Open tomvothecoder opened 3 days ago
Thank you for creating this issue. I think we can use v3 simulation data for the complete run. There is a bash file I created for the 2024 tutorial which include the required input data for all sets. https://github.com/E3SM-Project/e3sm_diags/blob/14612912b66fb71ac0e82504034ded2e8e2aba71/examples/tutorials2024/e3sm_diags_extended.bash
Is your feature request related to a problem?
The current integration test uses image checking to compare pixels. These tests often break due to very subtle differences that are introduced externally through different versions of packages, especially
matplotlib
.Describe the solution you'd like
The CDAT migration regression testing script adapts the
run__...py
script to run the sets. We then manually update a regression testing notebook to compare the results against the latestmain
(right now CDAT-based).Note: These are rough details that need to be ironed out. What we need to do is:
main
main
(aftercdat-migration-fy24
is merged) and store results on Perlmutter/LCRCpytest
and GH Actions build workflowDescribe alternatives you've considered
No response
Additional context
No response