ConfigSpace

What it is and what it does

ConfigSpace is a Python library for defining and working with configuration spaces—structured representations of algorithm parameters and hyperparameters. It provides a domain-specific language to declare parameters as continuous ranges, discrete integers, or categorical choices, and then sample valid configurations from that space. The library is commonly used in automated machine learning and hyperparameter optimization workflows, where you need to systematically explore parameter combinations.

The package depends on numpy, scipy, and pyparsing for numerical operations and parsing, plus typing_extensions and more_itertools for compatibility and iteration utilities. It supports Python 3.9 and later, is actively maintained, and carries no known security vulnerabilities. The library is permissive-licensed under BSD 3-clause, with a subset of code (the nx subpackage) also under BSD from NetworkX.

Use it for:

• Define hyperparameter spaces for machine learning model tuning in AutoML pipelines
• Sample random or structured configurations for algorithm benchmarking and empirical evaluation
• Represent parameter constraints and dependencies in black-box optimization frameworks like SMAC
• Build configuration search spaces for neural architecture search or algorithm selection tasks

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

ConfigSpace defines and manipulates configuration spaces for hyperparameter optimization and algorithm configuration, supporting continuous, discrete, and categorical parameters with sampling and constraint handling.

Yes. ConfigSpace is a stable, actively maintained library with low install friction, no security issues, and a permissive license. It is purpose-built for hyperparameter optimization and algorithm configuration workflows. Install it if you are working on AutoML, hyperparameter tuning, or algorithm configuration tasks and need a structured way to define and sample parameter spaces.

Install

configspace on PyPI

pip install configspace

uv add configspace

poetry add configspace

Installing ConfigSpace

Before you install

Low friction installation with a pure-Python wheel and five common scientific dependencies (numpy, scipy, pyparsing, typing_extensions, more_itertools). Actively maintained with recent commits and no known vulnerabilities.

License in practice

BSD 3-clause permissive license with an exception: files in ConfigSpace.nx are copied from networkx and licensed under a separate BSD license. Both are permissive and suitable for commercial and private use.

Quickstart

from ConfigSpace import ConfigurationSpace

cs = ConfigurationSpace(
    name="myspace",
    space={
        "a": (0.1, 1.5),
        "b": (2, 10),
        "c": ["mouse", "cat", "dog"],
    },
)
configs = cs.sample_configuration(2)

Verify before relying

• Whether ConfigSpace integrates with specific hyperparameter optimization frameworks beyond SMAC
• Performance characteristics when handling very large configuration spaces
• Support for conditional parameters or hierarchical constraints

Package facts

Evidence: configspace-1.2.2-py3-none-any.whl

Keywords: algorithm, configuration, hyperparameter, optimization, empirical, evaluation, black, box

Tags

More Software Development packages