Stratax v0.3 is now available on PyPI.
v0.3.0 introduced the latest set of features, and v0.3.1 adds cross-platform support for Windows, Linux, and macOS.
Install with:
pip install stratax
I've also been improving the documentation and overall project structure as Stratax continues to grow.
Next, I'm doing a full review of the codebase before moving on to the next major feature.
#Python #CPP #OpenSource #ScientificComputing
Broadcasting is one of those features that looks like magic from the outside.
(2, 3) + (3,)
works without copying data or writing special-case code.
The interesting part isn't the arithmetic; it's mapping the output indices back to their respective input tensors.
That's what I'm building next for Stratax.
#cpp #cplusplus #scientificcomputing #opensource
This week I've been working on implementing reductions for Stratax.
Functions like sum, mean, and argmax seem simple from the outside, but supporting reductions along arbitrary axes is an interesting problem.
A reduction doesn't just compute a value; it changes the tensor's shape while preserving the remaining dimensions.
Building these features from scratch has given me a much deeper appreciation for the engineering behind numerical libraries. And the mathematics that makes them possible.
Still a work in progress, but I'm excited to share the progress and hear feedback from other developers.
GitHub: https://t.co/bdZnXrFmrk
#cpp #python #opensource #numericalcomputing
Stratax v0.2.0 is now available!
This release brings the Python API to feature parity with the C++ API (with the dtype system still planned for a future release).
Highlights:
• Feature complete C++ and Python APIs
• Consistent interface across both languages
• Improved Python bindings and wrappers
• Documentation and general improvements
With both APIs now aligned, future releases can focus on expanding Stratax's numerical capabilities rather than catching one language up to the other.
Feedback, feature requests, and contributions are always welcome!
Repo: https://t.co/LEqRHpy51o
PyPI Package: https://t.co/dF5OK51dvD
#cpp #python #opensource #scientificcomputing #numericalcomputing
What is Stratax?
Stratax is a modern scientific computing library for C++ and Python.
The goal is to provide fast, intuitive multidimensional arrays with a clean API for scientific computing, numerical analysis, and engineering applications.
I'm building it as an open-source project from the ground up, with a C++20 core and a Python interface.
This is just the beginning; there's still plenty planned, including reductions, statistics, broadcasting, and linear algebra.
I'd love to hear what features you'd find most useful in a scientific computing library.
#cpp #python #opensource #scientificcomputing #numericalcomputing
Excited to share the first public release of Stratax (v0.1.0).
Stratax is an open-source scientific computing library for C++ and Python focused on vectors, matrices, tensors, and multidimensional arrays.
It's still early, and I'm looking forward to improving it through community feedback.
GitHub Page: https://t.co/bdZnXrFUgS