# Multimodal Universe: Carnegie Supernova Project DR3 Type Ia Light Curves (134 SNe, HATS)

## Facts

- Provider: Multimodal Universe Collaboration (Polymathic AI, UniverseTBD); Carnegie Supernova Project
- License: CC-BY-4.0 (https://csp.obs.carnegiescience.edu/data)
- Size: 195 files
- Format: PNG, Parquet (in sampled files)
- Stored at: Hugging Face, https://huggingface.co/datasets/UniverseTBD/mmu_csp_csp
- Access: public, readable without an account
- Category: Astronomy & Space
- Source: https://arxiv.org/abs/2412.02527
- Page: https://scigantic.com/archive/3e4b4dfa-68ae-5735-ab0c-cc7f7be0d903
- Keywords: Astronomy & Space, spectroscopy, time series, PNG, Parquet, Hugging Face

## About

Multi-band light curves of 134 spectroscopically confirmed Type Ia supernovae from the third data release (DR3) of the first stage of the Carnegie Supernova Project (CSP-I), observed 2004 to 2009 with the Swope 1 m and du Pont 2.5 m telescopes at Las Campanas Observatory, packaged as part of the Multimodal Universe collection. The modality is time series photometry in the optical and near-infrared ugriBVYJH bands: each supernova row carries a nested lightcurve (band, time in MJD, magnitude and magnitude error, in the natural CSP photometric system), plus redshift, ra, dec, spectroscopic class and object_id. No preprocessing is applied beyond repackaging. The README lists 134 rows in 122 HEALPix partitions; about 104 MB stored (measured 2026-11-16), main catalog mmu_csp_csp/ and margin mmu_csp_csp_10arcs/. Data are a HATS catalog: Apache Parquet partitioned by HEALPix tile, with a 10 arcsecond margin catalog for cross-matching. Open with LSDB (lsdb.open_catalog on the Hugging Face URL, with cone searches and crossmatch), or read the Parquet with pandas, pyarrow, dask or DuckDB. Uses include SN Ia standardisation and cosmology, light-curve template fitting (SNooPy, SALT), near-infrared supernova studies and light-curve inpainting. The CSP publishes all its data under a Creative Commons Attribution licence; cite Krisciunas et al. 2017 (CSP DR3) and the Multimodal Universe paper (arXiv:2412.02527).

## Load the data

```python
from datasets import load_dataset
ds = load_dataset("UniverseTBD/mmu_csp_csp")
print(ds)
```

## Data structure

File types in sampled files: PNG, Parquet

Example file: `mmu_csp_csp/dataset/Norder=10/Dir=0/Npix=255.parquet`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/3e4b4dfa-68ae-5735-ab0c-cc7f7be0d903
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/3e4b4dfa-68ae-5735-ab0c-cc7f7be0d903
- Over MCP: `get_schema_card` with id `3e4b4dfa-68ae-5735-ab0c-cc7f7be0d903` at https://api.scigantic.com/mcp

## Related topics

- [Parquet datasets](https://scigantic.com/topics/parquet)
- [Time series datasets](https://scigantic.com/topics/time-series)
- [Astronomy & Space datasets](https://scigantic.com/topics/astronomy-and-space)
- [Spectroscopy datasets](https://scigantic.com/topics/spectroscopy)
- [PNG datasets](https://scigantic.com/topics/png)
