Dataset for "Machine learning for a finite size correction in periodic coupled cluster theory calculations"

This data set contains input and output files used for generating CCSD-FS-GPR energy predictions for the uniform electron gas at rs=5 (density parameter), and the bcc phase of lithium and sodium metals. The input data includes the transition structure factors used to train Gaussian process regression (GPR) models, as well as the electron repulsion integrals, momentum transfer vector magnitudes, and momentum transfer counts needed to calculate coupled cluster singles and doubles (CCSD) correlation energies. Output files include GPR predicted structure factors with the corresponding momentum transfer vector magnitudes on both discrete and continuous domains. These outputs also contain energy predictions calculated using the GPR-predicted structure factors. Energy predictions extrapolated to zero on a 1/N domain (where N is electron number) or a 1/k^3 domain (where k refers to the k-points used for real systems). These extrapolations are referred to as CCSD-FS-GPR. CCSD and finite-size corrected CCSD (CCSD-FS) are provided as our benchmarks for comparison.