Peter Schwartz
Peter Schwartz
Made a PR that addresses this Issue. Point of note, the function we want is `tiny` instead of `epsilon` . `epsilon` provides the round-off error for a data type, so...
Since the new maint-3.0 has been created, I can start merging this PR this week. I'll submit tests to see how the developer test suite is impacted.
Yes, the abort is still there though the PR makes further changes to the cutoff. Here is the full if-else block modified by this PR. ```fortran ! check for negative...
Testing on Chrysalis with this PR had `ERS.r05_r05.IELM.chrysalis_intel.elm-lnd_rof_2way` FAIL ``` 118: Error: Negative channel storage found! -9420.09754789568 118: ERROR: mosart: negative channel storage 118: Image PC Routine Line Source 118:...
Expected DIFFs from `e3sm_developer`: ``` ERS.r05_r05.IELM.chrysalis_intel.elm-V2_ELM_MOSART_features (Overall: DIFF) details: SMS.MOS_USRDAT.RMOSGPCC.chrysalis_intel.mosart-unstructure (Overall: DIFF) details: ```
Thanks I'll test it out today merged to next for the mosart tests. Quick question, outside of this test are the elm and mosart timesteps properly synced by compset or...
the `lnd_rof_2way` test reported back a DIFF after merging it to next, so this should be ready to go monday.
Thanks for the reminder -- been a long day. merged to next
merged to master, submitted bless requests for 3 mosart tests
Finally looking into this. There is an unused option to create a stream that reads the entire file, which seems like the simplest fix to pursue. I'll try to have...