juliensimon/hipparcos-catalog

---
license: cc-by-4.0
pretty_name: "Hipparcos Star Catalog"
language:
- en
description: "The ESA Hipparcos space astrometry mission catalog containing 118,218 of the brightest stars in the sky with precise positions, parallaxes, and proper motions. The Hipparcos satellite (1989–1993) was"
task_categories:
- tabular-classification
- tabular-regression
tags:
- space
- hipparcos
- star
- astrometry
- parallax
- astronomy
- open-data
- tabular-data
- parquet
size_categories:
- 100K<n<1M
configs:
- config_name: default
data_files:
- split: train
path: data/hipparcos.parquet
default: true
---

# Hipparcos Star Catalog

<div align="center">
<img src="banner.jpg" alt="A youthful globular star cluster observed by the Hubble Space Telescope" width="400">
<p><em>Credit: NASA/ESA/Hubble</em></p>
</div>

*Part of a [dataset collection](https://huggingface.co/collections/juliensimon/astronomy-datasets-69c24caf2f17e36128946743) on Hugging Face.*

## Dataset description

The ESA Hipparcos space astrometry mission catalog containing 118,218 of the brightest stars in the sky with precise positions, parallaxes, and proper motions.

The Hipparcos satellite (1989–1993) was ESA's pioneering space astrometry mission. It measured the positions, parallaxes, and proper motions of stars with unprecedented precision, creating the first high-accuracy stellar reference frame from space. Hipparcos parallaxes remain the gold standard for nearby star distances and are the foundation for the cosmic distance ladder.

Hipparcos achieved milliarcsecond-level astrometry — a factor of 100 improvement over ground-based catalogs — by observing from above Earth's atmosphere. The mission's key deliverable, trigonometric parallax, provides the most direct and model-independent method of measuring stellar distances: a star at 1 parsec subtends a parallax of 1 arcsecond, and distance in parsecs is simply 1/parallax. With typical parallax uncertainties of 1 mas, Hipparcos yielded distances accurate to 10% out to about 100 pc.

The scientific legacy of Hipparcos extends far beyond simple distance measurement. Proper motions from the catalog revealed the kinematic structure of nearby stellar streams and moving groups. Combined with radial velocities, Hipparcos data enabled full three-dimensional space velocity determinations for thousands of stars. Although Gaia has since surpassed Hipparcos in depth and precision, the catalog retains enduring value: it provides an independent epoch (J1991.25) for long-baseline proper motion studies, and its bright-star astrometry benchmarks Gaia solutions at the bright end.

## Schema

| Column | Type | Description | Sample | Null % |
|--------|------|-------------|--------|--------|
| `hip_id` | Int64 | Hipparcos Input Catalog identifier; integer in range 1–120404; the standard cross-reference identifier for stars brighter than ~12 mag observed by the Hipparcos satellite (1989–1993); still widely used to cross-match with Gaia and Tycho-2 | 1 | 0.0% |
| `v_magnitude` | float64 | Johnson V-band apparent magnitude; higher values are fainter; catalog covers roughly V = 2–12.4 mag; null for stars with only Hp-band photometry | 9.1 | 0.0% |
| `ra_deg` | float64 | Right ascension in degrees, ICRS at reference epoch J1991.25 (the astrometric midpoint of the Hipparcos mission, not J2000.0); range 0–360; differs from J2000.0 by a small proper-motion correction that grows with stellar velocity | 0.00091185 | 0.2% |
| `dec_deg` | float64 | Declination in degrees, ICRS at reference epoch J1991.25; range −90 to +90; positive north of the celestial equator | 1.08901332 | 0.2% |
| `parallax_mas` | float64 | Trigonometric parallax in milliarcseconds; convert to distance via distance_pc = 1000 / parallax_mas; Hipparcos precision ~1 mas (cf. Gaia ~0.02 mas); negative values are physically meaningful measurement noise for distant stars where the true parallax is near zero | 3.54 | 0.2% |
| `proper_motion_ra_mas_yr` | float64 | Proper motion in right ascension in milliarcseconds per year, with the cos(dec) factor already applied so this is the true angular rate on the sky (not the coordinate rate); positive = eastward motion | -5.2 | 0.2% |
| `proper_motion_dec_mas_yr` | float64 | Proper motion in declination in milliarcseconds per year; positive = northward motion; combined with proper_motion_ra_mas_yr gives the full tangential velocity vector on the sky | -1.88 | 0.2% |
| `parallax_error_mas` | float64 | 1-sigma formal uncertainty on the parallax in milliarcseconds; stars where parallax_error_mas > 0.5 × parallax_mas have uncertain distances (signal-to-noise < 2); use with caution for distance-dependent analyses | 1.39 | 0.2% |
| `color_bv` | float64 | Johnson B−V color index in magnitudes; more positive values indicate redder, cooler stars (e.g. B−V ≈ −0.3 for hot O/B stars, +1.5 for cool M giants); null for stars lacking B-band photometry | 0.482 | 1.1% |
| `spectral_type` | object | MK (Morgan–Keenan) spectral classification from catalog cross-references, e.g. 'G2V' (Sun-like), 'K0III' (red giant); encodes temperature class (O B A F G K M), luminosity class (I–V), and sometimes peculiarity flags; null for ~30% of stars, especially fainter objects | F5 | 2.6% |
| `distance_pc` | float64 | Heliocentric distance in parsecs, derived as 1000 / parallax_mas; null when parallax_mas ≤ 0 (unphysical noise-dominated measurements); treat values with large parallax_error_mas as highly uncertain | 282.4858757062147 | 3.8% |

## Quick stats

- **118,218** stars
- **117,955** with measured parallax
- **113,710** with derived distance (parallax > 0)
- **115,184** with spectral type classification
- Median V magnitude: **8.4**

## Usage

```python
from datasets import load_dataset
import matplotlib.pyplot as plt

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

# Hertzsprung-Russell diagram: color vs. absolute magnitude
# Compute absolute magnitude from distance modulus: M = V - 5*log10(d/10)
import numpy as np
valid = df.dropna(subset=["color_bv", "v_magnitude", "distance_pc"])
valid = valid[valid["distance_pc"] > 0]
valid["abs_mag"] = valid["v_magnitude"] - 5 * np.log10(valid["distance_pc"] / 10)

plt.figure(figsize=(8, 10))
plt.scatter(valid["color_bv"], valid["abs_mag"], s=0.2, alpha=0.3, c="steelblue")
plt.gca().invert_yaxis()
plt.xlabel("B-V Color Index (bluer ← → redder)")
plt.ylabel("Absolute Magnitude Mv (brighter ↑)")
plt.title("Hipparcos HR Diagram")
plt.tight_layout()
plt.show()

# Nearest stars within 10 parsecs
nearby = df[df["distance_pc"] < 10].sort_values("distance_pc")
print(nearby[["hip_id", "v_magnitude", "distance_pc", "spectral_type"]].head(20))
```

## Data source

https://vizier.cds.unistra.fr/viz-bin/VizieR-3?-source=I/239/hip_main

## Related datasets

- [juliensimon/gcvs-variable-stars](https://huggingface.co/datasets/juliensimon/gcvs-variable-stars)

- [juliensimon/open-star-clusters](https://huggingface.co/datasets/juliensimon/open-star-clusters)

- [juliensimon/pulsar-catalog](https://huggingface.co/datasets/juliensimon/pulsar-catalog)

## Citation

```bibtex
@dataset{hipparcos_catalog,
title = {Hipparcos Star Catalog},
author = {juliensimon},
year = {2026},
url = {https://huggingface.co/datasets/juliensimon/hipparcos-catalog},
publisher = {Hugging Face}
}
```

## License

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