Hugging Face's logo

Datasets:
juliensimon
/
mcgill-magnetar-catalog

Tasks:
Tabular Classification
Tabular Regression
Modalities:
Tabular
Text
Formats:
parquet
Languages:
English
Size:
< 1K
Tags:
space
magnetars
neutron-stars
x-ray
astronomy
open-data
Libraries:
Datasets
pandas
Polars
License:
Dataset card Data Studio Files
xet
 Community
mcgill-magnetar-catalog / README.md
juliensimon's picture
juliensimon
Update McGill magnetar catalog: 31 magnetars
760258f verified
preview code
|
Raw
History Blame Contribute Delete
10.9 kB
metadata 
license: cc-by-4.0
pretty_name: McGill Online Magnetar Catalog
language:
  - en
description: >-
  All known magnetars — neutron stars with extreme magnetic fields (10^13-10^15
  G) — from the McGill Online Magnetar Catalog. Includes spin parameters,
  magnetic field strengths, X-ray properties, and as
task_categories:
  - tabular-classification
  - tabular-regression
tags:
  - space
  - magnetars
  - neutron-stars
  - x-ray
  - astronomy
  - open-data
  - tabular-data
  - parquet
size_categories:
  - n<1K
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/mcgill_magnetar_catalog.parquet
    default: true

McGill Online Magnetar Catalog

Illustration of different types of neutron stars

Credit: NASA/JPL-Caltech

Part of a dataset collection on Hugging Face.

Dataset description

All known magnetars — neutron stars with extreme magnetic fields (10^13-10^15 G) — from the McGill Online Magnetar Catalog. Includes spin parameters, magnetic field strengths, X-ray properties, and associations.

Magnetars are isolated neutron stars powered by the decay of their ultra-strong magnetic fields, rather than by rotation (like normal pulsars) or accretion. They manifest as Soft Gamma Repeaters (SGRs) and Anomalous X-ray Pulsars (AXPs), producing dramatic bursts and flares in X-rays and gamma-rays.

Magnetar magnetic fields — reaching 10^14 to 10^15 Gauss, a thousand times stronger than ordinary pulsars — are the strongest known in the universe. These fields exceed the quantum electrodynamic critical field at which the vacuum itself becomes birefringent. The decay of these colossal fields powers persistent X-ray emission at luminosities of 10^33-36 erg/s, far exceeding what rotational energy alone can supply. During outbursts, magnetars can release up to 10^46 erg in giant flares, rivaling the luminosity of the entire Galaxy.

The magnetar population bridges several areas of astrophysics. Their connection to fast radio bursts (FRBs) was dramatically confirmed in 2020 when SGR 1935+2154 emitted a millisecond radio burst bright enough to be detected at extragalactic distances. Magnetars are also candidate central engines for some gamma-ray bursts and super-luminous supernovae.

Schema

Column Type Description Sample Null %
name object Magnetar designation (e.g. 'SGR 1806-20', '1E 2259+586'); SGR = Soft Gamma Repeater, AXP = Anomalous X-ray Pulsar; both classes are now understood to be magnetars CXOU J010043.1-721134 0.0%
period_s float64 Spin period in seconds; magnetars: 2-12 s (far slower than recycled millisecond pulsars); null for sources where timing has not been achieved 8.020392 16.1%
period_err_s float64 1-sigma uncertainty on spin period (s) 9e-06 16.1%
period_derivative float64 Spin-down rate dP/dt in s/s; magnetars: ~10^-11 s/s, among the fastest-spinning-down neutron stars; drives inferred magnetic field and characteristic age 1.88e-11 19.4%
period_derivative_err float64 1-sigma uncertainty on period derivative (s/s) 8e-13 35.5%
magnetic_field_g float64 Dipole surface magnetic field strength in Gauss, inferred as B ~ 3.2e19 * sqrt(P * Pdot); magnetars: 10^14-10^15 G, roughly 1000x stronger than normal pulsars; null if period or period derivative is unmeasured 393000000000000.0 19.4%
spin_down_luminosity_erg_s float64 Rotational energy loss rate Edot = -4pi^2I*Pdot/P^3 in erg/s; for magnetars typically 10^32-10^34 erg/s, lower than their observed X-ray luminosity (evidence for magnetic field powering) 1.44e+33 19.4%
characteristic_age_yr float64 Characteristic spin-down age tau = P/(2*Pdot) in years; magnetars: ~10^3-10^4 yr (very young neutron stars); this is an upper limit on true age for initially fast rotators 6760.0 19.4%
column_density_cm2 float64 Interstellar hydrogen column density N_H in cm^-2, fit from soft X-ray absorption; used to estimate visual extinction and constrain distance; null if no X-ray spectrum available 6.3e+20 29.0%
column_density_err_up float64 Upper 1-sigma uncertainty on column density (cm^-2) 2e+20 29.0%
column_density_err_down float64 Lower 1-sigma uncertainty on column density (cm^-2) 1.6e+20 29.0%
photon_index float64 Photon index Gamma of the hard X-ray power-law spectral component (flux proportional to E^-Gamma); magnetars: Gamma ~ 2-4 in quiescence; null if power-law component not required by the spectrum 3.88 45.2%
photon_index_err_up float64 Upper 1-sigma uncertainty on photon index 0.01 45.2%
photon_index_err_down float64 Lower 1-sigma uncertainty on photon index 0.01 45.2%
blackbody_kt_kev float64 Temperature kT in keV of the soft X-ray blackbody spectral component; magnetars: kT ~ 0.3-0.7 keV; null if spectrum not well-fitted by a blackbody 0.3 45.2%
blackbody_kt_err_up float64 Upper 1-sigma uncertainty on blackbody kT (keV) 0.02 45.2%
blackbody_kt_err_down float64 Lower 1-sigma uncertainty on blackbody kT (keV) 0.02 45.2%
xray_flux_erg_cm2_s float64 Unabsorbed 2-10 keV X-ray flux in erg/cm^2/s from quiescent-state observations; magnetars: ~10^-12-10^-11 erg/cm^2/s; null for transient magnetars in quiescence below detection limits 1.4e-13 19.4%
xray_flux_err_up float64 Upper 1-sigma uncertainty on X-ray flux (erg/cm^2/s) 1.4e-15 61.3%
xray_flux_err_down float64 Lower 1-sigma uncertainty on X-ray flux (erg/cm^2/s) 1e-15 61.3%
distance_kpc float64 Distance in kpc; null for the majority of magnetars (reliable distances are rare — methods include HI absorption, SNR associations, and maser parallaxes) 62.4 19.4%
distance_err_up_kpc float64 Upper 1-sigma uncertainty on distance (kpc) 1.6 45.2%
distance_err_down_kpc float64 Lower 1-sigma uncertainty on distance (kpc) 1.6 45.2%
xray_luminosity_erg_s float64 Quiescent X-ray luminosity in erg/s computed from flux and distance; magnetars: 10^33-10^36 erg/s; null where distance is unknown 6.499999999999999e+34 25.8%
association object Name of associated supernova remnant or star cluster (e.g. 'CTB 109', 'Westerlund 1'); null for isolated magnetars without identified associations SMC 35.5%
optical_ir_counterpart object Whether an optical or infrared counterpart has been detected; null if no counterpart search has been published maybe 0.0%
observed_bands object Observational coverage codes: H=hard X-ray (>10 keV), X=soft X-ray, O=optical, I=infrared, R=radio, G=gamma-ray; null if not tabulated X 9.7%
activity_flags object Burst/flare activity type codes: B=bursts, G=giant flare, F=flare, T=transient outburst, A=anti-glitch; null for sources with no recorded activity BG 9.7%
ra_hms object Right ascension in sexagesimal format (HH MM SS.s), ICRS J2000 01 00 43.14 0.0%
ra_err_arcsec float64 1-sigma positional uncertainty in RA in arcseconds; null for sources without a precise X-ray or radio position 0.13 3.2%
dec_dms object Declination in sexagesimal format (+/-DD MM SS.s), ICRS J2000 -72 11 33.8 0.0%
dec_err_arcsec float64 1-sigma positional uncertainty in Dec in arcseconds; null for sources without a precise X-ray or radio position 0.6 3.2%
is_candidate bool True for unconfirmed magnetar candidates (marked with # in the McGill catalog); candidate status may change as new observations are published False 0.0%
type object Historical source class: 'SGR' (detected via gamma-ray bursts) or 'AXP' (detected as anomalous X-ray pulsar); distinction is observational, not physical AXP 0.0%
ra_deg float64 Right ascension in decimal degrees (ICRS J2000.0); derived from ra_hms 15.179749999999999 0.0%
dec_deg float64 Declination in decimal degrees (ICRS J2000.0); derived from dec_dms -72.19272222222223 0.0%
magnetic_field_g_is_limit bool True when the magnetic field value is an upper or lower limit rather than a detection False 0.0%
spin_down_luminosity_erg_s_is_limit bool True when the spin-down luminosity value is an upper or lower limit False 0.0%
characteristic_age_yr_is_limit bool True when the characteristic age value is an upper or lower limit False 0.0%
xray_flux_erg_cm2_s_is_limit bool True when the X-ray flux value is an upper or lower limit False 0.0%
xray_luminosity_erg_s_is_limit bool True when the X-ray luminosity value is an upper or lower limit False 0.0%
distance_kpc_is_limit bool True when the distance value is an upper or lower limit False 0.0%

Quick stats

  • 31 magnetars (24 confirmed, 7 candidates)
  • 13 Soft Gamma Repeaters, 18 Anomalous X-ray Pulsars
  • 26 with measured spin periods (0.33--11.8 s)
  • 25 with inferred magnetic fields (6.10e+12--1.96e+15 G)
  • 20 associated with supernova remnants or star clusters

Usage 

from datasets import load_dataset

ds = load_dataset("juliensimon/mcgill-magnetar-catalog", split="train")
df = ds.to_pandas()

# Confirmed magnetars only
confirmed = df[~df["is_candidate"]]

# P-Pdot diagram (period vs. period derivative)
import matplotlib.pyplot as plt
import numpy as np

valid = confirmed.dropna(subset=["period_s", "period_derivative"])
plt.figure(figsize=(8, 6))
plt.scatter(valid["period_s"], valid["period_derivative"], s=50, c="crimson", edgecolors="k")
plt.xscale("log")
plt.yscale("log")
plt.xlabel("Spin Period (s)")
plt.ylabel("Period Derivative (s/s)")
plt.title("Magnetar P-Pdot Diagram")
plt.tight_layout()
plt.show()

# Strongest magnetic fields
strongest = confirmed.sort_values("magnetic_field_g", ascending=False).head(5)
print(strongest[["name", "type", "magnetic_field_g", "period_s"]])

Data source

http://www.physics.mcgill.ca/~pulsar/magnetar/main.html

Related datasets

Citation 

@dataset{mcgill_magnetar_catalog,
  title = {McGill Online Magnetar Catalog},
  author = {juliensimon},
  year = {2026},
  url = {https://huggingface.co/datasets/juliensimon/mcgill-magnetar-catalog},
  publisher = {Hugging Face}
}

License

CC-BY-4.0