morebs2 0.1.38

data structures to aid in numerical data generation and clustering

MOREBS

Is a collection of methods and classes written in Python to aid in data generation, specifically vector data. Important classes include:

• BallComp
• ResplattingSearchSpaceIterator

For the source code to this project's latest version, go to https://github.com/changissz/morebs.

For a mathematics paper on the BallComp algorithm, see the link
https://github.com/changissnz/ballcomp.

Documentation for project can be found at _build/html. Documentation first needs to be built. The library Sphinx for generating the documentation is required to be installed.

Updates For Project On and Off

Update: 5/25/25 #3

Deleted the project DER from Github. The project was the original work I did before I refactored it into morebs, and has been sitting dead on Github for a while.

Update: 5/25/25 #2

So the new version is up on Github (0.1.1). I also took the step to delete the s*.txt files that were present from a few commits back. The files were relatively large, and I must have forgotten to exclude the files from being committed to Github.

Update: 5/25/25

I have not done much serious work on this project since February of 2023. Recently, I was working with directed graphs and decided to contribute some code for that topic to this project. I still remember the ideas that started this project, geometric intersection and data generation. On geometric intersection, the Schopenhauer wrote about it in his book, The World as Will and Representation. Even though he did not go into mathematical detail on it, his words left an inspiring impact on my computing thought. The topic of data generation is a pretty big field in computing. Cloud computing, especially, has been a big driver for big data analytics, the counter-active field to data generation. Now that there are present and emerging regulations regarding the "fair" and "benign" use of data in artificial intelligence and related fields, data generation has become very important to some enterprises that wish to train their artificial intelligence systems, but do not have authentic datasets in adequate quantity. I'm not surprised that no one has decided to help contribute code to this project. Not to mean any insult, but big data, machine learning, that kind of stuff really is not a normal person's interest (sorry, populists). Besides, most open-source projects that really take off are heavily funded. I have been out of the academic environment for almost half a decade now. Big data, machine learning, that kind of stuff, was mainly an academic business. It still is, pretty much, because all I ever read about from technology corporations is their business products, consumer-side.

I was reviewing some of the code in this project. The project definitely needs more thorough documentation as well as unit-testing. morebs was originally a project solely for the BallComp algorithm and a data generator, one that travels along an n-dimensional lattice, called NSDataInstruction. NSDataInstruction uses the ResplattingSearchSpaceIterator as the data structure that outputs preliminary values. Then I added a basic polynomial factorization algorithm (PolyFactorEst) and an n-dimensional delineation algorithm (see file deline.py). This was back in January of 2023. Not every algorithm is thoroughly tested, as a reminder.

Right now, I am working on directed graphs, so that is the topic of the new code content for the next version of morebs2 on pypi.org.