A computational framework and software tool for generating complex survival data with predefined censoring rates in simulation studies

An increasing number of survival analysis methods require validation under complex simulation scenarios. However, existing data generation tools are often constrained in practical applications: they typically simulate simplified censoring mechanisms and lack the flexibility to achieve predefined censoring rates when dealing with complex survival distributions. This study aims to develop a comprehensive computational framework and reproducible software to generate simulated data that meets more diverse needs. Rooted in the concept of numerical inversion, our framework employs automated algorithms to solve closed-form equations for the key parameters of censoring distributions. This enables the precise derivation of parameters required to achieve researcher-specified censoring rates, thereby generating simulated datasets that faithfully mirror complex survival distributions. This tool integrates four core functionalities: constructing complex survival distributions based on finite mixture models; simultaneously handling Type I and Type III censoring data; precisely calibrating user-defined censoring rates; and incorporating covariate effects within survival time distributions. We validated the tool across 64 simulated scenarios, systematically varying survival distributions, censoring distributions, predefined censoring rates, and covariate structures. Simulation results across 64 scenarios demonstrated the robustness and precision of the proposed toolkit. The average observed censoring rates from 1000 replications showed negligible deviation from the predefined targets. The stability of the generation process was confirmed, with the standard deviations of total censoring rates consistently remaining below 1.7% across all simulated settings. The study provides a standardized computational framework by automating the generation of complex survival data with precise censoring control, thereby enabling researchers to conduct simulation studies across diverse survival analysis scenarios and providing essential support for the rigorous validation of statistical methods.