BeautyInStats

An explainable agentic workflow as analysis copilot for LHCb OpenData and dissemination of best-practice methods for the apt evaluation of systematic uncertainties.

Development Guide

Setup

1. Prerequisites

python3 -m pip install --user pipx
python3 -m pipx ensurepath
source ~/.bashrc # or any other default shell

2. Install Poetry (package manager):

pipx install poetry

3. Clone the repository and install dependencies

git clone https://github.com/reallyblaised/beauty-in-stats.git
cd beauty-in-stats
poetry install

4. Verify the correct installation of the Poetry environment

poetry env list
>>> beautyinstats-U3Bi8mYg-py3.10 (Activated)

5. Load the Poetry shell to load the library environment

poetry self add poetry-plugin-shell
poetry shell

Build the LHCb paper corpus

# Get all papers
build-corpus

# Get specific number of papers
build-corpus --max-papers 10

# Get papers from date range
build-corpus --start-date 2020-01-01 --end-date 2023-12-31

# Show additional logging
build-corpus --verbose

Downloaded files are organized in the data/ directory:

data/pdfs/: PDF versions of papers
data/source/: LaTeX source files
data/expanded_tex/: Expanded LaTeX files
data/abstracts/: Paper abstracts

Dependencies

• Python ≥ 3.9
• latexpand (for processing LaTeX sources)