Reduce secondary mirror iterations to 1 for RICH1 in quartic reconstruction

This MR is based on top of !884 (merged). The only change w.r.t. that MR is to reduce the (max) number of iterations inside the quartic photon reconstruction for RICH1 to 1. This iteration is designed to deal with the non-flat secondary mirrors, which is really only an issue for RICH2.

This MR is separate to !884 (merged) as whilst that MR introduces no changes in the output, this will will cause some changes. Hence I want to keep them separate, and allow !884 (merged) to be first tested and merged, before this one is unWIP'ed.

@cattanem I have not added any slot flags. If you are happy to test this in one particular slot now, just let me know (or add the label yourself).

added 1 commit

• daab4149 - Update example options

Compare with previous version

added 1 commit

• 23e0cdbd - Improve some doxygen comments

Compare with previous version

@jonrob I added lhcb-gaudi-head label, you can watch that nightly as from tomorrow

added 13 commits

• 23e0cdbd...23ed59d4 - 12 commits from branch master
• 2d66418b - Merge branch 'master' of ssh://gitlab.cern.ch:7999/lhcb/Rec into…

Compare with previous version

changed the description

• [2018-02-04 00:17] Validation started with lhcb-gaudi-head#1746
• [2018-02-05 00:18] Validation started with lhcb-gaudi-head#1747
• [2018-02-05 09:59] Validation started with lhcb-gaudi-head#1748
• [2018-02-06 00:17] Validation started with lhcb-gaudi-head#1749
• [2018-02-07 00:15] Validation started with lhcb-gaudi-head#1750
• [2018-02-08 00:16] Validation started with lhcb-gaudi-head#1751
• [2018-02-08 13:52] Validation started with lhcb-gaudi-head#1752
• [2018-02-09 00:19] Validation started with lhcb-gaudi-head#1753
• [2018-02-09 13:31] Validation started with lhcb-gaudi-head#1754
• [2018-02-09 17:23] Validation started with lhcb-gaudi-head#1755
• [2018-02-10 00:11] Validation started with lhcb-clang-test#857
• [2018-02-10 00:12] Validation started with lhcb-gaudi-head-py3#60
• [2018-02-10 00:14] Validation started with lhcb-tdr-test#55
• [2018-02-10 00:15] Validation started with lhcb-head#1750
• [2018-02-10 00:22] Validation started with lhcb-gaudi-head#1756
• [2018-02-10 12:51] Validation started with lhcb-gaudi-merge#412

@jonrob As expected, this gives small differences in (many) Brunel tests, see https://lhcb-nightlies.cern.ch/logs/tests/nightly/lhcb-gaudi-head/1746/x86_64-slc6-gcc62-opt/Brunel/. As you have other MRs in the pipeline, how do you want to proceed:

• Update references and merge one by one
• Test all the MRs together and update references in one go? Latter is obviously faster and less work, but does not easily let you see any side effects of the individual MRs

Yes, I submitted the other MRs as I specifically wanted to double check each on its own. My idea was to create an MR for Brunel for the reference updates, and update it one at a time once I am happy with one Rec MR, allowing a test of the next MR in the lhcb-gaudi-head slot over night. So for instance once I am happy with the diffs due to this MR, I will create the Brunel MR with the updated refs, and label !888 (merged) to be tested next in lhcb-gaudi-head.

OK, perfect. So once you have the Brunel MR for this one you can un-WIP it and remove label. Please remember that the brunel2012magdown.ref.new produced in the nightlies is no good if the repro2012magdown test has also failed, as the latter wins, so you have to regenerate it by re-running the brunel2012magdown test locally (make test ARGS="-R brunel2012magdown"). Also, the reference blocks for ProtoANNPID may need to be updated.

mentioned in merge request Brunel!357 (merged)

I've submitted the MR for the changes here. Will hold off setting the another MR to build in lhcb-gaudi-head until I am sure the refs are correct.

mentioned in merge request !888 (merged)

@jonrob The Brunel tests in https://lhcb-nightlies.cern.ch/logs/tests/nightly/lhcb-gaudi-head/1748/x86_64-centos7-gcc62-opt/Brunel/ look fine. I think you can move on to the next one.

unmarked as a Work In Progress

changed title from WIP : Reduce secondary mirror iterations to 1 for RICH1 in quartic reconstruction to Reduce secondary mirror iterations to 1 for RICH1 in quartic reconstruction

merged

mentioned in commit 22723167