Home » Groups » STAR Protected

L2/emcCheck Comparison

Updated on Fri, 2013-10-04 15:17. Originally created by dolv719 on 2013-10-04 10:41.

When doing the QA of the BEMC using the status tables from L2, it was noticed that there were low statistics for approx. the last third of run 13.  All of the the information can be found here, and the instructions I created to do the QA are here.  The solution to this problem (we hope) is to produce the plots from the emcCheck runs in a similar matter to the EndCap.  This was done by Alice and Mike, and the new code can be found in the following directory /ldaphome/onlmon/emcstatus2013/pp500test/.

Kevin then wrote some code that is used to compare the two methods mentioned above. (/ldaphome/onlmon/emcstatus2013/compareMethods).

After the code was run, it needs to be compared to see if the two different methods to produce the plots gives similar results.  To see the plots use this blog. Alice also created another blog that nicely summarizes when the two different gives different statuses.  Note: That the second row of plots and the table below are for all of run 13, and the issue for the emcCheck status=1 and L2 status=0 has already been corrected.   In l2status.py some lines of code were added, which would start with the runnumber of the emc-check run, and then increment the runnumber by 1 until it finds an L2 file, and then read that L2 file to figure out which towers are masked.  After this was corrected the number dropped from 831 to 10, all for F17434.

The following statuses were compared: Note: All decisions in L2 for Fill 17568 and on (last third of run13) should be taken lightly, these is where there were low statistics so differentiating the statuses in L2 is all but impossible.  All spreadsheets used are attached. Plots for L2 and emc

  • L2=1 emc=18
    • Almost complete agreement.  The one exception is Fill 17415.
  • L2=18 emc=1
    • All plots marked 18 in L2 looked 18 in the plots
    • There wasnice agreement in emc except for the following towers: 1701, 2092(became 18 around Fill 17340)
    • Tower 4058 was 18 in L2, but looked 2 in emc.  After another round on emc QA, hardcode 4058 status 2?
  • L2=1 emc=2
    • For fills 17269 & 17479: the tower listed appear to be status 2.  No agreement in L2 tho.
    • For fill 17568: the towers are marked 2 in emc, but they look the same as tho marke status 1.  The same is true for fill 17566.  It only appears that this problem appears in those two fills, points to a possible problem with the emcCheck file?  Algorithm to find status 2 seems for work for the other fills.
  • L2=2 emc=1
    • All towers marked 2 in L2 are right, but only 1612 agrees and maybe 916
    • Again another problem in 17568 see above
  • L2=1 emc=254
    • All towers for emc are marked 254, but seem to be fine
    • Almost all L2 towers are good
    • For fill 17479 a few towers marked 1 appear to be 18.
    • Only seem to have this prolem for the middle fills 17479-17526

Based on the above I feel that the emcCheck runs agree well enough with the L2 plots, and for run 13 emcCheck runs should be used to produce the BEMC status tables.






 

Groups: