Data and code for "A Prototype Differential Atom Interferometer for Fundamental Physics"

Data and code for “A Prototype Differential Atom Interferometer for Fundamental Physics” 10.5281/zenodo.19592552 Data and plots curated by Charles Baynham [e-mail] - for authorship contributions see the paper. Reproducible figure-generation code for the Differential Atom Interferometry (DAI) paper from the AION experiment at Imperial College London. This repository contains the data and figure scripts for arXiv:2504.09158. Figures Notebook Description Figure 1.ipynb Black-hole merger sensitivity projections Figure 4.ipynb Composite figure: Allan deviation, signal recovery, and noise analysis Figure 5a.ipynb Maximum-likelihood frequency scan results Figure 5b.ipynb Amplitude histogram analysis across signal strengths Repository Structure ├── pyproject.toml # Python dependencies ├── uv.lock # Pinned dependency lockfile ├── utilities/ # Stripped-down analysis utilities │ ├── plots.py # Plotting helpers and styles │ ├── adevtools.py # Allan deviation tools │ ├── datasets/paths.py # Data path resolution (local paths) │ └── analysis/ │ └── signal_extraction.py ├── data/ # All data bundled (no network access needed) │ ├── intermediate_data/ # Pre-processed experimental data (CSVs, NPYs) │ ├── precomputed_true_signals/ # Pre-computed signal extraction results (NPZs) │ ├── 2026-01-20-DAI-Analysis/ # Monte Carlo and timeseries data │ ├── 2026-02-24-Amplitude-DAI-MC/ # Amplitude Monte Carlo results │ └── 2025-02-12-black-holes-plot/ # Black hole merger sensitivity data ├── figures/ # Notebooks and generated outputs │ ├── Figure 1.ipynb │ ├── Figure 4.ipynb │ ├── Figure 5a.ipynb │ └── Figure 5b.ipynb └── icl_experiments/ # Trimmed ARTIQ experiment repository ├── repository/ # Experiment code (DifferentialClockInterferometryWithNoiseAndSignalFrag) ├── device_db_config/ # Device database configuration ├── hdf5_data/ # Raw ARTIQ HDF5 datasets └── pyproject.toml # Experiment dependencies Running the Notebooks Dependencies are managed with uv. The uv.lock file pins every transitive dependency to exact versions. Download this repository and extract the files. Open a terminal in the extracted directory. Install uv if you don’t have it: macOS / Linux: curl -LsSf https://astral.sh/uv/install.sh | sh Windows (PowerShell): powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex" For more options, see the UV installation guide. Create a virtual environment and install all dependencies: uv sync Open any notebook in figures/: uv run jupyter notebook To update or add dependencies, edit pyproject.toml and run uv lock to regenerate the lockfile. All data is bundled in data/ — no network access or external data stores are required. Data Provenance Experimental data was collected using the ARTIQ control system on the AION strontium cold-atom interferometer at Imperial College London. Raw datasets are stored on the ICL Research Data Store (RDS) and were pre-processed into the CSV/NPY/NPZ files included here. intermediate_data/: Excitation fraction timeseries extracted from ARTIQ HDF5 datasets, plus MLE fitting results. The CSV filenames encode the 5-digit run IDs (RIDs) of the source HDF5 files. precomputed_true_signals/: Frequency and phase extraction results computed via NDScanDataset.from_rid() and extract_frequency_and_phase() — bundled as NPZ files to remove the ARTIQ/RDS dependency 2026-01-20-DAI-Analysis/: Monte Carlo Allan deviation results and residual timeseries 2026-02-24-Amplitude-DAI-MC/: Monte Carlo simulation results for three signal amplitude levels 2025-02-12-black-holes-plot/: Gravitational wave sensitivity curves and black hole merger population data icl_experiments/: Trimmed ARTIQ experiment repository retaining only the code needed to run DifferentialClockInterferometryWithNoiseAndSignalFrag. Raw HDF5 datasets are stored in icl_experiments/hdf5_data/.