fastdigest
A fast t-digest library for Python built on Rust.
Decision gist · record as of 2026-08-14
Yes. fastdigest is actively maintained, has no security vulnerabilities, carries a permissive MIT license, and offers substantial speed gains (480x faster than tdigest in the documented benchmark) with a straightforward API. Install it if you need streaming quantile estimation or online statistics on large datasets and want a fast, dependency-free solution.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Medium install friction due to compiled wheels; prebuilt binaries are available for Python 3.7–3.14 across macOS, Linux (including musl), and Windows architectures.
- Active maintenance with recent commits; no runtime dependencies.
License · maintenance · safety
permissive license (permissive) — MIT license (permissive); you may use, modify, and distribute this package freely in commercial and private projects with minimal restrictions.
last release 2026-03-15 (152 days) · last repo commit 2026-03-15 · 20 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 152,643 downloads/mo, #10,897 on PyPI
Alternatives
Verify before relying
pip install fastdigest
from fastdigest import TDigest
digest = TDigest.from_values(range(1001))
print(digest.quantile(0.99)) # 99th percentile
print(digest.cdf(990)) # rank of value 990- Accuracy bounds or error guarantees for quantile estimates compared to exact computation
- Memory consumption scaling with digest size or data distribution
- Whether the package supports Python 3.14 and 3.13 in practice or only declares support
What it is and what it does
fastdigest is a Python wrapper around a Rust implementation of the t-digest data structure, designed for computing approximate quantiles and other online statistics on streaming or distributed data. It trades exact computation for speed and memory efficiency, making it suitable for scenarios where you need to estimate percentiles, medians, or trimmed means from large datasets without storing all values in memory.
The package provides methods to initialize a digest from values, incrementally update it with new data (with optional weighting), merge multiple digests together, and query statistics like quantiles, cumulative distribution function values, means, and median absolute deviation. It supports serialization to/from dictionaries and pickle format, and aims for API compatibility with the tdigest library to ease migration.
Use it for
- Estimate percentiles from streaming sensor or log data without storing all raw values
- Aggregate quantile statistics across distributed compute jobs via digest merging
- Detect anomalies by tracking CDF shifts or trimmed means over time
- Compute real-time analytics dashboards that need fast quantile updates on large datasets
- Replace slower pure-Python t-digest implementations in existing codebases
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes.
fastdigest is actively maintained, has no security vulnerabilities, carries a permissive MIT license, and offers substantial speed gains (480x faster than tdigest in the documented benchmark) with a straightforward API. Install it if you need streaming quantile estimation or online statistics on large datasets and want a fast, dependency-free solution.
Install
fastdigest on PyPI
Before you install
Medium install friction due to compiled wheels; prebuilt binaries are available for Python 3.7–3.14 across macOS, Linux (including musl), and Windows architectures. Active maintenance with recent commits; no runtime dependencies.
License in practice
MIT license (permissive); you may use, modify, and distribute this package freely in commercial and private projects with minimal restrictions.
Quickstart
pip install fastdigest
from fastdigest import TDigest
digest = TDigest.from_values(range(1001))
print(digest.quantile(0.99)) # 99th percentile
print(digest.cdf(990)) # rank of value 990Verify before relying
- Accuracy bounds or error guarantees for quantile estimates compared to exact computation
- Memory consumption scaling with digest size or data distribution
- Whether the package supports Python 3.14 and 3.13 in practice or only declares support
Package facts
| License | permissive license permissive |
| Python support | Supports the current Python release >=3.7 |
| Install friction | Medium. Platform-specific wheel |
| Runtime dependencies | None |
| Maintenance | Actively maintained 152 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 152,643 / month, #10,897 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Development Status :: 4 - BetaIntended Audience :: DevelopersIntended Audience :: Science/ResearchLicense :: OSI Approved :: MIT LicenseOperating System :: OS IndependentProgramming Language :: Python :: 3Programming Language :: Python :: 3.10Programming Language :: Python :: 3.11Programming Language :: Python :: 3.12Programming Language :: Python :: 3.13Programming Language :: Python :: 3.14Programming Language :: Python :: 3.7Programming Language :: Python :: 3.8Programming Language :: Python :: 3.9Programming Language :: Python :: Free ThreadingProgramming Language :: Python :: Implementation :: CPythonProgramming Language :: RustTopic :: Scientific/EngineeringTopic :: Scientific/Engineering :: Information AnalysisTopic :: Scientific/Engineering :: MathematicsTopic :: Software Development :: Libraries :: Python Modules |
Evidence: fastdigest-0.12.0-cp310-cp310-macosx_10_12_x86_64.macosx_11_0_arm64.macosx_10_12_universal2.whl; fastdigest-0.12.0-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl; fastdigest-0.12.0-cp310-cp310-manylinux_2_17_armv7l.manylinux2014_armv7l.whl; fastdigest-0.12.0-cp310-cp310-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl; fastdigest-0.12.0-cp310-cp310-manylinux_2_17_s390x.manylinux2014_s390x.whl; fastdigest-0.12.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl; fastdigest-0.12.0-cp310-cp310-manylinux_2_5_i686.manylinux1_i686.whl; fastdigest-0.12.0-cp310-cp310-musllinux_1_2_aarch64.whl; fastdigest-0.12.0-cp310-cp310-musllinux_1_2_armv7l.whl; fastdigest-0.12.0-cp310-cp310-musllinux_1_2_i686.whl; fastdigest-0.12.0-cp310-cp310-musllinux_1_2_x86_64.whl; fastdigest-0.12.0-cp310-cp310-win32.whl; fastdigest-0.12.0-cp310-cp310-win_amd64.whl; fastdigest-0.12.0-cp311-cp311-macosx_10_12_x86_64.macosx_11_0_arm64.macosx_10_12_universal2.whl; fastdigest-0.12.0-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl; fastdigest-0.12.0-cp311-cp311-manylinux_2_17_armv7l.manylinux2014_armv7l.whl; fastdigest-0.12.0-cp311-cp311-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl; fastdigest-0.12.0-cp311-cp311-manylinux_2_17_s390x.manylinux2014_s390x.whl; fastdigest-0.12.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl; fastdigest-0.12.0-cp311-cp311-manylinux_2_5_i686.manylinux1_i686.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “t-digest quantile estimation”
- fastdigestfastdigest provides a Rust-backed t-digest implementation for…
- tdigestImplements Ted Dunning's t-digest data structure for efficient…
- quantile-forestQuantile regression forests for estimating conditional quantiles and…
Give your agent the search over MCP, or paste the wish link into any chat.
More Scientific/Engineering packages
NumPy provides an N-dimensional array object and a comprehensive suite of mathematical, linear algebra, Fourier transform, and random number functions for scientific computing in Python.
pandas provides fast, flexible data structures (Series and DataFrame) for loading, cleaning, transforming, and analyzing labeled or relational data in Python.
scipy provides numerical algorithms for mathematics, science, and engineering—including optimization, integration, linear algebra, Fourier transforms, signal and image processing, and ODE solvers—built on numpy arrays.
scikit-learn provides a comprehensive Python library for supervised and unsupervised machine learning, including classification, regression, clustering, dimensionality reduction, and model evaluation tools built on NumPy and SciPy.
Install it if you need to train, evaluate, or deploy supervised or unsupervised learning models.
dill extends Python's pickle module to serialize and deserialize a much wider range of Python objects, including functions, lambdas, classes, and interpreter sessions, to byte streams for storage or network transmission.
Multiprocess is an enhanced fork of Python's standard multiprocessing library that uses dill for better serialization, allowing you to spawn processes with a threading-like API and share complex objects between them.
Install it if you use multiprocessing and encounter pickle serialization limits with lambdas or complex objects.
See also tdigest · ddsketch · datasketches · accumulation-tree · crick · quantile-forest · tensor-grep · phonors · moviepilot-rust · runstats