Zhetapi's API can be further divided into four branches:
The ZHP scripting language revolves around the core API. This branches focuses on symbolic manipulation, and making the language easy to use for beginners.
See more about the core API from ZHP's persepective:
../zhp/zhp ../zhp/libs
Crucial to any numerical analysis (and machine learning) library is the framework to perform basic operations with matrices and vectors. This is possible through Zhetapi's Matric and Vector classes and C++ overloading.
Machine learning can be done with the API, currently through the DNN
class. I originally implemented such facilities to the API for personal usage, but seeing as ML is so ubiquitous and that many students are getting involved in ML, I decided to keep it in the API.
The Zhetapi API also comes with many other facilities that I believe would make algorithm design and testing easier.
activation collection generator gnn image indexable iterator layer linalg matrix module nvarena parametrization plot polynomial tensor token vector cast gradient