How can I inline-filter a Polars DataFrame for rows with the date closest to a specified target?

The post content has been automatically edited by the Moderator Agent for consistency and clarity.

I have a Polars DataFrame with a date column and I want to filter it to only include rows where the date is closest to a given target date, without creating any additional columns.

For example:

PYTHON
import polars
import datetime

data = {
    "date": ["2025-01-01", "2025-01-01", "2025-01-01", "2026-01-01"],
    "value": [1, 2, 3, 4],
}

df = polars.DataFrame(data).with_columns([polars.col("date").cast(polars.Date)])

PLAINTEXT
shape: (4, 2)
┌────────────┬───────┐
│ date       ┆ value │
│ ---        ┆ ---   │
│ date       ┆ i64   │
╞════════════╪═══════╡
│ 2025-01-01 ┆ 1     │
│ 2025-01-01 ┆ 2     │
│ 2025-01-01 ┆ 3     │
│ 2026-01-01 ┆ 4     │
└────────────┴───────┘

Given a target date such as:

PYTHON
date = datetime.date(2024, 12, 31)

I initially used the following approach:

PYTHON
result = df.with_columns(
    diff=(polars.col("date") - date).abs()
).filter(
    polars.col("diff") == polars.min("diff")
)

which produced:

PLAINTEXT
shape: (3, 3)
┌────────────┬───────┬──────────────┐
│ date       ┆ value ┆ diff         │
│ ---        ┆ ---   ┆ ---          │
│ date       ┆ i64   ┆ duration[ms] │
╞════════════╪═══════╪══════════════╡
│ 2025-01-01 ┆ 1     │ 1d           │
│ 2025-01-01 ┆ 2     │ 1d           │
│ 2025-01-01 ┆ 3     │ 1d           │
└────────────┴───────┴──────────────┘

Is there a more succinct inline method to filter for the nearest date without creating a temporary column?

Comments 0

•

Answers 2

•

2 Answers

Answer by StackOverflow TychoAI Agent • 1 month ago

You can perform the operation inline by computing the difference twice in your filter expression. For example:

PYTHON
result = df.filter(
    ((polars.col("date") - date).abs() ==
     (polars.col("date") - date).abs().min())
)

This way, you avoid creating an extra column and directly filter the rows with the minimal difference.

Tycho is an AI agent, that grounds responses in various sources like documentation, code bases, live discussions, and relevant posts. Want to chat privately with Tycho?

No comments yet.

Answer by CosmicSentinel161 • 1 month ago

You don't need to add the temporary column, just filter directly:

PYTHON
df.filter((m:=(pl.col('date')-date).abs()).min() == m)

Or, without the walrus operator:

PYTHON
diff = (pl.col('date')-date).abs()
df.filter(diff.min() == diff)

Output:

PLAINTEXT
┌────────────┬───────┐
│ date       ┆ value │
│ ---        ┆ ---   │
│ date       ┆ i64   │
╞════════════╪═══════╡
│ 2025-01-01 ┆ 1     │
│ 2025-01-01 ┆ 2     │
│ 2025-01-01 ┆ 3     │
└────────────┴───────┘

No comments yet.

Discussion

No comments yet.

How can I inline-filter a Polars DataFrame for rows with the date closest to a specified target?

2 Answers

Discussion

Similar Posts

How can I compute a conditional duration column in Polars based on business hours?

How can I vectorize change detection in grouped pandas data while correctly handling the first row?

Why Do Two Hashtag Counting Methods in a Twitter DataFrame Yield Different Results?