juliensimon/pdg-particle-properties
PDG Particle Properties Credit: NASA/JPL-Caltech/STScI/CXC/SAO Part of a dataset collection on Hugging Face. Dataset description Properties of every known particle from the Particle Data Group (PDG) -- THE definitive reference for particle physics, used by every particle physicist worldwide. The PDG, based at Lawrence Berkeley National Laboratory, has published its Review of Particle Physics since 1957 -- the single most cited publication in high-energy… See the full description on the dataset page: https://huggingface.co/datasets/juliensimon/pdg-particle-properties.
PDG Particle Properties
<div align="center"> <img src="banner.jpg" alt="Cassiopeia A supernova remnant in X-ray, optical, and infrared light" width="400"> <p><em>Credit: NASA/JPL-Caltech/STScI/CXC/SAO</em></p> </div>
Part of a [dataset collection](https://huggingface.co/collections/juliensimon/physics-datasets-69c2d4682d37dfdb77447bd7) on Hugging Face.
Dataset description
Properties of every known particle from the Particle Data Group (PDG) -- THE definitive reference for particle physics, used by every particle physicist worldwide.
The PDG, based at Lawrence Berkeley National Laboratory, has published its Review of Particle Physics since 1957 -- the single most cited publication in high-energy physics. The PDG compiles and critically evaluates measurements from thousands of experiments at facilities like CERN's Large Hadron Collider, Fermilab, KEK, and SLAC to produce world-average values for particle masses, widths, lifetimes, and quantum numbers.
The dataset covers the full spectrum of known particles: the six quarks, six leptons, and gauge bosons of the Standard Model; the Higgs boson discovered at CERN in 2012; hundreds of mesons (quark-antiquark bound states) and baryons (three-quark bound states) organized by flavor quantum numbers; and light nuclei.
This machine-readable version supports automated analysis pipelines in high-energy physics, Monte Carlo event generators, detector simulation frameworks, and educational tools. It is also valuable for machine learning applications in particle identification and anomaly detection in collider data.
This dataset is suitable for tabular classification, tabular regression tasks.
Schema
Quick stats
- 6,506 particles in the database
- 6,431 with measured mass
- 522 with measured width
- 88 self-conjugate particles
- Heaviest particle: Og294 (274,059 MeV)
Usage
from datasets import load_dataset
ds = load_dataset("juliensimon/pdg-particle-properties", split="train")
df = ds.to_pandas()from datasets import load_dataset
ds = load_dataset("juliensimon/pdg-particle-properties", split="train")
df = ds.to_pandas()
# All mesons (PDG ID 100-999)
mesons = df[(df["pdg_id"].abs() >= 100) & (df["pdg_id"].abs() < 1000)]
# Mass spectrum plot
import matplotlib.pyplot as plt
masses = df[df["mass_mev"].notna()]["mass_mev"]
plt.hist(masses[masses < 5000], bins=100, log=True)
plt.xlabel("Mass (MeV/c^2)")
plt.ylabel("Count")
plt.title("Particle Mass Spectrum")
plt.show()
# Stable particles (no measured width)
stable = df[df["width_mev"].isna() & df["mass_mev"].notna()]
print(f"{len(stable)} stable or long-lived particles")Data source
https://pdg.lbl.gov/
Related datasets
If you find this dataset useful, please consider giving it a like on Hugging Face. It helps others discover it.
About the author
Created by Julien Simon — AI Operating Partner at Fortino Capital. Part of the Space Datasets collection.
Citation
@dataset{pdg_particle_properties,
title = {PDG Particle Properties},
author = {Simon, Julien},
year = {2026},
url = {https://huggingface.co/datasets/juliensimon/pdg-particle-properties},
publisher = {Hugging Face},
note = {Derived from Particle Data Group, Lawrence Berkeley National Laboratory, https://pdg.lbl.gov/}
}