2019 Workshop:Hackathon

From CedarWiki
Revision as of 08:25, 18 June 2019 by Hirsch.m (Talk | contribs)

Jump to: navigation, search


2019 CEDAR Heliophysics Hackathon

Convenors: Russell Stoneback, Matthew Grawe, Michael Hirsch


About 40 people were present at any one time from about 7pm to 9pm. Many people were present by 6:40 and over 10 people were still present by 10 pm.

For 2020 CEDAR Hackathon, thoughts include:

  • make list of specific topics e.g. GitHub Issues at all skill levels. From updating docs to integrating Python C/Fortran extensions, type hinting / PEP8, add new instruments to PySat, add NetCDF4 file interface, etc.
  • make pairs or threes of people, where someone with moderate experience is paired with someone newer to Python (or at least the task area they're interested in
  • strongly preferred to have in a traditional conference room with tables vs. the scattered seating in the coffee shop
  • good to have pizza

Time, Location, Policies

  • Location and Time: Coffee Shop Room near Eldorado, 17 June, 18:30-midnight
    • To get to the coffee shop you exit the front of the hotel and turn right. It is on the right at the end of the block (109 N Guadalupe St Santa Fe, New Mexico)

Coffee Shop Street small.jpg

  • Room Capacity: 50
  • A limited amount of pizza will be provided during the event. We highly encourage participants to bring their own snacks and (non-alcoholic) beverages.
  • This event is for all proficiency levels. Tasks, however 'simple', are appropriate. Help will be available if you need it!

Event Overview

The event focuses on fostering a community around the use and development of open source tools to solve leading questions in space science, particularly those for the CEDAR community. There is a broad range of expertise across CEDAR. Concentrating even a small amount of our community’s expertise and effort around a set of common tools can achieve great things. We propose an effort to support loading, processing, and utilization of all data sources from across CEDAR through the pysat package (https://pysat.readthedocs.io/en/latest/). Readily available access to all CEDAR data (loaded data, not just file availability) through a single consistent interface would provide an invaluable resource as CEDAR pursues system science. The introduction concisely describes the library:

Every scientific instrument has unique properties though the general process for science data analysis is independent of platform. Find and download the data, write code to load the data, clean the data, apply custom analysis functions, and plot the results. The Python Satellite Data Analysis Toolkit (pysat) provides a framework for this general process that builds upon these commonalities to simplify adding new instruments, reduce data management overhead, and enable instrument independent analysis routines. Though pysat was initially designed for in-situ satellite based measurements it aims to support all instruments in space science.

Projects/activities should fall into one or more of the following categories:

  • data visualization / presentation using pysat as a data interface
  • data wrangling: adding support to pysat for dataset(s) you work with in your research not already supported, or adding netCDF support to existing data products / software
  • data science: super/unsupervised machine learning projects/code using pysat as a data interface

The emphasis is to produce code that others in the community can use in the future. The geospace data science session can accommodate 1-2 short presentations showcasing any fruitful results falling into the data science category produced during the event.

Getting Started

The Anaconda distribution is an easy way to quickly install Python (along with many useful data visualization, wrangling, and science libraries). Useful resources exist online that explain how to get started with Python. Included with Anaconda are the NumPy, SciPy, and matplotlib packages, which offer a powerful suite of tools for scientific computing and visualization. Finally, more useful resources exist that help users get started using pysat.

Submitting Progress

All participants are encouraged to submit progress summaries online for anything they work on during the event using the following form (a sign-in is required because the file attachment widget doesn’t allow anonymous submissions):

Submit progress report

Again, we stress that this event is for all skill levels. Any and all progress is worthy of reporting!