Open smalers opened 3 years ago
I updated the test results for StateCU 14.0.1, which fixed the problem of BD1
files not being created when isuply=0
. I also reran the South Platte and updated the numbers. I may not have properly recorded those results. The South Platte comparison has more and larger differences than other datasets.
https://github.com/OpenCDSS/cdss-app-statecu-fortran-test/blob/master/test-results-doc/README.md
Automated testing would benefit from smaller datasets to verify calculations, similar to StateMod examples. Then it is would be possible to confirm model results for each feature.
This issue can be closed by the State when they are satisfied with the results.
Attached are the summaries of South Platte dataset differences. SP2016_Restricted-summary-differences.txt SP2016-summary-differences.txt rg2012_FactorSoUMeter-summary-differences.txt
I have created a README in the repository summarizing 14.0.0 comparison with 13.10. This page can be updated over time to record software version testing.
https://github.com/OpenCDSS/cdss-app-statecu-fortran-test/blob/master/test-results-doc/README.md
Below is an example of larger differences for the
rg2012
scenario.Below is an example from the
rg2012_FactorSoUMeter
scenario.Below is the bit difference for
SP2016
:It seems like maybe the latest version is handling input differently in some cases?