Difference between revisions of "Staff Priorities - Masha"
From SCECpedia
Jump to navigationJump to search| Line 10: | Line 10: | ||
* Post CSEP status meeting minutes, send next meeting announcement | * Post CSEP status meeting minutes, send next meeting announcement | ||
* GEOL440 class at 11:00am | * GEOL440 class at 11:00am | ||
| − | * csep-op: investigate failure on csep-op for Global 1-year forecasts: Exception occured <class 'ValueError'> = Data has no positive values, and therefore can not be log-scaled. | + | * csep-op: investigate failure on csep-op for Global 1-year forecasts: Exception occured <class 'ValueError'> = Data has no positive values, and therefore can not be log-scaled: negative rate values in TripleS Global 1-year forecast with start date of 2016/1/1. Sent email to Jeremy reporting the problem. |
* Trac ticket #372: Support Bayesian ensemble models for one-day forecasts | * Trac ticket #372: Support Bayesian ensemble models for one-day forecasts | ||
Revision as of 22:39, 18 February 2016
Current Activities
Yesterday
- CSEP status call at noon
- Webcast on Pandas at noon
- csep-op: debugging failure on csep-op for Global 1-year forecasts: Exception occured <class 'ValueError'> = Data has no positive values, and therefore can not be log-scaled.
- Trac ticket #372: Support Bayesian ensemble models for one-day forecasts
Today
- Post CSEP status meeting minutes, send next meeting announcement
- GEOL440 class at 11:00am
- csep-op: investigate failure on csep-op for Global 1-year forecasts: Exception occured <class 'ValueError'> = Data has no positive values, and therefore can not be log-scaled: negative rate values in TripleS Global 1-year forecast with start date of 2016/1/1. Sent email to Jeremy reporting the problem.
- Trac ticket #372: Support Bayesian ensemble models for one-day forecasts
Blocked: No Followups: No
TODO
- csep-op failure: evaluation was issued on the very first day of new one-year Global forecasts - should not be the case
- Trac ticket #374: Remove temporary input catalog generated by CA ETASV12_10 model
- Trac ticket #353: Optimize RELM N-test by avoiding event to bin look up when masking bit is set for the whole forecast
