# Cd- and Zn-Doped CsPbI3 Perovskite Energetics: DFT Training Set for GNN Stability Prediction (Eremin et al., AIRI), ColabFit Conversion

## Facts

- Provider: Roman Eremin, Innokentiy Humonen, Alexey Kazakov, Vladimir Lazarev, Anatoly Pushkarev, Semen Budennyy (AIRI); ColabFit conversion
- License: MIT (https://github.com/AIRI-Institute/doped_CsPbI3_energetics/blob/main/LICENSE)
- Size: 6 files
- Format: Parquet (in sampled files)
- Stored at: Hugging Face, https://huggingface.co/datasets/colabfit/doped_CsPbI3_energetics_train_validate
- Access: public, readable without an account
- Category: Chemistry & Materials
- Source: https://doi.org/10.1016/j.commatsci.2023.112672
- Page: https://scigantic.com/archive/5b4cc201-c104-539a-8c84-ce28e5727ec1
- Keywords: Chemistry & Materials, DFT, graph learning, Parquet, Hugging Face

## About

DFT energies, formation energies and atomic forces for gamma-CsPbI3, the inorganic lead halide perovskite, with lead partly substituted by cadmium or zinc, in two polymorphic modifications. Doping CsPbI3 is a route to a stable semiconductor for optoelectronic devices; the authors treat the doped crystals as disordered structures, compute low-substitution configurations with DFT and train graph neural networks (GNN) to predict the structural stability and energetics of more heavily doped structures. This repository is the training plus validation split; the sibling ColabFit repository holds the test split. ColabFit's card counts 140 configurations and 22,400 atoms, labelled with energy, formation energy and atomic forces; about 3 MB. Files are the ColabFit Exchange conversion: Parquet files under co/ hold one row per configuration (atomic numbers, positions, cell, computed properties and metadata), ds.parquet holds dataset-level information, and cs/ groups configurations into named sets where defined; read them with pyarrow, pandas, polars or Hugging Face datasets. The AIRI-Institute GitHub repository is released under the MIT License; cite Eremin et al., Computational Materials Science 232, 112672 (2024).

## Load the data

```python
from datasets import load_dataset
ds = load_dataset("colabfit/doped_CsPbI3_energetics_train_validate")
print(ds)
```

## Data structure

File types in sampled files: Parquet

Example file: `co/co_0.parquet`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/5b4cc201-c104-539a-8c84-ce28e5727ec1
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/5b4cc201-c104-539a-8c84-ce28e5727ec1
- Over MCP: `get_schema_card` with id `5b4cc201-c104-539a-8c84-ce28e5727ec1` at https://api.scigantic.com/mcp

## Related topics

- [Parquet datasets](https://scigantic.com/topics/parquet)
- [Chemistry & Materials datasets](https://scigantic.com/topics/chemistry-and-materials)
- [DFT datasets](https://scigantic.com/topics/dft)
- [Graph learning datasets](https://scigantic.com/topics/graph-learning)
