Difference between revisions of "Staff Priorities - Kevin"

From SCECpedia
Jump to navigationJump to search
 
(97 intermediate revisions by the same user not shown)
Line 1: Line 1:
9/24/15
+
8/23/16
  
 
== Yesterday: ==
 
== Yesterday: ==
* major refactoring of MCEr calculations
+
* met with Tom
** finished map generation code (validated against old maps)
+
** needed help with SCEC-VDO figures for intern proposal
** built cache enabled calculators and file formats for saving GMPE results between map calculations
+
* worked on transparent polygon render order in VTK, not yet successful
** halfway done getting spectrum plotting code back online
+
* CyberShake call
** everything much simpler and cleaner
+
* UCERF3 Science paper figure work:
* corresponded with Ketih Richards-Dinger on RSQSim file formats, sounds like they want to move away from their internal file format and use the Barall format so it may not be worth implementing parsers for their format as I had planned
+
** Finished SCEC-VDO figure work, generated first draft of all Mojave M7 SCEC-VDO figures
 +
** discovered that 10k Mojave M7 simulations isn't enough for stable spatial distribution of probability gain. Submitted more calcs at Stampede which ran over the weekend, now up to 100k
  
 
== Today: ==
 
== Today: ==
* plan powell center travel (10/18-22)
+
* work on web location for UCERF3-ETAS simulations:
* finish MCEr refactoring
+
** one location that includes only the "final" simulations
* implement CB's proposed CyberShake/GMPE weighting algorithm:
+
** add descriptions of plots, hide plots not featured in paper
** average UCERF3 based GMPE calculations with UCERF2 based CyberShake calcs
+
** will be similar to: http://opensha.usc.edu/ftp/UCERF3.3/
** 1s: full GMPE weight, 10s: full CyberShake weight. Will linearly interpolate between for now but make it easy to swap in other weighting functions
+
* meet with Tom
** Apply weighted averaging directly to hazard curves (before RTGM calc) for probabilistic calcs, apply to spectrum for deterministic calculations.
+
* UCERF3 Science paper figure work:
** will use branch averaged UCERF3 for now, might eventually use partial logic tree if needed
+
** stage back, merge 100k Mojave M7 simulations, redo plots
* if time, continue mapping of historical earthquakes to UCERF3 ruptures in order to run ETAS simulations of aftershock sequences
+
** likely plot updates after meeting with Tom
 +
* work on figures for loss modeling paper
  
 
== Blocked: ==
 
== Blocked: ==
Line 23: Line 25:
  
 
== Followups: ==
 
== Followups: ==
no
+
Tom
 
 
  
 
== Lower priority or longer term tasks ==
 
== Lower priority or longer term tasks ==
* Fill in missing site/rupture distances in CyberShake DB for ERF 36 (200m)
+
* once new server is finished, add page on UCERF3-TD supplement corrections and add bullet linking to it from http://www.wgcep.org/UCERF3
* finish mapping of historical earthquakes to UCERF3 ruptures in order to run ETAS simulations of aftershock sequences
 

Latest revision as of 17:10, 23 August 2016

8/23/16

Yesterday:

  • met with Tom
    • needed help with SCEC-VDO figures for intern proposal
  • worked on transparent polygon render order in VTK, not yet successful
  • CyberShake call
  • UCERF3 Science paper figure work:
    • Finished SCEC-VDO figure work, generated first draft of all Mojave M7 SCEC-VDO figures
    • discovered that 10k Mojave M7 simulations isn't enough for stable spatial distribution of probability gain. Submitted more calcs at Stampede which ran over the weekend, now up to 100k

Today:

  • work on web location for UCERF3-ETAS simulations:
  • meet with Tom
  • UCERF3 Science paper figure work:
    • stage back, merge 100k Mojave M7 simulations, redo plots
    • likely plot updates after meeting with Tom
  • work on figures for loss modeling paper

Blocked:

no

Followups:

Tom

Lower priority or longer term tasks

  • once new server is finished, add page on UCERF3-TD supplement corrections and add bullet linking to it from http://www.wgcep.org/UCERF3