⚡

polars

Name: polars
Author: K-Dense-AI

安全 ⚙️ 外部命令🔑 環境變數⚡ 包含腳本🌐 網路存取

Work with Polars DataFrames efficiently

也可從以下取得: davila7

Process large datasets in memory with Polars, the high-performance DataFrame library. Features lazy evaluation, parallel execution, and Apache Arrow backend for 10x faster operations than pandas.

支援: Claude Codex Code(CC)

📊 70 充足

下載技能 ZIP

在 Claude 中上傳

前往設定 → 功能 → 技能 → 上傳技能

開啟並開始使用

測試它

正在使用「polars」。 Load a CSV file and filter rows where age is greater than 25

預期結果:

Created DataFrame with columns: name, age, city
Filtered to 2 rows where age > 25
Columns selected: name, age

正在使用「polars」。 Group sales data by product category and calculate total and average sales

預期結果:

Grouped by product_category
Calculated sum and mean of sales_amount
Result includes: category, total_sales, avg_sales

正在使用「polars」。 Read a Parquet file using lazy evaluation and collect only needed columns

預期結果:

Used scan_parquet for lazy loading
Selected only required columns early
Collected with predicate pushdown optimization

安全審計

安全

v4 • 1/17/2026

This skill contains ONLY markdown documentation files with Python code examples. All 690 static findings are FALSE POSITIVES. The analyzer misidentified markdown code blocks, Python syntax, and Polars library methods as security threats. No executable code, shell commands, credential access, or network operations exist.

已掃描檔案

7,074

分析行數

發現項

審計總數

風險因素

⚙️ 外部命令 (647)

🔑 環境變數 (9)

references/io_guide.md:270 references/io_guide.md:271 references/io_guide.md:272 references/io_guide.md:287 references/io_guide.md:288 references/io_guide.md:301 references/io_guide.md:270 references/io_guide.md:271 references/io_guide.md:301

⚡ 包含腳本 (1)

references/operations.md:532

🌐 網路存取 (3)

skill-report.json:6 skill-report.json:21 SKILL.md:4

審計者: claude 查看審計歷史 →

品質評分

架構

可維護性

內容

社群

100

安全

規範符合性

你能建構什麼

Build ETL pipelines

Create efficient data pipelines with lazy evaluation for memory optimization and parallel execution.

Transform and aggregate data

Filter, group, and aggregate large datasets with expression-based syntax and window functions.

Replace pandas with faster alternative

Migrate existing pandas code to Polars for significant performance improvements on medium datasets.

試試這些提示

Load and explore data

Load a CSV file with Polars and show the first rows, column types, and basic statistics.

Filter and select

Filter rows where a column meets a condition and select specific columns using Polars expressions.

Group and aggregate

Group data by one or more columns and compute aggregations like mean, sum, and count.

Lazy optimization

Convert this DataFrame operation to use lazy evaluation and explain the performance benefits.

最佳實務

Use scan_csv or scan_parquet with lazy evaluation for large datasets to enable query optimization
Filter and select columns early in your pipeline to reduce memory usage and improve performance
Prefer native Polars expressions over Python functions to enable parallel execution

避免

Avoid using read_csv on large files when lazy evaluation would suffice
Do not apply Python functions inside hot paths when Polars expressions can accomplish the same task
Avoid loading entire datasets into memory when streaming with collect(streaming=True) would work

常見問題

How is Polars different from pandas?

Polars has no index, uses strict typing, offers lazy evaluation, and parallelizes operations by default. It is faster for medium-sized datasets.

When should I use lazy evaluation?

Use lazy evaluation for large datasets, complex pipelines, or when performance matters. It enables query optimization before execution.

What data sizes work best with Polars?

Polars excels with datasets from 1MB to 100GB that fit in RAM. Use dask or vaex for larger-than-memory data.

Can I migrate from pandas easily?

Yes, the migration is straightforward with similar concepts but different syntax. Most operations map directly to Polars equivalents.

Does Polars support cloud storage?

Yes, Polars reads and writes to S3, Azure Blob Storage, and Google Cloud Storage with proper credentials configured.

What file formats does Polars support?

Polars supports CSV, Parquet, JSON, Excel, IPC/Arrow, and databases via connectors. Parquet is recommended for performance.

開發者詳情

作者

K-Dense-AI

授權

https://github.com/pola-rs/polars/blob/main/LICENSE

儲存庫

https://github.com/K-Dense-AI/claude-scientific-skills/tree/main/scientific-skills/polars

引用

main

檔案結構

📁 references/

📄 best_practices.md

📄 core_concepts.md

📄 io_guide.md

📄 operations.md

📄 pandas_migration.md

📄 transformations.md

📄 SKILL.md