Unidatas Python Efforts Report to Users Committee 16 September 2014 - - PowerPoint PPT Presentation

unidata s python efforts
SMART_READER_LITE
LIVE PREVIEW

Unidatas Python Efforts Report to Users Committee 16 September 2014 - - PowerPoint PPT Presentation

Unidatas Python Efforts Report to Users Committee 16 September 2014 Community Involvement Attend SciPy regularly Iris, Cartopy, Biggus IPython -> Project Jupyter coLaboratory Conda/Binstar Community Involvement


slide-1
SLIDE 1

Unidata’s Python Efforts

Report to Users Committee 16 September 2014

slide-2
SLIDE 2

Community Involvement

  • Attend SciPy regularly
  • Iris, Cartopy, Biggus
  • IPython -> Project Jupyter
  • coLaboratory
  • Conda/Binstar
slide-3
SLIDE 3

Community Involvement (cont.)

  • Hosting two-day Python workshop
  • Occasional bug fixes and

enhancements to Matplotlib

  • Animation fixes
  • Skew-T support
slide-4
SLIDE 4

NetCDF4-Python

  • Continuing to host on our GitHub
  • Help out occasionally on support
  • Would like to enhance to full

support of NetCDF-4 extended data model

slide-5
SLIDE 5

NetCDF4-Python (cont.)

  • Add support for CDM Remote

access

  • Still relying extensively on Jeff

Whitaker

slide-6
SLIDE 6

PyUDL

  • Python Unidata Library
  • Current collection of Python library

and recipes for working with Unidata technology

  • Main code comes from workshop
slide-7
SLIDE 7

Cloud Computing

  • IPython/Project Jupyter is a clear

solution for using Python for cloud computing

  • Wakari is an out of the box solution
slide-8
SLIDE 8

Cloud Computing (cont.)

  • Florita Rodriguez’s NHC archive

browser written using IPython

  • Potential to use IPython in the cloud

for data-proximate analysis

slide-9
SLIDE 9

SciTools (UKMet)

  • The UKMet office has fully

embraced Python

  • IRIS data analysis library
  • Cartopy mapping project
  • Biggus library for lazy array

manipulations

slide-10
SLIDE 10

SciTools (cont.)

  • These are extremely interesting

projects that we want to investigate and possibly collaborate upon

  • Unfortunately no such

investigations have taken place yet

slide-11
SLIDE 11

PyCDM

  • Implementing CDM in Python has

been proposed internally many times

  • Have been approached by Martin

Schultz with an early prototype

slide-12
SLIDE 12

PyCDM (cont.)

  • Not clear how this fits or duplicates

Iris

  • Have not examined prototype

implementation yet

slide-13
SLIDE 13

Closing

  • Unidata is actively involved in the

Python community

  • Unidata has kept up to date on many

exciting areas of development

  • Very limited dedicated resources