Response and Resolution Studies for the Trigger Level Analysis at the ATLAS Experiment

Abstract

When particles collide at the Large Hadron Collider, the invariant mass of new particles can be derived by detecting their decay products. The Trigger Level Analysis at the ATLAS experiment uses partly reconstructed particle collision events for its analysis. The smaller data size allows the Trigger Level Analysis to obtain a much larger number of events and a smaller statistical uncertainty. This notebook presents the data processing and analysis used to plot the invariant mass response. The response is the ratio of invariant mass as determined by particles simulated to interact with the detector, divided by the invariant mass as determined by the same particles, without interaction with the detector. Doing this after each step of the Trigger Level Analysis calibration allows us to see if any calibration step introduces signal-like features in the analysis. The insights gained from the response studies are also used to estimate the invariant mass resolution of the analysis and with that derive an optimal histogram binning.

Running on Binder (Recommended for Windows users):

1. Klick here:

• The build takes a long time ~10 min
• If it gets stuck at "Successfully pushed XXXXX", refresh the page and it should find an already built image and launch within a minute

2. When the Binder server is running, before opening the jupyter notebook, go to the drop-down menu on the right called "New". Click it and select "Terminal". A new browser tab will open with a command line.
3. Type 7z x data.zip into the terminal (Copy-paste will not work). When prompted, enter the password which I have emailed to you. This will extract and decrypt the data for the notebook. If you have trouble with this, email me at [email protected]
4. Close the terminal browser tab.
5. Go back to the initial tab, open the jupyter notebook responseStudies.ipynb, and follow the instructions there.

• In the notebook, you can run each cell of code/text one by one by using shift+enter. This will give you a chance to read the text and understand the code.
• To run everything from top to bottom in one go, press ⏩ in the top toolbar to restart and run the whole notebook

Running Locally (Linux):

To run this notebook locally you will need Conda or miniconda, as well as p7zip.

• To install miniconda follow the instructions found here: https://docs.conda.io/en/latest/miniconda.html
• To install p7zip do: sudo apt-get install p7zip-full

1. Clone this repository
   git clone https://github.com/teokem/responseStudies-alex-exook.git
2. Enter the newly cloned directory
   cd responseStudies-alex-exook
3. Download the encrypted data
   wget --load-cookies /tmp/cookies.txt "https://docs.google.com/uc?export=download&confirm=$(wget --quiet --save-cookies /tmp/cookies.txt --keep-session-cookies --no-check-certificate 'https://docs.google.com/uc?export=download&id=19Dv4JkJEqi-tzT33wjAvtufNqcwi8Lr0' -O- | sed -rn 's/.*confirm=([0-9A-Za-z_]+).*/\1\n/p')&id=19Dv4JkJEqi-tzT33wjAvtufNqcwi8Lr0" -O data.zip && rm -rf /tmp/cookies.txt
4. Decrypt the encrypted data
   7z x data.zip
5. Create the python environment using conda. This will take ~60 min due to the ROOT package
   conda env create -f environment.yml
6. Activate the newly created environment
   conda activate responseStudies
7. Open the notebook responseStudies.ipynb, and follow the instructions there
   jupyter notebook

• In the notebook, you can run each cell of code/text one by one by using shift+enter. This will give you a chance to read the text and understand the code. To run everything from top to bottom in one go, press ⏩ in the top toolbar to restart and run the whole notebook

Environment Packages

• python=3.8 is required to interprate the python code in this notebook
• notebook=6.2 is required to run the notebook
• root=6.20.00 is required to read and fit the data with the ROOT framework tools developed at CERN
• numpy=1.20 is used for the numpy array data structure
• pandas=1.2 is used for the data frame data structure, which is crucial for keeping the different data series, and slices of data series organized
• scipy=1.6 is used for curve fitting
• matplotlib=3.3 is used for plotting
• bokeh=2.3 is used for interactive plotting
• mplhep=0.2.18 is required for the ATLAS experiment specific plot styling