HONR269L - Research

Research Projects - Basic Overview

Top Quark Properties at LHC

    The top quark is the heaviest of the fundamental particles, and it is commonly a key player in new models of physics, such as a concept called technicolor. Furthermore, the top quark is also involved in supersymmetry, and, being the heaviest fundamental particle, it is suggestive that the top quark could be involved in beyond the standard model physics (Review of Top Quark Physics).

IceCube UMD Group

    IceCube is a high-energy neutrino observatory and is located in the south pole. This observatory will look into previously unexplored areas for astronomy and into regions such as the PeV regions. It can potentially answer questions about the origin of high energy photons coming in from space as well as potentially answering questions about / detecting dark matter particles (such as cold dark matter particles) and neutrinos of higher energy than those produced in our particle accelerators (icecube.umd.edu).

LZ Next Generation Dark Matter Experiment

    The LZ dark matter experiment is a combination of the LUX and ZEPLIN experiments. This experiment is between 30 and 100 times more sensitive than LUX was before. LZ is specifically designed for detection of heavy WIMPS (Weakly Interacting Massive Particles), but it also has capabilities for detection of other candidates and processes of dark matter (sanfordlab.org). This project makes use of data analysis techniques with Python and ROOT.

Segmented Crystal Calorimeters for Future Colliders

    Crystal calorimeters are key in improving high resolution EM calorimetry, which is necessary for the detection of particles from collisions in the LHC. As we continue to push the bounds of energy ranges and  collider technology, we require more and more sensitive instruments to effectively gather data. Now, potential calorimeter designs are suggesting that we can optimize them for the detection of both photons/electrons and hadrons; previously it was simply one or the other.

CMS LHC Group

    This project is involved in the search for beyond the standard model physics (related to W and Z boson physics). There are various models which can be explored, including "dark top", charged Higgs, F-mesons, and folded SUSY. Each of these models seems to be looking to fit theoretical work with experimental data, in essence, figuring out which BSM theories can potentially work or can be ruled out. Furthermore, these analyses can also simply set limits on the various theories, after which they can be modified and reevaluated. The specific models that this project is involved with relate to the different modes of producing W and Z bosons in LHC.

Research Topic: Top Quark Properties at LHC

Working with Yuqin Wang under Dr. Kaustubh Agashe

Background, Project Overview

Standard Model Information

Simulations ~ Getting Started Walkthrough

I created a step by step guide with pictures and possible errors / confusions, which can be found at: https://colab.research.google.com/drive/1MXjalSo1rShW7rhKzb185JcClj0-xCoV?usp=sharing

Simulations ~ MADGRAPH

Personal Work

I spent some time to make my life easier by creating some Shell scripts to automate common tasks. So far these include running cmsenv inside the CMSSW directory, and then changing to the madgraph directory and running madgraph, as well as a workaround for an issue we are having with madgraph not detecting a web browser.

Startup of MADGRAPH, cmsenv:

Another script I made was for automation of the lhe2root.py process we used last semester:

The workaround I mentioned is for viewing the HTML output files for a given event. Since madgraph is not detecting firefox, I setup a script to compress the event's output files and another script on my local machine to scp and extract the compressed files into a directory in my desktop so that I can use my firefox to open them:

p p →z h Event

    This week we have been tasked with revisiting what we did with MADGRAPH last semester, that is, to recreate the p p > z h, z  > l+ l-, (h > z l+ l-, z > l+ l-) process using madgraph and then generate the correct plots.

Here are Feynman Diagrams for two processes which satisfy this process:

madanalysis is run independently of madgraph from within madgraph directory: /home/abeisaw1/CMSSW_8_0_22/src/MCProduction/MG5_aMC_v2_6_0

run madanalysis with: ./madanalysis5/bin/ma5

Once madanalysis is open:

    import /home/abeisaw1/CMSSW_8_0_22/src/MCProduction/MG5_aMC_v2_6_0/pp_ttbar2/Events/run_01/unweighted_events.lhe

    plot E(b) 250 0 300 [logX]

    plot E(b~) 250 0 300 [logX]

    submit

first number is # of bins

second number is xmin

third is xmax

I tried and replicated this error with xmin at 0 and 1

Yuqin also gets this error with same command sequence

Log Scale Plotting Error: Resolution

We met with Dr. Jabeen to discuss this problem and try to come up with some resolution.

Looking closely at the output of the non-log scale plots, which were generated by:

    import /path/to/the/lhe/file/lhefile.lhe

    plot E(b) 250 0 300

    plot E(b~) 250 0 300

    submit

We found that in the respective output directory there is a C file generated. From madgraph directory:

    cd ANALYSIS_#/Output/Histos/

In here we find selection_0.C which is responsible for creating the histogram from the datasets used in madanalysis. We find a line in the "Finalizing TCanvas" section:

    canvas->SetLogx(0);

Changing the 0 to a 1 overrides the logX command in madanalysis and applies it afterwards, resulting in the proper log scaled histogram. After this line we can save the resulting histo as we please in the "Saving the image" section.

Since this is all done outside madanalysis, we should manually run this C file using ROOT:

    root -l selections_0.C

Result:

This plot is not finalized as the x range must be extended a bit and we want to normalize to one to better see everything since there are many data sets being displayed together, but the log scaling is apparent nonetheless.

Thank you Dr. Jabeen for this helpful workaround

Another way to get around this issue was to install ROOT locally and copy the same selections C files from the cluster to my local machine. From there I was able to use the ROOT editor GUI to manually add in annotations on the plots, change the scales, etc.

For our final presentations, I decided to regenerate all datasets and plots to ensure everything was created the same and as intended. This was done using a script for madgraph and madanalysis:

generate p p > t t~ > w+ b w- b~

output research2021Final

launch research2021Final -n pp_630gev

       set nevents 100000

       set ebeam1 315

       set ebeam2 315

launch -n pp_1980gev

       set nevents 100000

       set ebeam1 990

       set ebeam2 990

launch -n pp_7tev

       set nevents 100000

       set ebeam1 3500

       set ebeam2 3500

launch -n pp_13tev

       set nevents 100000

       set ebeam1 6500

       set ebeam2 6500

launch -n pp_33tev

       set nevents 100000

       set ebeam1 16500

       set ebeam2 16500

launch -n pp_100tev

       set nevents 100000

       set ebeam1 50000

       set ebeam2 50000

launch -n ppbar_630gev

       set lpp2 -1

       set nevents 100000

       set ebeam1 315

       set ebeam2 315

launch  -n ppbar_1980gev

       set lpp2 -1

       set nevents 100000

       set ebeam1 990

       set ebeam2 990

launch -n ppbar_7tev

       set lpp2 -1

       set nevents 100000

       set ebeam1 3500

       set ebeam2 3500

launch -n ppbar_13tev

       set lpp2 -1

       set nevents 100000

       set ebeam1 6500

       set ebeam2 6500

launch -n ppbar_33tev

       set lpp2 -1

       set nevents 100000

       set ebeam1 16500

       set ebeam2 16500

launch -n ppbar_100tev

       set lpp2 -1

       set nevents 100000

       set ebeam1 50000

       set ebeam2 50000

launch -n *name* creates the next run with specified name

set lpp2 determines whether proton-proton (set lpp2 1) or proton-antiproton (set lpp2 -1). No change is needed for proton-proton.

set nevents tells madgraph how large of a dataset to create, in this case we used 100,000 events each time

set ebeam1, set ebeam2 determine the energy of each beam in the collider, summing to the total collider energy in GeV

The MadAnalysis script is:

First for unzipping our lhe.gz files:

gunzip -c ./research2021Final/Events/pp_630gev/unweighted_events.lhe.gz > ./research2021Final/Events/pp_630gev/unweighted_events.lhe

gunzip -c ./research2021Final/Events/pp_1980gev/unweighted_events.lhe.gz > ./research2021Final/Events/pp_1980gev/unweighted_events.lhe

gunzip -c ./research2021Final/Events/pp_7tev/unweighted_events.lhe.gz > ./research2021Final/Events/pp_7tev/unweighted_events.lhe

gunzip -c ./research2021Final/Events/pp_13tev/unweighted_events.lhe.gz > ./research2021Final/Events/pp_13tev/unweighted_events.lhe

gunzip -c ./research2021Final/Events/pp_33tev/unweighted_events.lhe.gz > ./research2021Final/Events/pp_33tev/unweighted_events.lhe

gunzip -c ./research2021Final/Events/pp_100tev/unweighted_events.lhe.gz > ./research2021Final/Events/pp_100tev/unweighted_events.lhe

gunzip -c ./research2021Final/Events/ppbar_630gev/unweighted_events.lhe.gz > ./research2021Final/Events/ppbar_630gev/unweighted_events.lhe

gunzip -c ./research2021Final/Events/ppbar_1980gev/unweighted_events.lhe.gz > ./research2021Final/Events/ppbar_1980gev/unweighted_events.lhe

gunzip -c ./research2021Final/Events/ppbar_7tev/unweighted_events.lhe.gz > ./research2021Final/Events/ppbar_7tev/unweighted_events.lhe

gunzip -c ./research2021Final/Events/ppbar_13tev/unweighted_events.lhe.gz > ./research2021Final/Events/ppbar_13tev/unweighted_events.lhe

gunzip -c ./research2021Final/Events/ppbar_33tev/unweighted_events.lhe.gz > ./research2021Final/Events/ppbar_33tev/unweighted_events.lhe

gunzip -c ./research2021Final/Events/ppbar_100tev/unweighted_events.lhe.gz > ./research2021Final/Events/ppbar_100tev/unweighted_events.lhe

Then once in MadAnalysis, we use:

For p p

import ~/CMSSW_8_0_22/src/MCProduction/MG5_aMC_v2_6_0/research2021Final/Events/pp_630gev/unweighted_events.lhe as _630GeV

import ~/CMSSW_8_0_22/src/MCProduction/MG5_aMC_v2_6_0/research2021Final/Events/pp_1980gev/unweighted_events.lhe as _1980GeV

import ~/CMSSW_8_0_22/src/MCProduction/MG5_aMC_v2_6_0/research2021Final/Events/pp_7tev/unweighted_events.lhe as _7TeV

import ~/CMSSW_8_0_22/src/MCProduction/MG5_aMC_v2_6_0/research2021Final/Events/pp_13tev/unweighted_events.lhe as _13TeV

import ~/CMSSW_8_0_22/src/MCProduction/MG5_aMC_v2_6_0/research2021Final/Events/pp_33tev/unweighted_events.lhe as _33TeV

import ~/CMSSW_8_0_22/src/MCProduction/MG5_aMC_v2_6_0/research2021Final/Events/pp_100tev/unweighted_events.lhe as _100TeV

plot E(b) 317 0 300 [stack normalize2one]

plot PT(b) 317 0 250 [stack normalize2one]

plot E(b~) 317 0 300 [stack normalize2one]

plot PT(b~) 317 0 250 [stack normalize2one]

submit bbar_Energy_PT_Distribution

For p p~

import ~/CMSSW_8_0_22/src/MCProduction/MG5_aMC_v2_6_0/research2021Final/Events/ppbar_630gev/unweighted_events.lhe as _630GeV

import ~/CMSSW_8_0_22/src/MCProduction/MG5_aMC_v2_6_0/research2021Final/Events/ppbar_1980gev/unweighted_events.lhe as _1980GeV

import ~/CMSSW_8_0_22/src/MCProduction/MG5_aMC_v2_6_0/research2021Final/Events/ppbar_7tev/unweighted_events.lhe as _7TeV

import ~/CMSSW_8_0_22/src/MCProduction/MG5_aMC_v2_6_0/research2021Final/Events/ppbar_13tev/unweighted_events.lhe as _13TeV

import ~/CMSSW_8_0_22/src/MCProduction/MG5_aMC_v2_6_0/research2021Final/Events/ppbar_33tev/unweighted_events.lhe as _33TeV

import ~/CMSSW_8_0_22/src/MCProduction/MG5_aMC_v2_6_0/research2021Final/Events/ppbar_100tev/unweighted_events.lhe as _100TeV

plot E(b) 317 0 300 [stack normalize2one]

plot PT(b) 317 0 250 [stack normalize2one]

plot E(b~) 317 0 300 [stack normalize2one]

plot PT(b~) 317 0 250 [stack normalize2one]

submit bbar_FromPPBar_Energy_PT_Distribution

And now we have generated our html report / selections .C files to use with ROOT