# The Well: Acoustic Scattering with Random Inclusions

## Facts

- Provider: Polymathic AI (The Well); Michael McCabe
- License: CC-BY-4.0 (https://arxiv.org/html/2412.00568)
- Size: 47 files
- Format: HDF5 (in sampled files)
- Stored at: Hugging Face, https://huggingface.co/datasets/polymathic-ai/acoustic_scattering_inclusions
- Access: public, readable without an account
- Category: Physics
- Source: https://arxiv.org/abs/2412.00568
- Page: https://scigantic.com/archive/fd7db04b-03e6-5f1f-9f32-1c25c6324d90
- Keywords: Physics, HDF5, Hugging Face

## About

2D simulations of the variable-coefficient linear acoustic equations, which describe a pressure wave propagating through materials with different density and sound speed. Pressure waves from a variable number of point sources travel through a domain split by a discontinuous interface and seeded with randomly generated inclusions of very different density, a setting relevant to inverse scattering in geology. Solved with Clawpack on a uniform grid over [-1, 1] by [-1, 1], open in y with reflective walls in x; each trajectory has 101 steps of 256 by 256 images over t from 0 to 2. Fields are pressure, material density, material sound speed and velocity. The datasheet lists 4,000 trajectories. Listing data/ (measured 2026-11-13) shows train/ (28 files, 227 GB), valid/ and test/ (4 files, 32.5 GB each), stored as chunked files of many trajectories; about 308 GB in the repository. Files are HDF5 in the common The Well format; open them with the_well package (WellDataModule with base path hf://datasets/polymathic-ai/) or directly with h5py. Part of The Well (Ohana et al., NeurIPS 2024 Datasets and Benchmarks, arXiv:2412.00568); released under CC-BY-4.0, cite The Well and the simulation's source papers.

## Load the data

```python
from huggingface_hub import hf_hub_download
path = hf_hub_download(repo_id="polymathic-ai/acoustic_scattering_inclusions", filename="data/test/acoustic_scattering_inclusions_chunk_36.hdf5", repo_type="dataset")
print(path)
```

## Data structure

File types in sampled files: HDF5

Example file: `data/test/acoustic_scattering_inclusions_chunk_36.hdf5`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/fd7db04b-03e6-5f1f-9f32-1c25c6324d90
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/fd7db04b-03e6-5f1f-9f32-1c25c6324d90
- Over MCP: `get_schema_card` with id `fd7db04b-03e6-5f1f-9f32-1c25c6324d90` at https://api.scigantic.com/mcp

## Related topics

- [HDF5 datasets](https://scigantic.com/topics/hdf5)
- [Physics datasets](https://scigantic.com/topics/physics)
