SOL-C01 Question #1: Real Exam Question with Answer & Explanation

Question

You have a Python script running in a Snowflake Notebook that retrieves data from a Snowflake table, performs some complex calculations, and then visualizes the results using Matplotlib. The script is running slowly, even after optimizing the SQL query. Which of the following steps would MOST likely improve the performance of the Python script within the Snowflake Notebook environment?

Options

• AIncrease the size of the virtual warehouse associated with the Snowflake session.
• BVectorize the calculations using NumPy instead of looping through the data row by row.
• CUse the '%%osql' magic command to execute the calculations directly in Snowflake SQL.
• DStore the intermediate results in a Snowflake temporary table and retrieve them later.
• EUse a smaller data sample by adding 'LIMIT 1 00' in the SQL query to speed up the process.

Explanation

Option B, vectorizing calculations using NumPy, is the most likely to improve performance. Python loops are generally slower than vectorized operations performed by NumPy. Vectorization allows NumPy to perform calculations on entire arrays at once, significantly speeding up the process. Increasing the warehouse size (A) primarily improves SQL query performance, not Python code execution. Using (C) would offload the calculation to SQL, which might be faster if the calculation can be expressed efficiently in SQL, but the question states that the SQL query has already been optimized. Storing intermediate results in a temporary table (D) might be helpful in some cases, but doesn't directly address the slow Python calculations. Option E reduces the data to improve performance, but it doesn't solve the underlying issue of slow python calculation.

No community discussion yet for this question.