Profiling a lazy query

Before adding a new query to a daily workflow, the team wants to see where time is actually spent. Profiling runs the lazy query like .collect() does, but also returns a timings DataFrame with one row per stage.

This exercise is part of the course

Scaling and Optimizing Data Pipelines with Polars

View Course

Exercise instructions

Filter library to checkouts where use is "Physical".
Keep the 10 longest titles by title_len using .top_k().
Execute the query with profiling so you also get a timings DataFrame.

Hands-on interactive exercise

Have a go at this exercise by completing this sample code.

result, timings = (
    library
    # Filter to physical checkouts
    .filter(pl.col("use") == "____")
    .with_columns(pl.col("title").str.len_chars().alias("title_len"))
    # Keep the 10 longest titles
    .top_k(10, by="____")
    # Run with profiling to capture per-stage timings
    .____()
)
print(timings)

Edit and Run Code