DOE Utilities

Overview

Use drcutils.doe.generate_doe to create and summarize full factorial, Latin hypercube, and fractional 2-level designs.

Quick Start

from drcutils.doe import generate_doe

result = generate_doe(
    kind="lhs",
    factors={"temperature": (20.0, 80.0), "pressure": (1.0, 3.0)},
    n_samples=12,
    seed=42,
    randomize=True,
)

design = result["design"]
print(result["summary"])
print(result["interpretation"])

Optional Extras

For larger fractional factorial workflows, install optional DOE support:

pip install drcutils[doe]

CLI

drc-doe --kind lhs --factors-json '{"temperature": [20, 80], "pressure": [1, 3]}' --n-samples 12 --out artifacts/doe.csv

API Reference

High-level DOE generation workflows.

drcutils.doe.generate.generate_doe(*, kind, factors, n_samples=None, seed=0, center_points=0, replicates=1, randomize=True)[source]

Generate and summarize a DOE table.

Parameters:

kind – Design kind: full, lhs, or frac2.
factors – Factor specification map.
n_samples – Number of samples for LHS.
seed – Random seed.
center_points – Number of center points to append.
replicates – Number of full design replicates.
randomize – Whether to randomize run order.

Returns:

Structured dictionary with design, summary, interpretation, and warnings.