Notice: Please be sure to update your copy of Torch7 prior to work with bot7.

# bot7 # __bot7__ is a framework for Bayesian Optimization implemented in Torch7. For it's companion package on Gaussian Processes, please refer to [gpTorch7](https://github.com/j-wilson/gpTorch7). Please feel free to pass along suggestions, requests, or feedback. ## Package Content ##

Directory	Content
bots	Automated experiment runners
models	Target function models
scores	Acquisition functions (incl. EI, UCB)
grids	Candidate grids
samplers	Sampling methods (e.g. slice sampling)
utils	Utility methods
nnTools	Convenience methods for neural networks
benchmarks	Benchmarking functions
examples	Usage examples

Installation and Usage

You can install bot7 via luarocks provided by the torch distribution.

git clone https://github.com/j-wilson/bot7
cd bot7
luarocks make

After installing bot7, you can load the package inside Torch via:

require('bot7');
th> bot7.
bot7.grids.       bot7.models.      bot7.samplers.    
bot7.bots.        bot7.scores.      bot7.utils.

Developers' Notes

Further documentation coming soon.