New & Upgraded Version of NumPy is Here

One of the most popular Python libraries has received a new update, NumPy v1.24.0. This new release is said to be faster and supports Python versions 3.8-3.11. The new upgraded version will enhance the handling and promotion of data types, improve execution speed, and clarify documentation. However, there are also many new and expired deprecations due to changes in promotion and cleanups which can be called a deprecation release.

You can learn more about the update here. 

NumPy is one of the core Python libraries that is renowned for its dynamic features, including high-level syntax, support for various hardware and computing platforms, numerical computing tools, the flexibility of Python with the speed of compiled code, and more. NumPy is also the essential Python package for data scientists. 

Important highlights of the release:

New deprecations: Deprecate fastCopyAndTranspose and PyArray_CopyAndTranspos, Conversion of out-of-bound Python integers and Deprecate msort are a few of the new deprecations that are added. 

Expired deprecations: The normed keyword argument has been removed. Ragged array creation will now always raise a ValueError unless dtype=object is passed. Visual Studio 2015 and the Windows Interix POSIX interop layer are no longer supported. Many others have also been removed. 

New F2PY features and fixes: It includes new features like a new attribute symbol added to polynomial classes. The polynomial classes in the package now have a new symbol attribute which will be used to represent the indeterminate of the polynomial. This can be used to change the value of the variable when printing. 

New "dtype" and "casting" keywords for stacking functions—F2PY now supports wrapping Fortran functions with arguments, including passing Python unicode strings as Fortran character string arguments. The casting and dtype keyword arguments are now available for numpy.stack, numpy.vstack, numpy.hstack.