candidate-generator

Generate inorganic crystal structures by providing seed inputs like spacegroup and composition, then run the candidate generation pipeline to produce diverse structures for DFT screening and ML dataset construction.

Yes, you can explore chemical space using substitutions, ion exchanges, and defect generation to augment ensembles of crystal structures, with modular integration supporting pymatgen and ASE workflows.

Build a high-throughput dataset by generating seed structures from composition, expanding the candidate pool through chemical space exploration and ensemble augmentation, and producing ready-to-store structures validated for DFT screening.

The workflow integrates modularly with pymatgen, ASE, and Materials Project to produce ready-to-store inorganic crystal structures, enforcing safe defaults and input validation across high-throughput and ML training pipelines.

Generate seed structures from spacegroup and composition when you need to initiate a candidate pool for inorganic crystal discovery before applying substitutions, defect generation, and ensemble augmentation.

Disorder resolution is supported within the candidate generation pipeline to produce diverse inorganic structures, but the workflow enforces safe defaults and input validation, so complex disorder cases may require manual review before DFT screening.