Jupyter notebooks are a way of sharing executable experiments. They're widely used as standard in data science and machine learning, as well as being used more broadly for education, documentation and experimentation. This is a collection of notebooks on assorted topics in artificial intelligence, computer science and mathematics.
The notebooks are mainly in Python but other programming languages may be used as well.
| Link | Description | |
|---|---|---|
| OCR | Comparing different modes of Tesseract OCR and Google Vision API models for text detection and recognition |  |
| Object Detection | Comparing different methods for detecting checkboxes on a printed form | |
- Python 3
- Virtualenv
- Pip
You'll need to set up Virtualenv (this is optional but it makes it easier to manage Python dependencies) and install the dependencies. Once you've checked out the repository, run these commands in the root directory:
n.b. Depending on how Python is configured on your system it might be named python or python3. Same goes for pip/pip3
python3 -m venv env
env/bin/activate
pip3 install -r requirements.txtVirtualenv creates an env directory where Python dependencies for the project will be stored. This directory is ignored in Git.
Now you can start Jupyter by running jupyter notebook. This should open a tab in your web browser allowing you to browse Jupyter files, but if not then you can try visiting http://localhost:8888.
If you've already set up Virtualenv and installed dependencies then you can reactivate the virtual environment at any point by running env/bin/activate.
Instead of using Jupyter in the browser you can embed it directly into everybody's favourite text editor. There are a couple of approaches to doing this, but we'll stick with the Emacs Ipython Notebook extension. After following the installation instructions on the website you'll be ready to either:
- Run a Jupyter process within Emacs with
M-x ein:run - Connect to an existing Jupyter process with
M-x ein:login