Fix with_custom for heating #492
Conversation
|
I just realized this is failing one of the "with_custom" Jones benchmark case 0 tests. I may be being obstinate, but I think this test SHOULD have been failing before and (possibly) not falling after this pull request. I just confirmed that the simulation under tests/custom_reference/augment/main_input results in a heating function being zero because of the adjust_reference_heating issue. Presumably this case was meant to test what happens with nonzero heating in the Jones benchmark? Unless I am missing something, I would then suggest replacing the file tests/custom_reference/augment/Benchmark_Reports/00000200 which would presumably make this pull request pass the test. |
|
I'll try to look at this in detail soonish. Might be a couple of days, but I'm aware of the PR. |
illorenzo7
force-pushed
the
fix_withcustom_for_heating
branch
from
945c371 to
2e50d6a
Compare
|
Because of the latest approved pull requests, I removed the duplicate commit from #490 (the divide by N^2=0 pull request) and then rebased this off "main" to make the analysis of the specific change here more easily readable. |
illorenzo7
force-pushed
the
fix_withcustom_for_heating
branch
from
2e50d6a to
74228db
Compare
The with_custom framework successfully modifies the heating by a custom reference file. However, the flag "adjust_reference_heating" has been set to .true. earlier (for any heating_type > 0) and so the user's choices will possibly be overwritten in the boundary conditions file.
This pull request changes adjust_reference_heating back to .false. if the user sets a custom heating profile using the with_custom framework.