load-atoms

raw JSON →
0.3.10 verified Mon Apr 27 auth: no python

Large Open Access Datasets for Atomistic Materials Science (LOAD-AtoMS). Provides simple, reliable access to curated atomistic datasets (e.g., structures, trajectories) for materials science and machine learning. Current version 0.3.10, active development, frequent releases.

pip install load-atoms
error ModuleNotFoundError: No module named 'load_atoms'
cause Library not installed.
fix
Run pip install load-atoms
error ModuleNotFoundError: No module named 'load_atoms' when trying to import after installation.
cause Package name is 'load-atoms' but import is 'load_atoms' (underscore vs hyphen). Installed correctly, import may fail if environment is wrong.
fix
Ensure you installed the correct package and Python environment is consistent. Use pip list | grep load-atoms to confirm.
error ConnectionError: Failed to download dataset 'QM9'
cause Network issue or dataset server down.
fix
Check internet connection or try again later. Alternatively, set cache_dir to a path with existing download.
error ValueError: No dataset with id 'qm9'
cause Dataset ID is case-sensitive; 'qm9' is not recognized (correct is 'QM9').
fix
Use list_ids() to see available IDs, then call with exact string.
gotcha Some datasets are large (e.g., QM9) and may take time to download and cache on first use. Subsequent calls use the cache.
fix Be patient on first load; use `cache_dir` parameter to control cache location.
deprecated Support for Python 3.8 may be dropped in future releases.
fix Upgrade to Python >=3.9 when possible.
gotcha The 'Plinder' and 'OpenBabel' features are optional extras. Installing without them may cause import errors for certain datasets.
fix Install with `pip install load-atoms[plinder]` or `pip install load-atoms[openbabel]` as needed.
gotcha Dataset IDs are case-sensitive. For example, 'QM9' works but 'qm9' may not.
fix Use `list_ids()` to get exact dataset IDs.
pip install load-atoms[ase]
pip install load-atoms[plinder]
pip install load-atoms[openbabel]

List datasets, load one, and access atoms.

from load_atoms import load_dataset, list_datasets

# List available datasets
print(list_datasets())

# Load a dataset (e.g., 'QM9')
dataset = load_dataset('QM9')
print(f"Loaded {len(dataset)} structures")

# Access first structure
atoms = dataset[0]
print(atoms)