# PRBench Task Configuration
task_id: task_li_2014

# --- Paper Info ---
paper:
  title: "Classical-Quantum Correspondence for Above-Threshold Ionization"
  author: "Min Li, Ji-Wei Geng, Hong Liu, Yongkai Deng, Chengyin Wu, Liang-You Peng, Qihuang Gong, Yunquan Liu"
  doi: "10.1103/PhysRevLett.112.113002"
  year: 2014
  paper_file: "qtmc.md"

# --- Files ---
instruction_file: "instruction.md"
metadata_file: "metadata.md"
ground_truth_data_dir: "data"
ground_truth_code_dir: "reproduce"

# --- Expected Outputs ---
expected_outputs:
  analysis:
    - "reproduction/ANALYSIS.md"
  code:
    - "reproduction/qtmc_simulation.py"
  data:
    - "data/fig1d.csv"

# --- Docker Environment ---
docker:
  image: "python:3.11-slim"
  memory_limit: "4g"
  timeout: 7200
  pip_install:
    - numpy
    - scipy
    - matplotlib

# --- Grading Rubric ---
grading:
  dimensions:
    - name: "methodology_understanding"
      weight: 0.20
      description: >
        Does the agent's ANALYSIS.md correctly describe the QTMC methodology:
        ADK tunneling theory for initial conditions, Newton's equations with
        Coulomb potential for classical propagation, classical action / phase
        accumulation for quantum interference, and Kepler transformation for
        asymptotic momenta? Compare against metadata.md.
    - name: "code_correctness"
      weight: 0.15
      description: >
        Are all required code files present? Does the code correctly implement:
        ADK tunneling rate and initial condition sampling, trapezoidal laser
        pulse envelope, Newton's equations with Coulomb + laser field,
        phase/action integration along trajectories, Kepler transformation
        for asymptotic momentum, and complex amplitude accumulation on
        momentum grid?
    - name: "data_accuracy"
      weight: 0.50
      description: >
        How well does the generated CSV data match the ground truth?
        Due to Monte Carlo stochasticity, use statistical comparison:
        correlation coefficient, KL divergence, structural feature matching
        (ATI rings, Coulomb focusing, interference carpet). Normalize both
        distributions before comparison. Allow for statistical fluctuations
        inherent to Monte Carlo sampling.
    - name: "completeness"
      weight: 0.15
      description: >
        Are the CSV data file, code files, and ANALYSIS.md present?
        Are the files non-trivial (not empty/placeholder)?

# --- Code-Only Evaluation Mode ---
code_only:
  instruction_suffix: |
    ## IMPORTANT: Code-Only Mode

    In this task you are in **code-only mode**. You must:
    1. Read and analyze the paper
    2. Write `reproduction/ANALYSIS.md` with your methodology analysis and all formulas
    3. Implement ALL code files under `reproduction/`
    4. **DO NOT** run the simulation or generate CSV data files
    5. Focus on writing correct, complete, and well-structured code

    Your deliverables are ONLY the analysis document and code files.
    Do NOT produce any CSV data files.

  expected_outputs:
    analysis:
      - "reproduction/ANALYSIS.md"
    code:
      - "reproduction/qtmc_simulation.py"

  grading:
    dimensions:
      - name: "methodology_understanding"
        weight: 0.25
        description: >
          Does ANALYSIS.md correctly describe the QTMC methodology including
          ADK theory, classical trajectory propagation with Coulomb potential,
          phase accumulation, and interference? Compare against metadata.md.
      - name: "code_correctness"
        weight: 0.50
        description: >
          Read every code file carefully. Check physical and algorithmic correctness:
          - ADK tunneling rate formula and initial condition sampling
          - Trapezoidal laser pulse envelope and electric field
          - Newton's equations of motion with Coulomb + laser field
          - Classical action integration (kinetic, Coulomb, Ip terms)
          - Kepler transformation for asymptotic momentum
          - Complex amplitude accumulation and momentum grid binning
          - Correct parameter values (E1=0.0534, w1=0.057, Ip=0.4456, Z=1)
          Compare against ground truth formulas in metadata.md.
      - name: "code_quality"
        weight: 0.10
        description: >
          Is the code well-structured, readable, and does it follow
          reasonable modular organization?
      - name: "completeness"
        weight: 0.15
        description: >
          Are all code files and the ANALYSIS.md present and non-trivial?