Snowflake SPS-C01 Exam Syllabus Topics:

Snowflake Certified SnowPro Specialty - Snowpark Sample Questions:

1. You are developing a Snowpark application that requires secure access to Snowflake. You need to establish a session using Key Pair authentication. You have stored your private key in an encrypted file and have the passphrase available. Assume you've correctly configured your Snowflake user with the public key. Which of the following methods can be used to load the private key, assuming that 'PRIVATE KEY PATH' stores path to encrypted private key and 'PRIVATE KEY PASSPHRASE stores passphrase?

A)

B)

C)

D)

E)

2. Consider the following Snowpark Python code snippet designed to calculate the moving average of sales data'. You've identified that the code is performing poorly and suspect the window function is a bottleneck. How can you optimize this code for better performance?

A) Ensure that the data is pre-sorted according to the ordering specified in the window function before creating the Snowpark DataFrame.
B) Replace the 'Window.orderBy' with 'Window.partitionBV on a highly cardinal column to distribute the window calculations across multiple nodes.
C) Rewrite the window function logic as a series of aggregation queries to improve performance on very large datasets.
D) Explicitly specify a range-based window frame (e.g., 'rowsBetweeri) instead of a rows-based window frame (e.g., 'rangeBetweeri) if appropriate for the calculation.
E) Use the method on the Snowpark DataFrame before applying the window function to avoid re-reading the data multiple times.

3. You have a large CSV file containing product descriptions that you need to analyze using a sentiment analysis UDF. The CSV file is too large to fit in memory on your local machine. You want to stream the data directly from a Snowflake stage to your UDF for processing, avoiding the need to download the entire file. Which of the following approaches allows you to achieve this using Snowpark?

A) Use the 'snowflake.connector' library within the UDF to establish a new connection to Snowflake, read the CSV file from the stage, and analyze the product descriptions.
B) Load the CSV file into a Snowflake internal stage. Then, within a Snowpark UDF, directly query the stage using SQL to retrieve and analyze the product descriptions.
C) Use the method within the UDF to read the CSV file line by line and pass each product description to the sentiment analysis function.
D) Create an external table pointing to the CSV file on the stage. Use a Snowpark DataFrame to select the product descriptions from the external table and then apply the sentiment analysis UDF.
E) Read the CSV file directly into a Pandas DataFrame using within the UDF and perform sentiment analysis on the DataFrame.

4. You have a Snowpark DataFrame named 'products_df' with columns 'product_id' (INT), 'product_name' (VARCHAR), and 'price' (FLOAT). You want to create a new DataFrame called 'discounted_products df that includes all columns from 'products_df' plus a new column named 'discounted_price', which is calculated as the original price minus a discount percentage specified by the variable 'discount_rate' (e.g., 0.1 for 10%). The 'discount_rate' is stored in the database table named 'discount_table'. You want to load the rate to variable. Choose the correct ways to achieve this. (Select all that apply)

A)

B)

C)

D)

E)

5. You are tasked with creating a Snowpark Python stored procedure that reads data from a Snowflake table, performs a complex data transformation using a 3rd party Python library (e.g., pandas, scikit-learn), and writes the transformed data to another Snowflake table.
The data transformation requires significant memory. You need to register this stored procedure in Snowflake. Which of the following approaches is the MOST appropriate for registering the stored procedure and managing the dependencies?

A) Use the '@sproc' decorator without specifying any dependencies, assuming that the necessary libraries are pre-installed on the Snowflake worker nodes.
B) Install the required Python libraries directly on the Snowflake compute warehouse using a SQL command.
C) Create a Snowflake stage, upload the Python libraries as .zip files to the stage, and specify the stage path in the '@sproc' decorator's 'imports' parameter.
D) Create a conda environment file ('environment.yml') specifying the dependencies, upload it to a stage, and then use the '@sproc' decorator with the 'packages' argument referencing the conda environment.
E) Use the function to add the required Python libraries before registering the stored procedure with the '@sproc' decorator.

Solutions: