Why Snowpark brings code to the data in Snowflake - See It in Action

Why Snowpark Brings Code to the Data

📖 Scenario: You work at a company that stores large amounts of data in Snowflake. You want to process this data efficiently without moving it around.

🎯 Goal: Build a simple Snowpark program that shows how to bring code to the data inside Snowflake, instead of moving data out for processing.

📋 What You'll Learn

Create a Snowpark session connected to Snowflake

Create a DataFrame from an existing table called sales_data

Define a simple transformation function using Snowpark

Apply the function to the DataFrame and collect results

💡 Why This Matters

🌍 Real World

Companies use Snowpark to run data processing close to where data lives in Snowflake. This reduces data movement, speeds up processing, and improves security.

💼 Career

Understanding Snowpark is valuable for data engineers and cloud architects who build scalable data pipelines and analytics solutions on Snowflake.

1

Create a Snowpark session

Write code to create a Snowpark session called session using the Session.builder.configs() method with a dictionary connection_parameters containing your Snowflake account details.

Snowflake

# Your code here to create a Snowpark session named session

Need a hint?

Use Session.builder.configs(connection_parameters).create() to create the session.

2

Create a DataFrame from the sales_data table

Use the session.table() method to create a DataFrame called df from the existing table named sales_data.

Snowflake

session = Session.builder.configs(connection_parameters).create()
# Your code here to create a DataFrame named df from sales_data

Need a hint?

Use df = session.table("sales_data") to create the DataFrame.

3

Define a transformation function using Snowpark

Define a function called add_discount that takes a DataFrame df and returns a new DataFrame with a new column discounted_price calculated as price * 0.9.

Snowflake

df = session.table("sales_data")
# Your code here to define add_discount(df) function

Need a hint?

Use with_column to add the new column inside the function.

4

Apply the function and collect results

Call the add_discount function with df and assign the result to df_discounted. Then use df_discounted.collect() to fetch the results into a variable called results.

Snowflake

def add_discount(df):
    return df.with_column("discounted_price", df["price"] * 0.9)
# Your code here to apply add_discount and collect results

Need a hint?

Assign the function result to df_discounted and then call collect() to get results.