Intermittent test failures for wind

USDA-ARS-NWRC / smrf

SMRF was designed to increase the flexibility of taking measured weather data, or atmospheric models, and distributing the data across a watershed.

Other

12 stars 4 forks source link

Intermittent test failures for wind #180

Closed jomey closed 4 years ago

jomey commented 4 years ago

Tests wind in a threaded setup fail intermittently with a missing time step. Example: https://travis-ci.com/github/USDA-ARS-NWRC/smrf/jobs/359503602

scotthavens commented 4 years ago

This probably still has something to do with PR #178 where the thread variables for the wind class aren't being set properly so some of them persist from test to test.

jomey commented 4 years ago

I am wondering, why the time comparison has one less hour than expected. How are these related with the thread variables?

scotthavens commented 4 years ago

Not all the thread variables for the Winstral model are used for Wind Ninja. However, they still need to be set for precipitation so if they aren't set then they can fail. It typically fails at 2 timesteps because that is the max number in the queue. Changing max queue to 1 would make it fail at 1 I bet.

scotthavens commented 4 years ago

Fixed with #183