Open GallLeo opened 5 months ago
@HansOlsson Is this due to #4151?
@GallLeo Please add the appropriate labels for library and version.
@HansOlsson Is this due to #4151?
@GallLeo Please add the appropriate labels for library and version.
I'm a bit lost. What is the error? Clicking on the links just give a bunch of green texts.
Try https://www.ltx.de/download/MA/Compare_MSL_v4.1.0/Compare/Modelica/20240129232335/Modelica.Blocks.Examples.Noise.NormalNoiseProperties/CSVCompare/NormalNoiseProperties_report.html where the mean output error is the reason for result failure.
Try https://www.ltx.de/download/MA/Compare_MSL_v4.1.0/Compare/Modelica/20240129232335/Modelica.Blocks.Examples.Noise.NormalNoiseProperties/CSVCompare/NormalNoiseProperties_report.html where the mean output error is the reason for result failure.
Ah, yes, the protected variable mu
is now completely different due to PR #4151 - but y
should be the same; so just reference update (or removing from comparison).
Yes, removing from comparison might be good approach since protected anyway.
@GallLeo Please add the appropriate labels for library and version.
I'm a bit lost. What is the error? Clicking on the links just give a bunch of green texts.
Sorry, I removed the automatic browser links. They linked to the test case, not to the comparison report. Only the comparison links count.
@GallLeo Please add the appropriate labels for library and version.
I'm a bit lost. What is the error? Clicking on the links just give a bunch of green texts.
Sorry, I removed the automatic browser links. They linked to the test case, not to the comparison report. Only the comparison links count.
Note that the first two seems to be reference updates, unrelated to #4151
Possibly different random sequence for the first two.
Possibly different random sequence for the first two.
This needs some more evidence as I am not aware on RNG changes.
I run git bisect on Modelica.Blocks.Examples.Noise.ActuatorWithNoise and got:
a2cce87f7ab1a2a59f5263867c507e31241c23fd is the first bad commit
This points to #3727/#3728 and even @maltelenz' comment https://github.com/modelica/ModelicaStandardLibrary/issues/3727#issuecomment-763449287 where it is stated that the MSL 4.0.0 reference result is invalid.
This was discussed at the latest MAP-Lib meeting (2024-05-14), but the ones that I should update already have a PR linked to this one - it is just waiting to be reviewed and merged.
This was discussed at the latest MAP-Lib meeting (2024-05-14),
I would help if the outcome of this kind of discussion somehow gets back to the originating issues/PRs.
@GallLeo everything should be ready for the new results generation
The following models fail in result comparison. Status: 7ac79067 (2024-01-30)
[ ] Modelica.Blocks.Examples.Noise.ActuatorWithNoise
[ ] Modelica.Blocks.Examples.Noise.DrydenContinuousTurbulence
[ ] Modelica.Blocks.Examples.Noise.NormalNoiseProperties
please back-port #4151 to maint/4.1.x and link the PR here; Back-port #4297 which is done in #4416[ ] Modelica.Blocks.Examples.Noise.UniformNoiseProperties
please back-port #4151 to maint/4.1.x and link the PR here; Back-port #4297 which is done in #4416