HH → bb̄WW

This is the documentation for the HH→bb̄WW analysis — the search for Higgs-boson pair production in the bb̄WW final state, built on the FLAF framework.

The common workflow lives in the FLAF docs

HH→bb̄WW runs on FLAF, so installation, the task pipeline (NanoAOD → anaTuples → histograms → plots), the configuration system, storage, eras and CI are shared with every FLAF analysis and are documented once, in the FLAF documentation:

• Prerequisites & installation
• Your first run
• Full workflow walkthrough
• Command arguments

This site covers only what is specific to HH→bb̄WW.

What this analysis adds on top of FLAF

Ingredient	Purpose
DeepHME	Neural mass reconstruction of the HH system (the bb̄WW counterpart to SVfit).
B-tag-shape cache	Per-event b-tag shape weights, pre-computed via AnalysisCacheTask.
Resolved & boosted categories	Two topologies analysed (low- and high-p_T HH).
StatInference	Datacards and limits (x_hh_bbww_run3.yaml).

Setup is in Setup; analysis-specific run notes in Running the analysis; the statistics step in Statistical inference.

Quickstart

git clone --recursive git@github.com:cms-flaf/HH_bbWW.git
cd HH_bbWW
git lfs pull                                   # HH_bbWW ships large files via git LFS
source env.sh                                  # first time builds the environment
voms-proxy-init -voms cms -rfc -valid 192:00
law index --verbose

Run git lfs pull after cloning

HH→bb̄WW stores some large files with git LFS. If git lfs pull is skipped, those files are placeholder pointers and the analysis will fail in confusing ways.

Then smoke-test the chain (see FLAF → first run):

law run FLAF.Analysis.tasks.HistPlotTask \
  --version my_first_run --period Run3_2022 --workflow local --test 1000

New to FLAF? Read Key terms and Concepts first.

Eras

HH→bb̄WW runs over the Run 3 eras Run3_2022, Run3_2022EE, Run3_2023 and Run3_2023BPix. See FLAF → Eras.