ROGP

ROGP is a tool for including the mean and covariance function of Gaussian Processes in Pyomo models. It was developed for the work outlined in the following paper: A robust approach to warped Gaussian process-constrained optimization

Also see our Youtube Video on this project.

If you use this tool, please cite our paper as:

@misc{Wiebe2020robust,
      title={A robust approach to warped Gaussian process-constrained optimization}, 
      author={Johannes Wiebe and Inês Cecílio and Jonathan Dunlop and Ruth Misener},
      year={2020},
      eprint={2006.08222},
      archivePrefix={arXiv},
      primaryClass={math.OC}
}

Installation

python -m pip install git+https://github.com/cog-imperial/rogp.git

Usage

# Normalize training set x, y
norm = rogp.Normalizer()
norm.scale_by(x, y)
X, Y = norm.normalize(x, y)

# Train GP using GPy
kernel = GPy.kern.RBF(input_dim=1, variance=1., lengthscale=1.)
gp = GPy.models.GPRegression(X, Y, kernel)
gp.optimize(messages=True)

# Create ROGP object
gp = rogp.from_gpy(gp, norm=norm)

# Create Pyomo model and variable
m = pe.ConcreteModel()
m.x = pe.Var(range(n))

# Generate Pyomo expressions
xvar = rogp.pyomo_to_np(m.x)
mu = gp.predict_mu(xvar)
cov = gp.predict_cov(xvar)

Acknowledgements

This work was funded by the Engineering & Physical Sciences Research Council (EPSRC) Center for Doctoral Training in High Performance Embedded and Distributed Systems (EP/L016796/1) and an EPSRC/Schlumberger CASE studentship (EP/R511961/1, voucher 17000145).

Author: Johannes Wiebe