GPmp-contrib

gpmp-contrib extends gpmp with computer-experiment objects, multi-output model containers, Matérn model classes, sequential design procedures, set estimation tools, plots, and relaxed Gaussian-process utilities.

Use gpmp directly for core GP models, covariance functions, numerical backend operations, and low-level parameter selection. Use gpmp-contrib when a script needs a ComputerExperiment, a ModelContainer, a sequential strategy, a test problem, or reGP.

Main components

• Model containers and Matérn classes:
  • Model_ConstantMean_Maternp_ML
  • Model_ConstantMean_Maternp_REML
  • Model_ConstantMean_Maternp_REMAP
  • Model_ConstantMean_Maternp_REMAP_logsigma2
  • Model_ConstantMean_Maternp_REMAP_logsigma2_and_logrho_prior
  • Model_Noisy_ConstantMean_Maternp_REML
• Prior access on REMAP classes with priors:
  • get_prior(...)
  • set_prior(...)
• Sequential strategies:
  • fixed candidate sets with SequentialStrategyGridSearch
  • SMC particle sets with SequentialStrategySMC
  • BSS-style particle sets with SequentialStrategyBSS
• Optimization and set-estimation modules:
  • expected improvement in gpmpcontrib.optim.expectedimprovement
  • excursion sets in gpmpcontrib.optim.excursionset
  • set inversion and Pareto utilities in gpmpcontrib.optim
• reGP utilities in gpmpcontrib.regp.
• Parameter posterior sampling through ModelContainer.sample_parameters(...).

Package layout

• gpmpcontrib/models/: Matérn model container classes.
• gpmpcontrib/modelcontainer.py: multi-output model container.
• gpmpcontrib/sequentialprediction.py: observation storage and prediction updates.
• gpmpcontrib/sequentialstrategy.py: sequential decision strategies.
• gpmpcontrib/optim/: EI, excursion-set, set-inversion, and Pareto tools.
• gpmpcontrib/regp/: relaxed Gaussian-process utilities.
• examples/: scripts using the public objects.
• docs/: Sphinx documentation.

Requirements

• Python >=3.9
• gpmp >= 0.9.37
• numpy
• scipy>=1.12.0
• matplotlib

Installation

Install the released package from PyPI:

pip install gpmp-contrib

This installs gpmp and the other runtime dependencies declared in pyproject.toml.

For development, clone the repository and install it in editable mode:

git clone https://github.com/gpmp-dev/gpmp-contrib.git
cd gpmp-contrib
pip install -e .

When testing against a local gpmp checkout, install gpmp first, then install gpmp-contrib in editable mode.

Minimal example

import gpmpcontrib as gpc

problem = gpc.ComputerExperiment(
    1,
    [[-1.0], [1.0]],
    single_function=lambda x: x**2,
)

The full documentation starts with docs/source/getting_started.rst and then continues through the user guide. The examples section documents model construction, noisy observations, expected improvement, excursion sets, set inversion, and reGP.

Documentation

The documentation is available at https://gpmp-dev.github.io/gpmp-contrib/.

To build it locally, install the documentation dependencies and build the HTML pages:

pip install -r docs/requirements.txt
cd docs
sphinx-build -M html source _build -E

Generate the static example figures with:

cd docs
python make_example_results.py

Authors

See AUTHORS.md.

How to cite

If you use GPmp-contrib in your research, please cite it as follows:

@software{gpmpcontrib2026,
  author       = {Emmanuel Vazquez},
  title        = {GPmp-contrib},
  year         = {2026},
  url          = {https://github.com/gpmp-dev/gpmp-contrib},
  note         = {Version 0.9.37},
}

Update the version number when citing another release.

Copyright

Copyright (C) 2022-2026 CentraleSupelec

License

GPmp-contrib is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

GPmp-contrib is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with GPmp-contrib. If not, see http://www.gnu.org/licenses/.