README.md

---
license: mit
---

# Offline Data Lab v1.0

The Offline Data Lab is a frozen, locally hosted data science toolkit for macOS. A simple double-click launches a fully configured JupyterLab workspace in a uv virtual environment.<br>
This is a Thumb-Drive App (TDA).

- Self contained. All dependencies are stored locally (970 MB) - including the uv binary, the wheels for the packages and the Python 3.10 interpreter.
- Easy to copy and share using a USB Drive or via AirDrop. All dependencies travel with the copy.
- No system-wide installation required.
- Runs completely offline. 
- Includes nine commonly used Python data science packages.

<img src="images/image1.png" alt="User interface" height="300">

## What packages are included?

pandas<br>
numpy<br>
scipy<br>
matplotlib<br>
scikit-learn<br>
xgboost<br>
lightgbm<br>
catboost<br>
jupyter<br>
jupyterlab<br>
tqdm

## What problem does this solve?

It solves two often overlooked problems: dependency on a network connection and the difficulty in sharing sofware directly from one person to another. These features are helpful in places where there's no internet or where the internet connection is metered and expensive. A satellite internet connection is one example.

Whether you're on a plane, in the field, or at a workshop venue with unreliable Wi-Fi, the Offline Data Lab just works. Everything it needs is already there.

## Who is this for?

It's well-suited for students and data analysts who want a local JupyterLab environment for tabular data and standard machine learning.

It's especially helpful to:

- Workshop attendees — a trainer can hand out USB sticks or share a download link. Everyone will be up and running in 5 minutes.

- Field researchers — people doing data collection in remote locations without a reliable internet connection.

- Occasional users — someone who only needs data science tools occasionally and doesn't want a permanent Anaconda installation.

## What is a Thumb-Drive App (TDA)?

- A portable folder-based AI app.
- Runs completely offline.
- No installation required.
- All dependencies are bundled.
- Created to be run on a Silicon Mac.
- Easy to share using AirDrop or a thumb-drive. 
- All dependencies travel with the copy.
- Can be run on an external SSD.
- Double-click to launch.

Designed to bring the power of AI to places where there's no internet.

## How to Use

System Requirements:

- Operating System: MacOS
- Computer: Apple Silicon Mac (M1, M2, M3, etc.)
- RAM: 8 GB
- Free disk Space: 970 MB



1. Download the Offline-Data-Lab-v1.0-TDA.zip file and unzip

2. Right click on the Offline-Data-Lab-v1.0-TDA folder and select: <strong>New Terminal at Folder</strong>
3. MacOS often quarantines downloaded files. To make the launch script executable, paste this command into the terminal and press Enter:<br>

```
cat start-mac-app.command > temp && mv temp start-mac-app.command && chmod +x start-mac-app.command
```
You only need to do this once, during the setup.

4. Open the Offline-Data-Lab-v1.0-TDA folder and double-click this file: <strong>start-mac-app.command</strong><br>
If a MacOS security popup appears, click: "Allow"


5. JupyterLab will open. This may take a few seconds. Please be patient if you don't see anything happening. During normal use the app will open much faster.

## One Command. Complete Control.

You would have noticed that, to make this app executable, this command needed to be typed in the terminal:

```
cat start-mac-app.command > temp && mv temp start-mac-app.command && chmod +x start-mac-app.command
```

To remove this requirement the start-mac-app.command file would need to be code signed. Most macOS apps distributed outside the App Store are code signed using an Apple Developer ID certificate. Code signing requires an active Apple Developer Program membership.

This project does not use code signing because it introduces a point of failure. If the Apple membership lapses, or if Apple revokes the certificate for any reason, signed apps will stop launching cleanly for new users. This risk is unacceptable.

In exchange for the minor inconvenience of typing one command into the terminal, you get a reliable app that has no dependency on any third party. You have complete control.


## How to modify the environment

This app uses a self-contained, frozen environment. The uv binary, Python 3.10 interpreter, and all necessary package wheels are stored locally. To modify the environment, please refer to ENV_SETUP_GUIDE.md in the Offline-Data-Lab-v1.0-TDA/app directory. This guide provides step-by-step instructions for a fresh setup. It can be viewed in any text editor.

## References

- Thumb-Drive App Concept<br>
https://github.com/vbookshelf/Thumb-Drive-App-Concept



## Revision History

Version 1.0<br>
16-March-2026<br>
Prototype. Released for testing.

<br>

## Lessons Learned

- A uv virtual environment cannot simply be copied from one Mac to another, due to the Mac security setup. A new venv has to be created on the new machine.
- When creating shareable software, a local copy of the Python runtime also needs to be shared. This prevents issues that will arise if the user does not have the required Python version available on their machine.
- JupyterLab saves your notebook in a hidden file named .ipynb_checkpoints. This saved hidden personal data needs to be considered when building secure, privacy-first applications.
- To display all hidden files on a Mac press three keys: Shift Cmd . [Shift Command Full-stop]
- To display hidden files in the terminal: ls -a
- Mac has a tmp folder that's cleaned each time the system reboots. When storing things there keep in mind that the large datasets used in data science work often result in the local machine freezing, and then needing to be restarted.
- It's takes a little bit of thinking and downloading but its entirely possible to create an offline, shareable data science envoronment where all dependencies are stored locally.
- Deleting the hidden .venv folder will significantly reduce the file size when sharing a copy of the app.