The release of the Python library for scientific computing, NumPy 1.22, has become available. It is aimed at working with multidimensional arrays and matrices and provides a large collection of functions implementing various algorithms related to matrix operations. NumPy is one of the most sought-after libraries used for scientific calculations. The project code is written in Python with optimizations in C and is distributed under the BSD license.
In the new version:
- The work on defining annotations for the main namespace has been completed.
- A preliminary version of the Array API has been proposed, corresponding to the Python Array API standard and implemented in a separate namespace. The new API aims to prepare standard functions for working with arrays that can also be used in applications based on other libraries, such as CuPy and JAX.
- A DLPack backend has been implemented, providing support for the namesake format for exchanging content of arrays (tensors) between different frameworks.
- A set of methods for functions related to the concepts of quantile and percentile has been added.
- A new customizable memory manager (numpy-allocator) has been introduced.
- Work on optimizing functions and platforms using SIMD vector instructions has continued.
- Support for Python 3.7 has been discontinued; versions 3.8-3.10 are required for operation.
Source: opennet.ru
