Apache Superset

Apache Superset is an open source data visualization and business intelligence web application designed for modern data teams. Originally created at Airbnb and now an Apache Software Foundation top-level project, Superset provides a rich set of data exploration, visualization, and dashboarding capabilities that make it a popular self-hosted alternative to commercial BI tools.

The chart library in Apache Superset is extensive, covering a wide range of visualization types. Available charts include line charts, area charts, bar charts, scatter plots, pie charts, histograms, heatmaps, box plots, treemaps, sunburst charts, sankey diagrams, chord charts, funnel charts, waterfall charts, geographic maps (choropleth, point, and path), time series charts, and pivot tables. The Echarts integration provides additional high-quality interactive visualizations. Custom chart types can be added through the plugin system.

The dataset model in Superset provides a semantic layer between the raw database tables and the visualizations. A dataset can be a physical database table, a view, or a virtual dataset defined by a SQL query. Calculated columns using SQL expressions can be added to datasets, enabling metric definitions, unit conversions, and derived dimensions. Metrics (aggregate measures) can be defined on datasets and referenced in charts, ensuring that business logic is defined once and applied consistently.

The chart builder (Explore) is Superset's interactive visualization editor. Users select a dataset and chart type, then configure dimensions, metrics, filters, and visual settings through a panel-based interface. The generated SQL query is visible in the editor, allowing users to understand what data is being fetched. Queries can be exported as SQL for use in other tools.

The SQL Lab is a powerful browser-based SQL editor for writing and executing queries against any connected database. It supports multiple database connections in separate tabs, query history, results export to CSV, chart creation from query results, and saving queries as datasets for further exploration and dashboarding.

Dashboard creation in Superset uses a drag-and-drop layout system where charts are positioned on a grid canvas. Dashboards support cross-filtering, where clicking on a value in one chart automatically filters related charts on the same dashboard, enabling highly interactive data exploration. Dashboard filters and native filters allow viewers to slice data by specific dimensions.

Superset supports connecting to virtually any SQL-speaking database through SQLAlchemy. Compatible databases include PostgreSQL, MySQL, SQL Server, SQLite, Google BigQuery, Amazon Redshift, Snowflake, ClickHouse, Presto, Trino, Druid, Hive, Impala, and many more. Superset also integrates with Apache Pinot for real-time analytics.

The security model in Superset is granular and role-based. Access can be controlled at the database, schema, table, row, and metric level. Row-level security rules restrict which rows of data each user role can see. The permission system integrates with OAuth, LDAP, OpenID Connect, SAML, and other enterprise identity providers.

Superset can be deployed using Docker Compose or Kubernetes and scales to support large data teams with many users, numerous data sources, and high dashboard usage. The Preset managed cloud service provides a hosted version with enterprise support.