The Future of PyData

from a pandas perspective

Pandas is a middleware layer

Core Principles

  • API
    • stable
    • consistent & intuitive
  • Performant
  • Interoperable
    • CSV,SQL, HDF5, excel,html
    • matplotlib, bokeh, seaborn
  • Extensible

 

  • Multi-core / distributed integration
    • Ipython.parallel
    • bcolz
  • Plotting
    • matplotlib interface
  • Extensions
    • libdynd
    • Index
    • Virtual Dtypes

  • Don't try to do everything

 

Where do we go from here?

Made with Slides.com