Overview of the Differences: How to Use Snowpark With and Without Ascend
- Data Loading: To effectively use Snowpark, users must become acquainted with the Snowpark Python library. This entails loading data from particular tables into Snowpark DataFrames.
- Data Transformations: Data manipulation in Snowpark involves diving into the snowflake.snowpark.functions library, necessitating familiarity with a slew of functions.
- Examining & Storing Data: Snowpark provides the option to inspect query execution plans using the explain() function, introducing another layer of complexity. After these operations, users save the transformed data into Snowflake tables, with an option to return the data as Snowpark DataFrame.
- Advanced Deployment: Snowpark offers the ability to deploy the entire worksheet as a Python Stored Procedure. This introduces benefits like scheduling tasks via Snowflake. However, it also adds another layer of intricacy to the process.
- Ingesting Data: No longer do you have to grapple with multiple data ingestion steps. It’s direct and straightforward.
- Writing Business Logic: Once your data is ingested, all you need to focus on is your core business logic.
- Schema/table creation
- Code deployment
- Job scheduling
- Warehouse selection
- Data loading
- and more!
A Practical Demonstration
Let’s get our hands dirty and see how to use Snowpark in two steps in action.