Welcome To Wei Zou's Work Space


Peking University, Beijing, China

Working Notes(Update everyday in Fermilab)

I will write the working notes everyday in the Fermilab. The note is the summary of daily working activity and thinking about the physics. Enjoy writing and thinking!!!

Fermilab 2011/09/24-2011/09/25

These two days I finish running the muon channel data samples and checking the distributions. Using the old weight factor, Data agrees with MC very well. And I then begin to extract the QCD shape from the Data. Electron channel QCD shape can be exacted from two ways: 1. gamma conversion 2. anti-relisolation. I am still working on that. The main idea to use the data driven method is we want to get a better distribution about the QCD multiplicity events. I am still working on that.

Fermilab 2011/09/21-2011/09/23

These days I begin to process the new Muon and electron data. I am not very familiar with the trigger menu, while the CMS trigger menu changed very rapidly. And I begin to to use Fan module(This module can deal with wild-card in the trigger name) and update the json file. For the Rereco dataset, I use the special rereco json file and use the latest Prompt Golden json file. And then I use the lumiCalc2.py to get the Luminosity. The main problem is that the Prompt_v4 dataset covers two trigger menu(1e33 and 1point4e33). At the beginning, I didn't consider these effects. Later I found these problems, so I update all the muon and electron channel data samples. Then Data agrees with the MC very well. We have 2 /fb data now.

The other reason is the electron trigger. The single electron trigger pt is so high(65GeV). So we begin to use the inclusive "W" trigger and electron+2jet+PFMHT trigger. The man problem is that this inclusive trigger has the MT cut. While in the MC samples, they didn't have that kind of trigger. So we have to get grid of the trigger impact. We need add the W boson transverse mass in our analysis cut(Mt > 40 GeV). Then the problem is sovled. The muon channel is updated too.

Fermilab 2011/09/19-2011/09/20

These two days I study the new cuts which may help to improve the S/B. The new cut is study the angular correlation between the top quark and the leading jet(which should be decaying from the bottom quark). Top quark is the best combination whose invariant mass is closet to top quark mass. And I study the cut: deltaR(top, b), deltaPhi(top,b), leading jet pt / top pt and top pt. It seems to these angular cut will improve the S/B. I think the top pt is the most promising cut. This top pt is simple and robust. Because it's the kinematic cut.

Fermilab 2011/09/14-2011/09/18

These days I finish reading the CMS trigger menu from 5e33 to 3e33 peak lumi. Through these readings, I find some proper and useful trigger for our analysis. We will update our data to 2 /fb both on muon channel and electron channel. The muon channel should use the HLT_IsoMu24 trigger which is not prsscaled under the 3e33 trigger menu. And for the electron channel, the trigger is not trivial. Because the single electron pt is so high. Then We think we should consider the inclusive "W" trigger(Electron PT + Electron ID + PFMHT) and ELectron+ hadron trigger(Electron + Jets (+ Btag)). We will discuss it to choose the appropriate trigger for the electron channel.

There are several useful trigger Web for future checking.

1. SWGuideHighLevelTrigger

2. HowToFindPrescales

3. HLTriggerTools

4. TriggerTables

5. FNAL4Tops And I also check the standard modulr "HLTHighLevel". This module can work with wild-card trigger name. It's very useful to use that module.

Next I read some papers on the EXO, Super Symmetry and EWK papers to get some useful information.This EXO use the latest nuisance parameter setting. I think we should use that. And then I use the THETA package to get Electron channel temporary mass limit plot. It seems a little different with muon channel distribution. We will check to find the reason.

And I also read some statistic material and begin to produce the analysis systematic error samples for two channels.

And I need to deal with some distribution and make the right combination to get the top mass. We should measure the second leading jet index under the closest top mass hypothesis. And then I will check the report on the BSM meeting to plot the distributions of DeltaPhi(MET,lepton), DeltaR(top,b quark jet).etc.

Fermilab 2011/09/09-2011/09/13

I begin to learn the THETA package software and at the same reading the book about statistic knowledge. This book's name is "Statistical methods in experimental physics" written by James.

The books contains a lot of statistic theory knowledge and many useful statistic distributions. And THETA package can deal with the nuisance parameter and systematic templates. The nuisance parameter error should cite the latest CMS measurement result. This result can found on the latest CMS approve result especially related with our analysis. We will use THETA package with settle our major systematic error: 1. Jet energy Scale Up and Down 2. B tag systematic Scale Factor Up and Down. I can find the reference configure setting on the THETA web.

Fermilab 2011/09/08

Today I analysis the btag systematic error and Jet energy correction Scale Up and Scale Down template. This btag systematic error is : For b tag 9%, c tag 18%. So I store four each templates: at least 1TCHPM btag: Up and Down; 2 TCHPM btag: Up and Down. And I begin to learn the theta package to set the mass limit.

Fermilab 2011/09/07

Today I update the new verison V8_1 electron channel files and put in the ttmuj common space.

This update is focusing two changes:

1. Store the Generator MC information to check the W' decaying.

2. Store the Electron isolation information.

And I make the new Electron Wprime Mass template which is based on the new B tag weight factor. And I make the tex(latex) file to produce the Electron channel cutflow. Next day I will focus on the electron trigger we will use.

Fermilab 2011/09/06

Today I give a talk on the Update Electron channel Wprime in the FNAL local meeting. The talk is using the pile-up re-weight factor and the average btag weight factor. At last I require 2 TCHPM btag jet, the left background is ttbar. So I think we should study some cut to reduce the ttbar--our major background. Because requiring 2 btag jet will suppress the Up mass limit too much. I think asking one b tag jet is better and asking the other cut to reduce the ttbar background. Next task is Updating the electron to 1.2 /fb. So I will do something about the Electron HLT trigger.

Fermilab 2011/09/05

Today is USA Labor's Day, enjoy playing!!!!

My Interested Topics

There are several topics I am interested in, especially the top-quark related physics and Higgs physics.

W'-> tb

WW WZ Cross Section Measurement


Many Useful Links

1 TopLikeBSMSpring2011

2 2010Top Group Muon Channel selection

3 SimpleCutBasedEleID

4 PileupMCReweightingUtilities

5 TriggerTables

6 LumiCalc

Software Install

Using the CVS service at the FNAL CMSLPC machine:

1. If you are using FNAL machines, be sure your .cshrc file includes:

           source /uscmst1/prod/sw/cms/cshrc prod

2. you can set up by typing:

For a write-read session:

           project CMSSW


For anonymous log in:

          cmscvsroot CMSSW
          cvs login 

3. Then you can use the CVS command (be sure including "-d $CVSROOT"):

          step 1:   cvs -d $CVSROOT update -Ad
          step 2:   cvs -d $CVSROOT add 
          step 3:   cvs -d $CVSROOT commit
          step 4: cvs -d $CVSROOT tag Vxxxx

          How to create a cvs branch:
          step 1: cvs -d $CVSROOT tag -b < new_branch_name >
          step 2: cvs -d $CVSROOT update -r < new_branch_name >
          step 3: CVS -d $CVSROOT	commit

          To add a new file to an existing CVS branch:

         cvs -d $CVSROOT add -m "blah" file1


Install Python and ROOT to build the PyROOT

PyROOT is very important for me(like python very much) to use the ROOT utility and python advantages. 

   1. Down the python from the Web site: http://www.python.org/download/releases/2.6.7/python. using the command: wget + link of python file

   2. Compile the python, the most important note is that: be sure to --enable-shared. 
      This option is to ensure that the shared python library will be generated for the root to compile. 
      Then root can use this shared library to generate the position-independent code(-fPIC).
      (In particular, the static library will contain object files using position-independent code (PIC) on platforms where PIC flagsare needed for the shared library.)
      cd (python directory)   
      ./configure --enable-shared --prefix=(the python bin and bin directory you want python to build in)
      make install

   3. Be sure to export or setenv the shell variable before compile the root file in you ~/.bashrc or ~/.tcshrc:
       export PYTHONDIR=/home/cms/zouw/mypython2
       export ROOTSYS=/home/cms/zouw/Myroot1/root
       export PATH=$ROOTSYS/bin:$PYTHONDIR/bin:$PATH
       export PYTHONPATH=$ROOTSYS/lib:$PYTHONPATH       

   4. Download the root file and unzip the tar file. 
       gzip -dc roottarfile | tar -xvf -
       cd root
        make install 

    Then you are succeed.
    5. When you run the python script, you need to change the head: #!/usr/bin/python to #!/PYTHONPATH/python
        Then you can use your own root and python.

Install Cygwin

Cygwin is a Open-Source software which can make Windows feel like Linux or Unix. It's very convenient to use when you don't want to install another Linux or Unix system. Using the Cygwin, you can even compile root source(just Linux or Unix)and using Linux editors like Vi, emacs,etc.

1. Download the Cygwin from here and follow the instructions of the Cygwin User Guide. And choose the nearest mirror website, Be Careful, choosing the website is very important which will impact the download speed crucially.

2. The basic category we will use is the X11 category. We need choose xorg-server, xinit, xorg-docs, X-start-menu-icons. And we also need openssh package. If you want to compile the root source file, you need to select gcc, g++, gdb, make in the Devel category. Moreover we can also download vim(gvim), emacs in the editor category or even python!!!! The default action for the Cygwin is "Skip", you need click the "Skip" and then choose the latest version.

3. Then click "next" and wait the Cygwin to download and install the package we need.(Now you can go and do something else)

4. At last keep the Cygwin as the short icon on the PC Desk. Then you can run!!! How to use the Cygwin, you can also find the good reference on the User Guide.

5. I often use startwin or startw to start the X server. Then enjoy using the Cygwin!!!

Configure the Cygwin for FNAL Kerberized SSH

Default Setting of Cygwin is not enough for you to use the "ssh" to login in the FNAL CMSLPC Unix machine. You need to configure some special files under the /ect, /ssh and /bin directory.

1. First, you should mv /bin/ssh.exe /bin/ssh_standard.exe, download and unzip the executables and put them under the /bin directory. These exe file had been modified to include the kerberize utility(kinit, klist, kdestroy, kpasswd) and versions of ssh, ftp, and scp.

2. Second, Get the latest Fermilab krb5.conf and put it in /etc directory.

3. Third, Create a directory called .ssh in your home area and put the ssh config file there. And then change the protections on the ssh config file:

chmod 600 ~/.ssh/config

4. Fourth, Create a directory ssh under the /etc directory. And put the config file under the /etc/ssh directory

5. Add the following lines to your ~/.bashrc (assuming BASH is your default shell)

export KRB5_CONFIG=/etc/krb5.conf

export KRB5CCNAME=FILE:/var/tmp/krb5_cc_cache

6. Fifth, you can open an X sever and login in the FNAL CMSLPC machine:

/bin/kinit.exe -A -f username@FNAL.GOV


/bin/ssh.ete username@cmslpc-sl5NOSPAMPLEASE.fnal.gov

7. Successful, you can enjoy working with the Cygwin!!!!!!!!!!!

Edit | Attach | Watch | Print version | History: r22 < r21 < r20 < r19 < r18 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r22 - 2012-06-09 - WeiZou
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    Main All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2019 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback