This lesson's interactive features are locked, please to keep using them

Basic Operations

Unlike Pandas, Polars leans hard on an expression-based API. You usually build expressions with the pl.col() function, which means "use this column in an operation."

Filter Rows

To filter for devices with temperatures above 80, use df.filter():

df.filter(pl.col("temperature_c") > 80)

Select Columns

To select the device_id and temperature_c columns, use df.select():

df.select(["device_id", "temperature_c"])
df.select(pl.col("device_id"), pl.col("temperature_c"))

Create Derived Columns

To add or replace columns, use df.with_columns(). For example, to convert Celsius readings to Fahrenheit:

df.with_columns(((pl.col("temperature_c") * 9 / 5) + 32).alias("temperature_f"))

The .alias() method names the new column.

Assignment

SnackStack's billing team needs a clean report of high-value transactions.

Complete the prepare_transaction_report function. It accepts a Polars DataFrame of transactions and returns a new, report-ready DataFrame.

Filter for rows where amount is greater than 100, referencing the column in an expression.
Add a column named processing_fee equal to 3% of amount.
Select and return only the transaction_id, customer_id, amount, and processing_fee columns.

Tips

You can build this report one step at a time, creating a variable for the return DataFrame and updating it with each operation.
Expressions often need to be nested. For example, pl.col() expressions will go inside the .filter() and .with_columns() calls.