Building a correlation matrix

Now that you've ranked albums, the strategy team wants to understand which metrics move together. The Spotify dataset has been enriched with streams_billions and streams_per_listener columns. Build a correlation matrix to spot strong and weak relationships between these numeric features.

polars is loaded as pl. The DataFrame spotify with additional columns is preloaded for you.

This exercise is part of the course

Data Transformation with Polars

View Course

Exercise instructions

Pick five numeric columns: streams_billions, monthly_listeners, streams_per_listener, duration_ms, and popularity, and compute the correlation.
Add a metric column to the result for better readability.

Hands-on interactive exercise

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

# Build a correlation matrix from selected columns
corr = spotify.____(
    "streams_billions",
    "monthly_listeners",
    "streams_per_listener",
    "duration_ms",
    "____",
).____()

# Add a metric column for row labels
result = corr.with_columns(pl.Series("____", corr.columns))

print(result)

Edit and Run Code