Difference between revisions of "Staff Priorities - Kevin"

Revision as of 16:59, 15 October 2015

10/15/15

While profiling running simulations on stampede, found major bug in Ned's latest ETAS code that was causing simulations to hang. Basically a simulation could get stuck in an infinite loop if a rupture below a certain depth was triggered, which is rare but possible when running many simulations.
- implemented temporary non-ideal fix for current runs
- Ned working on long term fix
plots for latest runs
edits to XSEDE proposal

help Ned with fix for EQ depth/infinite loop issue
post process/generate plots for 100yr runs which finished, send off to Morgan/Nicholas
merge in latest ETAS code from Ned for hopefully final batch of runs prior to Powell Center meeting
post process results for scenario ETAS runs yesterday, send to Ned

no

no

Fill in missing site/rupture distances in CyberShake DB for ERF 36 (200m)
finish mapping of historical earthquakes to UCERF3 ruptures in order to run ETAS simulations of aftershock sequences

@@ Line 1: / Line 1: @@
-/13/15
+/15/15
 == Yesterday: ==
-* lots of ETAS runs and post processing for Ned and Morgan
+* While profiling running simulations on stampede, found major bug in Ned's latest ETAS code that was causing simulations to hang. Basically a simulation could get stuck in an infinite loop if a rupture below a certain depth was triggered, which is rare but possible when running many simulations.
-* XSEDE progress report and request for Phil
+** implemented temporary non-ideal fix for current runs
-* solicited John's help with keeping opensha.usc.edu working smoothly during disk downtime
+** Ned working on long term fix
+* plots for latest runs
+* edits to XSEDE proposal
 == Today: ==
-* more ETAS calculations for Ned in advance of Powell Center meeting next week
+* help Ned with fix for EQ depth/infinite loop issue
-* mine logs to see if my FastMPJ failure detection script is working. no un-caught failures so far, so we've either been lucky or the script is working
+* post process/generate plots for 100yr runs which finished, send off to Morgan/Nicholas
-* help Ned assemble plots needed for meeting next week
+* merge in latest ETAS code from Ned for hopefully final batch of runs prior to Powell Center meeting
-* work on figuring out runaway sequences
+* post process results for scenario ETAS runs yesterday, send to Ned
 == Blocked: ==