Galaxy Zoo DESI: Detailed Morphology Classifications
Detailed morphological classifications of 8.7 million galaxies from the DESI Legacy Imaging Survey, produced with deep learning trained on Galaxy Zoo volunteers.
- Size
- 11.5 GB
- Format
- CSV, Parquet
- License
- CC-BY-4.0
- Access
- Open Access
- Source
- Zenodo (record 8360385)
Your edits to the notebook only affect your own copy in the Hub. Changed it and want the original back? Restore it here.
Challenge
Does galaxy shape track galaxy colour and distance the way theory predicts? Across all 8.7 million galaxies, test whether smooth (elliptical) galaxies are redder than featured/spiral ones — and check how much of the trend with redshift is real versus an observational bias in a magnitude-limited survey.
- Where to start
- Never load these multi-GB files whole. Use DuckDB to query the Parquet on disk: join gz_desi_deep_learning_catalog_friendly.parquet (the smooth-or-featured and has-spiral-arms vote fractions) to external_catalog.parquet on dr8_id, pulling only mag_g_desi, mag_r_desi and redshift. Aggregate and bin inside the SQL so only small summary tables reach memory.
- What to share
- One or two plots — e.g. mean smooth vs. featured vote fraction against g−r colour, and the featured/spiral fraction against redshift — plus a short paragraph separating the genuine colour–morphology relation from the distance/brightness selection effect.
💬 Discuss this dataset, ask questions and share your results in its discussion thread. Challenges are open-ended — there's no single right answer.