Dataset for: Improving parton shower predictions via precision moments of energy flow polynomials

Event samples and derived per-event tables for the paper "Improving parton shower predictions via precision moments of energy flow polynomials" (Assi, Lee, Thaler). Contents: per-event EFP tables (214 observables by 10^6 events) for the Sherpa target and the three degraded priors (central alpha_s and plus or minus 20 percent), event-shape and N-jettiness tables, all trained Lagrange-multiplier sets, the 10^6-event Herwig 7.3 sample in HepMC3 with its run card and trained weights, heavy-hemisphere substructure tables covering Durham subjets, the Lund plane, charged multiplicity, and fragmentation, and the degree-6 EFP naming calibration map. Together with the code at https://github.com/benleo12/efp_maxent these reproduce all figures in the paper.