---
id: crick
version: "0.0.8"
license: BSD-3-Clause
license_treatment: permissive
maintenance: aging
---

# crick — High performance approximate and streaming algorithms

License: permissive · Maintenance: aging · Downloads: 330.8K/mo

## What it is and what it does

Crick is a library of approximate and streaming algorithms designed for high-performance computation on data streams. It allows you to compute statistics and aggregations on large datasets without loading all data into memory at once, using techniques that trade exact answers for speed and memory efficiency.

The library is distributed as pre-compiled wheels for Python 3.10+ across Windows, macOS, and Linux platforms. It has no runtime dependencies, making installation straightforward. The project originated in 2016 and received its last update in December 2024, placing it in aging maintenance—suitable for stable use cases but not actively developed.

Use it for:

- Computing approximate quantiles or percentiles on streaming data without storing all values
- Estimating cardinality (distinct element count) in large datasets with bounded memory
- Aggregating statistics over time-series or event streams in real-time systems
- Reducing memory overhead in data pipelines that process high-volume continuous data

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Crick provides approximate and streaming algorithms optimized for performance, enabling efficient computation on large data streams without storing all data in memory.

Yes, if you need approximate or streaming algorithms for memory-constrained or real-time scenarios. The library is stable, has no vulnerabilities, and covers a specific niche well. However, verify that the specific algorithms you need are implemented, and be aware that active development has slowed—suitable for production use of existing features but not for packages expecting ongoing feature additions.

## Install

pip install crick
uv add crick
poetry add crick

## Installing crick

Before you install: Medium install friction due to compiled wheels; available for current Python versions (3.10+) across major platforms. Repository is archived but maintained, with last commit in September 2025. Aging maintenance status suggests limited active development.

License in practice: BSD-3-Clause is permissive; you may use, modify, and distribute crick freely in commercial and private projects provided you retain the license notice.

Quickstart:

pip install crick

import crick
# Crick provides streaming algorithm implementations for approximate computation

Requires Python 3.10 or later.

Verify before relying:

- Specific algorithms available in the library (e.g., quantile sketches, cardinality estimation, frequency counting)
- Performance characteristics and accuracy guarantees of the approximate algorithms
- Whether the library is actively maintained or in maintenance-only mode despite aging status

## Package facts

- License: BSD-3-Clause (permissive)
- Python support: supports_current
- Install friction: medium
- Maintenance: aging
- Downloads: 330.8K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

streaming algorithms library, approximate algorithms, fast data stream processing, memory-efficient aggregation, approximate statistics, online algorithms, streaming data structures, streaming-algorithms, approximate-computing, data-structures

[View on SkillFed](https://skillfed.io/packages/crick) · [View on PyPI](https://pypi.org/project/crick/)