The NumPy paper is out!

https://www.nature.com/articles/s41586-020-2649-2">https://www.nature.com/articles/...

NumPy provides an easily readable, expressive, high-level API for array programming. It takes care of the underlying mechanics that make operations fast.

The array programming foundation provided by NumPy, combined with a rich surrounding ecosystem of tools — inside of IPython or Jupyter — forms an interactive environment ideally suited to exploratory data analysis.

NumPy includes protocols to facilitate interoperability with external libraries like PyTorch, Dask, and JAX.

Through such features, NumPy provides a standard API for tensor computation and is a central coordinating mechanism between array technologies in Python.

For more history on NumPy & SciPy, see the background section of the "SciPy 1.0" paper: https://www.nature.com/articles/s41592-019-0686-2">https://www.nature.com/articles/...

NumPy plays a leading role in scientific computing, and continues to evolve in response to the changing landscape of data science.

To build a NumPy that meets the needs of the next decade of data science, we welcome your help!

https://numpy.org/contribute/ ">https://numpy.org/contribut...