juliensimon/fermi-4fgl-dr4

---
license: cc-by-4.0
pretty_name: "Fermi LAT 4FGL-DR4 Gamma-Ray Source Catalog"
language:
- en
description: "The 14-year all-sky gamma-ray source catalog from the Fermi Large Area Telescope — the deepest survey of the gamma-ray sky."
task_categories:
- tabular-classification
- tabular-regression
tags:
- space
- gamma-ray
- fermi
- lat
- nasa
- astronomy
- high-energy
- open-data
size_categories:
- 1K<n<10K
---

# Fermi LAT 4FGL-DR4 Gamma-Ray Source Catalog


The Fourth Fermi Large Area Telescope Source Catalog, Data Release 4 (4FGL-DR4),
based on 14 years of all-sky gamma-ray survey data. Currently **7,195** sources.

## Dataset description

This dataset contains every gamma-ray source detected by the Fermi LAT instrument
in its 14-year all-sky survey. The 4FGL-DR4 is the deepest catalog of the
gamma-ray sky ever produced, with sources spanning blazars, pulsars, supernova
remnants, globular clusters, starburst galaxies, and many unidentified sources.

Each record includes sky position, spectral properties, flux measurements,
variability information, and source associations with counterparts at other
wavelengths.

## Schema

| Column | Type | Description |
|--------|------|-------------|
| `source_name` | string | 4FGL source name (e.g. "4FGL J0001.2+3738") |
| `ra_deg` | float64 | Right Ascension J2000 (degrees) |
| `dec_deg` | float64 | Declination J2000 (degrees) |
| `glon_deg` | float64 | Galactic longitude (degrees) |
| `glat_deg` | float64 | Galactic latitude (degrees) |
| `significance` | float64 | Detection significance (sigma) |
| `flux_1000_mev` | float64 | Photon flux above 1 GeV (photons/cm2/s) |
| `energy_flux_100_mev` | float64 | Energy flux above 100 MeV (erg/cm2/s) |
| `spectrum_type` | string | Spectral model type (PowerLaw, LogParabola, etc.) |
| `variability_index` | float64 | Variability index (>18.48 = variable at 99%) |
| `source_class` | string | Source classification (e.g. "bll", "psr", "fsrq") |
| `association` | string | Associated source name at other wavelengths |
| `redshift` | float64 | Redshift (where available) |
| `flags` | Int64 | Analysis flags |
| `pivot_energy_mev` | float64 | Pivot energy (MeV) |
| `power_law_index` | float64 | Power-law spectral index |
| `log_parabola_index` | float64 | Log-parabola spectral index |
| `log_parabola_beta` | float64 | Log-parabola curvature parameter |
| `is_variable` | bool | Variability flag (variability_index > 18.48) |

## Quick stats

- **7,195** gamma-ray sources
- **3,051** variable sources (99% confidence)
- **0** sources with measured redshift
- Top source classes:
- **bcu**: 1,622
- **bll**: 1,468
- **fsrq**: 776
- **MSP**: 139
- **PSR**: 137
- **unk**: 137
- **spp**: 126
- **rdg**: 47
- **FSRQ**: 44
- **glc**: 41

## Usage

```python
from datasets import load_dataset

ds = load_dataset("juliensimon/fermi-4fgl-dr4", split="train")
df = ds.to_pandas()

# Brightest sources by significance
brightest = df.sort_values("significance", ascending=False).head(20)

# Variable blazars
variable_blazars = df[
(df["is_variable"] == True) &
(df["source_class"].isin(["bll", "fsrq", "BLL", "FSRQ"]))
]

# Sources with known redshift
with_z = df[df["redshift"].notna()].sort_values("redshift", ascending=False)

# Sky distribution
import matplotlib.pyplot as plt
fig, ax = plt.subplots(subplot_kw={"projection": "aitoff"})
import numpy as np
l = np.radians(df["glon_deg"].values)
l[l > np.pi] -= 2 * np.pi
b = np.radians(df["glat_deg"].values)
ax.scatter(l, b, s=0.1, alpha=0.3)
```

## Data source

[Fermi LAT collaboration](https://fermi.gsfc.nasa.gov/ssc/data/access/lat/14yr_catalog/),
4FGL-DR4 (Abdollahi et al. 2022, updated). Based on 14 years of Fermi LAT
Pass 8 data.

## Update schedule

Annual (January 1) via [GitHub Actions](https://github.com/juliensimon/space-datasets).

## Related datasets

- [grb-catalog](https://huggingface.co/datasets/juliensimon/grb-catalog) -- Gamma-ray burst catalog
- [snr-catalog](https://huggingface.co/datasets/juliensimon/snr-catalog) -- Supernova remnant catalog
- [pulsar-catalog](https://huggingface.co/datasets/juliensimon/pulsar-catalog) -- Pulsar catalog

## Pipeline

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

## Citation

```bibtex
@dataset{fermi_4fgl_dr4,
author = {Simon, Julien},
title = {Fermi LAT 4FGL-DR4 Gamma-Ray Source Catalog},
year = {2026},
publisher = {Hugging Face},
url = {https://huggingface.co/datasets/juliensimon/fermi-4fgl-dr4},
note = {Based on Fermi LAT 4FGL-DR4 catalog (Abdollahi et al. 2022, updated)}
}
```

## License

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

license: 知识共享署名4.0（CC-BY-4.0）
pretty_name: "费米大视场望远镜4FGL-DR4伽马射线源目录（Fermi LAT 4FGL-DR4 Gamma-Ray Source Catalog）"
language:
- 英语
description: "本数据集源自费米大视场望远镜（Fermi Large Area Telescope, LAT）的14年全天巡天数据，是迄今最深的伽马射线天区巡天源目录。"
task_categories:
- 表格分类（tabular-classification）
- 表格回归（tabular-regression）
tags:
- 空间科学
- 伽马射线
- 费米
- LAT
- 美国国家航空航天局（National Aeronautics and Space Administration, NASA）
- 天文学
- 高能物理
- 开放数据
size_categories:
- 1000<n<10000

# 费米大视场望远镜4FGL-DR4伽马射线源目录（Fermi LAT 4FGL-DR4 Gamma-Ray Source Catalog）


第四版费米大视场望远镜源目录数据发布4（4FGL-DR4），基于14年的全天伽马射线巡天数据构建，当前共收录**7195**个伽马射线源。

## 数据集说明

本数据集收录了费米LAT仪器在14年全天巡天中探测到的全部伽马射线源。4FGL-DR4是迄今发布的最深伽马射线天区目录，收录的源涵盖耀变体、脉冲星、超新星遗迹、球状星团、星暴星系以及大量未证认源。

每条数据记录包含天区位置、光谱特性、流量测量结果、变源信息以及与其他波段对应天体的关联信息。

## 数据结构

| 列名 | 类型 | 描述 |
|--------|------|-------------|
| `source_name` | 字符串 | 4FGL源名称（例如"4FGL J0001.2+3738"） |
| `ra_deg` | 64位浮点型 | 赤经J2000（度） |
| `dec_deg` | 64位浮点型 | 赤纬J2000（度） |
| `glon_deg` | 64位浮点型 | 银经（度） |
| `glat_deg` | 64位浮点型 | 银纬（度） |
| `significance` | 64位浮点型 | 探测显著性（σ） |
| `flux_1000_mev` | 64位浮点型 | 1GeV以上光子流量（光子/cm²/s） |
| `energy_flux_100_mev` | 64位浮点型 | 100MeV以上能量流量（erg/cm²/s） |
| `spectrum_type` | 字符串 | 光谱模型类型（如幂律模型、对数抛物线模型等） |
| `variability_index` | 64位浮点型 | 变源指数（>18.48代表99%置信度下为变源） |
| `source_class` | 字符串 | 源分类（例如"bll"、"psr"、"fsrq"） |
| `association` | 字符串 | 其他波段关联天体名称 |
| `redshift` | 64位浮点型 | 红移（如有可用数据） |
| `flags` | 64位整型 | 分析标记 |
| `pivot_energy_mev` | 64位浮点型 | 转折能量（MeV） |
| `power_law_index` | 64位浮点型 | 幂律光谱指数 |
| `log_parabola_index` | 64位浮点型 | 对数抛物线光谱指数 |
| `log_parabola_beta` | 64位浮点型 | 对数抛物线曲率参数 |
| `is_variable` | 布尔型 | 变源标记（变源指数>18.48时为真） |

## 快速统计

- **7195**个伽马射线源
- **3051**个变源（99%置信度）
- **0**个拥有测量红移的源
- 主要源分类：
- **bcu**: 1622
- **bll**: 1468
- **fsrq**: 776
- **MSP**: 139
- **PSR**: 137
- **unk**: 137
- **spp**: 126
- **rdg**: 47
- **FSRQ**: 44
- **glc**: 41

## 使用示例

python
from datasets import load_dataset

ds = load_dataset("juliensimon/fermi-4fgl-dr4", split="train")
df = ds.to_pandas()

# 按探测显著性排序的最亮源
brightest = df.sort_values("significance", ascending=False).head(20)

# 变耀变体
variable_blazars = df[
(df["is_variable"] == True) &
(df["source_class"].isin(["bll", "fsrq", "BLL", "FSRQ"]))
]

# 拥有已知红移的源
with_z = df[df["redshift"].notna()].sort_values("redshift", ascending=False)

# 天区分布
import matplotlib.pyplot as plt
fig, ax = plt.subplots(subplot_kw={"projection": "aitoff"})
import numpy as np
l = np.radians(df["glon_deg"].values)
l[l > np.pi] -= 2 * np.pi
b = np.radians(df["glat_deg"].values)
ax.scatter(l, b, s=0.1, alpha=0.3)

## 数据来源

[费米LAT合作组](https://fermi.gsfc.nasa.gov/ssc/data/access/lat/14yr_catalog/)，4FGL-DR4（Abdollahi等人，2022，已更新）。基于费米LAT第8版数据分析的14年巡天数据。

## 更新计划

每年更新一次（1月1日），通过[GitHub Actions](https://github.com/juliensimon/space-datasets)实现。

## 相关数据集

- [grb-catalog](https://huggingface.co/datasets/juliensimon/grb-catalog)——伽马射线暴目录
- [snr-catalog](https://huggingface.co/datasets/juliensimon/snr-catalog)——超新星遗迹目录
- [pulsar-catalog](https://huggingface.co/datasets/juliensimon/pulsar-catalog)——脉冲星目录

## 数据流程

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

## 引用格式

bibtex
@dataset{fermi_4fgl_dr4,
author = {Simon, Julien},
title = {Fermi LAT 4FGL-DR4 Gamma-Ray Source Catalog},
year = {2026},
publisher = {Hugging Face},
url = {https://huggingface.co/datasets/juliensimon/fermi-4fgl-dr4},
note = {Based on Fermi LAT 4FGL-DR4 catalog (Abdollahi et al. 2022, updated)}
}

## 许可证

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