Fix some Tesla tests

At the moment this just gets a couple of the Tesla tests running again. They don't pass for me locally, and I don't expect to in the nightlies either, but it's a better place to start debugging from. I can update this MR with any progress, so I'd say let's leave it open for a week and hopefully I'll have some more news soon.

OK, sounds good. Thanks a lot.

• [2018-02-16 00:27] Automatic merge failed in lhcb-2017-patches#296
• [2018-02-17 00:21] Validation started with lhcb-2017-patches#297
• [2018-02-17 00:54] Validation started with lhcb-2017-patches#297
• [2018-02-17 01:25] Validation started with lhcb-2017-patches#297
• [2018-02-17 16:39] Validation started with lhcb-2017-patches#297
• [2018-02-17 17:06] Validation started with lhcb-2017-patches#297
• [2018-02-17 17:32] Validation started with lhcb-2017-patches#297
• [2018-02-18 00:18] Validation started with lhcb-2017-patches#298
• [2018-02-19 00:18] Validation started with lhcb-2017-patches#299
• [2018-02-20 00:19] Validation started with lhcb-2017-patches#300
• [2018-02-21 00:29] Validation started with lhcb-2017-patches#301
• [2018-02-22 01:20] Validation started with lhcb-2017-patches#302
• [2018-02-23 01:20] Validation started with lhcb-2017-patches#303
• [2018-02-24 01:19] Validation started with lhcb-2017-patches#304
• [2018-02-25 01:21] Validation started with lhcb-2017-patches#305
• [2018-02-26 01:18] Validation started with lhcb-2017-patches#306
• [2018-02-27 01:27] Validation started with lhcb-2017-patches#307
• [2018-02-28 01:21] Validation started with lhcb-2017-patches#308
• [2018-03-01 00:28] Validation started with lhcb-2017-patches#309
• [2018-03-02 00:24] Validation started with lhcb-2017-patches#310
• [2018-03-03 00:20] Validation started with lhcb-2017-patches#311
• [2018-03-04 00:21] Validation started with lhcb-2017-patches#312
• [2018-03-05 00:28] Validation started with lhcb-2017-patches#313
• [2018-03-06 00:22] Validation started with lhcb-2017-patches#314

mentioned in merge request !149 (closed)

• [2018-02-17 17:16] Validation started with lhcb-2017-patches#297
• [2018-02-17 17:21] Validation started with lhcb-2017-patches#297
• [2018-02-17 17:31] Validation started with lhcb-2017-patches#297
• [2018-02-17 17:44] Validation started with lhcb-2017-patches#297
• [2018-02-17 17:47] Validation started with lhcb-2017-patches#297
• [2018-02-17 17:49] Validation started with lhcb-2017-patches#297

added 1 commit

• 22a8c0e7 - Update of refs for new input file for tests tesla.qms/z_FSRstream_*

Compare with previous version

Hi @apearce, I updated the refs for 2 of the files, to continue the clean-up/updates. These updates are necessary given that the input file changed. Let me know if you disagree, of course. There are still a few more tests to sort out, though ...

@erodrigu The luminosity change does not correspond to the change in calibration, so I am very weary of this update. I will take a look at the test

I did see the changes in numbers. But how did you immediately figure that it was not expected based on the change in calibration? Would be good to know, thanks.

@erodrigu basically, the change in scale 1.71e-11 -> 1.577e-11 does not correspond to the change in luminosity

Indeed the ratio of ratii is very far from 1. The only difference in the discrepancy could come from the sample size. If the new input file is supposed to contain the same events as before, then this is indeed very worrisome. Can you confirm this point about the input file, @apearce ?

I remember seeing these differences quite some time ago, but didn't have time to investigate further. I don't think the input file is the problem.

changed this line in version 3 of the diff

@apearce, the plan is to release a full stack on this branch asap. It would be ideal to have this MR issues sorted once and for all. Is there any hope for this given time constraints?

The changes to the configuration can be merged, but the changes to lumi statistics in the reference files I'd rather not include, as we don't understand the differences. I won't be able to look at that this week, so I suggest proceeding with the release.

added 1 commit

• 9cf591dd - Revert "Update of refs for new input file for tests tesla.qms/z_FSRstream_*"

Compare with previous version

Fair enough. I've just reverted back to where we were before my commit. This MR can then be accepted ... But we should not forget these differences still observed, to try and clean-up the tests ...

mentioned in commit 3178526d

merged