Package to generate numbers in the unit hypercube subject to constraints.
Sampling a high dimensional space under non-linear constraints.
Initially composed as a technical challenge for Citrine.
As a python script
Install from PyPy
Scaling and future work
- Security considerations of
- Special cases have analytic solutions (linear constraints defining a simplex)
- The current method works best in continuous spaces.
- Consider limits of time and memory
- Improved uniformnity on sampling after certain number of steps (autocorrelation function) D. Chandler, Introduction to Modern Statistical Mechanics (Oxford University Press, 1987
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Hashes for constrained_sampler-maffettone-0.0.2.tar.gz