Customizing Data in IrisX Analytics - Best Practices

The IrisX folder in your catalog contains all provided base data. While you can't modify this original data or add customized tables directly to the IrisX catalog, you can create new tables derived from this data for your specific use cases.
Store your custom tables in catalogs labeled with your name, following the dev/test/prod environment structure for proper development progression.

Add new data to your environment:

1. Write code in a notebook that performs your desired calculations.
2. Save results in your designated catalog.
3. Choose the appropriate storage format.
4. For regular updates to your custom tables, create a workflow to run your notebooks automatically on a schedule.

Different storage formats:

A table is a structured dataset that physically stores data. Choose this when:

• You need to query or modify the dataset repeatedly
• Performance is critical for frequently accessed data
• The dataset serves as input for other analytics processes

For example, Trackunit provides its base data as tables to ensure optimal performance and reliability.

A view is a virtual table defined by a SQL query that doesn't store physical data. Views are ideal when:

• You need a consistent interface to dynamically generated results
• The underlying data changes frequently
• You want to simplify complex queries for business users
• Storage efficiency is important

Materialized views combine aspects of both tables and views by:

• Storing the results of a query physically (like tables)
• Maintaining a connection to the source query for refreshes
• Offering improved query performance while ensuring data currency