Pandas GroupBy and Aggregation

Split–apply–combine

groupby splits rows into groups, applies a function to each, and combines the results.

# average order value per customer
df.groupby("customer_id")["amount"].mean()

# multiple aggregations at once
df.groupby("category").agg(
    total=("amount", "sum"),
    orders=("amount", "count"),
    avg=("amount", "mean"),
)

Why it matters for ML

Aggregations are how you build features: counts, averages, and recency per entity often predict more than raw rows. Master groupby and you can engineer most tabular features quickly.