juliensimon/globular-star-clusters

---
license: cc-by-4.0
pretty_name: "Milky Way Globular Star Clusters"
language:
- en
description: "Comprehensive catalog of 170 Milky Way globular clusters merging the Harris (2010) and Baumgardt databases. Includes positions, distances, metallicities, masses, velocity dispersions, structural parameters, and photometry."
task_categories:
- tabular-classification
- tabular-regression
tags:
- space
- globular-clusters
- stars
- milky-way
- astronomy
- open-data
- tabular-data
size_categories:
- n<1K
configs:
- config_name: default
data_files:
- split: train
path: data/globular_star_clusters.parquet
default: true
---

# Milky Way Globular Star Clusters

*Part of the [Astronomy Datasets](https://huggingface.co/collections/juliensimon/astronomy-datasets-67ac2ada12aceb39f8feca3b) collection on Hugging Face.*

A comprehensive catalog of **170** Milky Way globular clusters, merging two authoritative sources:
the [Harris (2010 edition)](https://physics.mcmaster.ca/~harris/mwgc.dat) catalog for metallicities and
photometry, and the [Baumgardt globular cluster database](https://people.smp.uq.edu.au/HolgerBaumgardt/globular/)
for dynamical masses, velocity dispersions, and structural parameters from N-body model fits to
Gaia DR3 proper motions and HST data.

## Dataset description

Globular clusters are ancient, gravitationally bound collections of stars orbiting the Milky Way.
They are among the oldest objects in the Galaxy (10--13 Gyr), with typical masses of
10⁴--10⁶ M☉ and half-light radii of a few parsecs. Their metallicities, dynamics,
and spatial distribution encode the formation and assembly history of the Milky Way.

This dataset combines Harris (2010) photometric and chemical data with Baumgardt's dynamical
parameters derived from N-body fits to modern astrometric and spectroscopic data, providing
the most complete per-cluster view available.

## Schema

| Column | Type | Description |
|--------|------|-------------|
| `name` | string | Cluster name (e.g. "NGC 104", "Pal 5") |
| `ra_deg` / `dec_deg` | float64 | Position in decimal degrees (J2000) |
| `distance_kpc` | float64 | Distance from the Sun (kpc) |
| `distance_err_kpc` | float64 | Distance uncertainty (kpc) |
| `distance_gc_kpc` | float64 | Galactocentric distance (kpc) |
| `metallicity_fe_h` | float64 | Metallicity [Fe/H] (dex) |
| `reddening_e_bv` | float64 | Foreground reddening E(B−V) |
| `apparent_mag_v` | float64 | Apparent integrated V magnitude |
| `absolute_mag_v` | float64 | Absolute integrated V magnitude |
| `color_u_b` / `color_b_v` / `color_v_r` / `color_v_i` | float64 | Integrated color indices |
| `spectral_type` | string | Integrated spectral type |
| `ellipticity` | float64 | Projected ellipticity e = 1−b/a |
| `mass_msun` | float64 | Total cluster mass (M☉) |
| `mass_err_msun` | float64 | Mass uncertainty |
| `mass_to_light_v` | float64 | V-band mass-to-light ratio (M☉/L☉) |
| `log_initial_mass_msun` | float64 | Log₁₀ initial mass (M☉) |
| `dissolution_time_gyr` | float64 | Predicted dissolution time (Gyr) |
| `core_radius_pc` | float64 | Core radius (pc) |
| `half_light_radius_pc` | float64 | Projected half-light radius (pc) |
| `half_mass_radius_pc` | float64 | 3D half-mass radius (pc) |
| `tidal_radius_pc` | float64 | Tidal radius (pc) |
| `log_central_density_msun_pc3` | float64 | Log₁₀ central density (M☉/pc³) |
| `log_half_mass_density_msun_pc3` | float64 | Log₁₀ half-mass density (M☉/pc³) |
| `log_half_mass_relaxation_time_yr` | float64 | Log₁₀ half-mass relaxation time (yr) |
| `velocity_dispersion_km_s` | float64 | Central velocity dispersion (km/s) |
| `escape_velocity_km_s` | float64 | Central escape velocity (km/s) |
| `radial_velocity_km_s` | float64 | Heliocentric radial velocity (km/s) |
| `anisotropy_central` / `anisotropy_half_mass` | float64 | Velocity anisotropy (η) |
| `rotation_amplitude_km_s` | float64 | Rotation amplitude (km/s) |
| `rotation_probability_pct` | float64 | Probability of significant rotation (%) |
| `mass_function_slope` | float64 | Present-day mass function slope (α) |
| `n_radial_velocity_stars` | int | Stars with radial velocity measurements |
| `n_proper_motion_stars` | int | Stars with proper motion measurements |
| `core_collapsed` | bool | Core-collapsed cluster (Harris) |
| `concentration_harris` | float64 | King-model concentration c = log(r_t/r_c) |

## Quick stats

- **170** Milky Way globular clusters
- **167** with dynamical mass estimates (7.46e+02–3.94e+06 M☉)
- **152** with metallicity measurements (-2.37 to 0.00 dex)
- **167** with central velocity dispersions
- **29** identified as core-collapsed

## Usage

```python
from datasets import load_dataset

ds = load_dataset("juliensimon/globular-star-clusters", split="train")
df = ds.to_pandas()

# Most massive clusters
massive = df.nlargest(10, "mass_msun")[["name", "mass_msun", "metallicity_fe_h"]]

# Metal-poor vs metal-rich populations
metal_poor = df[df["metallicity_fe_h"] < -1.5]
metal_rich = df[df["metallicity_fe_h"] >= -1.5]

# Core-collapsed clusters
cc = df[df["core_collapsed"]]

# Mass-metallicity relation
import matplotlib.pyplot as plt
plt.scatter(df["metallicity_fe_h"], df["mass_msun"].apply(lambda x: x if x else None))
plt.xlabel("[Fe/H]"); plt.ylabel("Mass (M☉)"); plt.yscale("log")
```

## Data sources

1. **Harris (2010 edition)**: [McMaster Globular Cluster Catalog](https://physics.mcmaster.ca/~harris/mwgc.dat).
Please cite [Harris (1996), AJ 112, 1487](https://ui.adsabs.harvard.edu/abs/1996AJ....112.1487H) — 2010 edition.

2. **Baumgardt Globular Cluster Database**: [https://people.smp.uq.edu.au/HolgerBaumgardt/globular/](https://people.smp.uq.edu.au/HolgerBaumgardt/globular/).
Please cite [Baumgardt & Hilker (2018), MNRAS 478, 1520](https://ui.adsabs.harvard.edu/abs/2018MNRAS.478.1520B).

## Related datasets

- [open-star-clusters](https://huggingface.co/datasets/juliensimon/open-star-clusters) — Milky Way open clusters
- [stellar-streams](https://huggingface.co/datasets/juliensimon/stellar-streams) — Tidal stellar streams
- [pulsars](https://huggingface.co/datasets/juliensimon/pulsars) — ATNF Pulsar Catalogue

## Pipeline

Source code: [juliensimon/space-datasets](https://github.com/juliensimon/space-datasets)

## Citation

```bibtex
@dataset{globular_star_clusters,
author = {Simon, Julien},
title = {Milky Way Globular Star Clusters},
year = {2026},
publisher = {Hugging Face},
url = {https://huggingface.co/datasets/juliensimon/globular-star-clusters},
note = {Merged from Harris (1996, 2010 edition) and Baumgardt et al. globular cluster databases}
}
```

## License

[CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/)

许可证：CC-BY-4.0
美观名称："银河系球状星团"
语言：
- 英语
描述："整合Harris（2010）与Baumgardt数据库的170个银河系球状星团综合目录，包含位置、距离、金属丰度、质量、速度弥散、结构参数与测光数据。"
任务类别：
- 表格分类
- 表格回归
标签：
- 太空
- 球状星团
- 恒星
- 银河系
- 天文学
- 开放数据
- 表格数据
数据规模类别：
- n<1K
配置项：
- 配置名称：default
数据文件：
- 拆分：train
路径：data/globular_star_clusters.parquet
默认启用：true

# 银河系球状星团

*隶属于Hugging Face上的[Astronomy Datasets](https://huggingface.co/collections/juliensimon/astronomy-datasets-67ac2ada12aceb39f8feca3b)数据集合集。*

## 数据集概述
球状星团是围绕银河系运行的古老引力束缚恒星集合，是银河系中最古老的天体之一（100亿至130亿年），典型质量为10⁴至10⁶倍太阳质量（M☉），半光半径为数秒差距（pc）。它们的金属丰度、动力学特性与空间分布蕴含着银河系形成与演化的历史信息。

本数据集整合了Harris（2010版）的测光与化学数据，以及Baumgardt基于N体拟合现代天体测量与光谱数据得到的动力学参数，提供了目前最为完备的单星团视角。

## 数据Schema
| 列名 | 数据类型 | 描述 |
|--------|------|-------------|
| `name` | 字符串 | 星团名称（例如"NGC 104"、"Pal 5"） |
| `ra_deg` / `dec_deg` | float64 | J2000坐标系下的位置（单位：十进制度） |
| `distance_kpc` | float64 | 日心距离（单位：千秒差距，kpc） |
| `distance_err_kpc` | float64 | 距离不确定度（单位：kpc） |
| `distance_gc_kpc` | float64 | 银心距离（单位：kpc） |
| `metallicity_fe_h` | float64 | 金属丰度[Fe/H]（单位：dex） |
| `reddening_e_bv` | float64 | 前景红化E(B−V) |
| `apparent_mag_v` | float64 | 综合视V星等 |
| `absolute_mag_v` | float64 | 综合绝对V星等 |
| `color_u_b` / `color_b_v` / `color_v_r` / `color_v_i` | float64 | 综合色指数 |
| `spectral_type` | 字符串 | 综合光谱型 |
| `ellipticity` | float64 | 投影椭率 e = 1−b/a |
| `mass_msun` | float64 | 星团总质量（单位：M☉） |
| `mass_err_msun` | float64 | 质量不确定度 |
| `mass_to_light_v` | float64 | V波段质量光度比（单位：M☉/L☉） |
| `log_initial_mass_msun` | float64 | 初始质量的以10为底对数（单位：M☉） |
| `dissolution_time_gyr` | float64 | 预测的瓦解时间（单位：Gyr） |
| `core_radius_pc` | float64 | 核心半径（单位：pc） |
| `half_light_radius_pc` | float64 | 投影半光半径（单位：pc） |
| `half_mass_radius_pc` | float64 | 三维半质量半径（单位：pc） |
| `tidal_radius_pc` | float64 | 潮汐半径（单位：pc） |
| `log_central_density_msun_pc3` | float64 | 中心密度的以10为底对数（单位：M☉/pc³） |
| `log_half_mass_density_msun_pc3` | float64 | 半质量密度的以10为底对数（单位：M☉/pc³） |
| `log_half_mass_relaxation_time_yr` | float64 | 半质量弛豫时间的以10为底对数（单位：yr） |
| `velocity_dispersion_km_s` | float64 | 中心速度弥散（单位：km/s） |
| `escape_velocity_km_s` | float64 | 中心逃逸速度（单位：km/s） |
| `radial_velocity_km_s` | float64 | 日心径向速度（单位：km/s） |
| `anisotropy_central` / `anisotropy_half_mass` | float64 | 速度各向异性（η） |
| `rotation_amplitude_km_s` | float64 | 旋转振幅（单位：km/s） |
| `rotation_probability_pct` | float64 | 显著旋转的概率（单位：%） |
| `mass_function_slope` | float64 | 现今质量函数斜率（α） |
| `n_radial_velocity_stars` | 整数 | 拥有径向速度测量数据的恒星数量 |
| `n_proper_motion_stars` | 整数 | 拥有自行测量数据的恒星数量 |
| `core_collapsed` | 布尔值 | 是否为核坍缩星团（来自Harris目录） |
| `concentration_harris` | float64 | King模型浓度 c = log(r_t/r_c) |

## 快速统计
- **170**个银河系球状星团
- **167**个拥有动力学质量估计（质量范围：7.46×10² 至 3.94×10⁶ M☉）
- **152**个拥有金属丰度测量值（范围：-2.37 至 0.00 dex）
- **167**个拥有中心速度弥散数据
- **29**个被鉴定为核坍缩星团

## 使用示例
python
from datasets import load_dataset

# 加载数据集训练拆分并转换为Pandas DataFrame
ds = load_dataset("juliensimon/globular-star-clusters", split="train")
df = ds.to_pandas()

# 筛选质量前十的星团
massive = df.nlargest(10, "mass_msun")[["name", "mass_msun", "metallicity_fe_h"]]

# 划分贫金属与富金属星团群体
metal_poor = df[df["metallicity_fe_h"] < -1.5]
metal_rich = df[df["metallicity_fe_h"] >= -1.5]

# 筛选核坍缩星团
cc = df[df["core_collapsed"]]

# 绘制质量-金属丰度关系图
import matplotlib.pyplot as plt
plt.scatter(df["metallicity_fe_h"], df["mass_msun"].apply(lambda x: x if x else None))
plt.xlabel("[Fe/H]"); plt.ylabel("质量（M☉）"); plt.yscale("log")

## 数据来源
1. **Harris（2010版）**：[麦克马斯特大学球状星团目录](https://physics.mcmaster.ca/~harris/mwgc.dat)。请引用[Harris (1996), AJ 112, 1487](https://ui.adsabs.harvard.edu/abs/1996AJ....112.1487H) — 2010版。
2. **Baumgardt球状星团数据库**：[https://people.smp.uq.edu.au/HolgerBaumgardt/globular/](https://people.smp.uq.edu.au/HolgerBaumgardt/globular/)。请引用[Baumgardt & Hilker (2018), MNRAS 478, 1520](https://ui.adsabs.harvard.edu/abs/2018MNRAS.478.1520B)。

## 相关数据集
- [开放星团](https://huggingface.co/datasets/juliensimon/open-star-clusters) — 银河系开放星团
- [恒星流](https://huggingface.co/datasets/juliensimon/stellar-streams) — 潮汐恒星流
- [脉冲星](https://huggingface.co/datasets/juliensimon/pulsars) — ATNF脉冲星目录

## 代码流程
源代码：[juliensimon/space-datasets](https://github.com/juliensimon/space-datasets)

## 引用格式
bibtex
@dataset{globular_star_clusters,
author = {Simon, Julien},
title = {Milky Way Globular Star Clusters},
year = {2026},
publisher = {Hugging Face},
url = {https://huggingface.co/datasets/juliensimon/globular-star-clusters},
note = {Merged from Harris (1996, 2010 edition) and Baumgardt et al. globular cluster databases}
}

## 许可证
[知识共享署名4.0协议（CC-BY-4.0）](https://creativecommons.org/licenses/by/4.0/)