When writing PySpark code to load data from an Azure Cosmos DB analytical store, what format should be specified?

When loading data from an Azure Cosmos DB analytical store using PySpark, the appropriate format to specify is 'cosmos.olap'. This format is specifically designed for integrating with the analytical capabilities of Cosmos DB, enabling efficient data queries and processing through the underlying architecture.

Azure Cosmos DB's analytical store is optimized for large-scale analytical queries and allows you to run complex operations on your data. By using the 'cosmos.olap' format, PySpark can effectively interact with this analytical store, leveraging its capabilities to perform data transformations and analytical processing efficiently.

The other options, while potentially resembling valid formats, do not align with the specific requirements for accessing data from Cosmos DB's analytical store. Utilizing the correct format ensures optimal performance and compatibility when working with the data, making 'cosmos.olap' the correct choice in this context.