DataNvim is a custom Neovim distribution prepared for Data Science and to easily run code and interact with Jupyter Notebooks without ever leaving your terminal. It also provides a well structured config that makes it an easy starting point to extend it as you want.

• GCC
• Kitty 0.28+
• Neovim 0.9.4+
• Python 3.10+
• Ripgrep 11.0+
• Quarto
• Npm 10.8.1+
• ImageMagick
• Lua 5.1
• Required python packages:
  • pynvim
  • jupyter_client
  • cairosvg
  • plotly and kaleido
  • pyperclip
  • nbformat
  • pillow
  • ipykernel

• Carefully designed configuration file structure for easy extension and modification
• Beautiful colorscheme from catppuccin
• Powerful file explorer with nvim-tree
• Fuzzy search and preview files, words and gitfiles thanks to telescope
• Preconfigured highlighting and Lsp Autocompletion for Python thanks to treesitter and nvim-lsp-config
• Extensible formatting, diagnostics, code actions and completion with none-ls and ruff
• Complete Jupyter Notebook interaction set up thanks to molten.nvim, quarto, and jupytext
• Simple but beautiful status bar thanks to lualine.nvim
• Easy buffer navigation with bufferline.nvim

• Pull DataNvim's image from ghcr.io/noopeeks/data-nvim:latest

docker run -it ghcr.io/noopeeks/data-nvim:latest /bin/bash
# Once inside prompt, run nvim
nvim

• Try it out, there is a demo .ipynb file in the home directory!

• Install all the listed dependencies
• Install Kitty
• Install luarocks, Lua language's package manager.
• Install the magick rock with: luarocks --local --lua-version=5.1 install magick
• Create a virtual environment for Neovim in ~/.virtualenvs/

mkdir ~/.virtualenvs
cd ~/.virtualenvs
python -m venv neovim

• Activate the virtual environment and install the beforementioned python packages with:

source ~/.virtualenvs/neovim/bin/activate
pip install ipykernel jupytext pynvim jupyter_client cairosvg plotly kaleido pyperclip nbformat pillow
python -m ipykernel
deactivate

• Install ipykernel and jupytext in your project-scoped virtual environment

cd whatever/directory/your/project/is/in
source venv/bin/activate
pip install ipykernel jupytext
python -m ipykernel install --user --name project_name

• Backup your current Neovim configuration

mv ~/.config/nvim ~/config/nvim.bak

• Install DataNvim's configuration

git clone https://github.com/NoOPeEKS/DataNvim.git ~/.config/nvim && nvim

• Run the following commands:

:UpdateRemotePlugins
:MasonInstallAll

• Open your notebook with the virtual environment activated and load the kernel with :MoltenInit project_name or Space + m + i.
• Start executing cells with Keybindings

DataNvim is a very straightforward Neovim distribution that aims to be a batteries-included starting point for Data Science in Neovim. Thus, its directory structure for configuration is meant to be easy to understand:

├── init.lua                <- Neovim's entrypoint. You should probably not modify anything here. Just require your own files.
├── lua                     <- Configuration for DataNvim goes here.
│   ├── plugins.lua         <- Empty table. It will contain everything inside plugins/
│   ├── vim-options.lua     <- Vim options and general movement keybindings.
│   ├── plugins             <- This folder contains all of your plugins. To install new ones, create a file and make it return your plugin.
│   │   └── plugin_name.lua <- Returns a table with the plugin to install and its configuration.