Hello Olivier,
I agree we should release 3.0.1 before the Christmas break.
concerning the side effect, i reproduced it, and it goes away if I do the following change
class TestAMCatNLOEW(unittest.TestCase):
""" a suite of extra tests for the ew stuff """
The other bogus behavious was related to the order settings in t-channel single top, which we discussed in a separate email.
I have added an error message that should warn the user to specify the coupling orders in that case (rev 958)
Cheers,
Marco
> On 17 Dec 2018, at 09:23, Olivier Mattelaer <email address hidden> wrote:
>
> Hi,
>
> I would like to put this quite high priority for this week (i.e. for me Wednesday-friday).
>
> I have rerun the tests and found that some tests are crashing when run via the full script but are working if run alone. So I have run (the very slow script to determine which of the tests had such side effects and the culprint is:
> tests.unit_tests.fks.test_extra_ew.TestAMCatNLOEW.test_generate_fks_2to1_no_finalstate_confs
> which has a side effect that lead to the following crash:
>
> FAIL: test_generate_fks_ew_dijet_qed2qcd2_qcd (tests.unit_tests.fks.test_extra_ew.TestAMCatNLOEW)
> check that the generate command works as expected.
> ----------------------------------------------------------------------
> Traceback (most recent call last):
> File "/Users/omattelaer/Documents/workspace/3.0.1/tests/unit_tests/fks/test_extra_ew.py", line 869, in test_generate_fks_ew_dijet_qed2qcd2_qcd
> self.assertEqual(real.fks_infos[0]['extra_cnt_index'], -1)
> AssertionError: 0 != -1
>
> Might be interesting that you take a look. But I think that we should release this version this week independently.
>
> Any objection? comment?
>
> Olivier
>
>
> --
> https://code.launchpad.net/~maddevelopers/mg5amcnlo/3.0.1/+merge/356133
> You are reviewing the proposed merge of lp:~maddevelopers/mg5amcnlo/3.0.1 into lp:~maddevelopers/mg5amcnlo/FKS_EW_granny.
Hello Olivier, unittest. TestCase) :
I agree we should release 3.0.1 before the Christmas break.
concerning the side effect, i reproduced it, and it goes away if I do the following change
class TestAMCatNLOEW(
""" a suite of extra tests for the ew stuff """
- interface = mgcmd.MasterCmd()
+ def setUp(self):
+ self.interface = mgcmd.MasterCmd()
I have pushed the fix into rev 957.
The other bogus behavious was related to the order settings in t-channel single top, which we discussed in a separate email.
I have added an error message that should warn the user to specify the coupling orders in that case (rev 958)
Cheers,
Marco
> On 17 Dec 2018, at 09:23, Olivier Mattelaer <email address hidden> wrote: tests.fks. test_extra_ ew.TestAMCatNLO EW.test_ generate_ fks_2to1_ no_finalstate_ confs fks_ew_ dijet_qed2qcd2_ qcd (tests. unit_tests. fks.test_ extra_ew. TestAMCatNLOEW) ------- ------- ------- ------- ------- ------- ------- ------- ------- omattelaer/ Documents/ workspace/ 3.0.1/tests/ unit_tests/ fks/test_ extra_ew. py", line 869, in test_generate_ fks_ew_ dijet_qed2qcd2_ qcd l(real. fks_infos[ 0]['extra_ cnt_index' ], -1) /code.launchpad .net/~maddevelo pers/mg5amcnlo/ 3.0.1/+ merge/356133
>
> Hi,
>
> I would like to put this quite high priority for this week (i.e. for me Wednesday-friday).
>
> I have rerun the tests and found that some tests are crashing when run via the full script but are working if run alone. So I have run (the very slow script to determine which of the tests had such side effects and the culprint is:
> tests.unit_
> which has a side effect that lead to the following crash:
>
> FAIL: test_generate_
> check that the generate command works as expected.
> -------
> Traceback (most recent call last):
> File "/Users/
> self.assertEqua
> AssertionError: 0 != -1
>
> Might be interesting that you take a look. But I think that we should release this version this week independently.
>
> Any objection? comment?
>
> Olivier
>
>
> --
> https:/
> You are reviewing the proposed merge of lp:~maddevelopers/mg5amcnlo/3.0.1 into lp:~maddevelopers/mg5amcnlo/FKS_EW_granny.