This dataset contains the TPC-DS benchmark data generated for evaluating DPJoin, a cost-based, replication-aware query optimizer for distributed analytical workloads. The dataset includes three scale factors: SF1 (~0.4 GB), SF10 (~4 GB), and SF50 (~16 GB), each generated using the official TPC-DS data generation toolkit (https://www.tpc.org/tpcds/) without modification. These datasets were used to run all performance experiments in the DPJoin study, including comparisons with RelJoin, AQE, ShuffleHashJoin, and other distributed join strategies. The provided files reproduce the exact experimental environment of the paper, enabling full transparency, independent verification, and reusability. All scale-factor directories preserve the original TPC-DS schema and table formats. The data is suitable for research on distributed query optimization, join algorithms, replication strategies, cost modeling, and large-scale analytics.