Mixed data mis-modeling

Problem

v4j mass / pt and eta mismodeled in mixed data

  • not there in 3b
  • Seen in pz and non-pz mixing strategies
  • Indpendentent of ttbar subtraction.

Logs

20 May 2026 Wednesday

  • Making the correlation plots with the 3b data
    • These dont include the PS-tagged jets
  • Using cand jets instead of tagged jets (Will fix my 3b issue)
  • Adding 2d correlation plots for all jets and other jets

  • Running 4b and 3b noFvT 

    ./run_container snakemake --snakefile coffea4bees/workflows/Snakefile_Run3.smk output/Run3_quadjet_run2/plots_wJCM/plots_done.txt --config mode=quadjet_run2 datasets=[data] compute_hemi_mixing_diagnostics=true --allowed-rules create_histogram_config_wJCM make_histograms_wJCM merge_histograms_wJCM make_plots_wJCM --forceall  --logger snkmt --cores 4
  • Running the mixed data no MvD
./run_container snakemake --snakefile coffea4bees/workflows/Snakefile_Run3MvD.smk output/Run3_MvD_quadjet_run2_rank0/histAll_Run3MvD_quadjet_run2.coffea --config mode=quadjet_run2 dataset_name=mixeddata_all_rank0 compute_hemi_mixing_diagnostics=true --allowed-rules create_friends_wSvB create_histogram_config_wSvB make_histograms_mixeddata merge_histograms --forceall --cores 4
  • Rerunning.. node died
  • OK making lot of intereisting plots …
  • Adding plots for the inclusive selections… Seeing a lot of artifact form SB|SR
  • Rerunning 3b and 4b with the inclusive selection

19 May 2026 Tuesday

  • checking output from 15 May 2026 Friday

  • Runing on mixed-correlations 

    ./run_container snakemake --snakefile coffea4bees/workflows/Snakefile_Run3MvD.smk output/Run3_MvD_quadjet_run2_rank0/plots_MvD/plots_done.txt --config mode=quadjet_run2 dataset_name=mixeddata_all_rank0 compute_hemi_mixing_diagnostics=true --allowed-rules create_friends_MvD create_histogram_config_MvD make_histograms_data_MvD make_histograms_mixeddata_MvD merge_histograms_MvD make_plots_MvD --forceall -n -p
  • Finished… Just realized that this is with the MvD weights… I want without

  • Running without the MvD weights 

    ./run_container snakemake --snakefile coffea4bees/workflows/Snakefile_Run3MvD.smk output/Run3_MvD_quadjet_run2_rank0/histAll_Run3MvD_quadjet_run2.coffea --config mode=quadjet_run2 dataset_name=mixeddata_all_rank0 compute_hemi_mixing_diagnostics=true --allowed-rules create_friends_wSvB create_histogram_config_wSvB make_histograms_mixeddata merge_histograms --forcea
ll --logger snkmt --cores 4
  • Finished… now running to get 3b with no FvT
./run_container snakemake --snakefile coffea4bees/workflows/Snakefile_Run3.smk output/Run3_quadjet_run2/plots_wJCM/plots_done.txt --config mode=quadjet_run
2 datasets=[data] compute_hemi_mixing_diagnostics=true --allowed-rules create_histogram_config_wJCM make_histograms_wJCM merge_histograms_wJCM make_plots_wJCM --forceall --logger snkmt --cores 4

15 May 2026 Friday

  • Running with diagonstic histograms
  • submitted on cmslpc367
  • Running data… still need to run mixed

13 May 2026 Wednesday

  • Making a Mixing Toy
  • Got it going.
  • Looks like I can reproduce the "problem" seen in data.
  • The mixing breaks the correlation when the matching is not perfect

12 May 2026 Tuesday

  • [ ] Not enough Priority given to b-cand matching ?
  • Claude session re:variance -> bias in joint hemisphere variables
    • [ ] Compute some covariances

Backlinks