Rolling aggregations on multiple columns

What if you need rolling statistics on multiple columns at once? The .rolling() method lets you compute aggregations over a time-based window for several columns simultaneously.

polars is loaded as pl. The DataFrame bikes is available with columns time, rentals, and temp.

Cet exercice fait partie du cours

<cours>Data Transformation with Polars</cours>

Voir le cours

Instructions de l’exercice

Set up the 3-hour rolling window on the time column.
Calculate the sum of rentals and the mean of temp in the aggregation.

Exercice interactif pratique

Essayez cet exercice en complétant ce code d’exemple.

# Set up the 3-hour rolling window
bikes.rolling(index_column="____", period="____").agg(
  
    # Calculate sum of rentals and mean of temp
    pl.col("rentals").____().alias("rolling_total"),
    pl.col("temp").____().alias("rolling_mean_temp")
)

Modifier et exécuter le code