Skip to content

Home

psyphy logo

Active-learning-driven adaptive experimentation in psychophysics

Documentation | Examples | Contributing

Install

psyphy supports Python 3.10 and later. We recommend installing psyphy under a virtual environment. Once you've created a virtual environment for psyphy and activated it, you can install psyphy using pip:

1
pip install psyphy

If you're a developer or want the latest features, install from GitHub: 

1
2
3
git clone https://github.com/flatironinstitute/psyphy.git
cd psyphy
pip install -e .

What is in psyphy?

psyphy is an open-source, JAX-based package for psychophysics research. It fits Bayesian models of perceptual discrimination, currently the Wishart Process Psychophysical Model (WPPM), on CPU or GPU. A fit gives either a single maximum a posteriori (MAP) estimate or samples from the full posterior with the No-U-Turn Sampler (NUTS). The aim is to use that posterior to choose the next trial in an adaptive experiment.

The examples so far are on human color perception, but the model applies to any stimulus space where discrimination noise varies smoothly.