaic plots use position and area to encode data
Treemaps primarily use area and nesting to represent hierarchical relationships