pysurrogate

A unified surrogate-modeling toolkit for Python — sampling, fitting, and model selection in one place.

pysurrogate consolidates a DACE Kriging engine and a multi-backend model layer into a single, consistently-styled package:

• pysurrogate.dace — the DACE Kriging engine: the Dace model with pluggable regression trends, correlation kernels, and theta optimizers.
• (more backends — RBF, SVR, KNN, IDW, ... — and the model-selection layer are being folded in.)

Install

pip install pysurrogate

Quick start

import numpy as np
from pysurrogate import Dace
from pysurrogate.dace import Gaussian, ConstantRegression

X = np.random.random((20, 2))
y = (X ** 2).sum(axis=1, keepdims=True)

model = Dace(regr=ConstantRegression(), corr=Gaussian(), theta=1.0, thetaL=1e-3, thetaU=20.0)
model.fit(X, y)

pred = model.predict(np.random.random((5, 2)), mse=True)
print(pred.y, pred.mse)

Development

This project is driven by pyclawd. See AGENTS.md for the command contract; the short version:

pyclawd check     # format + lint + typecheck + descriptions + tests
pyclawd test      # run the suite