Apache Superset is an open-source, web-based platform for exploring data and building interactive business-intelligence dashboards without writing application code. It connects to databases such as DuckDB through SQLAlchemy, provides ready-made visualizations, and supports custom JavaScript visualizations when needed. section 9.2.1 section 9.3
Build a dashboard through a layered workflow:
- Connect a database, such as DuckDB, and let Superset detect its schemas and tables. section 9.2.2 section 9.2.3
- Create datasets from tables or saved SQL queries. section 9.2.2 section 9.2.4
- Create charts from datasets, then combine those charts into dashboards for interactive metric exploration. section 9.2.2 section 9.2.3
SQL dataset with a trendline
A saved query can identify Grand Slam winners aged 30 or older and calculate a cumulative count. Turn that dataset into a Big Number with Trendline chart by using tourney_date on the X-axis, Day as the time grain, and SUM(cumulative_count) as a custom metric; then add it to an existing dashboard. section 9.2.4
Figure 9.12 Terms used by Superset and their relationships
Use the relationship diagram to navigate from databases and saved queries through datasets and charts to dashboards. section 9.2.2
tree dashboard_export_20231203T162310dashboard_export_20231203T162310 ├── charts │ ├── Grand_Slam_Winners__30_11.yaml │ ├── Grand_Slam_winners__25_9.yaml │ ├── Lowest_Ranked_Grand_Slam_Winners_8.yaml │ ├── Matches_Played_7.yaml │ └── Multi_Title_Winners_6.yaml ├── dashboards │ └── ATP_Dashboard_1.yaml ├── databases │ └── DuckDB.yaml ├── datasets │ └── DuckDB │ ├── Grand_Slam_Winners__30.yaml │ ├── Multi_Title_Winners.yaml │ ├── Young_Grand_Slam_Winners.yaml │ └── matches.yaml └── metadata.yaml
Inspect the exported dashboard structure: YAML configuration separates databases, datasets, charts, dashboards, and metadata for reuse elsewhere. section 9.2.5
