Details
-
Spike
-
Not Assigned
-
None
-
Data Processing
-
-
-
3
-
3
-
26.667
-
Team_NZAPP, Team_SCHAAP, Team_SIM, Team_YANDA
-
Sprint 5
-
-
-
-
10.4
-
Stories Completed, Outcomes Reviewed, NFRS met, Demonstrated, Satisfies Acceptance Criteria, Accepted by FO
-
SPO-880
Description
Size estimate based on 3-4 teams contributing 2-3 people each for 2 days of effort.
In PI9 we would like to set the groundwork for developing MID and LOW continuum imaging pipelines as a step towards a having an initial set of SDP workflows by the end of PI12 that will be built on and scale in future Feature Sets.
In order to bootstrap this work, this spike is intended to take the form of a workshop or meeting between SDP teams intending to contribute to the goal of maintaining alignment between MID and LOW solutions. This spike should:
- Identify the testing strategy
- Identify pipeline steps (with consideration for state-of-the-art solutions from precursor instruments)
- Map steps to existing software solutions (including those already identified in work to look at state-of-the ART DD imaging and calibration algorithms)Â
- Consider variations and alternatives
- Consider the pipeline in the context of the SDP architecture (future integration as an SDP workflow, decomposition into execution engine stages and processing components)
- Identify steps needed to ensure that the developed pipelines will be a solution that can easily be run and developed further by other SDP teams and where in future PIs we indeed to replacing key steps in the pipeline to achieve scaling at ~10 node level.
- Identify how pipelines are to be benchmarked for validation of outputs and for compute performance
- Identify input data (simulation or from precursor/pathfinders) that can be used to test these pipelines
The outcome of this spike is expected to be one or more confluence pages (possibly in the Solution Intent space) describing the workflows to be implemented.
Attachments
Issue Links
- blocks
-
SP-1331 Initial implementation of (continuum) imaging pipeline(s)
- Done
- followed by
-
SP-1693 Onboard and test complete set continuum imaging pipelines using simulation data (without DD effects)
- Done
- relates to
-
SP-929 Yandasoft Regularised Bandpass Solver - robustness to spectral ripple
- Funnel
-
SP-1329 Exploration of resource availability for SDP simulations and workflows
- Funnel
-
SP-987 Yandasoft Regularised Bandpass Solver - Enhanced smoothing constraints
- Done
-
SP-988 Yandasoft Regularised Bandpass Solver Investigations
- Done
-
SP-1328 Developer portal (top level) group for SDP workflow and simulation projects
- Done
-
SP-1334 Identify categories and test scenarios for RFI simulation and mitigation
- Done
-
SP-1335 RFI Simulations for testing future SDP imaging workflows
- Done
-
SP-1411 Enhance the GPU imaging/gain calibration pipeline to support visibility weighting schemes
- Done
-
SP-1433 Onboard LEAP and explore Scope for Parallelization
- Done
-
SP-1447 Demonstrate yandasoft peeling on ASKAP precursor data
- Done
-
SP-1448 Heterogenous Beam Simulations: WSClean
- Done
-
SP-1449 Heterogenous Beam Simulations: DDFacet
- Done
-
SP-1336 Identify and collect data sets with RFI matching criteria
- Discarded
-
SP-1450 Heterogenous Beam Simulations: Comparison Report
- Discarded
-
SP-732 Describe and demonstrate fast Kolmogorov phase screen generation script
- Done
-
SP-1235 Improve direct estimation of screens for DD corrections
- Done
-
SP-1331 Initial implementation of (continuum) imaging pipeline(s)
- Done
-
SP-1448 Heterogenous Beam Simulations: WSClean
- Done
- mentioned in
-
Page Loading...
-
Page Loading...
-
Page Loading...
-
Page Loading...
-
Page Loading...
-
Page Loading...
-
Page Loading...
-
Page Loading...
-
Page Loading...
-
Page Loading...
-
Page Loading...
-
Page Loading...
-
Page Loading...
-
Page Loading...
-
Page Loading...
-
Page Loading...