This repo, the King Tutte Datamap Codex is a Jupyter Book (v2) full of datamapping notebooks design to run on Google's Colab Jupyter service.
This repo can stand alone by itself but it is also used as a part of the King Tutte Scrolls. Together the two are the King Tutte Datmap SDK. The Scrolls is an Agent Skill, as defined by Anthropic. This Codex is included as a submodule in the Scrolls.
The core software (UMAP, HDBSCAN, Toponomy, EVōC, DataMapPlot, etc.) is technology out of the Tutte Institute, including most of the notebooks. Note that this project and its lighthearted tone are the product of Connoiter; The Tutte Institute had nothing to do with the Codex (except having done all the hard work).
The main value of this project is to gather various pre-existing datamap generating sample code into one currated resource (the Codex -- this repo -- which can be built into a Jupyter Book) and ensure that they all run as notebooks on Colab.
- If a pre-existing resource is a Jupyter notebook (*.ipynb file), it is modified as needed to get it to work out of the box on Colab.
- If a pre-existing resource is a Python script (*.py file) the code is repackaged as a Jupyter notebook and tweaked for execution on Colab.
This Codex was developed for use during conference workshops but also serves well for self-study of how to make King Tutte pipelines, which produce datamaps. For use of the Codex within an Agent Skill, see [the King Tutte Datamap Scrolls(https://github.com/Connoiter/king_tutte_scrolls) which includes this Codex as a git submodule.
The Scrolls project started before the Codex repo. Actually the latter was forked out of the former.
The name, Scrolls, being plural implies multiple separate scrolls because that project is essentially just an agglomeration of multiple separate repos (each a scroll in this metaphor) into one unit which can be packaged as an Agent Skill.
The Codex is a Jupyter Book ergo the name, Codex. (A codex is an ancient type of book wherein individual sheets of paper are bound on one side, as distinct from modern books which might have signatures, that is, pages folded in half and then bound to seem like two pages, etc.). Both scrolls and codices could be made of papyrus.
"A book in a collection of scrolls? Wha?" If really necessary then the metaphor could be tortured by saying a codex can be unbound and then the pages rebound as a scroll. Actually, in the case of the Scrolls the Jupyter notebooks (*.ipynb files) are repurposed into markdown files (the default file type of Agent Skills. So take that, you pedants :P