---
id: dask-glm
version: "0.4.0"
license: BSD-3-Clause
license_treatment: permissive
maintenance: aging
---

# dask-glm — Generalized Linear Models with Dask

License: permissive · Maintenance: aging · Downloads: 110.4K/mo

## What it is and what it does

Dask-glm is a library for fitting generalized linear models in a distributed setting using Dask, the parallel computing framework. It integrates with Dask's task scheduler to enable model training across multiple machines or cores, leveraging dependencies like scikit-learn for model definitions, numpy and scipy for numerical computation, and numba for JIT-compiled performance. The package targets scientific and machine-learning workflows where dataset size or computational cost makes single-machine fitting impractical.

However, the package's own documentation explicitly warns that it is not ready for use, and it has not received recent updates. The Beta development status and aging maintenance signal suggest the project is dormant or abandoned. Installation is straightforward (low friction, pure-Python wheel), but adopting it for production work carries substantial risk given the explicit non-readiness statement and lack of recent activity.

Use it for:

- Training logistic regression or other GLMs on datasets too large to fit in a single machine's memory using Dask clusters.
- Prototyping distributed statistical models in research or exploratory analysis where the library's experimental status is acceptable.
- Learning how to structure distributed machine-learning code using Dask's primitives and task graphs.
- Benchmarking or comparing distributed GLM implementations across different parallelization frameworks.

## Worth the install?

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

Distributed generalized linear model fitting using Dask for parallel computation across clusters or multi-core systems.

No. The package explicitly states it is not ready for use, has not been updated recently, and carries Beta status with an aging maintenance signal. While the install friction is low and the license is permissive, the lack of readiness and dormant development make it unsuitable for production or even reliable experimental work. Consider alternatives with active maintenance.

## Install

pip install dask-glm
uv add dask-glm
poetry add dask-glm

## Installing dask-glm

Before you install: Low friction install with a pure-Python wheel, but the package is explicitly marked as not ready for use and has not been updated recently despite being in Beta status. The maintenance signal is aging.

License in practice: BSD-3-Clause permissive license allows commercial and private use with minimal restrictions, requiring only license and copyright notice inclusion.

Quickstart:

pip install dask-glm
import dask_glm
# See documentation at https://dask-glm.readthedocs.io for usage examples

Package explicitly states 'This library is not ready for use' in its own description; production use is not recommended.

Verify before relying:

- Current API stability and whether the 'not ready for use' warning has been lifted since the last release.
- Whether distributed and numba dependencies are optional or required for all use cases.
- Specific GLM models and solvers currently implemented and their performance characteristics.

## Package facts

- License: BSD-3-Clause (permissive)
- Python support: supports_current
- Install friction: low
- Maintenance: aging
- Downloads: 110.4K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

distributed glm fitting, dask generalized linear models, parallel logistic regression, scalable statistical modeling, dask machine learning, distributed statistical inference, large-scale glm training, distributed-computing, generalized-linear-models, dormant

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