Spyglass Workshop 2026

A hands-on workshop covering practical scientific computing tools and the Spyglass neuroscience data framework.

---

Contents

• Pre-Workshop Requirements
• On the Day
• Schedule
• Resources

---

Pre-Workshop Requirements

Please complete all of the following before arriving.

1. Create a GitHub account if you do not already have one.

2. Bring laptop. Linux or macOS preferred. Spyglass may have issues on Windows; Windows users are welcome to attend and report any problems they encounter.

3. Install VS Code and Miniconda (or Mamba, recommended).

4. Install Spyglass using its installer script:

   May take up to 45 minutes on a new machine.

   git clone https://github.com/LorenFrankLab/spyglass
   cd spyglass
   python scripts/install.py --minimal # full install not required

   This creates a spyglass conda environment containing Spyglass and its core dependencies.

   Skip the database setup step — credentials will be provided on the day.

5. Verify the install:

   conda activate spyglass
   python -c "import spyglass; print('OK')"

   You should see OK. If you see an ImportError, re-run the installer or ask for help before the workshop day: Chris.Broz[at]ucsf.edu

6. Brush up on Python basics: functions, classes, for loops, if/else, try/except, and shell commands (pip install, conda activate).

---

On the Day

1. Fork and clone this repository

Click Fork at the top-right of this page → Create fork, then:

git clone https://github.com/<your-username>/SpyglassWorkshop2026
cd SpyglassWorkshop2026

2. Add the workshop to your Spyglass environment

conda activate spyglass
pip install -e .
pre-commit install

This installs the workshop package and its tools (ruff, pytest, pre-commit, jupytext) into the spyglass environment you set up before the workshop.

3. Open VS Code

code .

When VS Code opens, accept any prompts to install the recommended extensions listed in .vscode/extensions.json. Select the spyglass conda environment as your Python interpreter (Ctrl+Shift+P → Python: Select Interpreter).

---

Schedule

G6, Hugh Robson Building, University of Edinburgh

Time	Session
10:00 – 11:30	Session 1: Tools for Scientific Computing
11:30 – 12:30	Break
12:30 – 14:00	Session 2: Spyglass & DataJoint Infrastructure

Session 1: Tools for Scientific Computing (90 min)

Practical tools and workflows for scientific Python development:

• Overview of GitHub: repositories, branches, forks, pull requests, and issues
• Setting up VS Code for Python and Jupyter development
• Working interactively with Jupyter notebooks
• Code style and quality tools: ruff (replaces black and isort)
• Debugging your first installable Python package

Session 2: Spyglass & DataJoint Infrastructure (120 min)

Introduction to relational databases and the Spyglass framework:

• DataJoint infrastructure: Python–MySQL connection
• Table tiers and declaration syntax
• Spyglass design patterns: Parameter, Selection, Analysis, and Merge tables
• Ingesting data into Spyglass
• Designing a custom Spyglass pipeline

---

Resources

• Spyglass documentation
• DataJoint documentation
• VS Code Python docs
• conda cheatsheet
• Jupyter shortcuts
• Markdown syntax
• VSCode shortcuts
• GitHub: forking · branches · pull requests