# The Well: Gray-Scott Reaction-Diffusion Turing Patterns

## Facts

- Provider: Polymathic AI (The Well); Daniel Fortunato, Flatiron Institute CCM and CCB
- License: CC-BY-4.0 (https://arxiv.org/html/2412.00568)
- Size: 27 files
- Format: HDF5 (in sampled files)
- Stored at: Hugging Face, https://huggingface.co/datasets/polymathic-ai/gray_scott_reaction_diffusion
- Access: public, readable without an account
- Category: Physics
- Source: https://arxiv.org/abs/2412.00568
- Page: https://scigantic.com/archive/6faf177f-dd95-5a08-86a4-86ff5d51878b
- Keywords: Physics, HDF5, Hugging Face

## About

2D simulations of the Gray-Scott equations, coupled reaction-diffusion equations for two chemical species A and B whose feed rate f and kill rate k select qualitatively different patterns: gliders, bubbles, mazes, worms, spirals and spots. Solved in MATLAB with Chebfun using a Fourier spectral method in space and ETDRK4 time stepping on a periodic [-1, 1] by [-1, 1] domain; each trajectory has 1,001 time steps of 128 by 128 images to t = 10,000. Fields are the concentrations of A and B. The datasheet lists 1,200 trajectories (6 parameter sets times 200 initial conditions) and flags trajectories that reach a stationary state. Listing data/ (measured 2026-11-13) shows train/ (6 files, 126 GB, one per pattern regime, named by f and k), valid/ and test/ (6 files, 15.9 GB each); about 158 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/gray_scott_reaction_diffusion", filename="data/test/gray_scott_reaction_diffusion_bubbles_F_0.098_k_0.057.hdf5", repo_type="dataset")
print(path)
```

## Data structure

File types in sampled files: HDF5

Example file: `data/test/gray_scott_reaction_diffusion_bubbles_F_0.098_k_0.057.hdf5`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/6faf177f-dd95-5a08-86a4-86ff5d51878b
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/6faf177f-dd95-5a08-86a4-86ff5d51878b
- Over MCP: `get_schema_card` with id `6faf177f-dd95-5a08-86a4-86ff5d51878b` at https://api.scigantic.com/mcp

## Related topics

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