Looker Visualization Components

This skill enables agents to assist users in building custom, high-performance data experiences using Looker's React-based visualization components.

Overview

Looker Visualization Components offer a suite of React primitives to render Looker charts client-side. Unlike iframe-based embedding, these components offer:

• Performance: No iframe overhead; faster rendering and interactions.
• Customization: Deep control over chart configuration and the ability to swap default charts with custom React components.
• Seamless Integration: Fully integrates with Looker's design system (@looker/components).

1. Installation

Install the core library along with its peer dependencies.

npm install @looker/visualizations @looker/components-data @looker/components @looker/sdk styled-components react react-dom

2. Core Architecture

The system relies on a data-loading layer and a rendering layer.

• DataProvider: Context provider that supplies an authenticated Looker SDK instance.
• Query: Handles data fetching. Supports fetching by query (ID or slug) or dashboard ID.
• Visualization: Renders the chart using metadata returned by the query.

3. Basic Usage

Wrap your application in DataProvider and use Query to fetch data.

import { Looker40SDK } from '@looker/sdk'
import { DataProvider } from '@looker/components-data'
import { Query, Visualization } from '@looker/visualizations'

const sdk = new Looker40SDK(authSession)

export const AnalyticsDashboard = () => (
  <DataProvider sdk={sdk}>
    {/* Fetch by Query Slug */}
    <Query query="evomfl66xHx1jZk2Hzvv1R">
      <Visualization />
    </Query>

    {/* Fetch by Dashboard ID (fetches the first query on the dashboard) */}
    <Query dashboard="123">
      <Visualization config={{ type: 'area' }} />
    </Query>
  </DataProvider>
)

4. Internationalization (i18n)

Localization must be initialized once at the application root.

import { i18nInit, frFR } from '@looker/visualizations'

// Initialize with a specific locale
i18nInit(frFR)

5. Customization & Overrides

Chart Type Overrides

Force a specific chart type regardless of the Looker-saved configuration.

<Visualization config={{ type: 'sparkline', series: [{ color: '#2196f3' }] }} />

Custom Component Mapping

Replace a standard chart type with your own implementation using chartTypeMap.

const CustomTable = ({ data, fields }) => {
  return <table>...</table>
}

<Visualization chartTypeMap={{ table: CustomTable }} />

Supported Types

• line, area, scatter, sparkline
• bar, column
• pie
• table (Standard and Pivot support)
• single_value

6. Advanced Patterns

• Data Transformations: The Query component automatically applies transformations like sortByDateTime, nullValueZero, and xAxisReversed.
• Manual Data Handling: For full control, you can use hooks from @looker/components-data (e.g., useQueryData, useVisConfig) directly instead of the Query component.
• Extension Framework: When used inside a Looker Extension, the SDK is typically provided via ExtensionContext.

7. Troubleshooting

• Theme Errors: Visualization components require a ThemeProvider. If not provided by the app, they will automatically wrap themselves in a ComponentsProvider (Looker's default theme).
• Date Measures: Measures of type date are currently not supported in standard visualization components.
• Missing Peer Deps: Ensure styled-components version is ^5.3.1 or higher to avoid context mismatches.