Name: two_percent_DESI
Owner: desihub
Description: Simulating 2% of the DESI survey
Created: 2017-03-30 15:43:30.0
Updated: 2017-11-02 19:08:29.0
Pushed: 2018-04-17 20:06:18.0
Homepage: null
Size: 21976
Language: Jupyter Notebook
GitHub Committers
| User | Most Recent Commit | # Commits |
|---|
Other Committers
| User | Most Recent Commit | # Commits |
|---|
This repo has files related to the 2017 data challenges of processing 2% of DESI observations.
Spring 2017.
The original “2% sprint” that turned out to take quite awhile. We succeeded at processing 2%, but not in a clean, easy, documented, and reproducible way.
Summer 2017.
Post-dc17a, establishing the steps for processing just 5 tiles end-to-end and fixing dataflow issues along the way. It turns out that 2% of DESI is an unwieldy amount of data for testing. Even 5 tiles = 25k spectra isn't trivial.
Not started yet. This will be a redo of dc17a, using what we learned/fixed from the dc17a and the mini-sprint to redo 2% of DESI with the latest code in a more streamlined and documented manner.
dc17a and dc17b are 2% of observations, which cover more than 2% of the area but to only a single layer of depth. We could also do a test with 2% of the area to full depth, which would be especially useful for Lyman-alpha studies.
Do this including pixel-level simulations + extractions instead of spectral-level simulations.
Then onwards to 5%, 20%, 100%…