Details
-
Feature
-
Must have
-
None
-
Data Processing
-
-
-
Intra Program
-
5
-
5
-
13
-
2.6
-
Team_ORCA
-
Sprint 5
-
-
-
-
18.3
-
Stories Completed, Accepted by FO
-
-
SPO-1592 SPO-1593
Description
Who?
- SDP pipeline developers, Commissioning scientists, (eventually) wider radio-astronomy community
What?
- The first iteration of a SKA self-cal pipeline that is composed of SKA processing functions with clean levels of abstraction between the processing functions and Execution Engine application. (As title writes, would want to have this distributed in Dask. However first first thing is to have a non-distributed but Python version using the SDP Proc Fns)
- Consideration for scaling to multiple cores, accelerators, and eventually nodes
- Consideration for test data in the form of simulations and precursor data for focused quick CI testing and working towards being able to process an ~AA2 size data set.
Why?
- Demonstrates and explores approaches to SDP pipeline development, leveraging on work done in previous PIs and in PI14 on processing functions.