Inserting missing columns

One year's extracted file is missing the pub column (publisher), but the team still wants to scan both files as one dataset. Pick the right argument so Polars inserts null where a column is missing instead of failing.

polars is loaded as pl, and the directory is in DRIFT_DIR. The header of each file is printed for you, so you can see the schema difference.

This exercise is part of the course

Scaling and Optimizing Data Pipelines with Polars

View Course

Exercise instructions

Use a glob pattern to scan every seattle_*.csv file in DRIFT_DIR.
Add the right argument so Polars inserts nulls for columns that are missing in some files.

Hands-on interactive exercise

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

# Scan both yearly files as one combined dataset
combined = pl.scan_csv(
    str(DRIFT_DIR / "____"),
    try_parse_dates=True,
    # Insert missing columns instead of failing on schema differences
    ____="____",
)

result = combined.select("date", "format", "title", "pub").collect()

print("First rows (from 2023 file):")
print(result.head(3))
print("\nLast rows (from 2024 file):")
print(result.tail(3))

Edit and Run Code