Aggregating several ways in Python

group_by().agg() takes a list of expressions, each named.

import polars as pl

frame = pl.DataFrame({
    "lang": ["python", "python", "rust", "rust", "sql"],
    "tpm": [104, 91, 98, 88, 79],
    "errors": [1, 4, 0, 2, 3],
})

summary = frame.group_by("lang").agg(
    pl.len().alias("runs"),
    pl.col("tpm").max().alias("best"),
    pl.col("tpm").mean().round(1).alias("average"),
    (pl.col("errors") == 0).sum().alias("clean"),
).sort("average", descending=True)

print(summary)

How it works

Every aggregate is an expression, so it can filter inside itself.
pl.len counts rows in the group.
The result has one row per group and a column per expression.

Keywords and builtins used here

as
print

The run, in numbers

Lines: 16
Characters to type: 399
Tokens: 182
Three-star pace: 110 tpm

At the three-star pace of 110 tokens a minute, this run takes about 99 seconds.

Type this snippet

Step 2 of 3 in Aggregation, step 13 of 32 in Polars.

← Previous Next →