Using Superset in a Space

Owned by Vytas Glambinskas
Nov 12, 2023
4 min read

To use Superset components in a Space you must

  1. Activate a space session via the secure desktop.

  2. Select the Superset tool icon.

  3. When the tool opens, login with the required credentials (Username : superset / Password : superset).

 

The core capabilities provided by Superset are :

  • Interactive data exploration.

  • Interactive chart design and generation.

  • Dashboard design and creation.

  • Graphical SQL query development.

  • Interactive SQL development environment.

 

Objects created in Superset are persisted and shared across your collaboration for iterative insight development across your project team.

Superset components

The superset tool is made up of 4 main components which are outlined below.

Data

Superset charts and visualizations are built from connections to hosted data sets. In the case of Spaces these are the data products added to the space which are exposed via the HIVE metastore.

As part of the Space configuration a Superset ‘Data Source’ called Spaces Metastore is created that provides the connection to the data products.

Additionally a Superset ‘Dataset’ is also created for every table within the data product data bases. These data sets describe each table associated with the Space, including the field names and data types. These actions are taken as part of Space configuration so that the datasets are fully available at the start of a Space session without any manual configuration.

The Superset dataset definitions are refreshed during each Space activation so that any updates to the data products associated with a Spaces will be reflected.

Whilst Data Product tables are automatically configured during Space activation, custom tables and outputs created in collaborate_db and publish_db are added to Superset by users as required by their projects and the dynamic creation of these objects within Space sessions.

To add a table in collaborate_db or publish_db to Superset do :

→ Data

→ Datasets

→ “+ Dataset”

→ Select Data Source (Spaces Metastore will be default, pre-selected)

→ Select collaborate_db or publish_db from the Schema drop down

→ Select the table you wish to add from the Tables drop down

→ Click Add

This will add the table that been created in one of these data bases as a superset dataset, including inferring all the column names and data types.

If you are using an environment with the Presto query engine available an additional Data Source called ‘Fast Spaces Metastore’ will also be displayed in Superset. It is recommended to utilise this ‘Fast Spaces Metastore’ when it’s available to experience an up to 10x performance increase in query execution and chart rendering within Superset.

Charts

This component allows you to explore data and create charts and visualizations from the Superset Datasets. Charts are visually configured by selecting elements such as :

  • Chart type.

  • Fields of interest.

  • Required metrics (e.g. count, sum, min, max, custom calculations etc.)

  • Visual configurations (colour schemes, labelling, titles etc.)

To create a chart in superset do :

→ Charts

→ ‘+ CHART'

→ Choose a Dataset from the top left hand side drop down

→ Select a chart type via either the categories or by searching for chart types

→ Select ‘CREATE NEW CHART’

Full guidance on charts available, and configuration options for each type can be found through the open source Superset documentation described in .

Once finalized charts can be saved for use in future sessions and Dashboards.

As all superset visualizations are based off of SQL logic that is executed against the data sources, charts can also be used as a visual SQL code development tool. Once a chart has been created showing the metrics and insights required, the underlying SQL logic can be accessed by clicking the three bars button in the top right hand control bar and selecting ‘View query’ or ‘Run in SQL Lab' :

This displays the underlying logic which can be executed within the Spaces SQL tools to create outputs in collaborate_db or publish_db, and used to facilitate Automated Tasks to create the output metrics on an on-going basis.

Dashboards

This component allows you to create dashboards by combining the charts you have produced into reporting views and insight analyses. In addition to your charts other formatting objects such as tabs, rows, columns and markdown text boxes can be included in dashboards to provide richer insights and context to visualizations.

To create a dashboard in Superset do:

→ Dashboards

→ ‘+ DASHBOARD’

→ Update dashboard name in top left of pane

→ ‘SAVE’

Once you have saved your dashboard you can start to add structuring elements and data visualizations by dragging and dropping the components required from the right hand ‘COMPONENTS’ and ‘CHARTS’ panes within the design view.

Use the ‘SAVE’ button to save progress as dashboards are incrementally developed. When saved Dashboards are available to all collaborators within a Space.

Full guidance on the creation and customization of Dashboards can be found through the open source Superset documentation described in .

SQL Lab

This component provides a SQL development environment that can be used for interacting with Data Products as described in

In addition to providing a SQL development environment, SQL Lab integrates with the wider Superset feature set as described above. This facilitates both the extraction of the SQL logic used to create charts and the direct exploration of the results of queries executed within SQL Lab within the superset charting functionality.

References and FAQs

Related Pages