regression-tests issueshttps://gitlab.psi.ch/OPAL/regression-tests/-/issues2019-07-18T09:33:28+02:00https://gitlab.psi.ch/OPAL/regression-tests/-/issues/69Restart tests fail due to missing quotes2019-07-18T09:33:28+02:00frey_mRestart tests fail due to missing quotesConnected to https://gitlab.psi.ch/OPAL/src/merge_requests/143.Connected to https://gitlab.psi.ch/OPAL/src/merge_requests/143.frey_mfrey_mhttps://gitlab.psi.ch/OPAL/regression-tests/-/issues/68Failed kicker tests after kicker update2019-08-01T06:07:27+02:00snuverink_jjochem.snuverink@psi.chFailed kicker tests after kicker updateAfter merging src!101 and fixing src#37, the kicker regression tests fail with small error margins: http://amas.web.psi.ch/opal/regressionTests/master/results_2019-05-24.xmlAfter merging src!101 and fixing src#37, the kicker regression tests fail with small error margins: http://amas.web.psi.ch/opal/regressionTests/master/results_2019-05-24.xmlkrauskraushttps://gitlab.psi.ch/OPAL/regression-tests/-/issues/67GasStripping test fails2019-04-23T08:19:50+02:00snuverink_jjochem.snuverink@psi.chGasStripping test failsThe `GasStripping` test fails from time to time with a too large difference for the surviving number of particles.
The `GasStripping` test checks how many particles are lost, in the reference these are 1e5 - 99623 = 377. Since this is a stochastic process, the expected error is about sqrt(377) = 20. The allowed error right now in the test is 25. [The GasStripping test failed with a difference of 42](http://amas.web.psi.ch/opal/regressionTests/OPAL-2.0/results_2019-04-16.xml). To avoid that the test fails randomly from time to time, I propose to set the allowed error to at least 2 sigma=40, perhaps even better 3 sigma=60 (also because the reference of 377 might actually not be a good estimate for the expected number).The `GasStripping` test fails from time to time with a too large difference for the surviving number of particles.
The `GasStripping` test checks how many particles are lost, in the reference these are 1e5 - 99623 = 377. Since this is a stochastic process, the expected error is about sqrt(377) = 20. The allowed error right now in the test is 25. [The GasStripping test failed with a difference of 42](http://amas.web.psi.ch/opal/regressionTests/OPAL-2.0/results_2019-04-16.xml). To avoid that the test fails randomly from time to time, I propose to set the allowed error to at least 2 sigma=40, perhaps even better 3 sigma=60 (also because the reference of 377 might actually not be a good estimate for the expected number).gsellext-calvo_ppedro.calvo@ciemat.esgsellhttps://gitlab.psi.ch/OPAL/regression-tests/-/issues/66Degrader-1 test broken2019-08-01T06:08:15+02:00krausDegrader-1 test brokenThe necessary changes to the source code to make the reference particle lose energy in a degrader (OPAL/src#298) produce different results in this test. Will have to remake the reference.The necessary changes to the source code to make the reference particle lose energy in a degrader (OPAL/src#298) produce different results in this test. Will have to remake the reference.krauskraushttps://gitlab.psi.ch/OPAL/regression-tests/-/issues/65AWAGun not running2018-12-28T21:55:55+01:00krausAWAGun not runningWith the latest changes this regression test isn't running. I'm working on it.With the latest changes this regression test isn't running. I'm working on it.krauskraushttps://gitlab.psi.ch/OPAL/regression-tests/-/issues/64MAP-Circular test failing2019-02-15T08:51:23+01:00snuverink_jjochem.snuverink@psi.chMAP-Circular test failingThe MAP-Circular test is failing since its introduction.
@gsell : locally this test works for me, but it fails on the nightly without specific information (perhaps a timeout?). Do you have a log file with some more information why this test fails?The MAP-Circular test is failing since its introduction.
@gsell : locally this test works for me, but it fails on the nightly without specific information (perhaps a timeout?). Do you have a log file with some more information why this test fails?snuverink_jjochem.snuverink@psi.chsnuverink_jjochem.snuverink@psi.chhttps://gitlab.psi.ch/OPAL/regression-tests/-/issues/63Regression tests for 2.02018-08-23T16:33:18+02:00snuverink_jjochem.snuverink@psi.chRegression tests for 2.0Regression tests for the 2.0 branch are not yet running.Regression tests for the 2.0 branch are not yet running.gsellgsellhttps://gitlab.psi.ch/OPAL/regression-tests/-/issues/62Degrader-1 RNG2019-04-18T10:30:27+02:00adelmannDegrader-1 RNGCheck: https://gitlab.psi.ch/OPAL/regression-tests/commit/dca0692ea675bb25ed3425587db708c2976a398c
We should use a portable rng for this test.
Maybe string RNGTYPE RANDOM (default), Quasi-random number generators: HALTON, SOBOL, NIEDERREITER (Gsl ref manual 18.5)
just use one of the Quasi-random number generators.Check: https://gitlab.psi.ch/OPAL/regression-tests/commit/dca0692ea675bb25ed3425587db708c2976a398c
We should use a portable rng for this test.
Maybe string RNGTYPE RANDOM (default), Quasi-random number generators: HALTON, SOBOL, NIEDERREITER (Gsl ref manual 18.5)
just use one of the Quasi-random number generators.adelmannadelmannhttps://gitlab.psi.ch/OPAL/regression-tests/-/issues/60EGunCTF3-1, EGunCTF3-2 and PSIGUN-12018-04-11T10:32:41+02:00adelmannEGunCTF3-1, EGunCTF3-2 and PSIGUN-1EGunCTF3-1, EGunCTF3-2 and PSIGUN-1 is working fine on merlin: not broken and results are ok.EGunCTF3-1, EGunCTF3-2 and PSIGUN-1 is working fine on merlin: not broken and results are ok.adelmannadelmannhttps://gitlab.psi.ch/OPAL/regression-tests/-/issues/59Regression tests not reported/running anymore2018-03-26T09:56:44+02:00snuverink_jjochem.snuverink@psi.chRegression tests not reported/running anymoreSince a few days the [regression tests on master](http://amas.web.psi.ch/opal/regressionTests/master/) are not reported anymore. This might have to do with OPAL/src@0e771d47?
Also the [regression tests on 1.6](http://amas.web.psi.ch/opal/regressionTests/OPAL-1.6/) are not updated since 4-1-2018. Is this on purpose?Since a few days the [regression tests on master](http://amas.web.psi.ch/opal/regressionTests/master/) are not reported anymore. This might have to do with OPAL/src@0e771d47?
Also the [regression tests on 1.6](http://amas.web.psi.ch/opal/regressionTests/OPAL-1.6/) are not updated since 4-1-2018. Is this on purpose?gsellgsellhttps://gitlab.psi.ch/OPAL/regression-tests/-/issues/57RestartTest-2 broken2018-02-01T14:54:37+01:00krausRestartTest-2 brokenThis test ran yesterday, it failed but wasn't broken. However I don't see an apparent cause in the committed changes.This test ran yesterday, it failed but wasn't broken. However I don't see an apparent cause in the committed changes.frey_mfrey_mhttps://gitlab.psi.ch/OPAL/regression-tests/-/issues/56reference stat file of SAAMG-Test-1 broken2018-01-26T23:30:24+01:00gsellreference stat file of SAAMG-Test-1 brokenThe reference stat file of the SAAMG-Test-1 is broken due to an unresolved conflict before committingThe reference stat file of the SAAMG-Test-1 is broken due to an unresolved conflict before committinghttps://gitlab.psi.ch/OPAL/regression-tests/-/issues/55Presentation of results2018-01-23T22:44:32+01:00krausPresentation of resultsWith a little use of javascript we could present the results in a tidier way: @all please look at the pictures below and comment!
![results1](/uploads/1cacbe9e37140691f666ef54e119427d/results1.png)
![results2](/uploads/804d25f50dca616c37c00c3c787ce179/results2.png)
![results3](/uploads/05f6c823ab3e545d1c6c581ea49437da/results3.png)With a little use of javascript we could present the results in a tidier way: @all please look at the pictures below and comment!
![results1](/uploads/1cacbe9e37140691f666ef54e119427d/results1.png)
![results2](/uploads/804d25f50dca616c37c00c3c787ce179/results2.png)
![results3](/uploads/05f6c823ab3e545d1c6c581ea49437da/results3.png)krausgsellsnuverink_jjochem.snuverink@psi.chkraushttps://gitlab.psi.ch/OPAL/regression-tests/-/issues/53Include unit tests2019-05-08T18:03:24+02:00krausInclude unit testsThe unit tests should run (at least) every time the regression tests run. It should be easy to include the unit tests in the python script. A first step towards this is to compile the unit tests when compiling for the regression tests.The unit tests should run (at least) every time the regression tests run. It should be easy to include the unit tests in the python script. A first step towards this is to compile the unit tests when compiling for the regression tests.krausgsellkraushttps://gitlab.psi.ch/OPAL/regression-tests/-/issues/52Regression test for injected flat-top distribution2017-12-18T17:10:19+01:00snuverink_jjochem.snuverink@psi.chRegression test for injected flat-top distributionCurrently there is no regression test for an injected flat-top distribution.
Related to issue OPAL/src#194.Currently there is no regression test for an injected flat-top distribution.
Related to issue OPAL/src#194.https://gitlab.psi.ch/OPAL/regression-tests/-/issues/51opalimpact-1 failing after replacing OFFSETZ with ZSTART2019-02-15T08:51:23+01:00snuverink_jjochem.snuverink@psi.chopalimpact-1 failing after replacing OFFSETZ with ZSTARTSince we discussed to get the regression tests working, I had a more detailed look at one of them.
After replacing `OFFSETZ` with `ZSTART` in commit e2e902d8 the opalimpact-1 test is failing.
I believe this is because the starting s position is different in both cases, perhaps contrary to what one would expect(?).
In the reference .stat file there is (t,s):
```
1.000000000000000e+00 1.383007003378306e-02
2.000000000000001e+00 2.766014006756613e-02
```
So the reference starts at t=0.0, s=0.0
and the current version (t,z):
```
2.115406793793548e+00 2.883007003378255e-02
3.115406793793549e+00 4.266014006756509e-02
```
The current version starts at t=0.11 , s = 0.15 (value of `ZSTART`).
since the test has `ZSTOP = 1.0` the last line will be different and the test will fail.
One possible way to fix the test is to have `ZSTOP=1.15`.
I haven't checked the other tests that were changed in the same way, but these might also be affected in the same way.Since we discussed to get the regression tests working, I had a more detailed look at one of them.
After replacing `OFFSETZ` with `ZSTART` in commit e2e902d8 the opalimpact-1 test is failing.
I believe this is because the starting s position is different in both cases, perhaps contrary to what one would expect(?).
In the reference .stat file there is (t,s):
```
1.000000000000000e+00 1.383007003378306e-02
2.000000000000001e+00 2.766014006756613e-02
```
So the reference starts at t=0.0, s=0.0
and the current version (t,z):
```
2.115406793793548e+00 2.883007003378255e-02
3.115406793793549e+00 4.266014006756509e-02
```
The current version starts at t=0.11 , s = 0.15 (value of `ZSTART`).
since the test has `ZSTOP = 1.0` the last line will be different and the test will fail.
One possible way to fix the test is to have `ZSTOP=1.15`.
I haven't checked the other tests that were changed in the same way, but these might also be affected in the same way.kraussnuverink_jjochem.snuverink@psi.chkraushttps://gitlab.psi.ch/OPAL/regression-tests/-/issues/50some regression-tests use removed keyword DEBIN2019-02-15T08:51:23+01:00adelmannsome regression-tests use removed keyword DEBINDEBIN was removed from the DISTRIBUTION **BUT** the regressiontests are **NOT** updated. We should avoid this ....
Sorry was not able to find out who did this ....DEBIN was removed from the DISTRIBUTION **BUT** the regressiontests are **NOT** updated. We should avoid this ....
Sorry was not able to find out who did this ....https://gitlab.psi.ch/OPAL/regression-tests/-/issues/49Remake correct references after using P0 as momentum for reference particle2018-03-26T10:04:16+02:00snuverink_jjochem.snuverink@psi.chRemake correct references after using P0 as momentum for reference particleAfter commit [14df35e](https://gitlab.psi.ch/OPAL/src/commit/14df35ee805b6add1a1ef153ef39e1128d3aa3a4) more regression tests (in master) fail (from 94 to 39 passed). The reference files need to be remade.After commit [14df35e](https://gitlab.psi.ch/OPAL/src/commit/14df35ee805b6add1a1ef153ef39e1128d3aa3a4) more regression tests (in master) fail (from 94 to 39 passed). The reference files need to be remade.krauskraushttps://gitlab.psi.ch/OPAL/regression-tests/-/issues/48Remake correct references for those tests that are affected by the Hz -> MHz ...2017-10-27T13:49:42+02:00krausRemake correct references for those tests that are affected by the Hz -> MHz changethe references of some tests, e.g. EGunCTF3-1, have changed with 1a282f4c and are *wrong*.the references of some tests, e.g. EGunCTF3-1, have changed with 1a282f4c and are *wrong*.adelmannadelmannhttps://gitlab.psi.ch/OPAL/regression-tests/-/issues/47Fix units of BFREQ / BCURRENT2019-02-15T08:51:23+01:00krausFix units of BFREQ / BCURRENTIn the following tests it is unclear whether the units for BFREQ and BCURRENT are correct
Degrader-1
Distribution-Binomial-1
Distribution-Gauss-1
Distribution-Gauss-2
EGunCTF3-1
EGunCTF3-2
ExternalFieldTest
LaserEmission-1
OpalRingTest
ParallelPlateFurman1
ParallelPlateVaughan1
PSIGUN-1
RestartTest-1
RestartTest-2
RestartTest-3
RestartTest-4
RestartTest-5
RestartTest-6
RestartTest-7
RestartTest-8
RingCyclotron
RingCyclotronMatched
RingCyclotronMTS
RingCyclotronSingleParticle
Scan-1In the following tests it is unclear whether the units for BFREQ and BCURRENT are correct
Degrader-1
Distribution-Binomial-1
Distribution-Gauss-1
Distribution-Gauss-2
EGunCTF3-1
EGunCTF3-2
ExternalFieldTest
LaserEmission-1
OpalRingTest
ParallelPlateFurman1
ParallelPlateVaughan1
PSIGUN-1
RestartTest-1
RestartTest-2
RestartTest-3
RestartTest-4
RestartTest-5
RestartTest-6
RestartTest-7
RestartTest-8
RingCyclotron
RingCyclotronMatched
RingCyclotronMTS
RingCyclotronSingleParticle
Scan-1adelmannadelmann