NuMojo v0.5 Release Highlights
NuMojo v0.5 introduces powerful features and optimizations for numerical computations. Explore the updates below, and check our updated documentation site.
---
- Complex Arrays: Added
- Type Coercion:
- Data Buffer Management: New
- Array Iterators: Use
- Tensor Conversion: Seamlessly convert between
- Linear Algebra: Added
---
- Array Manipulation: Improved memory layout flags and reshaping methods.
- Enhanced Methods: Safer
- Print Customization: New
---
- Replaced
- Removed
---
- Fixed negative index overflow in
- Improved
---
Visit our documentation site for:
- Guides for all features.
- API tutorials and examples.
- Details on the latest changes.
Enjoy NuMojo v0.5's features! Learn more at our documentation.
---
New Features
- Complex Arrays: Added
ComplexNDArray, ComplexSIMD, and CDType for advanced complex number operations.- Type Coercion:
TypeCoercion struct handles type conversions in math functions.- Data Buffer Management: New
OwnData type improves memory management in NDArray and Matrix.- Array Iterators: Use
NDIter for flexible array traversal with NDArray.nditer().- Tensor Conversion: Seamlessly convert between
NDArray and MAX Tensor.- Linear Algebra: Added
det for determinants and Householder-based QR decomposition.---
Improvements
- Array Manipulation: Improved memory layout flags and reshaping methods.
- Enhanced Methods: Safer
load/store with boundary checks and expanded indexing options.- Print Customization: New
PrintOption for array formatting.---
Deprecations
- Replaced
order and datatype with flags and dtype.- Removed
offset and coefficient for simpler memory management.---
Fixes
- Fixed negative index overflow in
get and set.- Improved
__setitem__ for better handling of slices and integers.---
Documentation Updates
Visit our documentation site for:
- Guides for all features.
- API tutorials and examples.
- Details on the latest changes.
Enjoy NuMojo v0.5's features! Learn more at our documentation.
