...
completed | in progress | on hold | review | to do |
---|
64 |
|
|
|
|
(vs record 61 77 completed sprint 1950)
Epic | | Owner | Deliverables | |
---|
Validation | 1) 400m Whole NZ validation run (tectonic classification)2) Re-run advanced IMs 2)Generate gcmt faults to find correct orientation 3) Srf generation Subduction specific behaviour for GCMT (Robin to add more info)
| 1) RobinJonney 2) JonneyJames | - still in progress
- Plots generated and passed on to Robin
|
|
Ground Motion DB | 1) Ground Motion Extraction 2) Hikurangi Geometry 3) Tectonic Classification 4) Mw Reconciliation (comparison : John Ristau vs international dataset) 5) Comparison with VH2017 | 1)Viktor 2) James 3) Mike / James 4) ? 5) ? | - ?
- On hold for this sprint
- Mike continuing
| Hikurangi surface geometry |
Cybershake | 1) Cybershake Subduction (8 faults) : Complete BB after IM plot checks 2)Running 200m sims LF only - on Kisti / Maui (until Maui comm allocation leaves 100k) - lower south island / all rels (KISTI)
(Maui) - Central South island / all rels (Maui)
- all faults/1 rel (KISTI) (Maui)
- remaining faults/all rels (KISTI)
uses 500k) 2) 4) Empirical DS - DB calculation 53) Empirical / Cybershake / ratio hazard maps | Jason | 1) Jason150k used on Comm00213 since start of June. Using ~10-15k a day at the moment. 500/~4500 realisations 2) SungCompleted & plotted (with issues) need to investigate it. |
|
Slurm Workflow | - Generalize hacks for KISTI
- CH estimation based on linear regression
- Integrate with pre-processing (starting with VM)
- Separate db creation out of install
- Create generic slurm for the rest of install, make install operate on each fault
- Create a generic slurm for VM generation
- Pre-processing - GCMT to realisation – add some conditional logic about tectonic type (Brendon / Robin to consult about conditional parameters)
- Near-real time Simulation plan
- Cybershake Workflow Improvements
| - James
- Jason / Brendon Ethan
- Jonney
- James
- Jonney
- Jason / Jonney
| - Work mostly complete. Final tests on Kisti underway. Checks on tacc yet to begin.
- ?
- Deferred
-
- Done (from Jason side).
| 3. Adding Pre-processing into automated workflow |
Workflow Calc | 1) BA18 site amp | 1) Jason | 1) Started investigation. Read paper / code. Have done some quick scripting. Ready to do some testing. |
|
SeisTech | SeisTech | - GM Selection for Empirical
- Integrate workflow into seistech master
- Synthetic tests for GCIM between Empirical and Simulation (1 rupture, N ruptures)
- For Daniel: - what tasks for him to do?
- Testing, trial calculations
- Automate documentation
- Front-end
- Milestone #2
- authentication.
- user access to individual tab controlled by authentication
- User data managed by API, and additional user data kept in a separate db
- Data integrity and consistency between Auth0 and separate db maintained.
- Milestone #3
- API call framework - plan the API proxy
- Create Mock API's with canned JSON response
- Call mocked API from the front end and ingest JSON. To demonstrate an example of full functionality of the front end
- Milestone #4: API – 2
- Create an API proxy - including error messages for restricted access
- Call existing Core API’s from the front end replacing the mocks.
- JWT authorisation between the proxy and the core API
- Milestone #5: Deployment of 3 core environments
- Deploy DEV, TEST/EA, PROD versions
- Deployment is fully automated via scripts and config files.
- Update or switch-over is easily achieved utilizing deployment slots
- Disruption is kept minimal when updated.
4) Empirical DS - DB calculation- GM Selection for Empirical
- Integrate basic (single branch) empirical GMS
- Automate documentation
- Front-end
| - Daniel / Claudio
- Background task (Jason)
- Andy
| 1) Done (no testing) | Roadmap (scientific functionality list) Production - TODO (longer term tasks) Frontend |
IM Calc |
|
|
|
|
Bug fixes |
|
|
|
|
Seismic risk |
|
|
|
|
Machine Learning | - NN - GMM
- Implement an initial basic pipeline with some NN config + flexible feature selection & preprocessingSpatial plots
- Add hypo-depth to NN GMM dataset
- GM Classifier – see link
| ClaudioGM classifier - progess | 1) Added (generalised) functionality for spatial map plotting. Currently supports IM residual maps, few others to be added. 2) Worked on validation & PhaseNet evaluation with Mike, 70-80% done in terms of integrating PhaseNet as p-wave picker (will allow multiple |
|
Empirical engine |
|
|
|
|
Vs30 |
| Viktor |
| demo video on slack http://hypocentre:5088 documented github page |
Misc | - SimAtlas simulation+animation:
- Test auto workflow with batch 4. (total 100 faults)
- Once batch 4 is done. Keep fueling the fire
- Batch 5 - keeping continual work
- Data management policy
- (classification of data importance
. - NoisePy
- NeSI allocation: Chat with NeSI, keep subscription allocation separate)
- )
- Plan to incorporate this as part of business as usual – what data do we need to keep and what can be regenerated as needed
- NoisePy
- Velocity Model basins
- Dashboard
- Add comm00213 as a separate allocation
- Changes of storage (TB): With Fortnightly update info added
- Zero point at the start of the allocation (graph doesn't start midway)
- Vs30 ServerVelocity Model basins
| - Sung/Jonney
- Sung
- Viktor
- Jason
- Sung
- Viktor
| - Running
- Automated management code done.
- on hold
- -
- (a) done, (b-c) to be done
- Done
|
|