Closed rvizarreta closed 1 year ago
I create a local variable from my work area location for $ROCKMUONCALIBRATIONROOT. Now I have a new issue:
source master_s2s_script.sh
-- s2s script parameters --
-- data_processing directory /minerva/data/users/minervacal/data_processing_s2s_eroica
-- data_processing version v10r8pre
-- first run 846 2
-- last run 898 11
/minerva/data/users/minervacal/data_processing_s2s_eroica/
Traceback (most recent call last):
File "make_playlist.py", line 28, in
* Creating plex map **
bash: [: -le: unary operator expected bash: Gaudi.exe: command not found Could not make plex map! Connection to minervagpvm05.fnal.gov closed.
Chris comments about current issue:
OK, so I think master_s2s_script.sh is something that will automatically run through a bunch of IOVs (intervals of validity). This is intended to automate the process once everything is very stable Right now, I don't think you want to do that. You should basically look at what it is doing and do the steps one by one, by had, because you will get lots of errors So looking at my readme, basically you want to skip step 0 for now -- that is making the ntuple files. At some point we will want to integrate this with what Carlos is doing on the production side We probably aren't going to use IOVs for s2s, we will just do it once, so we don't need to do that step either I would basically go to where it says "details of Step 3" and follow that For alignment, you will want to use the output of the ReadNT step
I added a sample .root file to test this calibration, also I am sourcing manually instead of sourcing master_s2s_script.sh. Note that on make_s2s.sh I had to comment on each exit command because it closed my ssh connection. Now I am looking at the error generated by each step:
source make_s2s.sh /minerva/data/users/rvizarr/minervacal v10r8pre 847 2 898 11 -- s2s script parameters -- -- data_processing directory /minerva/data/users/rvizarr/minervacal -- data_processing version v10r8pre -- first run 847 2 -- last run 898 11 rundir:/minerva/data/users/rvizarr/minervacal/ 1 1 /minerva/data/users/rvizarr/minervacal/minervatest_01_01
* Creating plex map **
bash: Gaudi.exe: command not found Could not make plex map! mv: cannot stat ‘plex.root’: No such file or directory
* Reading Ntuples ***
bash: ./ReadNT: No such file or directory Could not read ntuples!
Creating summary tuple
bash: root: command not found Could not make summary tuple!
* Creating constants ****
bash: root: command not found
Could not make s2s constants!
bash: ./DoFitCorrection: No such file or directory
Could not make fit correction file!
cp: cannot stat ‘/minerva/data/calibrations/s2s_tables/minerva/v5/s2s_constants_gen2_prime_01.txt’: No such file or directory
Traceback (most recent call last):
File "IterateS2SConstants.py", line 1, in
***** Done! ****
I forgot to make the make
command before the start, that's why I was having some of the previous issues. The make
command will essentially run the Makefiel file. After doing this, the current issue is:
source make_s2s.sh /minerva/data/users/rvizarr/minervacal v10r8pre 847 2 898 11 -- s2s script parameters -- -- data_processing directory /minerva/data/users/rvizarr/minervacal -- data_processing version v10r8pre -- first run 847 2 -- last run 898 11 rundir:/minerva/data/users/rvizarr/minervacal/ 1 1 /minerva/data/users/rvizarr/minervacal/minervatest_01_01
* Creating plex map **
bash: Gaudi.exe: command not found Could not make plex map! mv: cannot stat ‘plex.root’: No such file or directory
* Reading Ntuples ***
Getting ntuples... Processing 657 tracks for 40001/1 Make time header... There are 1 files Filling summary histograms... Determining truncated mean with 8 iterations... Filling summary histograms for module 0... Filling summary histograms for module 10... Filling summary histograms for module 20... Filling summary histograms for module 30... Filling summary histograms for module 40... Filling summary histograms for module 50... Filling summary histograms for module 60... Filling summary histograms for module 70... Filling summary histograms for module 80... Filling summary histograms for module 90... Filling summary histograms for module 100... Filling summary histograms for module 110...
Creating summary tuple
root [0]
Processing MakeSummaryTuple.C...
Error in
* Creating constants ****
root [0]
Processing PrintConstants.C...
Error: illegal pointer to class object nts 0x0 1383 PrintConstants.C:31:
Interpreter error recovered
Processing 657 tracks for 40001 / 1
Determining truncated mean
Fitting...
cp: cannot stat ‘/minerva/data/calibrations/s2s_tables/minerva/v5/s2s_constants_gen2_prime_01.txt’: No such file or directory
Traceback (most recent call last):
File "IterateS2SConstants.py", line 1, in
***** Done! ****
It seems that now it is trying to read the tuples, but failing due to the plex map error. After the weekend I'll ask my team about how to run Gaudi. I also needed to point the $MINEVA_GEOMETRY environment variable to the following location: /minerva/app/users/rvizarr/cmtuser/Minerva_v22r1p1_2x2/Det/MinervaDDDB/DDDB/
After some discussion with Chris we came to the conclusion that just need to use the part of the code related to the alignment. This is bringing new issues as it seems that the ReadNT.root file is empty.
I am not sure if I can modify ReadNT.C as it is already compiled on a .o file. But I'll try to generate a new root file using the .C file. The problem seems to be definitely here, will do more exploration.
I created a copy of the original ntuple just to see if the problem was related to being only one tuple available, and the ReadNT.cxx file worked now. The root file produced is ReadNT.root now has some interesting plots:
The alignment constants file is being generated but it is empty. Now the DoAlignment.C file shows a different error:
root -l DoAlignment.C
root [0]
Processing DoAlignment.C...
Warning in
Take some time to clean your workspace, maybe create new folders dedicated only to the alignment. Is confusing to work with so many files.
The issue was related with a piece of code Chris omitted from his original code. He provided me with an oldest version that makes the 3D TProfile object: https://cdcvs.fnal.gov/cgi-bin/public-cvs/cvsweb-public.cgi/AnalysisFramework/Cal/RockMuonCalibration/scripts/s2s/Attic/ReadNT.C?rev=1.1.2.1;content-type=text%2Fplain;cvsroot=mnvsoft
Wait, it seems like p3ModLposBase (new object) is the same as p_moduleLposBase (current object). Ok I just changed the name on the DoAlignment.C file of the object and started plotting. It fails but now I can manage it.
On DoAlignment.C we iterate over the 240 planes. For calibration, do we need to use only the first module?
for (int modbin = 1; modbin <= 240; ++modbin)
. It seems that changing the loop won't run properly. For now let's just run everything and see the final output.
This is working now, we just need to run it again with more ntuples.
I got this issue after correcting the paths on
make_s2s_script.py
:source master_s2s_script.sh bash: cd: /scripts/s2s: No such file or directory -- s2s script parameters -- -- data_processing directory /minerva/data/users/minervacal/data_processing_s2s_eroica -- data_processing version v10r8pre -- first run 846 2 -- last run 898 11 /minerva/data/users/minervacal/data_processing_s2s_eroica/ Traceback (most recent call last): File "make_playlist.py", line 28, in
list = os.listdir(rundir)
OSError: [Errno 2] No such file or directory: '/minerva/data/users/minervacal/data_processing_s2s_eroica/'
* Creating plex map **
bash: cd: /options: No such file or directory cp: cannot stat ‘PlexWriter.opts’: No such file or directory bash: [: -le: unary operator expected bash: Gaudi.exe: command not found Could not make plex map! Connection to minervagpvm05.fnal.gov closed.