fwXmachina example: Jet energy resolution
Description
This dataset provides two ROOT files: - dijet_200_ntuple.root describes p p --> j j - hh4b_200_ntuple.root describes p p --> h h --> b b b b (dihiggs production in gluon-gluon fusion, with each higgs decaying to a bottom-antibottom pair) All events are generated with MadGraph5, decay and parton showering were performed with Pythia8, and detector simulation was performed with Delphes. Each ROOT file contains one TTree entitled "ntuple". In this ntuple, there are 100,000 events, organized into jagged arrays. Branches starting with "akt_" describe jets calculated by the particle flow anti-kt algorithm by Delphes. Branches starting with "sw_" describe jets calculated by a simplified sliding window algorithm. The primary data described here are transverse energy (ET) deposits in a 0.1x0.1 calorimeter, described in "rings", or "cones" around the center of the jets. For instance, "cone34" sums the ET of calorimeter cells between 0.3 < dR < 0.4 from the center of the jet. In addition, jet pT (as calculated by anti-kt), coordinates (eta/phi), and flavor are described. In the case of sliding window jets, if a sliding window jet falls dR < 0.3 from at least one anti-kt jet, it is noted as matched, and important values of the matching jet (such as anti-kt ET) are described as well. Further descriptions of this data, including a figure representing the "cones", a description of the sliding window algorithm, and an application of the dataset can be found in the paper: https://arxiv.org/abs/2507.16686 The full list of features is: - akt_cone{xx}{yyy} where xx can be (01, 04, 12, 23, 24, 34, 46) [for instance, 04 is from the center of the jet to R=0.4 away, and 23 is the ring from 0.2-->0.3 from the center of the jet] and yyy can be (Ehad, Eem, and Energy) for hadronic, electromagnetic, and total energy respectively. - akt_energy for the ET calculated with anti-kt - akt_eta - akt_phi - akt_flavor based on the flavor of the generator-level particle with the highest PDG value (for instance bottom=5, strange=3). - akt_isTruthMatched which is True if the akt jet is > 0.4 from a generator jet - akt_isPileup which is True if akt_flavor == 0 - akt_rho which is the local energy density calculated by anti-kt - akt_truthRecoDeltaR which is the distance to the nearest generator-level jet - sw_cone{xx}{yyy} using the same xx and yyy as above, with the jet seeded by the sliding window algorithm - sw_aktEnergy is the anti-kt ET of the nearest anti-kt jet (or 0.0 if R > 0.4 away) - sw_aktRecoDeltaR is the distance to the nearest anti-kt jet - sw_eta - sw_phi - sw_truthEnergy is the ET of the nearest generator-level jet - sw_truthRecoDeltaR is the distance to the nearest generator-level jet For those interested in evaluating this dataset, we recommend using uproot in Python to read the events into an awkward array. This array would be of length 100,000 (number of events) and composed of variable-length arrays (depending on the jets in each event).
Files
Institutions
- Westmont College
- University of Pittsburgh
- Saint Louis University