Bridging constrained random-phase approximation and linear response theory for computing Hubbard parameters

Accurately parameterizing the screened Coulomb interaction U is essential for predictive density-functional theory extensions, such as DFT+U and DFT plus dynamical mean-field theory (DFT+DMFT), which model strongly correlated electron systems. This dataset provides the comprehensive computational records for a systematic study mapping the physical equivalence between the two primary first-principles methods for determining this parameter: linear response theory (LRT) and the constrained random-phase approximation (cRPA). Using a unified framework of maximally localized Wannier functions applied to four prototypical transition-metal systems (CrO2, KCuF3, NiO, and Sr2FeO4), the work compares systematically the two methods for determining the U value.

To support the verification and extension of these physical insights, the repository contains the complete input and output files for the underlying quantum mechanical simulations. This includes cRPA calculations partitioned into distinct target orbital configurations, spanning d-only, d+p, and frontier Wannier subspaces, alongside LRT datasets computed both with and without exchange-correlation effects. The entire execution pipeline is preserved through shell scripts that document the automated workflows.