Installation Guide¶

This guide explains how to install pyGSK — the Generalized Spectral Kurtosis Toolkit — using different methods depending on your environment and purpose.
Whether you’re an end-user running SK analyses or a developer contributing to the codebase, installation is straightforward and fully cross-platform.

Requirements¶

Python ≥ 3.9
Operating systems: Linux, macOS, or Windows
Recommended tools:
pip, venv, or conda
git for source installation

Option 1: Install from PyPI (recommended)¶

The easiest way to install the latest stable release:

pip install pygsk

To verify installation:

python -m pygsk --version

You should see output similar to:

pyGSK 1.0.0

Option 2: Install from GitHub (latest development version)¶

If you prefer to use the most recent development branch:

pip install git+https://github.com/suncast-org/pygsk.git

This command automatically clones and installs the package in your environment.

Option 3: Clone and install locally (developer mode)¶

For contributors or advanced users:

git clone https://github.com/suncast-org/pygsk.git
cd pygsk
pip install -e .

The -e flag installs the package in editable mode, allowing immediate testing of code changes without re-installation.

Optional Dependencies¶

Some extra functionalities (e.g., plotting, testing) require additional packages:

Feature	Extra Packages	Install Command
Plotting	`matplotlib`, `numpy`	`pip install pygsk[plot]`
Development & Testing	`pytest`, `flake8`, `bumpver`	`pip install pygsk[dev]`

Quick Test¶

To confirm a working installation, run:

pygsk sk-test --M 128 --N 64 --pfa 1e-3

Expected output:

Computing SK thresholds...
Lower = 0.853, Upper = 1.185  (for pfa = 1.00e-03)

Troubleshooting¶

Issue	Possible Cause	Solution
`command not found: pygsk`	PATH not updated	Reopen terminal or use `python -m pygsk`
`ModuleNotFoundError`	Wrong environment	Activate your virtual environment before running
Permission errors on Windows	User privileges	Add `--user` flag to `pip install`

Uninstallation¶

To remove pyGSK:

pip uninstall pygsk

Next Steps¶

Learn how to run SK analyses in usage.md
Explore the command-line tools in cli_guide.md
Understand the theoretical background in theory.md