Difference between revisions of "Product Backlog - William"

From SCECpedia
Jump to navigationJump to search
 
(53 intermediate revisions by the same user not shown)
Line 1: Line 1:
 
William Savran's Project/Product Backlog
 
William Savran's Project/Product Backlog
  
== Near-Term ==
+
== Product Backlog ==
* implement fixes to CSEP outlined in [[CSEP_1_STEPCoulombQuickFix|RFC002]]
+
# research and develop new techniques for the rigorous statistical evaluation of stochastic event sets
** investigate grid issue between STEP and Coulomb
+
# reprocess missing one-day forecasts/evaluations identified from csep_db
** run test case to verify fix of coulomb model [waiting on benchmark from matt]
+
# provide open source codes for CSEP1 evaluations to begin modularizing current CSEP1 codebase
** update working group page to show additional fixes to the Coulomb part and step-java model
+
# create archived csep1 data set containing all data and meta-data from csep operational servers and provide to the community
* produce a list containing status of all new Zealand forecasts and evaluations between 2010 - current
+
# run 1-year u3etas and no-faultsETAS simulations from start of >Mw 2.5 catalog (~1985) until current day to provide data set for applying csep2 evaluations on stochastic event sets
** create data model
+
# prototype tests of forecast of finite-ruptures, for example, in the ground motion domain
** get sqlite tool
+
# develop a protocol to accept external forecasts for CSEP testing
** implement mvp of data model
+
# develop plan to ensure the continued prospective testing of csep1 models
*** build test data
+
# develop blueprint to describe how to implement current on-demand CSEP 1 processing and reprocessing.
*** import into data base using tool
+
# write data management plan for the csep2 ecosystem.
** write queries to test
+
# implement restAPI to provide access to csep2 data archives.
** publish code to github repository
 
** expand model to contain csep system
 
 
 
== Project/Product Backlog ==
 
* using parallel reprocessing capabilities recompute all new zealand forecasts from 2007-present
 
* provide curated tools for running CSEP evaluations and experiments outside of CSEP images
 
* provide method to access daily archived catalogs from SVN repository
 
 
 
== Detritus ==
 
* address SVN issue with localhost path to SVN server on csep-debug
 
* run forecast for global region on csep-debug to check how year breaks function
 
** run for 2007
 
** run for 2018
 
* brainstorm 5 year plans for CSEP2
 
* create scripts to find forecasts in data archives
 
* investigate how SVN works with EQ catalogs
 
** how do we query SVN for previous years using tags
 
 
 
== CSEP 5 Year Plan ==
 
<b>Immediate</b>
 
* Fix NZ STEP models. Target date: 1 March 2018.
 
* Verify CSEP results and data streams. Target date: 15 March 2018.
 
* Upgrade NZ CSEP version. Target date: 15 March 2018.
 
* Identify dependencies between NZ forecast groups. Target date: 15 March 2018.
 
* Identify dependencies between SCEC forecast groups. Target date: 15 March 2018.
 
<b>Near</b>
 
* Parallelise reprocessing. Target date: 15 April 2018.
 
* Complete reprocessing NZ and SCEC. Target date: 15 May 2018.
 
* Provide access tools to curated CSEP1.0 dataset. Target date: 1 June 2018.
 
<b>Far</b>
 
* Prototype (reprocessing) ensemble methods. Target date: 1 July 2018
 
* Process ensemble models in NZ and SCEC. Target date: 1 August 2018
 
* Publish results. Target submission date: 1 September 2018.
 
* Publish results. Target date: 1 November 2018
 
  
 
== Related Entries ==
 
== Related Entries ==
 
*[[CME Software Staff]]
 
*[[CME Software Staff]]
 
*[[CME Project]]
 
*[[CME Project]]

Latest revision as of 20:56, 11 January 2019

William Savran's Project/Product Backlog

Product Backlog

  1. research and develop new techniques for the rigorous statistical evaluation of stochastic event sets
  2. reprocess missing one-day forecasts/evaluations identified from csep_db
  3. provide open source codes for CSEP1 evaluations to begin modularizing current CSEP1 codebase
  4. create archived csep1 data set containing all data and meta-data from csep operational servers and provide to the community
  5. run 1-year u3etas and no-faultsETAS simulations from start of >Mw 2.5 catalog (~1985) until current day to provide data set for applying csep2 evaluations on stochastic event sets
  6. prototype tests of forecast of finite-ruptures, for example, in the ground motion domain
  7. develop a protocol to accept external forecasts for CSEP testing
  8. develop plan to ensure the continued prospective testing of csep1 models
  9. develop blueprint to describe how to implement current on-demand CSEP 1 processing and reprocessing.
  10. write data management plan for the csep2 ecosystem.
  11. implement restAPI to provide access to csep2 data archives.

Related Entries