spynal: Simple Python Neural Analysis Library

Tools for preprocessing and basic analysis of systems/cognitive neurophysiology data in Python

Covers typical preprocessing and basic analysis steps in neural analysis workflows. Intended users are anyone doing analysis of neurophysiology data, but we particularly aim to be accessible and easy-to-use for people migrating to Python from Matlab or other similar programming languages.

Features include:

Simplicity: Easy-to-use procedural interface; no OOP or complicated data/params structures
Consistency: Consistent function signature ~ analysis(data, labels, axis, extra_param=value)
Modularity: Can use specific functionality without buy-in to an entire processing chain
Foolproof: Extensive documentation and checking for proper function inputs
Flexibility: Detailed parameterization allows customization of analysis to users' needs
Vectorized analysis: Runs in parallel across all channels, no loops or data reshaping needed

Package includes the following modules:

matIO — Painless loading of MATLAB .mat data files into appropriate Python data structures
spikes — Preprocessing and analyses of spiking data
spectra — Spectral analysis and preprocessing of continuous data (LFP/EEG)
sync — Analysis of oscillatory neural synchrony (field-field and spike-field)
info — Measures of neural information about task/behavior variables
randstats — Nonparametric randomization, permutation, and bootstrap statistics
plots — Generation of common plot types; plotting utilities
utils — Numerical and general purpose utilities

Full documentation can be found at: https://spynal.readthedocs.io/en/latest/

Download & installation instructions

NOTE: Currently, spynal can only be installed against Python<=3.11, due to changes in the Import Manager in Python 3.12 that have not been accounted for in all dependent packages. For details, see this Stack Overflow answer: https://stackoverflow.com/a/77364602.

Users can either install the most recent stable release using pip or install the latest updates from the Github repository

Option 1: pip install

Recommended for users who prefer a simple installation and don't require the latest updates

- Open a terminal window, navigate to folder you want repository to live under (cd "parent directory")
- Run: pip install spynal

Option 2: Install from source (Github)

Recommended for developers and for users that would like access to the latest updates

2a: Download repository

- Open a terminal window, navigate to folder you want repository to live under (cd "parent directory")
- Run: git clone https://github.com/sbrincat/spynal.git

2b: Install package

- Navigate into the newly-created spynal folder (cd "parent directory"/spynal)
- For end users:  at command line, run: python setup.py install
- For developers: at command line, run: python setup.py develop
    (this will allow you to edit the library without reinstalling)

Usage

You should now be able to directly import spynal and its modules/functions in your code/notebooks like so:
import spynal
import spynal as spy
import spynal.spectra as spectra
from spynal.spectra import spectrogram

Contributions

We welcome any contributions of new/expanded functionality or more efficient implementations,
and all issues/bug reports.
Submit a pull request/issue on Github or email me at [Github handle] AT [institution] .edu