|
|
|
|
|
--- |
|
|
license: apache-2.0 |
|
|
size_categories: |
|
|
- 1B<n |
|
|
task_categories: |
|
|
- tabular-classification |
|
|
- feature-extraction |
|
|
tags: |
|
|
- astronomy |
|
|
- astrophysics |
|
|
- gaia-dr3 |
|
|
- stellar-data |
|
|
- space |
|
|
--- |
|
|
|
|
|
# Gaia-DR3: A Billion-Star Dataset for Galactic Analysis |
|
|
|
|
|
|
|
|
|
|
|
This dataset is a high-fidelity, pre-processed collection of over **1 Billion stellar records** derived from the **European Space Agency (ESA) Gaia Mission (Data Release 3)**. It is specifically curated for large-scale galactic archaeology, 3D mapping, and training advanced machine learning models like **SADIM-V2 77M**. |
|
|
|
|
|
### 1. Dataset Overview |
|
|
The dataset provides a comprehensive snapshot of the Milky Way, covering astrometric, kinematic, and photometric parameters. It has been optimized for high-performance computing and AI-driven astronomical frameworks. |
|
|
|
|
|
### 2. Feature Schema (13 Core Parameters) |
|
|
The dataset is structured with 13 essential features for understanding stellar dynamics: |
|
|
|
|
|
| Feature Name | Data Type | Scientific Description | |
|
|
| :--- | :--- | :--- | |
|
|
| **source_id** | `int64` | Unique Gaia DR3 identifier for each stellar source. | |
|
|
| **ra / dec** | `float64` | **Equatorial Coordinates:** Right Ascension & Declination. | |
|
|
| **l / b** | `float64` | **Galactic Coordinates:** Longitude & Latitude relative to the Galaxy. | |
|
|
| **pmra / pmdec** | `float64` | **Proper Motion:** Angular velocity of the star across the sky (mas/yr). | |
|
|
| **d_pc** | `float64` | **Distance:** Calculated distance from Earth in Parsecs ($1/parallax$). | |
|
|
| **x, y, z** | `float64` | **3D Cartesian:** Heliocentric position relative to the Sun. | |
|
|
| **abs_m** | `float64` | **Absolute Magnitude:** The intrinsic brightness of the star. | |
|
|
| **bp_rp** | `float32` | **Color Index:** Difference between BP and RP (Temperature indicator). | |
|
|
|
|
|
### 3. Usage & Access (Streaming Mode) |
|
|
**Note:** Due to the massive scale of this dataset (1B+ rows), downloading the full files to a local machine is not recommended. Use the **Streaming Mode** provided by the Hugging Face `datasets` library to process data on the fly: |
|
|
|
|
|
```python |
|
|
from datasets import load_dataset |
|
|
|
|
|
# Stream the dataset directly without downloading the full files |
|
|
dataset = load_dataset("samfatnassi/gaia-dr3", split="train", streaming=True) |
|
|
|
|
|
# Access a single stellar record |
|
|
star_record = next(iter(dataset)) |
|
|
print(star_record) |
|
|
|
|
|
4. Integration with SADIM-V2 |
|
|
This dataset serves as the foundational "knowledge base" for the SADIM-54M Model. While the dataset provides the raw observational facts, the model provides the analytical intelligence to predict and classify these stars. |
|
|
5. Research & Ethics (Open Science) |
|
|
This dataset is released under the Apache 2.0 License. It is provided as a contribution to Open Science and Humanity, encouraging researchers, students, and developers worldwide to explore the mysteries of our galaxy without boundaries. |
|
|
Data Source: European Space Agency (ESA) Gaia Mission |
|
|
Project Lead: KilmaAI / Sadim |
|
|
|
|
|
|
|
|
|
|
|
|
