You're tasked with creating a Snowpark UDF to calculate the Haversine distance between two sets of latitude and longitude coordinates (point A and point B). Which of the following statements about deploying and using this UDF is/are TRUE?

You have a Snowpark Python stored procedure that performs complex data transformations. This stored procedure needs to read data from a large table ('TRANSACTIONS) and write the transformed data to another table PROCESSED TRANSACTIONS'). You want to optimize the performance of this stored procedure by leveraging Snowpark's features for parallel processing. Which of the following approaches can significantly improve the performance of the stored procedure, assuming sufficient warehouse resources are available?

A data engineering team has developed a Snowpark Python application to process customer orders, enrich them with external data (e.g., geo location, weather) and update the Customer360 table. The application is deployed to a production environment. The application's latency has significantly increased over the last week. Your investigation reveals that the Snowflake warehouse used by the application is constantly switching between the 'Scaling Up' and 'Scaling Down' states. The team has set the Auto Suspend time to 5 minutes and Auto Resume to True. Assuming that the team hasn't changed the code, the external API or any parameter related to data ingestion, which combination of the following actions would MOST likely fix the warehouse instability issue and improve the performance of this Snowpark application in production without substantial cost increases?

You're designing a Snowpark application to process large CSV files containing sensor data'. Each CSV file has millions of rows, and you need to calculate aggregate statistics (e.g., average, min, max) for specific sensor readings. The processing needs to be highly parallelized for performance. You have the following code snippet (incomplete):

Which of the following code snippets, when inserted at the ' ??? Add code here to calculate aggregate statistics ??? ' marker, would correctly calculate the average, minimum, and maximum readings for a column named 'sensor value' and return the result in a new Snowpark DataFrame?

You are developing a Snowpark application that uses a UDTF written in Python to perform complex data transformations. The UDTF takes several input columns and returns multiple output columns. The data volume is very large. You observe performance bottlenecks during the UDTF execution. Which of the following strategies could you employ to optimize the performance of your UDTF? (Select TWO)