Event Generation

MadLAD turns a pre‑built container and a single YAML configuration file into a full event‑generation workflow. Below is a step‑by‑step guide, the configuration syntax and an example that produces all‑hadronic $t\bar t$ events.

---

1. Launching a Run

python -m madlad.generate --config-name=ttbar-allhad.yaml [options]

• --config-name – basename of the YAML file stored in processes folder without the .yaml extension.
• [options] – optional Hydra overrides, e.g. run.auto-launch=True.

---

2. Generation Configuration

A config is split into three logical blocks:

Block	Purpose
run	How MadLAD should start the container (image, path to MG5, etc.).
gen	All MadGraph‑specific options (run card, madspin, shower etc.).
post	Bash commands executed after the event generation (e.g. histogramming).

Below is a fully commented example (processes/ttbar-allhad.yaml).

# ────────────────────── run ───────────────────────
run:
  image:       "madlad-custom"            # Docker image to use
  mg5:         "/app/MG5_aMC/bin/mg5_aMC" # Path to MG5 inside the container
  launch-from: null                       # Optional: launch an existing madgraph process
  auto-launch: False                      # Whether generation should start automatically
  no-shower:   False                      # Skip parton showering

# ────────────────────── gen ───────────────────────
gen:
  # --- Model and process definition ---------------------------------
  block_model:
    model:          loop_sm-no_b_mass
    multiparticle:  ["p = g u c d s u~ c~ d~ s~ b b~",
                     "j = g u c d s u~ c~ d~ s~ b b~"]
    proc:           "p p > t t~ [QCD]"
    save_dir:       "Output"

  # --- Run card options ---------------------------------------------
  block_run:
    nevents:        1000
    nevt_job:       1000
    iseed:          0
    beamenergy:     6500
    pdlabel:        "lhapdf"
    lhaid:          260000
    parton_shower:  "PYTHIA8"
    maxjetflavor:   5
    reweight_scale: ".true."
    reweight_PDF:   ".true."
    PDF_set_min:    260001
    PDF_set_max:    260100
    muR_over_ref:   1.0
    muF1_over_ref:  1.0
    muF2_over_ref:  1.0
    dynamical_scale_choice: 10
    jetalgo:        -1          # Anti‑kT by default
    jetradius:      0.4
    ptj:            0.1
    req_acc:        0.001

  # --- MadSpin decay card -------------------------------------------
  block_madspin:
    multiparticle:  ["p = g u c d s u~ c~ d~ s~ b b~",
                     "j = g u c d s u~ c~ d~ s~ b b~"]
    decays:         ["t > w+ b, w+ > j j",
                     "t~ > w- b~, w- > j j"]
    bwcut:          15
    max_weight_ps_point: 500
    Nevents_for_max_weight: 500

  # --- Parameter card -------------------------------------------------
  block_param:
    param: "aMcAtNlo_param_card_loop_sm-no_b_mass.dat"

  # --- Custom scale definitions --------------------------------------
  block_sf:
    custom_scales: [
      "c         Q^2= mt^2 + 0.5*(pt^2+ptbar^2)",
      "          xm2=dot(pp(0,3),pp(0,3))",
      "          tmp=sqrt(xm2+0.5*(pt(pp(0,3))**2+pt(pp(0,4))**2))",
      "          temp_scale_id='mt**2 + 0.5*(pt**2+ptbar**2)'",
      ""
    ]

  # --- Delphes run card ----------------------------------------------
  block_delphes:
    delphes_card: "ATLAS_sim_FASTJET.tcl"

# ────────────────────── post ───────────────────────
post:
  - "echo 'Finished'"