michaelmackenzie/mu2eii_snowmass

Analysis tools for the Mu2e-II Snowmass 2021 sensitivity group

Mu2e-II Snowmass 2021 sensitivity tools

Analysis tools for the Mu2e-II Snowmass 2021 sensitivity group

Building

source /cvmfs/mu2e.opensciencegrid.org/setupmu2e-art.sh
cd /mu2e/app/users/${USER}/
mkdir mu2eii
cd mu2eii
git clone --single-branch --branch Mu2eII_SM21 [email protected]:Mu2e/Offline.git
git clone [email protected]:Mu2e/TrkAna.git
git clone [email protected]:michaelmackenzie/Mu2eII_Snowmass.git mu2eii
# Some Stntuple classes currently have issues due to Offline advancing, use a version patched for now
git clone /mu2e/app/users/mmackenz/muse/Stntuple/ Stntuple
./mu2eii/scripts/build_config_muse
muse setup
muse build -j12

Histogramming TrkAna trees

Histogramming is implemented using an object that takes in a TrkAna tree and writes out sets of histograms of the event/track parameters with different event/track selections.

root.exe -q -b mu2eii/ana/scripts/process_trkana.C

Optimizing the signal window

Currently this runs using the SU2020 histograms with the code written by Pasha and imported from Mu2e/su2020, with the code imported in the stat package. DIO and CE histograms can be partially imported from the Mu2e-II TrkAna histogramming by using a Mode bit of 100.

This evaluates the mean discovery R for scans in the track p vs t space. The histogram directory is taken from the .rootrc file, where these should be migrated from /mu2e/data/projects/su2020/hist/ to /mu2e/data/projects/mu2eii_snowmass/hist/. This path can be updated to a new histogram path if needed.

root.exe mu2eii/scripts/mumem_sensitivity.C
root> mumem ana(13); //See stat/constants.cc to find the Mode options
root> ana.scan_pmin(104.1, 104.9, 680, 1700, 6);
root> ana.scan_tmin(104.0, 104.9, 750, 1700, 10);
root> mumem ana_2(113); //Use the Mu2e-II DIO/CE histograms
root> ana_2.scan_pmin(104.5, 105., 680, 1650, 15);

Evaluating the median expected limit and discovery

This uses the calc package, similarly imported from Mu2e/su2020.

This has been updated to run in a compiled mode, and a new version has been implemented to create the Poisson model from a datacard that lists the sources of signals and backgrounds, the mean expected conttributions of each, and the systematic uncertainties on each, where each uncertainty is either 100% or 0% correlated with the other sources.

Original script based model implementation:

root.exe -q -b mu2eii/scripts/Mu2eII_model.C

Datacard based model implementation (using the datacard mu2eii/scripts/datacard.txt with estimates using SU2020 histograms and Mu2e-II parameters):

root.exe -q -b mu2eii/scripts/model_builder.C