This library wraps PyMC to estimate Weber ratios in psychology experiments.
This allows for statistics incorporating the reliability of measured Weber ratios, which appears to substantially influence the statistics. The reliability is computed via MCMC on the posterior distribution over Weber ratios, given the behavioral responses.
This tool is currently under heavy development by colala ( http://colala.bcs.rochester.edu/ )
weber-mcmc.py includes python wrappers for PyMC, including several natural priors on Weber ratios.
weber-mcmc.R includes a simple R wrapper. It requires WEBER_PY_PATH to point to where weber-mcmc.py lives
The simulation directory includes some tools to evaluate how well we can estimate Weber ratios from different experimental paradigms (e.g. staircasing)