HOWTO: Run Base QA Tracking Evaluation

We will evaluate the tracker performance in the following steps:

1) Simulate 200 GeV AuAu minbias events and reconstruct them
2) Run the base QA codes, creating many useful plots for evaluating tracking efficiency
3) Create a web page showing the results.

The TrackerBaseQA package provides all of the codes needed for these tasks.  You need to only edit them to suit your needs and run.   This document will show you how.

First you should checkout the package and setup symbolic links

$ cvs co offline/users/jwebb/TrackerBaseQA/
$ ln -s offline/users/jwebb/TrackerBaseQA/* .

You will also need to setup python to find ROOT.  Execute the following and (or) add this to your ~/.cshrc file (and start a new session).
 

$ setenv PYTHONPATH $ROOTSYS/lib:/opt/star/lib/python2.7/site-packages/


Next we want to run the simulations.  This is done by submitting the runHijingSimulation.xml macro to the STAR batch system.  Before running, you should edit the file.  At the top of the file there are several entities declared:
 

<!ENTITY WORKINGDIR "/star/simu/simu/jwebb/2017/11-28-2017-misalign-embedding/">
<!ENTITY LOGDIR     "/star/simu/simu/jwebb/2017/11-28-2017-misalign-embedding/logHI">
<!ENTITY OUTDIR     "/star/simu/simu/jwebb/2017/11-28-2017-misalign-embedding/recoHI">
<!ENTITY NEVENT     "10">
<!ENTITY NJOBS      "100">
<!ENTITY GEOMETRY   "y2017a"> 
<!ENTITY TIMESTAMP  "sdt20170214"> 
<!ENTITY ALIGNMENT  "ideal"> 

WORKINGDIR should be set to the directory where you are running the jobs from (your current working directory).
LOGDIR should be set to the directory where you want the log files copied to when the job finishes.
OUTDIR should be set to the directory where you want the MuDst files to be saved.  This should be in an area where you have sufficient space.  (Your home directory is a bad choice...)

NEVENT and NJOBS specify the number of events to generate per job, and the total number of jobs.  I recommend around 5000 total events. 

GEOMETRY specifies the version of the STAR detector.
TIMESTAMP should be set within the given RHIC run.

If you need more detailed control of either the simulation or reconstruction, you can edit the runHijingSimulation.C and runHijingReconstruction.C macros.

The code is run using

$ star-submit runHijingSimulation.xml

The result will be a number of MuDst files stored in the directory specified by OUTDIR.  The next step is to setup and run the runBaseQa.py python script.  

$ vi runBaseQa.py

Go to the end of the file.  Here you will see a setupExample() function.  There are two fields which you need to set.  One is qa.mudir, the second is qa.imgdir.  Essentially:

qa.mudir = "/path/to/your/mudst/files"
qa.imgdir = "/where/your/plots/will/go"

Once you have the code setup, you simply run it using
 

$ python runBaseQa.py

Images will be written to your imgdir.  Our next task is to create a web page for viewing the plots (there are, after all, alot of them).  You should create a directory in your web area where the files will reside.  You can find help for doing that on this pageIt is recommened that you setup AutoIndex using instructions found here.

Let's say you want to create a web page called study.html, and link it to three sets of QA histograms.  QA1, QA2 and QA3.
 

$ python BaseQaPage.py > /path/to/your/WWW/area/study.html
$ cp -R /path/to/QA1 /path/to/your/WWW/area/QA1
$ cp -R /path/to/QA2 /path/to/your/WWW/area/QA2
$ cp -R /path/to/QA3 /path/to/your/WWW/area/QA3

This will create an html page, accessible from http://www.star.bnl.gov/~username/area/study.html, which displays three columns of plots, corresponding to the image files created in the QA1 QA1 and QA3 directories.

You may, of course, edit BaseQa.py in order to point to different directory names, and to give each column an appropriate description.