---
dataset_info:
  features:
  - name: targetid
    dtype: int64
  - name: healpix
    dtype: int64
  - name: survey
    dtype: string
  - name: program
    dtype: string
  - name: redshift
    dtype: float64
  - name: RGB_image
    dtype: image
  - name: VIS_image
    sequence:
      sequence: float32
  - name: NISP_Y_image
    sequence:
      sequence: float32
  - name: NISP_J_image
    sequence:
      sequence: float32
  - name: NISP_H_image
    sequence:
      sequence: float32
  - name: spectrum
    struct:
    - name: error
      sequence: float32
    - name: flux
      sequence: float32
    - name: ivar
      sequence: float32
    - name: mask
      sequence: float32
    - name: targetid
      dtype: int64
    - name: wavelength
      sequence: float32
  splits:
  - name: train
    num_bytes: 34358672981.6
    num_examples: 31880
  - name: test
    num_bytes: 8589268242.93
    num_examples: 7970
  download_size: 33060048114
  dataset_size: 42947941224.53
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---

# DESI DR1 × Euclid Q1 Dataset

## Dataset Description

This dataset contains the cross-matched catalog between **DESI Data Release 1 (DR1)** and **Quick Data Release 1 (Q1)** data. 

### Dataset Summary

- **Total Sources**: 39,850 galaxies
- **Cross-match Radius**: 0.5 arcsec
- **Data Products**: 
  - Euclid VIS imaging
  - Euclid NISP imaging (Y, J, H bands)
  - DESI spectra

## Source Data

1) **DESI Sample Selection**: Applied quality cuts to DESI DR1:
   - ZCAT_PRIMARY == True
   - ZWARN == 0 or ZWARN == 4
   - COADD_FIBERSTATUS == 0
   - SPECTYPE != 'STAR'
   - Z > 0

2) **Euclid Sample Selection**: Applied quality cuts to Euclid MER Q1: 
   - VIS_DET == 1
   - gaia_id IS NULL
   - spurious_prob < 0.2
   - flux_detection_total > 0.0


## Data Splits

- **Train**: 80% (31,880 galaxies)
- **Test**: 20% (7,970 galaxies)

Split created using stratified random sampling with `random_state=42`.

## Usage

```python
from datasets import load_dataset

# Load the full dataset
dataset = load_dataset("msiudek/astroPT_euclid_Q1_desi_dr1_dataset")

# Access train/test splits
train_data = dataset['train']
test_data = dataset['test']

# Example: Access a single sample
sample = train_data[0]
vis_image = sample['vis_image']
spectrum = sample['spectrum']
redshift = sample['redshift']
```



## Metadata catalog

A companion metadata catalog (FITS table) with per-object properties and cross-match details is published as a separate dataset on Hugging Face:

- Metadata dataset: https://huggingface.co/datasets/msiudek/astroPT_euclid_Q1_desi_dr1_metadata

How to access the metadata FITS directly from the hub using Astropy:

```python
from huggingface_hub import hf_hub_download
from astropy.table import Table
import warnings
from astropy.units.core import UnitsWarning

# Optional: silence non-standard FITS unit warnings (e.g., 'nanomaggy', 'NA')
warnings.filterwarnings("ignore", category=UnitsWarning)

fits_path = hf_hub_download(
  repo_id="msiudek/astroPT_euclid_Q1_desi_dr1_metadata",
  repo_type="dataset",
  filename="DESI_DR1_Euclid_Q1_dataset_catalog_EM.fits",
)
tab = Table.read(fits_path, format="fits")
print(tab)
```

## Citation

If you use this dataset, please cite:

```bibtex
@dataset{euclid_desi_dr1_q1,
  author = {Siudek, M. and collaborators},
  title = {DESI DR1 × Euclid Q1 Cross-matched Dataset},
  year = {2025},
  url = {https://huggingface.co/datasets/msiudek/astroPT_euclid_Q1_desi_dr1_dataset}
}
```