Data Apis: Standardization of N-dimensional arrays and dataframes
summary
A consortium called Data Apis claims a working array spec with over 125 standardized functions, support for 11 data types and device-agnostic interchange (CPU/GPU), plus a beta, library‑independent test suite and tools: an API comparison tool and a Python record API. Numpy adoption via NEP 47 and experimental modules in NumPy and CuPy already exist. Dataframe work targets a minimal interchangeable table (chunked columns, 1‑D columns, missing support) while debates about indexing and multi‑indexing continue.
AI-written summary · about
chapters (11)
- 0:00Intro
- 0:28PYTHON DATA ECOSYSTEM
- 1:21A LITTLE BIT OF HISTORY.
- 5:46OBJECTIVES OF THE CONSORTIUM
- 7:08FIRST YEAR PROGRESS
- 10:11TOOLS
- 12:13ACTUAL STATE
- 14:22FUTURE WORK
- 18:09CHARACTERISTICS
- 18:36RESOURCES
- 20:07NEXT STEPS
related talks
28mA Hitchhiker's Guide to the Array API Standard Ecosystem
05mBackstage interview
41mWe don't dataframe shame: A love letter to dataframes
50mNumPy Versus APL
01h 22mUsing Dask for Parallel Computing in Python
40mThe Future of NumPy Indexing.
01h 20mAdvanced Data Storage
27mAPL-Inspired Techniques for Advanced NumPy