phy documentation

phy is an open-source graphical interface for visualizing and manually curating large-scale electrophysiology datasets. It is optimized for high-density probes, including Neuropixels, and provides both an interactive curation workflow and Python extension points for specialized labs.

Current stable release: phy 2.1.0. The source tree is developing phy 2.1.1. See the release notes and changelog.

Start here

New users should follow these pages in order:

1. Install phy in a fresh Python environment.
2. Prepare and validate a dataset.
3. Follow the first-ten-minutes quickstart.
4. Learn the GUI selection model and similarity-guided workflow.
5. Read how saving and output files work before curating valuable data.

Supported workflows

The Template GUI is the primary workflow for template-based spike sorters such as KiloSort and SpyKING CIRCUS. It reads NumPy arrays, TSV metadata, and params.py.

The Kwik GUI remains available for legacy klusta/Kwik datasets but is no longer the primary maintained workflow. The experimental Trace GUI opens a continuous raw recording directly. See the command-line reference for all entry points.

Customize and extend phy

Common settings—including spike sampling, caches, state, and keyboard shortcuts—are covered under Configuration and customization. Plugin authors and analysts can continue with the power-user guide, plugin examples, and Python API.