Project overview

This repository contains code and notebooks for numerical/simulation work on a two-multilevel-atom quantum Rabi–type model and QFI (quantum Fisher information) heatmap generation/processing, used in the paper 'Equilibrium thermometry in the multilevel quantum Rabi model' by Tabitha Doicin, Luis A. Correa, Jonas Glatthard, Andrew D. Armour, Gerardo Adesso.

The repo is organised as three largely independent folders, each with its own requirements.txt (kept folder-specific on purpose).

Folder layout

• Optimality_and_AA/

  • Jupyter notebooks for the “Optimality and AA” calculations/figures.
  • simulation_2MQRM.py: QuTiP-based simulation utilities/classes used by the notebooks.
  • requirements.txt: dependencies for this folder’s notebooks/scripts.

• QFIHeatmapGenerator/

  • HeatmapGenerator.py: script that generates QFI vs temperature data (CSV + pickle outputs).
  • requirements.txt: dependencies for running the generator.

• QFIHeatmapProcessing/

  • processing.ipynb: notebook(s) for post-processing/plotting generated data.
  • simulation_2MQRM.py: QuTiP-based utilities used by the processing notebook(s).
  • requirements.txt: dependencies for this folder’s notebooks/scripts.

Installation

Because requirements are per-folder, install from inside the folder you want to run.

Example (recommended: a fresh venv per folder):

cd QFIHeatmapGenerator
python -m venv .venv
source .venv/bin/activate  # Windows: .venv\Scripts\activate
pip install -r requirements.txt

Repeat the same pattern for Optimality_and_AA/ or QFIHeatmapProcessing/.

Note about mpc / mpfr lines

Two of the requirements.txt files include mpc=... and mpfr=..., which are commonly conda package names rather than pip package names. If pip install -r requirements.txt fails on those lines, the simplest fix is to use conda/mamba for that environment or remove those two lines for a pip-only environment.

Usage

Generate QFI heatmap data

From QFIHeatmapGenerator/:

python HeatmapGenerator.py

Outputs are written to the working directory (by default):

• qfidataframe.csv
• avgqfi.pkl
• darkbrightqfi.pkl
• brightbrightqfi.pkl
• tlist.pkl

Key run parameters (defined near the top of HeatmapGenerator.py) include:

• Dg, De (ground/excited degeneracies)
• wc, wa (frequencies)
• ep1, ep2 (detuning noise scales)
• minT, maxT, numT (temperature grid)
• totallines, totalsets, workers (parallel workload)
• theta (cutoff)
• Individuallynormalised (normalisation choice)

Process data

Open the notebook and generated files in QFIHeatmapProcessing/processing.ipynb (Jupyter, VS Code, etc.) after installing that folder’s requirements.

Run the simulation utilities

Both Optimality_and_AA/simulation_2MQRM.py and QFIHeatmapProcessing/simulation_2MQRM.py are intended to be imported by notebooks/scripts, e.g.:

from simulation_2MQRM import DoubleMultilevel

Reproducibility notes

• Several workflows rely on randomness (numpy.random, Python random). If you need deterministic results, set seeds at the start of your script/notebook.
• Parallel generation uses ProcessPoolExecutor, so each worker has its own RNG state unless you explicitly manage seeding.

Citation / license

(code and project files cleaned up with ChatGPT for enhanced readability)