torchcell.knowledge_graphs.build_time_projection.project_build_time

project_build_time(rates: list[AdapterRate], dataset_full_records: dict[str, int], subset_config: SubsetConfig, *, label: str = 'projection', measured_sec: float | None = None) → BuildTimeProjection

Project total build time for subset_config from calibrated rates.

projected_sec_i = rate_i * min(cap_i, full_records_i) summed over adapters. The breakdown is sorted by descending contribution. When measured_sec is given (e.g. the kg_full self-check), the percent error is reported.