---
id: pydoe3
version: "1.6.2"
license: BSD-3-Clause
license_treatment: permissive
maintenance: abandoned
---

# pyDOE3 — Design of experiments for Python

License: permissive · Maintenance: abandoned · Downloads: 78.3K/mo

## What it is and what it does

pyDOE3 is a Python package for constructing experimental designs used in scientific research, engineering, and statistics. It provides a comprehensive suite of design methods spanning factorial designs (full and fractional), response-surface designs (Box-Behnken, central-composite), randomized sampling (Latin hypercube, random k-means), low-discrepancy sequences (Sobol, Halton, Korobov), Taguchi designs for robust optimization, and advanced optimal design algorithms with multiple search strategies. The package depends on numpy and scipy for numerical computation.

Scientists and engineers use pyDOE3 to systematically plan experiments that explore parameter spaces efficiently, reduce the number of required trials, or optimize designs according to statistical criteria. It is typically used in the planning phase of research to generate design matrices that guide data collection.

Use it for:

- Generate full or fractional factorial designs to explore all factor combinations or their subsets in experimental studies
- Create Latin hypercube samples for efficient exploration of high-dimensional parameter spaces with uniform coverage
- Design response-surface experiments (Box-Behnken, central-composite) to model and optimize system behavior
- Apply Taguchi design methods to identify robust parameter settings that minimize sensitivity to noise
- Construct low-discrepancy sequences (Sobol, Halton) for quasi-random sampling in Monte Carlo simulations
- Use optimal design algorithms with various criteria (D, A, G, etc.) to maximize information from limited experimental runs

## Worth the install?

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

pyDOE3 provides functions to construct experimental designs for scientific and engineering research, including factorial, response-surface, randomized, low-discrepancy, sampling, Taguchi, optimal, and sparse grid designs.

Yes, if you need to construct experimental designs for research or engineering. The package is mature (marked Production/Stable), has low install friction, and covers a broad range of design methods. However, the archived repository and abandoned maintenance status mean you should not expect active bug fixes or compatibility updates; use it for stable, well-tested design generation but plan to fork or maintain locally if you encounter issues with future Python or dependency versions.

## Install

pip install pydoe3
uv add pydoe3
poetry add pydoe3

## Installing pyDOE3

Before you install: Low install friction with only numpy and scipy as runtime dependencies. Repository is archived and marked abandoned as of the fact sheet date, though the latest release is recent (2026-01-12); maintenance status is a concern for long-term support.

License in practice: BSD-3-Clause (permissive) allows commercial and private use with attribution and liability disclaimer; no restrictions on modification or redistribution.

Quickstart:

pip install pyDOE3

import numpy as np
from pyDOE3 import fullfact, bbdesign, lhs

# Full factorial design for 3 factors, 2 levels each
design = fullfact([2, 2, 2])

# Box-Behnken design for 3 factors
bb = bbdesign(3)

# Latin Hypercube Sampling for 5 factors, 10 samples
lhs_samples = lhs(5, samples=10)

Verify before relying:

- Whether the archived repository status will affect future bug fixes or compatibility with newer numpy/scipy versions
- Whether all design functions are actively tested against current Python 3.9–3.12 environments

## Package facts

- License: BSD-3-Clause (permissive)
- Python support: unspecified
- Install friction: low
- Maintenance: abandoned
- Downloads: 78.3K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

experimental design library, design of experiments DOE, factorial design generator, latin hypercube sampling, response surface methodology, optimal experimental design, sparse grid design, taguchi design orthogonal arrays, experimental-design, doe, optimization

[View on SkillFed](https://skillfed.io/packages/pydoe3) · [View on PyPI](https://pypi.org/project/pydoe3/)